Nothing too exciting in this lot :-) The following changes since commit ba0fa56bc06e563de68d2a2bf3ddb0cfea1be4f9: Merge remote-tracking branch 'remotes/vivier/tags/q800-for-6.2-pull-request' into staging (2021-09-29 21:20:49 +0100) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20210930 for you to fetch changes up to 1f4b2ec701b9d73d3fa7bb90c8b4376bc7d3c42b: hw/arm: sabrelite: Connect SPI flash CS line to GPIO3_19 (2021-09-30 13:44:13 +0100) ---------------------------------------------------------------- target-arm queue: * allwinner-h3: Switch to SMC as PSCI conduit * arm: tcg: Adhere to SMCCC 1.3 section 5.2 * xlnx-zcu102, xlnx-versal-virt: Support BBRAM and eFUSE devices * gdbstub related code cleanups * Don't put FPEXC and FPSID in org.gnu.gdb.arm.vfp XML * Use _init vs _new convention in bus creation function names * sabrelite: Connect SPI flash CS line to GPIO3_19 ---------------------------------------------------------------- Alexander Graf (2): allwinner-h3: Switch to SMC as PSCI conduit arm: tcg: Adhere to SMCCC 1.3 section 5.2 Peter Maydell (10): configs: Don't include 32-bit-only GDB XML in aarch64 linux configs target/arm: Fix coding style issues in gdbstub code in helper.c target/arm: Move gdbstub related code out of helper.c target/arm: Don't put FPEXC and FPSID in org.gnu.gdb.arm.vfp XML scsi: Replace scsi_bus_new() with scsi_bus_init(), scsi_bus_init_named() ipack: Rename ipack_bus_new_inplace() to ipack_bus_init() pci: Rename pci_root_bus_new_inplace() to pci_root_bus_init() qbus: Rename qbus_create_inplace() to qbus_init() qbus: Rename qbus_create() to qbus_new() ide: Rename ide_bus_new() to ide_bus_init() Tong Ho (9): hw/nvram: Introduce Xilinx eFuse QOM hw/nvram: Introduce Xilinx Versal eFuse device hw/nvram: Introduce Xilinx ZynqMP eFuse device hw/nvram: Introduce Xilinx battery-backed ram hw/arm: xlnx-versal-virt: Add Xilinx BBRAM device hw/arm: xlnx-versal-virt: Add Xilinx eFUSE device hw/arm: xlnx-zcu102: Add Xilinx BBRAM device hw/arm: xlnx-zcu102: Add Xilinx eFUSE device docs/system/arm: xlnx-versal-virt: BBRAM and eFUSE Usage Xuzhou Cheng (1): hw/arm: sabrelite: Connect SPI flash CS line to GPIO3_19 docs/system/arm/xlnx-versal-virt.rst | 49 ++ configs/targets/aarch64-linux-user.mak | 2 +- configs/targets/aarch64-softmmu.mak | 2 +- configs/targets/aarch64_be-linux-user.mak | 2 +- configs/targets/arm-linux-user.mak | 2 +- configs/targets/arm-softmmu.mak | 2 +- configs/targets/armeb-linux-user.mak | 2 +- include/hw/arm/xlnx-versal.h | 15 + include/hw/arm/xlnx-zynqmp.h | 5 + include/hw/ide/internal.h | 4 +- include/hw/ipack/ipack.h | 8 +- include/hw/nvram/xlnx-bbram.h | 54 ++ include/hw/nvram/xlnx-efuse.h | 132 +++++ include/hw/nvram/xlnx-versal-efuse.h | 68 +++ include/hw/nvram/xlnx-zynqmp-efuse.h | 44 ++ include/hw/pci/pci.h | 10 +- include/hw/qdev-core.h | 6 +- include/hw/scsi/scsi.h | 30 +- target/arm/internals.h | 7 + hw/arm/allwinner-h3.c | 2 +- hw/arm/sabrelite.c | 2 +- hw/arm/xlnx-versal-virt.c | 88 +++ hw/arm/xlnx-versal.c | 57 ++ hw/arm/xlnx-zcu102.c | 30 ++ hw/arm/xlnx-zynqmp.c | 49 ++ hw/audio/intel-hda.c | 2 +- hw/block/fdc.c | 2 +- hw/block/swim.c | 3 +- hw/char/virtio-serial-bus.c | 4 +- hw/core/bus.c | 13 +- hw/core/sysbus.c | 10 +- hw/gpio/bcm2835_gpio.c | 3 +- hw/hyperv/vmbus.c | 2 +- hw/i2c/core.c | 2 +- hw/ide/ahci.c | 2 +- hw/ide/cmd646.c | 2 +- hw/ide/isa.c | 2 +- hw/ide/macio.c | 2 +- hw/ide/microdrive.c | 2 +- hw/ide/mmio.c | 2 +- hw/ide/piix.c | 2 +- hw/ide/qdev.c | 4 +- hw/ide/sii3112.c | 2 +- hw/ide/via.c | 2 +- hw/ipack/ipack.c | 10 +- hw/ipack/tpci200.c | 4 +- hw/isa/isa-bus.c | 2 +- hw/misc/auxbus.c | 2 +- hw/misc/mac_via.c | 4 +- hw/misc/macio/cuda.c | 4 +- hw/misc/macio/macio.c | 4 +- hw/misc/macio/pmu.c | 4 +- hw/nubus/nubus-bridge.c | 2 +- hw/nvme/ctrl.c | 4 +- hw/nvme/subsys.c | 3 +- hw/nvram/xlnx-bbram.c | 545 +++++++++++++++++++ hw/nvram/xlnx-efuse-crc.c | 119 +++++ hw/nvram/xlnx-efuse.c | 280 ++++++++++ hw/nvram/xlnx-versal-efuse-cache.c | 114 ++++ hw/nvram/xlnx-versal-efuse-ctrl.c | 783 +++++++++++++++++++++++++++ hw/nvram/xlnx-zynqmp-efuse.c | 855 ++++++++++++++++++++++++++++++ hw/pci-host/raven.c | 4 +- hw/pci-host/versatile.c | 6 +- hw/pci/pci.c | 30 +- hw/pci/pci_bridge.c | 4 +- hw/ppc/spapr_vio.c | 2 +- hw/s390x/ap-bridge.c | 2 +- hw/s390x/css-bridge.c | 2 +- hw/s390x/event-facility.c | 4 +- hw/s390x/s390-pci-bus.c | 2 +- hw/s390x/virtio-ccw.c | 3 +- hw/scsi/esp-pci.c | 2 +- hw/scsi/esp.c | 2 +- hw/scsi/lsi53c895a.c | 2 +- hw/scsi/megasas.c | 3 +- hw/scsi/mptsas.c | 2 +- hw/scsi/scsi-bus.c | 6 +- hw/scsi/spapr_vscsi.c | 3 +- hw/scsi/virtio-scsi.c | 4 +- hw/scsi/vmw_pvscsi.c | 3 +- hw/sd/allwinner-sdhost.c | 4 +- hw/sd/bcm2835_sdhost.c | 4 +- hw/sd/pl181.c | 3 +- hw/sd/pxa2xx_mmci.c | 4 +- hw/sd/sdhci.c | 3 +- hw/sd/ssi-sd.c | 3 +- hw/ssi/ssi.c | 2 +- hw/usb/bus.c | 2 +- hw/usb/dev-smartcard-reader.c | 3 +- hw/usb/dev-storage-bot.c | 3 +- hw/usb/dev-storage-classic.c | 4 +- hw/usb/dev-uas.c | 3 +- hw/virtio/virtio-mmio.c | 3 +- hw/virtio/virtio-pci.c | 3 +- hw/xen/xen-bus.c | 2 +- hw/xen/xen-legacy-backend.c | 2 +- target/arm/gdbstub.c | 154 ++++++ target/arm/gdbstub64.c | 140 +++++ target/arm/helper.c | 262 --------- target/arm/psci.c | 35 +- gdb-xml/arm-neon.xml | 2 - gdb-xml/arm-vfp-sysregs.xml | 17 + gdb-xml/arm-vfp.xml | 2 - gdb-xml/arm-vfp3.xml | 2 - hw/Kconfig | 2 + hw/arm/Kconfig | 2 + hw/nvram/Kconfig | 19 + hw/nvram/meson.build | 8 + 108 files changed, 3806 insertions(+), 447 deletions(-) create mode 100644 include/hw/nvram/xlnx-bbram.h create mode 100644 include/hw/nvram/xlnx-efuse.h create mode 100644 include/hw/nvram/xlnx-versal-efuse.h create mode 100644 include/hw/nvram/xlnx-zynqmp-efuse.h create mode 100644 hw/nvram/xlnx-bbram.c create mode 100644 hw/nvram/xlnx-efuse-crc.c create mode 100644 hw/nvram/xlnx-efuse.c create mode 100644 hw/nvram/xlnx-versal-efuse-cache.c create mode 100644 hw/nvram/xlnx-versal-efuse-ctrl.c create mode 100644 hw/nvram/xlnx-zynqmp-efuse.c create mode 100644 gdb-xml/arm-vfp-sysregs.xml From: Alexander Graf <agraf@csgraf.de> The Allwinner H3 SoC uses Cortex-A7 cores which support virtualization. However, today we are configuring QEMU to use HVC as PSCI conduit. That means HVC calls get trapped into QEMU instead of the guest's own emulated CPU and thus break the guest's ability to execute virtualization. Fix this by moving to SMC as conduit, freeing up HYP completely to the VM. Signed-off-by: Alexander Graf <agraf@csgraf.de> Message-id: 20210920203931.66527-1-agraf@csgraf.de Fixes: 740dafc0ba0 ("hw/arm: add Allwinner H3 System-on-Chip") Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com> Tested-by: Niek Linnenbank <nieklinnenbank@gmail.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/allwinner-h3.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/arm/allwinner-h3.c b/hw/arm/allwinner-h3.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/allwinner-h3.c +++ b/hw/arm/allwinner-h3.c @@ -XXX,XX +XXX,XX @@ static void allwinner_h3_realize(DeviceState *dev, Error **errp) /* Provide Power State Coordination Interface */ qdev_prop_set_int32(DEVICE(&s->cpus[i]), "psci-conduit", - QEMU_PSCI_CONDUIT_HVC); + QEMU_PSCI_CONDUIT_SMC); /* Disable secondary CPUs */ qdev_prop_set_bit(DEVICE(&s->cpus[i]), "start-powered-off", -- 2.20.1 From: Alexander Graf <agraf@csgraf.de> The SMCCC 1.3 spec section 5.2 says The Unknown SMC Function Identifier is a sign-extended value of (-1) that is returned in the R0, W0 or X0 registers. An implementation must return this error code when it receives: * An SMC or HVC call with an unknown Function Identifier * An SMC or HVC call for a removed Function Identifier * An SMC64/HVC64 call from AArch32 state To comply with these statements, let's always return -1 when we encounter an unknown HVC or SMC call. Signed-off-by: Alexander Graf <agraf@csgraf.de> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/psci.c | 35 ++++++----------------------------- 1 file changed, 6 insertions(+), 29 deletions(-) diff --git a/target/arm/psci.c b/target/arm/psci.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/psci.c +++ b/target/arm/psci.c @@ -XXX,XX +XXX,XX @@ bool arm_is_psci_call(ARMCPU *cpu, int excp_type) { - /* Return true if the r0/x0 value indicates a PSCI call and - * the exception type matches the configured PSCI conduit. This is - * called before the SMC/HVC instruction is executed, to decide whether - * we should treat it as a PSCI call or with the architecturally + /* + * Return true if the exception type matches the configured PSCI conduit. + * This is called before the SMC/HVC instruction is executed, to decide + * whether we should treat it as a PSCI call or with the architecturally * defined behaviour for an SMC or HVC (which might be UNDEF or trap * to EL2 or to EL3). */ - CPUARMState *env = &cpu->env; - uint64_t param = is_a64(env) ? env->xregs[0] : env->regs[0]; switch (excp_type) { case EXCP_HVC: @@ -XXX,XX +XXX,XX @@ bool arm_is_psci_call(ARMCPU *cpu, int excp_type) return false; } - switch (param) { - case QEMU_PSCI_0_2_FN_PSCI_VERSION: - case QEMU_PSCI_0_2_FN_MIGRATE_INFO_TYPE: - case QEMU_PSCI_0_2_FN_AFFINITY_INFO: - case QEMU_PSCI_0_2_FN64_AFFINITY_INFO: - case QEMU_PSCI_0_2_FN_SYSTEM_RESET: - case QEMU_PSCI_0_2_FN_SYSTEM_OFF: - case QEMU_PSCI_0_1_FN_CPU_ON: - case QEMU_PSCI_0_2_FN_CPU_ON: - case QEMU_PSCI_0_2_FN64_CPU_ON: - case QEMU_PSCI_0_1_FN_CPU_OFF: - case QEMU_PSCI_0_2_FN_CPU_OFF: - case QEMU_PSCI_0_1_FN_CPU_SUSPEND: - case QEMU_PSCI_0_2_FN_CPU_SUSPEND: - case QEMU_PSCI_0_2_FN64_CPU_SUSPEND: - case QEMU_PSCI_0_1_FN_MIGRATE: - case QEMU_PSCI_0_2_FN_MIGRATE: - return true; - default: - return false; - } + return true; } void arm_handle_psci_call(ARMCPU *cpu) @@ -XXX,XX +XXX,XX @@ void arm_handle_psci_call(ARMCPU *cpu) break; case QEMU_PSCI_0_1_FN_MIGRATE: case QEMU_PSCI_0_2_FN_MIGRATE: + default: ret = QEMU_PSCI_RET_NOT_SUPPORTED; break; - default: - g_assert_not_reached(); } err: -- 2.20.1 From: Tong Ho <tong.ho@xilinx.com> This introduces the QOM for Xilinx eFuse, an one-time field-programmable storage bit array. The actual mmio interface to the array varies by device families and will be provided in different change-sets. Co-authored-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Co-authored-by: Sai Pavan Boddu <sai.pavan.boddu@xilinx.com> Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Signed-off-by: Sai Pavan Boddu <sai.pavan.boddu@xilinx.com> Signed-off-by: Tong Ho <tong.ho@xilinx.com> Message-id: 20210917052400.1249094-2-tong.ho@xilinx.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/nvram/xlnx-efuse.h | 132 ++++++++++++++++ hw/nvram/xlnx-efuse-crc.c | 119 +++++++++++++++ hw/nvram/xlnx-efuse.c | 280 ++++++++++++++++++++++++++++++++++ hw/nvram/Kconfig | 7 + hw/nvram/meson.build | 2 + 5 files changed, 540 insertions(+) create mode 100644 include/hw/nvram/xlnx-efuse.h create mode 100644 hw/nvram/xlnx-efuse-crc.c create mode 100644 hw/nvram/xlnx-efuse.c diff --git a/include/hw/nvram/xlnx-efuse.h b/include/hw/nvram/xlnx-efuse.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/nvram/xlnx-efuse.h @@ -XXX,XX +XXX,XX @@ +/* + * QEMU model of the Xilinx eFuse core + * + * Copyright (c) 2015 Xilinx Inc. + * + * Written by Edgar E. Iglesias <edgari@xilinx.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#ifndef XLNX_EFUSE_H +#define XLNX_EFUSE_H + +#include "sysemu/block-backend.h" +#include "hw/qdev-core.h" + +#define TYPE_XLNX_EFUSE "xlnx,efuse" +OBJECT_DECLARE_SIMPLE_TYPE(XlnxEFuse, XLNX_EFUSE); + +struct XlnxEFuse { + DeviceState parent_obj; + BlockBackend *blk; + bool blk_ro; + uint32_t *fuse32; + + DeviceState *dev; + + bool init_tbits; + + uint8_t efuse_nr; + uint32_t efuse_size; + + uint32_t *ro_bits; + uint32_t ro_bits_cnt; +}; + +/** + * xlnx_efuse_calc_crc: + * @data: an array of 32-bit words for which the CRC should be computed + * @u32_cnt: the array size in number of 32-bit words + * @zpads: the number of 32-bit zeros prepended to @data before computation + * + * This function is used to compute the CRC for an array of 32-bit words, + * using a Xilinx-specific data padding. + * + * Returns: the computed 32-bit CRC + */ +uint32_t xlnx_efuse_calc_crc(const uint32_t *data, unsigned u32_cnt, + unsigned zpads); + +/** + * xlnx_efuse_get_bit: + * @s: the efuse object + * @bit: the efuse bit-address to read the data + * + * Returns: the bit, 0 or 1, at @bit of object @s + */ +bool xlnx_efuse_get_bit(XlnxEFuse *s, unsigned int bit); + +/** + * xlnx_efuse_set_bit: + * @s: the efuse object + * @bit: the efuse bit-address to be written a value of 1 + * + * Returns: true on success, false on failure + */ +bool xlnx_efuse_set_bit(XlnxEFuse *s, unsigned int bit); + +/** + * xlnx_efuse_k256_check: + * @s: the efuse object + * @crc: the 32-bit CRC to be compared with + * @start: the efuse bit-address (which must be multiple of 32) of the + * start of a 256-bit array + * + * This function computes the CRC of a 256-bit array starting at @start + * then compares to the given @crc + * + * Returns: true of @crc == computed, false otherwise + */ +bool xlnx_efuse_k256_check(XlnxEFuse *s, uint32_t crc, unsigned start); + +/** + * xlnx_efuse_tbits_check: + * @s: the efuse object + * + * This function inspects a number of efuse bits at specific addresses + * to see if they match a validation pattern. Each pattern is a group + * of 4 bits, and there are 3 groups. + * + * Returns: a 3-bit mask, where a bit of '1' means the corresponding + * group has a valid pattern. + */ +uint32_t xlnx_efuse_tbits_check(XlnxEFuse *s); + +/** + * xlnx_efuse_get_row: + * @s: the efuse object + * @bit: the efuse bit address for which a 32-bit value is read + * + * Returns: the entire 32 bits of the efuse, starting at a bit + * address that is multiple of 32 and contains the bit at @bit + */ +static inline uint32_t xlnx_efuse_get_row(XlnxEFuse *s, unsigned int bit) +{ + if (!(s->fuse32)) { + return 0; + } else { + unsigned int row_idx = bit / 32; + + assert(row_idx < (s->efuse_size * s->efuse_nr / 32)); + return s->fuse32[row_idx]; + } +} + +#endif diff --git a/hw/nvram/xlnx-efuse-crc.c b/hw/nvram/xlnx-efuse-crc.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/nvram/xlnx-efuse-crc.c @@ -XXX,XX +XXX,XX @@ +/* + * Xilinx eFuse/bbram CRC calculator + * + * Copyright (c) 2021 Xilinx Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ +#include "qemu/osdep.h" +#include "hw/nvram/xlnx-efuse.h" + +static uint32_t xlnx_efuse_u37_crc(uint32_t prev_crc, uint32_t data, + uint32_t addr) +{ + /* A table for 7-bit slicing */ + static const uint32_t crc_tab[128] = { + 0x00000000, 0xe13b70f7, 0xc79a971f, 0x26a1e7e8, + 0x8ad958cf, 0x6be22838, 0x4d43cfd0, 0xac78bf27, + 0x105ec76f, 0xf165b798, 0xd7c45070, 0x36ff2087, + 0x9a879fa0, 0x7bbcef57, 0x5d1d08bf, 0xbc267848, + 0x20bd8ede, 0xc186fe29, 0xe72719c1, 0x061c6936, + 0xaa64d611, 0x4b5fa6e6, 0x6dfe410e, 0x8cc531f9, + 0x30e349b1, 0xd1d83946, 0xf779deae, 0x1642ae59, + 0xba3a117e, 0x5b016189, 0x7da08661, 0x9c9bf696, + 0x417b1dbc, 0xa0406d4b, 0x86e18aa3, 0x67dafa54, + 0xcba24573, 0x2a993584, 0x0c38d26c, 0xed03a29b, + 0x5125dad3, 0xb01eaa24, 0x96bf4dcc, 0x77843d3b, + 0xdbfc821c, 0x3ac7f2eb, 0x1c661503, 0xfd5d65f4, + 0x61c69362, 0x80fde395, 0xa65c047d, 0x4767748a, + 0xeb1fcbad, 0x0a24bb5a, 0x2c855cb2, 0xcdbe2c45, + 0x7198540d, 0x90a324fa, 0xb602c312, 0x5739b3e5, + 0xfb410cc2, 0x1a7a7c35, 0x3cdb9bdd, 0xdde0eb2a, + 0x82f63b78, 0x63cd4b8f, 0x456cac67, 0xa457dc90, + 0x082f63b7, 0xe9141340, 0xcfb5f4a8, 0x2e8e845f, + 0x92a8fc17, 0x73938ce0, 0x55326b08, 0xb4091bff, + 0x1871a4d8, 0xf94ad42f, 0xdfeb33c7, 0x3ed04330, + 0xa24bb5a6, 0x4370c551, 0x65d122b9, 0x84ea524e, + 0x2892ed69, 0xc9a99d9e, 0xef087a76, 0x0e330a81, + 0xb21572c9, 0x532e023e, 0x758fe5d6, 0x94b49521, + 0x38cc2a06, 0xd9f75af1, 0xff56bd19, 0x1e6dcdee, + 0xc38d26c4, 0x22b65633, 0x0417b1db, 0xe52cc12c, + 0x49547e0b, 0xa86f0efc, 0x8ecee914, 0x6ff599e3, + 0xd3d3e1ab, 0x32e8915c, 0x144976b4, 0xf5720643, + 0x590ab964, 0xb831c993, 0x9e902e7b, 0x7fab5e8c, + 0xe330a81a, 0x020bd8ed, 0x24aa3f05, 0xc5914ff2, + 0x69e9f0d5, 0x88d28022, 0xae7367ca, 0x4f48173d, + 0xf36e6f75, 0x12551f82, 0x34f4f86a, 0xd5cf889d, + 0x79b737ba, 0x988c474d, 0xbe2da0a5, 0x5f16d052 + }; + + /* + * eFuse calculation is shown here: + * https://github.com/Xilinx/embeddedsw/blob/release-2019.2/lib/sw_services/xilskey/src/xilskey_utils.c#L1496 + * + * Each u32 word is appended a 5-bit value, for a total of 37 bits; see: + * https://github.com/Xilinx/embeddedsw/blob/release-2019.2/lib/sw_services/xilskey/src/xilskey_utils.c#L1356 + */ + uint32_t crc = prev_crc; + const unsigned rshf = 7; + const uint32_t im = (1 << rshf) - 1; + const uint32_t rm = (1 << (32 - rshf)) - 1; + const uint32_t i2 = (1 << 2) - 1; + const uint32_t r2 = (1 << 30) - 1; + + unsigned j; + uint32_t i, r; + uint64_t w; + + w = (uint64_t)(addr) << 32; + w |= data; + + /* Feed 35 bits, in 5 rounds, each a slice of 7 bits */ + for (j = 0; j < 5; j++) { + r = rm & (crc >> rshf); + i = im & (crc ^ w); + crc = crc_tab[i] ^ r; + + w >>= rshf; + } + + /* Feed the remaining 2 bits */ + r = r2 & (crc >> 2); + i = i2 & (crc ^ w); + crc = crc_tab[i << (rshf - 2)] ^ r; + + return crc; +} + +uint32_t xlnx_efuse_calc_crc(const uint32_t *data, unsigned u32_cnt, + unsigned zpads) +{ + uint32_t crc = 0; + unsigned index; + + for (index = zpads; index; index--) { + crc = xlnx_efuse_u37_crc(crc, 0, (index + u32_cnt)); + } + + for (index = u32_cnt; index; index--) { + crc = xlnx_efuse_u37_crc(crc, data[index - 1], index); + } + + return crc; +} diff --git a/hw/nvram/xlnx-efuse.c b/hw/nvram/xlnx-efuse.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/nvram/xlnx-efuse.c @@ -XXX,XX +XXX,XX @@ +/* + * QEMU model of the EFUSE eFuse + * + * Copyright (c) 2015 Xilinx Inc. + * + * Written by Edgar E. Iglesias <edgari@xilinx.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "hw/nvram/xlnx-efuse.h" + +#include "qemu/error-report.h" +#include "qemu/log.h" +#include "qapi/error.h" +#include "sysemu/blockdev.h" +#include "hw/qdev-properties.h" +#include "hw/qdev-properties-system.h" + +#define TBIT0_OFFSET 28 +#define TBIT1_OFFSET 29 +#define TBIT2_OFFSET 30 +#define TBIT3_OFFSET 31 +#define TBITS_PATTERN (0x0AU << TBIT0_OFFSET) +#define TBITS_MASK (0x0FU << TBIT0_OFFSET) + +bool xlnx_efuse_get_bit(XlnxEFuse *s, unsigned int bit) +{ + bool b = s->fuse32[bit / 32] & (1 << (bit % 32)); + return b; +} + +static int efuse_bytes(XlnxEFuse *s) +{ + return ROUND_UP((s->efuse_nr * s->efuse_size) / 8, 4); +} + +static int efuse_bdrv_read(XlnxEFuse *s, Error **errp) +{ + uint32_t *ram = s->fuse32; + int nr = efuse_bytes(s); + + if (!s->blk) { + return 0; + } + + s->blk_ro = !blk_supports_write_perm(s->blk); + if (!s->blk_ro) { + int rc; + + rc = blk_set_perm(s->blk, + (BLK_PERM_CONSISTENT_READ | BLK_PERM_WRITE), + BLK_PERM_ALL, NULL); + if (rc) { + s->blk_ro = true; + } + } + if (s->blk_ro) { + warn_report("%s: Skip saving updates to read-only eFUSE backstore.", + blk_name(s->blk)); + } + + if (blk_pread(s->blk, 0, ram, nr) < 0) { + error_setg(errp, "%s: Failed to read %u bytes from eFUSE backstore.", + blk_name(s->blk), nr); + return -1; + } + + /* Convert from little-endian backstore for each 32-bit row */ + nr /= 4; + while (nr--) { + ram[nr] = le32_to_cpu(ram[nr]); + } + + return 0; +} + +static void efuse_bdrv_sync(XlnxEFuse *s, unsigned int bit) +{ + unsigned int row_offset; + uint32_t le32; + + if (!s->blk || s->blk_ro) { + return; /* Silent on read-only backend to avoid message flood */ + } + + /* Backstore is always in little-endian */ + le32 = cpu_to_le32(xlnx_efuse_get_row(s, bit)); + + row_offset = (bit / 32) * 4; + if (blk_pwrite(s->blk, row_offset, &le32, 4, 0) < 0) { + error_report("%s: Failed to write offset %u of eFUSE backstore.", + blk_name(s->blk), row_offset); + } +} + +static int efuse_ro_bits_cmp(const void *a, const void *b) +{ + uint32_t i = *(const uint32_t *)a; + uint32_t j = *(const uint32_t *)b; + + return (i > j) - (i < j); +} + +static void efuse_ro_bits_sort(XlnxEFuse *s) +{ + uint32_t *ary = s->ro_bits; + const uint32_t cnt = s->ro_bits_cnt; + + if (ary && cnt > 1) { + qsort(ary, cnt, sizeof(ary[0]), efuse_ro_bits_cmp); + } +} + +static bool efuse_ro_bits_find(XlnxEFuse *s, uint32_t k) +{ + const uint32_t *ary = s->ro_bits; + const uint32_t cnt = s->ro_bits_cnt; + + if (!ary || !cnt) { + return false; + } + + return bsearch(&k, ary, cnt, sizeof(ary[0]), efuse_ro_bits_cmp) != NULL; +} + +bool xlnx_efuse_set_bit(XlnxEFuse *s, unsigned int bit) +{ + if (efuse_ro_bits_find(s, bit)) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: WARN: " + "Ignored setting of readonly efuse bit<%u,%u>!\n", + object_get_canonical_path(OBJECT(s)), + (bit / 32), (bit % 32)); + return false; + } + + s->fuse32[bit / 32] |= 1 << (bit % 32); + efuse_bdrv_sync(s, bit); + return true; +} + +bool xlnx_efuse_k256_check(XlnxEFuse *s, uint32_t crc, unsigned start) +{ + uint32_t calc; + + /* A key always occupies multiple of whole rows */ + assert((start % 32) == 0); + + calc = xlnx_efuse_calc_crc(&s->fuse32[start / 32], (256 / 32), 0); + return calc == crc; +} + +uint32_t xlnx_efuse_tbits_check(XlnxEFuse *s) +{ + int nr; + uint32_t check = 0; + + for (nr = s->efuse_nr; nr-- > 0; ) { + int efuse_start_row_num = (s->efuse_size * nr) / 32; + uint32_t data = s->fuse32[efuse_start_row_num]; + + /* + * If the option is on, auto-init blank T-bits. + * (non-blank will still be reported as '0' in the check, e.g., + * for error-injection tests) + */ + if ((data & TBITS_MASK) == 0 && s->init_tbits) { + data |= TBITS_PATTERN; + + s->fuse32[efuse_start_row_num] = data; + efuse_bdrv_sync(s, (efuse_start_row_num * 32 + TBIT0_OFFSET)); + } + + check = (check << 1) | ((data & TBITS_MASK) == TBITS_PATTERN); + } + + return check; +} + +static void efuse_realize(DeviceState *dev, Error **errp) +{ + XlnxEFuse *s = XLNX_EFUSE(dev); + + /* Sort readonly-list for bsearch lookup */ + efuse_ro_bits_sort(s); + + if ((s->efuse_size % 32) != 0) { + error_setg(errp, + "%s.efuse-size: %u: property value not multiple of 32.", + object_get_canonical_path(OBJECT(dev)), s->efuse_size); + return; + } + + s->fuse32 = g_malloc0(efuse_bytes(s)); + if (efuse_bdrv_read(s, errp)) { + g_free(s->fuse32); + } +} + +static void efuse_prop_set_drive(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + DeviceState *dev = DEVICE(obj); + + qdev_prop_drive.set(obj, v, name, opaque, errp); + + /* Fill initial data if backend is attached after realized */ + if (dev->realized) { + efuse_bdrv_read(XLNX_EFUSE(obj), errp); + } +} + +static void efuse_prop_get_drive(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + qdev_prop_drive.get(obj, v, name, opaque, errp); +} + +static void efuse_prop_release_drive(Object *obj, const char *name, + void *opaque) +{ + qdev_prop_drive.release(obj, name, opaque); +} + +static const PropertyInfo efuse_prop_drive = { + .name = "str", + .description = "Node name or ID of a block device to use as eFUSE backend", + .realized_set_allowed = true, + .get = efuse_prop_get_drive, + .set = efuse_prop_set_drive, + .release = efuse_prop_release_drive, +}; + +static Property efuse_properties[] = { + DEFINE_PROP("drive", XlnxEFuse, blk, efuse_prop_drive, BlockBackend *), + DEFINE_PROP_UINT8("efuse-nr", XlnxEFuse, efuse_nr, 3), + DEFINE_PROP_UINT32("efuse-size", XlnxEFuse, efuse_size, 64 * 32), + DEFINE_PROP_BOOL("init-factory-tbits", XlnxEFuse, init_tbits, true), + DEFINE_PROP_ARRAY("read-only", XlnxEFuse, ro_bits_cnt, ro_bits, + qdev_prop_uint32, uint32_t), + DEFINE_PROP_END_OF_LIST(), +}; + +static void efuse_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->realize = efuse_realize; + device_class_set_props(dc, efuse_properties); +} + +static const TypeInfo efuse_info = { + .name = TYPE_XLNX_EFUSE, + .parent = TYPE_DEVICE, + .instance_size = sizeof(XlnxEFuse), + .class_init = efuse_class_init, +}; + +static void efuse_register_types(void) +{ + type_register_static(&efuse_info); +} +type_init(efuse_register_types) diff --git a/hw/nvram/Kconfig b/hw/nvram/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/nvram/Kconfig +++ b/hw/nvram/Kconfig @@ -XXX,XX +XXX,XX @@ config NMC93XX_EEPROM config CHRP_NVRAM bool + +config XLNX_EFUSE_CRC + bool + +config XLNX_EFUSE + bool + select XLNX_EFUSE_CRC diff --git a/hw/nvram/meson.build b/hw/nvram/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/nvram/meson.build +++ b/hw/nvram/meson.build @@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_AT24C', if_true: files('eeprom_at24c.c')) softmmu_ss.add(when: 'CONFIG_MAC_NVRAM', if_true: files('mac_nvram.c')) softmmu_ss.add(when: 'CONFIG_NPCM7XX', if_true: files('npcm7xx_otp.c')) softmmu_ss.add(when: 'CONFIG_NRF51_SOC', if_true: files('nrf51_nvm.c')) +softmmu_ss.add(when: 'CONFIG_XLNX_EFUSE_CRC', if_true: files('xlnx-efuse-crc.c')) +softmmu_ss.add(when: 'CONFIG_XLNX_EFUSE', if_true: files('xlnx-efuse.c')) specific_ss.add(when: 'CONFIG_PSERIES', if_true: files('spapr_nvram.c')) -- 2.20.1 From: Tong Ho <tong.ho@xilinx.com> This implements the Xilinx Versal eFuse, an one-time field-programmable non-volatile storage device. There is only one such device in the Xilinx Versal product family. This device has two separate mmio interfaces, a controller and a flatten readback. The controller provides interfaces for field-programming, configuration, control, and status. The flatten readback is a cache to provide a byte-accessible read-only interface to efficiently read efuse array. Co-authored-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Co-authored-by: Sai Pavan Boddu <sai.pavan.boddu@xilinx.com> Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Signed-off-by: Sai Pavan Boddu <sai.pavan.boddu@xilinx.com> Signed-off-by: Tong Ho <tong.ho@xilinx.com> Message-id: 20210917052400.1249094-3-tong.ho@xilinx.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/nvram/xlnx-versal-efuse.h | 68 +++ hw/nvram/xlnx-versal-efuse-cache.c | 114 ++++ hw/nvram/xlnx-versal-efuse-ctrl.c | 783 +++++++++++++++++++++++++++ hw/nvram/Kconfig | 4 + hw/nvram/meson.build | 3 + 5 files changed, 972 insertions(+) create mode 100644 include/hw/nvram/xlnx-versal-efuse.h create mode 100644 hw/nvram/xlnx-versal-efuse-cache.c create mode 100644 hw/nvram/xlnx-versal-efuse-ctrl.c diff --git a/include/hw/nvram/xlnx-versal-efuse.h b/include/hw/nvram/xlnx-versal-efuse.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/nvram/xlnx-versal-efuse.h @@ -XXX,XX +XXX,XX @@ +/* + * Copyright (c) 2020 Xilinx Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ +#ifndef XLNX_VERSAL_EFUSE_H +#define XLNX_VERSAL_EFUSE_H + +#include "hw/irq.h" +#include "hw/sysbus.h" +#include "hw/register.h" +#include "hw/nvram/xlnx-efuse.h" + +#define XLNX_VERSAL_EFUSE_CTRL_R_MAX ((0x100 / 4) + 1) + +#define TYPE_XLNX_VERSAL_EFUSE_CTRL "xlnx,versal-efuse" +#define TYPE_XLNX_VERSAL_EFUSE_CACHE "xlnx,pmc-efuse-cache" + +OBJECT_DECLARE_SIMPLE_TYPE(XlnxVersalEFuseCtrl, XLNX_VERSAL_EFUSE_CTRL); +OBJECT_DECLARE_SIMPLE_TYPE(XlnxVersalEFuseCache, XLNX_VERSAL_EFUSE_CACHE); + +struct XlnxVersalEFuseCtrl { + SysBusDevice parent_obj; + qemu_irq irq_efuse_imr; + + XlnxEFuse *efuse; + + void *extra_pg0_lock_spec; /* Opaque property */ + uint32_t extra_pg0_lock_n16; + + uint32_t regs[XLNX_VERSAL_EFUSE_CTRL_R_MAX]; + RegisterInfo regs_info[XLNX_VERSAL_EFUSE_CTRL_R_MAX]; +}; + +struct XlnxVersalEFuseCache { + SysBusDevice parent_obj; + MemoryRegion iomem; + + XlnxEFuse *efuse; +}; + +/** + * xlnx_versal_efuse_read_row: + * @s: the efuse object + * @bit: the bit-address within the 32-bit row to be read + * @denied: if non-NULL, to receive true if the row is write-only + * + * Returns: the 32-bit word containing address @bit; 0 if @denies is true + */ +uint32_t xlnx_versal_efuse_read_row(XlnxEFuse *s, uint32_t bit, bool *denied); + +#endif diff --git a/hw/nvram/xlnx-versal-efuse-cache.c b/hw/nvram/xlnx-versal-efuse-cache.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/nvram/xlnx-versal-efuse-cache.c @@ -XXX,XX +XXX,XX @@ +/* + * QEMU model of the EFuse_Cache + * + * Copyright (c) 2017 Xilinx Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "hw/nvram/xlnx-versal-efuse.h" + +#include "qemu/log.h" +#include "hw/qdev-properties.h" + +#define MR_SIZE 0xC00 + +static uint64_t efuse_cache_read(void *opaque, hwaddr addr, unsigned size) +{ + XlnxVersalEFuseCache *s = XLNX_VERSAL_EFUSE_CACHE(opaque); + unsigned int w0 = QEMU_ALIGN_DOWN(addr * 8, 32); + unsigned int w1 = QEMU_ALIGN_DOWN((addr + size - 1) * 8, 32); + + uint64_t ret; + + assert(w0 == w1 || (w0 + 32) == w1); + + ret = xlnx_versal_efuse_read_row(s->efuse, w1, NULL); + if (w0 < w1) { + ret <<= 32; + ret |= xlnx_versal_efuse_read_row(s->efuse, w0, NULL); + } + + /* If 'addr' unaligned, the guest is always assumed to be little-endian. */ + addr &= 3; + if (addr) { + ret >>= 8 * addr; + } + + return ret; +} + +static void efuse_cache_write(void *opaque, hwaddr addr, uint64_t value, + unsigned size) +{ + /* No Register Writes allowed */ + qemu_log_mask(LOG_GUEST_ERROR, "%s: efuse cache registers are read-only", + __func__); +} + +static const MemoryRegionOps efuse_cache_ops = { + .read = efuse_cache_read, + .write = efuse_cache_write, + .endianness = DEVICE_LITTLE_ENDIAN, + .valid = { + .min_access_size = 1, + .max_access_size = 4, + }, +}; + +static void efuse_cache_init(Object *obj) +{ + XlnxVersalEFuseCache *s = XLNX_VERSAL_EFUSE_CACHE(obj); + SysBusDevice *sbd = SYS_BUS_DEVICE(obj); + + memory_region_init_io(&s->iomem, obj, &efuse_cache_ops, s, + TYPE_XLNX_VERSAL_EFUSE_CACHE, MR_SIZE); + sysbus_init_mmio(sbd, &s->iomem); +} + +static Property efuse_cache_props[] = { + DEFINE_PROP_LINK("efuse", + XlnxVersalEFuseCache, efuse, + TYPE_XLNX_EFUSE, XlnxEFuse *), + + DEFINE_PROP_END_OF_LIST(), +}; + +static void efuse_cache_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + device_class_set_props(dc, efuse_cache_props); +} + +static const TypeInfo efuse_cache_info = { + .name = TYPE_XLNX_VERSAL_EFUSE_CACHE, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(XlnxVersalEFuseCache), + .class_init = efuse_cache_class_init, + .instance_init = efuse_cache_init, +}; + +static void efuse_cache_register_types(void) +{ + type_register_static(&efuse_cache_info); +} + +type_init(efuse_cache_register_types) diff --git a/hw/nvram/xlnx-versal-efuse-ctrl.c b/hw/nvram/xlnx-versal-efuse-ctrl.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/nvram/xlnx-versal-efuse-ctrl.c @@ -XXX,XX +XXX,XX @@ +/* + * QEMU model of the Versal eFuse controller + * + * Copyright (c) 2020 Xilinx Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "hw/nvram/xlnx-versal-efuse.h" + +#include "qemu/log.h" +#include "qapi/error.h" +#include "migration/vmstate.h" +#include "hw/qdev-properties.h" + +#ifndef XLNX_VERSAL_EFUSE_CTRL_ERR_DEBUG +#define XLNX_VERSAL_EFUSE_CTRL_ERR_DEBUG 0 +#endif + +REG32(WR_LOCK, 0x0) + FIELD(WR_LOCK, LOCK, 0, 16) +REG32(CFG, 0x4) + FIELD(CFG, SLVERR_ENABLE, 5, 1) + FIELD(CFG, MARGIN_RD, 2, 1) + FIELD(CFG, PGM_EN, 1, 1) +REG32(STATUS, 0x8) + FIELD(STATUS, AES_USER_KEY_1_CRC_PASS, 11, 1) + FIELD(STATUS, AES_USER_KEY_1_CRC_DONE, 10, 1) + FIELD(STATUS, AES_USER_KEY_0_CRC_PASS, 9, 1) + FIELD(STATUS, AES_USER_KEY_0_CRC_DONE, 8, 1) + FIELD(STATUS, AES_CRC_PASS, 7, 1) + FIELD(STATUS, AES_CRC_DONE, 6, 1) + FIELD(STATUS, CACHE_DONE, 5, 1) + FIELD(STATUS, CACHE_LOAD, 4, 1) + FIELD(STATUS, EFUSE_2_TBIT, 2, 1) + FIELD(STATUS, EFUSE_1_TBIT, 1, 1) + FIELD(STATUS, EFUSE_0_TBIT, 0, 1) +REG32(EFUSE_PGM_ADDR, 0xc) + FIELD(EFUSE_PGM_ADDR, PAGE, 13, 4) + FIELD(EFUSE_PGM_ADDR, ROW, 5, 8) + FIELD(EFUSE_PGM_ADDR, COLUMN, 0, 5) +REG32(EFUSE_RD_ADDR, 0x10) + FIELD(EFUSE_RD_ADDR, PAGE, 13, 4) + FIELD(EFUSE_RD_ADDR, ROW, 5, 8) +REG32(EFUSE_RD_DATA, 0x14) +REG32(TPGM, 0x18) + FIELD(TPGM, VALUE, 0, 16) +REG32(TRD, 0x1c) + FIELD(TRD, VALUE, 0, 8) +REG32(TSU_H_PS, 0x20) + FIELD(TSU_H_PS, VALUE, 0, 8) +REG32(TSU_H_PS_CS, 0x24) + FIELD(TSU_H_PS_CS, VALUE, 0, 8) +REG32(TRDM, 0x28) + FIELD(TRDM, VALUE, 0, 8) +REG32(TSU_H_CS, 0x2c) + FIELD(TSU_H_CS, VALUE, 0, 8) +REG32(EFUSE_ISR, 0x30) + FIELD(EFUSE_ISR, APB_SLVERR, 31, 1) + FIELD(EFUSE_ISR, CACHE_PARITY_E2, 14, 1) + FIELD(EFUSE_ISR, CACHE_PARITY_E1, 13, 1) + FIELD(EFUSE_ISR, CACHE_PARITY_E0S, 12, 1) + FIELD(EFUSE_ISR, CACHE_PARITY_E0R, 11, 1) + FIELD(EFUSE_ISR, CACHE_APB_SLVERR, 10, 1) + FIELD(EFUSE_ISR, CACHE_REQ_ERROR, 9, 1) + FIELD(EFUSE_ISR, MAIN_REQ_ERROR, 8, 1) + FIELD(EFUSE_ISR, READ_ON_CACHE_LD, 7, 1) + FIELD(EFUSE_ISR, CACHE_FSM_ERROR, 6, 1) + FIELD(EFUSE_ISR, MAIN_FSM_ERROR, 5, 1) + FIELD(EFUSE_ISR, CACHE_ERROR, 4, 1) + FIELD(EFUSE_ISR, RD_ERROR, 3, 1) + FIELD(EFUSE_ISR, RD_DONE, 2, 1) + FIELD(EFUSE_ISR, PGM_ERROR, 1, 1) + FIELD(EFUSE_ISR, PGM_DONE, 0, 1) +REG32(EFUSE_IMR, 0x34) + FIELD(EFUSE_IMR, APB_SLVERR, 31, 1) + FIELD(EFUSE_IMR, CACHE_PARITY_E2, 14, 1) + FIELD(EFUSE_IMR, CACHE_PARITY_E1, 13, 1) + FIELD(EFUSE_IMR, CACHE_PARITY_E0S, 12, 1) + FIELD(EFUSE_IMR, CACHE_PARITY_E0R, 11, 1) + FIELD(EFUSE_IMR, CACHE_APB_SLVERR, 10, 1) + FIELD(EFUSE_IMR, CACHE_REQ_ERROR, 9, 1) + FIELD(EFUSE_IMR, MAIN_REQ_ERROR, 8, 1) + FIELD(EFUSE_IMR, READ_ON_CACHE_LD, 7, 1) + FIELD(EFUSE_IMR, CACHE_FSM_ERROR, 6, 1) + FIELD(EFUSE_IMR, MAIN_FSM_ERROR, 5, 1) + FIELD(EFUSE_IMR, CACHE_ERROR, 4, 1) + FIELD(EFUSE_IMR, RD_ERROR, 3, 1) + FIELD(EFUSE_IMR, RD_DONE, 2, 1) + FIELD(EFUSE_IMR, PGM_ERROR, 1, 1) + FIELD(EFUSE_IMR, PGM_DONE, 0, 1) +REG32(EFUSE_IER, 0x38) + FIELD(EFUSE_IER, APB_SLVERR, 31, 1) + FIELD(EFUSE_IER, CACHE_PARITY_E2, 14, 1) + FIELD(EFUSE_IER, CACHE_PARITY_E1, 13, 1) + FIELD(EFUSE_IER, CACHE_PARITY_E0S, 12, 1) + FIELD(EFUSE_IER, CACHE_PARITY_E0R, 11, 1) + FIELD(EFUSE_IER, CACHE_APB_SLVERR, 10, 1) + FIELD(EFUSE_IER, CACHE_REQ_ERROR, 9, 1) + FIELD(EFUSE_IER, MAIN_REQ_ERROR, 8, 1) + FIELD(EFUSE_IER, READ_ON_CACHE_LD, 7, 1) + FIELD(EFUSE_IER, CACHE_FSM_ERROR, 6, 1) + FIELD(EFUSE_IER, MAIN_FSM_ERROR, 5, 1) + FIELD(EFUSE_IER, CACHE_ERROR, 4, 1) + FIELD(EFUSE_IER, RD_ERROR, 3, 1) + FIELD(EFUSE_IER, RD_DONE, 2, 1) + FIELD(EFUSE_IER, PGM_ERROR, 1, 1) + FIELD(EFUSE_IER, PGM_DONE, 0, 1) +REG32(EFUSE_IDR, 0x3c) + FIELD(EFUSE_IDR, APB_SLVERR, 31, 1) + FIELD(EFUSE_IDR, CACHE_PARITY_E2, 14, 1) + FIELD(EFUSE_IDR, CACHE_PARITY_E1, 13, 1) + FIELD(EFUSE_IDR, CACHE_PARITY_E0S, 12, 1) + FIELD(EFUSE_IDR, CACHE_PARITY_E0R, 11, 1) + FIELD(EFUSE_IDR, CACHE_APB_SLVERR, 10, 1) + FIELD(EFUSE_IDR, CACHE_REQ_ERROR, 9, 1) + FIELD(EFUSE_IDR, MAIN_REQ_ERROR, 8, 1) + FIELD(EFUSE_IDR, READ_ON_CACHE_LD, 7, 1) + FIELD(EFUSE_IDR, CACHE_FSM_ERROR, 6, 1) + FIELD(EFUSE_IDR, MAIN_FSM_ERROR, 5, 1) + FIELD(EFUSE_IDR, CACHE_ERROR, 4, 1) + FIELD(EFUSE_IDR, RD_ERROR, 3, 1) + FIELD(EFUSE_IDR, RD_DONE, 2, 1) + FIELD(EFUSE_IDR, PGM_ERROR, 1, 1) + FIELD(EFUSE_IDR, PGM_DONE, 0, 1) +REG32(EFUSE_CACHE_LOAD, 0x40) + FIELD(EFUSE_CACHE_LOAD, LOAD, 0, 1) +REG32(EFUSE_PGM_LOCK, 0x44) + FIELD(EFUSE_PGM_LOCK, SPK_ID_LOCK, 0, 1) +REG32(EFUSE_AES_CRC, 0x48) +REG32(EFUSE_AES_USR_KEY0_CRC, 0x4c) +REG32(EFUSE_AES_USR_KEY1_CRC, 0x50) +REG32(EFUSE_PD, 0x54) +REG32(EFUSE_ANLG_OSC_SW_1LP, 0x60) +REG32(EFUSE_TEST_CTRL, 0x100) + +#define R_MAX (R_EFUSE_TEST_CTRL + 1) + +#define R_WR_LOCK_UNLOCK_PASSCODE (0xDF0D) + +/* + * eFuse layout references: + * https://github.com/Xilinx/embeddedsw/blob/release-2019.2/lib/sw_services/xilnvm/src/xnvm_efuse_hw.h + */ +#define BIT_POS_OF(A_) \ + ((uint32_t)((A_) & (R_EFUSE_PGM_ADDR_ROW_MASK | \ + R_EFUSE_PGM_ADDR_COLUMN_MASK))) + +#define BIT_POS(R_, C_) \ + ((uint32_t)((R_EFUSE_PGM_ADDR_ROW_MASK \ + & ((R_) << R_EFUSE_PGM_ADDR_ROW_SHIFT)) \ + | \ + (R_EFUSE_PGM_ADDR_COLUMN_MASK \ + & ((C_) << R_EFUSE_PGM_ADDR_COLUMN_SHIFT)))) + +#define EFUSE_TBIT_POS(A_) (BIT_POS_OF(A_) >= BIT_POS(0, 28)) + +#define EFUSE_ANCHOR_ROW (0) +#define EFUSE_ANCHOR_3_COL (27) +#define EFUSE_ANCHOR_1_COL (1) + +#define EFUSE_AES_KEY_START BIT_POS(12, 0) +#define EFUSE_AES_KEY_END BIT_POS(19, 31) +#define EFUSE_USER_KEY_0_START BIT_POS(20, 0) +#define EFUSE_USER_KEY_0_END BIT_POS(27, 31) +#define EFUSE_USER_KEY_1_START BIT_POS(28, 0) +#define EFUSE_USER_KEY_1_END BIT_POS(35, 31) + +#define EFUSE_RD_BLOCKED_START EFUSE_AES_KEY_START +#define EFUSE_RD_BLOCKED_END EFUSE_USER_KEY_1_END + +#define EFUSE_GLITCH_DET_WR_LK BIT_POS(4, 31) +#define EFUSE_PPK0_WR_LK BIT_POS(43, 6) +#define EFUSE_PPK1_WR_LK BIT_POS(43, 7) +#define EFUSE_PPK2_WR_LK BIT_POS(43, 8) +#define EFUSE_AES_WR_LK BIT_POS(43, 11) +#define EFUSE_USER_KEY_0_WR_LK BIT_POS(43, 13) +#define EFUSE_USER_KEY_1_WR_LK BIT_POS(43, 15) +#define EFUSE_PUF_SYN_LK BIT_POS(43, 16) +#define EFUSE_DNA_WR_LK BIT_POS(43, 27) +#define EFUSE_BOOT_ENV_WR_LK BIT_POS(43, 28) + +#define EFUSE_PGM_LOCKED_START BIT_POS(44, 0) +#define EFUSE_PGM_LOCKED_END BIT_POS(51, 31) + +#define EFUSE_PUF_PAGE (2) +#define EFUSE_PUF_SYN_START BIT_POS(129, 0) +#define EFUSE_PUF_SYN_END BIT_POS(255, 27) + +#define EFUSE_KEY_CRC_LK_ROW (43) +#define EFUSE_AES_KEY_CRC_LK_MASK ((1U << 9) | (1U << 10)) +#define EFUSE_USER_KEY_0_CRC_LK_MASK (1U << 12) +#define EFUSE_USER_KEY_1_CRC_LK_MASK (1U << 14) + +/* + * A handy macro to return value of an array element, + * or a specific default if given index is out of bound. + */ +#define ARRAY_GET(A_, I_, D_) \ + ((unsigned int)(I_) < ARRAY_SIZE(A_) ? (A_)[I_] : (D_)) + +QEMU_BUILD_BUG_ON(R_MAX != ARRAY_SIZE(((XlnxVersalEFuseCtrl *)0)->regs)); + +typedef struct XlnxEFuseLkSpec { + uint16_t row; + uint16_t lk_bit; +} XlnxEFuseLkSpec; + +static void efuse_imr_update_irq(XlnxVersalEFuseCtrl *s) +{ + bool pending = s->regs[R_EFUSE_ISR] & ~s->regs[R_EFUSE_IMR]; + qemu_set_irq(s->irq_efuse_imr, pending); +} + +static void efuse_isr_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalEFuseCtrl *s = XLNX_VERSAL_EFUSE_CTRL(reg->opaque); + efuse_imr_update_irq(s); +} + +static uint64_t efuse_ier_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalEFuseCtrl *s = XLNX_VERSAL_EFUSE_CTRL(reg->opaque); + uint32_t val = val64; + + s->regs[R_EFUSE_IMR] &= ~val; + efuse_imr_update_irq(s); + return 0; +} + +static uint64_t efuse_idr_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalEFuseCtrl *s = XLNX_VERSAL_EFUSE_CTRL(reg->opaque); + uint32_t val = val64; + + s->regs[R_EFUSE_IMR] |= val; + efuse_imr_update_irq(s); + return 0; +} + +static void efuse_status_tbits_sync(XlnxVersalEFuseCtrl *s) +{ + uint32_t check = xlnx_efuse_tbits_check(s->efuse); + uint32_t val = s->regs[R_STATUS]; + + val = FIELD_DP32(val, STATUS, EFUSE_0_TBIT, !!(check & (1 << 0))); + val = FIELD_DP32(val, STATUS, EFUSE_1_TBIT, !!(check & (1 << 1))); + val = FIELD_DP32(val, STATUS, EFUSE_2_TBIT, !!(check & (1 << 2))); + + s->regs[R_STATUS] = val; +} + +static void efuse_anchor_bits_check(XlnxVersalEFuseCtrl *s) +{ + unsigned page; + + if (!s->efuse || !s->efuse->init_tbits) { + return; + } + + for (page = 0; page < s->efuse->efuse_nr; page++) { + uint32_t row = 0, bit; + + row = FIELD_DP32(row, EFUSE_PGM_ADDR, PAGE, page); + row = FIELD_DP32(row, EFUSE_PGM_ADDR, ROW, EFUSE_ANCHOR_ROW); + + bit = FIELD_DP32(row, EFUSE_PGM_ADDR, COLUMN, EFUSE_ANCHOR_3_COL); + if (!xlnx_efuse_get_bit(s->efuse, bit)) { + xlnx_efuse_set_bit(s->efuse, bit); + } + + bit = FIELD_DP32(row, EFUSE_PGM_ADDR, COLUMN, EFUSE_ANCHOR_1_COL); + if (!xlnx_efuse_get_bit(s->efuse, bit)) { + xlnx_efuse_set_bit(s->efuse, bit); + } + } +} + +static void efuse_key_crc_check(RegisterInfo *reg, uint32_t crc, + uint32_t pass_mask, uint32_t done_mask, + unsigned first, uint32_t lk_mask) +{ + XlnxVersalEFuseCtrl *s = XLNX_VERSAL_EFUSE_CTRL(reg->opaque); + uint32_t r, lk_bits; + + /* + * To start, assume both DONE and PASS, and clear PASS by xor + * if CRC-check fails or CRC-check disabled by lock fuse. + */ + r = s->regs[R_STATUS] | done_mask | pass_mask; + + lk_bits = xlnx_efuse_get_row(s->efuse, EFUSE_KEY_CRC_LK_ROW) & lk_mask; + if (lk_bits == 0 && xlnx_efuse_k256_check(s->efuse, crc, first)) { + pass_mask = 0; + } + + s->regs[R_STATUS] = r ^ pass_mask; +} + +static void efuse_data_sync(XlnxVersalEFuseCtrl *s) +{ + efuse_status_tbits_sync(s); +} + +static int efuse_lk_spec_cmp(const void *a, const void *b) +{ + uint16_t r1 = ((const XlnxEFuseLkSpec *)a)->row; + uint16_t r2 = ((const XlnxEFuseLkSpec *)b)->row; + + return (r1 > r2) - (r1 < r2); +} + +static void efuse_lk_spec_sort(XlnxVersalEFuseCtrl *s) +{ + XlnxEFuseLkSpec *ary = s->extra_pg0_lock_spec; + const uint32_t n8 = s->extra_pg0_lock_n16 * 2; + const uint32_t sz = sizeof(ary[0]); + const uint32_t cnt = n8 / sz; + + if (ary && cnt) { + qsort(ary, cnt, sz, efuse_lk_spec_cmp); + } +} + +static uint32_t efuse_lk_spec_find(XlnxVersalEFuseCtrl *s, uint32_t row) +{ + const XlnxEFuseLkSpec *ary = s->extra_pg0_lock_spec; + const uint32_t n8 = s->extra_pg0_lock_n16 * 2; + const uint32_t sz = sizeof(ary[0]); + const uint32_t cnt = n8 / sz; + const XlnxEFuseLkSpec *item = NULL; + + if (ary && cnt) { + XlnxEFuseLkSpec k = { .row = row, }; + + item = bsearch(&k, ary, cnt, sz, efuse_lk_spec_cmp); + } + + return item ? item->lk_bit : 0; +} + +static uint32_t efuse_bit_locked(XlnxVersalEFuseCtrl *s, uint32_t bit) +{ + /* Hard-coded locks */ + static const uint16_t pg0_hard_lock[] = { + [4] = EFUSE_GLITCH_DET_WR_LK, + [37] = EFUSE_BOOT_ENV_WR_LK, + + [8 ... 11] = EFUSE_DNA_WR_LK, + [12 ... 19] = EFUSE_AES_WR_LK, + [20 ... 27] = EFUSE_USER_KEY_0_WR_LK, + [28 ... 35] = EFUSE_USER_KEY_1_WR_LK, + [64 ... 71] = EFUSE_PPK0_WR_LK, + [72 ... 79] = EFUSE_PPK1_WR_LK, + [80 ... 87] = EFUSE_PPK2_WR_LK, + }; + + uint32_t row = FIELD_EX32(bit, EFUSE_PGM_ADDR, ROW); + uint32_t lk_bit = ARRAY_GET(pg0_hard_lock, row, 0); + + return lk_bit ? lk_bit : efuse_lk_spec_find(s, row); +} + +static bool efuse_pgm_locked(XlnxVersalEFuseCtrl *s, unsigned int bit) +{ + + unsigned int lock = 1; + + /* Global lock */ + if (!ARRAY_FIELD_EX32(s->regs, CFG, PGM_EN)) { + goto ret_lock; + } + + /* Row lock */ + switch (FIELD_EX32(bit, EFUSE_PGM_ADDR, PAGE)) { + case 0: + if (ARRAY_FIELD_EX32(s->regs, EFUSE_PGM_LOCK, SPK_ID_LOCK) && + bit >= EFUSE_PGM_LOCKED_START && bit <= EFUSE_PGM_LOCKED_END) { + goto ret_lock; + } + + lock = efuse_bit_locked(s, bit); + break; + case EFUSE_PUF_PAGE: + if (bit < EFUSE_PUF_SYN_START || bit > EFUSE_PUF_SYN_END) { + lock = 0; + goto ret_lock; + } + + lock = EFUSE_PUF_SYN_LK; + break; + default: + lock = 0; + goto ret_lock; + } + + /* Row lock by an efuse bit */ + if (lock) { + lock = xlnx_efuse_get_bit(s->efuse, lock); + } + + ret_lock: + return lock != 0; +} + +static void efuse_pgm_addr_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalEFuseCtrl *s = XLNX_VERSAL_EFUSE_CTRL(reg->opaque); + unsigned bit = val64; + bool ok = false; + + /* Always zero out PGM_ADDR because it is write-only */ + s->regs[R_EFUSE_PGM_ADDR] = 0; + + /* + * Indicate error if bit is write-protected (or read-only + * as guarded by efuse_set_bit()). + * + * Keep it simple by not modeling program timing. + * + * Note: model must NEVER clear the PGM_ERROR bit; it is + * up to guest to do so (or by reset). + */ + if (efuse_pgm_locked(s, bit)) { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: Denied setting of efuse<%u, %u, %u>\n", + object_get_canonical_path(OBJECT(s)), + FIELD_EX32(bit, EFUSE_PGM_ADDR, PAGE), + FIELD_EX32(bit, EFUSE_PGM_ADDR, ROW), + FIELD_EX32(bit, EFUSE_PGM_ADDR, COLUMN)); + } else if (xlnx_efuse_set_bit(s->efuse, bit)) { + ok = true; + if (EFUSE_TBIT_POS(bit)) { + efuse_status_tbits_sync(s); + } + } + + if (!ok) { + ARRAY_FIELD_DP32(s->regs, EFUSE_ISR, PGM_ERROR, 1); + } + + ARRAY_FIELD_DP32(s->regs, EFUSE_ISR, PGM_DONE, 1); + efuse_imr_update_irq(s); +} + +static void efuse_rd_addr_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalEFuseCtrl *s = XLNX_VERSAL_EFUSE_CTRL(reg->opaque); + unsigned bit = val64; + bool denied; + + /* Always zero out RD_ADDR because it is write-only */ + s->regs[R_EFUSE_RD_ADDR] = 0; + + /* + * Indicate error if row is read-blocked. + * + * Note: model must NEVER clear the RD_ERROR bit; it is + * up to guest to do so (or by reset). + */ + s->regs[R_EFUSE_RD_DATA] = xlnx_versal_efuse_read_row(s->efuse, + bit, &denied); + if (denied) { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: Denied reading of efuse<%u, %u>\n", + object_get_canonical_path(OBJECT(s)), + FIELD_EX32(bit, EFUSE_RD_ADDR, PAGE), + FIELD_EX32(bit, EFUSE_RD_ADDR, ROW)); + + ARRAY_FIELD_DP32(s->regs, EFUSE_ISR, RD_ERROR, 1); + } + + ARRAY_FIELD_DP32(s->regs, EFUSE_ISR, RD_DONE, 1); + efuse_imr_update_irq(s); + return; +} + +static uint64_t efuse_cache_load_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalEFuseCtrl *s = XLNX_VERSAL_EFUSE_CTRL(reg->opaque); + + if (val64 & R_EFUSE_CACHE_LOAD_LOAD_MASK) { + efuse_data_sync(s); + + ARRAY_FIELD_DP32(s->regs, STATUS, CACHE_DONE, 1); + efuse_imr_update_irq(s); + } + + return 0; +} + +static uint64_t efuse_pgm_lock_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxVersalEFuseCtrl *s = XLNX_VERSAL_EFUSE_CTRL(reg->opaque); + + /* Ignore all other bits */ + val64 = FIELD_EX32(val64, EFUSE_PGM_LOCK, SPK_ID_LOCK); + + /* Once the bit is written 1, only reset will clear it to 0 */ + val64 |= ARRAY_FIELD_EX32(s->regs, EFUSE_PGM_LOCK, SPK_ID_LOCK); + + return val64; +} + +static void efuse_aes_crc_postw(RegisterInfo *reg, uint64_t val64) +{ + efuse_key_crc_check(reg, val64, + R_STATUS_AES_CRC_PASS_MASK, + R_STATUS_AES_CRC_DONE_MASK, + EFUSE_AES_KEY_START, + EFUSE_AES_KEY_CRC_LK_MASK); +} + +static void efuse_aes_u0_crc_postw(RegisterInfo *reg, uint64_t val64) +{ + efuse_key_crc_check(reg, val64, + R_STATUS_AES_USER_KEY_0_CRC_PASS_MASK, + R_STATUS_AES_USER_KEY_0_CRC_DONE_MASK, + EFUSE_USER_KEY_0_START, + EFUSE_USER_KEY_0_CRC_LK_MASK); +} + +static void efuse_aes_u1_crc_postw(RegisterInfo *reg, uint64_t val64) +{ + efuse_key_crc_check(reg, val64, + R_STATUS_AES_USER_KEY_1_CRC_PASS_MASK, + R_STATUS_AES_USER_KEY_1_CRC_DONE_MASK, + EFUSE_USER_KEY_1_START, + EFUSE_USER_KEY_1_CRC_LK_MASK); +} + +static uint64_t efuse_wr_lock_prew(RegisterInfo *reg, uint64_t val) +{ + return val != R_WR_LOCK_UNLOCK_PASSCODE; +} + +static const RegisterAccessInfo efuse_ctrl_regs_info[] = { + { .name = "WR_LOCK", .addr = A_WR_LOCK, + .reset = 0x1, + .pre_write = efuse_wr_lock_prew, + },{ .name = "CFG", .addr = A_CFG, + .rsvd = 0x9, + },{ .name = "STATUS", .addr = A_STATUS, + .rsvd = 0x8, + .ro = 0xfff, + },{ .name = "EFUSE_PGM_ADDR", .addr = A_EFUSE_PGM_ADDR, + .post_write = efuse_pgm_addr_postw, + },{ .name = "EFUSE_RD_ADDR", .addr = A_EFUSE_RD_ADDR, + .rsvd = 0x1f, + .post_write = efuse_rd_addr_postw, + },{ .name = "EFUSE_RD_DATA", .addr = A_EFUSE_RD_DATA, + .ro = 0xffffffff, + },{ .name = "TPGM", .addr = A_TPGM, + },{ .name = "TRD", .addr = A_TRD, + .reset = 0x19, + },{ .name = "TSU_H_PS", .addr = A_TSU_H_PS, + .reset = 0xff, + },{ .name = "TSU_H_PS_CS", .addr = A_TSU_H_PS_CS, + .reset = 0x11, + },{ .name = "TRDM", .addr = A_TRDM, + .reset = 0x3a, + },{ .name = "TSU_H_CS", .addr = A_TSU_H_CS, + .reset = 0x16, + },{ .name = "EFUSE_ISR", .addr = A_EFUSE_ISR, + .rsvd = 0x7fff8000, + .w1c = 0x80007fff, + .post_write = efuse_isr_postw, + },{ .name = "EFUSE_IMR", .addr = A_EFUSE_IMR, + .reset = 0x80007fff, + .rsvd = 0x7fff8000, + .ro = 0xffffffff, + },{ .name = "EFUSE_IER", .addr = A_EFUSE_IER, + .rsvd = 0x7fff8000, + .pre_write = efuse_ier_prew, + },{ .name = "EFUSE_IDR", .addr = A_EFUSE_IDR, + .rsvd = 0x7fff8000, + .pre_write = efuse_idr_prew, + },{ .name = "EFUSE_CACHE_LOAD", .addr = A_EFUSE_CACHE_LOAD, + .pre_write = efuse_cache_load_prew, + },{ .name = "EFUSE_PGM_LOCK", .addr = A_EFUSE_PGM_LOCK, + .pre_write = efuse_pgm_lock_prew, + },{ .name = "EFUSE_AES_CRC", .addr = A_EFUSE_AES_CRC, + .post_write = efuse_aes_crc_postw, + },{ .name = "EFUSE_AES_USR_KEY0_CRC", .addr = A_EFUSE_AES_USR_KEY0_CRC, + .post_write = efuse_aes_u0_crc_postw, + },{ .name = "EFUSE_AES_USR_KEY1_CRC", .addr = A_EFUSE_AES_USR_KEY1_CRC, + .post_write = efuse_aes_u1_crc_postw, + },{ .name = "EFUSE_PD", .addr = A_EFUSE_PD, + .ro = 0xfffffffe, + },{ .name = "EFUSE_ANLG_OSC_SW_1LP", .addr = A_EFUSE_ANLG_OSC_SW_1LP, + },{ .name = "EFUSE_TEST_CTRL", .addr = A_EFUSE_TEST_CTRL, + .reset = 0x8, + } +}; + +static void efuse_ctrl_reg_write(void *opaque, hwaddr addr, + uint64_t data, unsigned size) +{ + RegisterInfoArray *reg_array = opaque; + XlnxVersalEFuseCtrl *s; + Object *dev; + + assert(reg_array != NULL); + + dev = reg_array->mem.owner; + assert(dev); + + s = XLNX_VERSAL_EFUSE_CTRL(dev); + + if (addr != A_WR_LOCK && s->regs[R_WR_LOCK]) { + qemu_log_mask(LOG_GUEST_ERROR, + "%s[reg_0x%02lx]: Attempt to write locked register.\n", + object_get_canonical_path(OBJECT(s)), (long)addr); + } else { + register_write_memory(opaque, addr, data, size); + } +} + +static void efuse_ctrl_register_reset(RegisterInfo *reg) +{ + if (!reg->data || !reg->access) { + return; + } + + /* Reset must not trigger some registers' writers */ + switch (reg->access->addr) { + case A_EFUSE_AES_CRC: + case A_EFUSE_AES_USR_KEY0_CRC: + case A_EFUSE_AES_USR_KEY1_CRC: + *(uint32_t *)reg->data = reg->access->reset; + return; + } + + register_reset(reg); +} + +static void efuse_ctrl_reset(DeviceState *dev) +{ + XlnxVersalEFuseCtrl *s = XLNX_VERSAL_EFUSE_CTRL(dev); + unsigned int i; + + for (i = 0; i < ARRAY_SIZE(s->regs_info); ++i) { + efuse_ctrl_register_reset(&s->regs_info[i]); + } + + efuse_anchor_bits_check(s); + efuse_data_sync(s); + efuse_imr_update_irq(s); +} + +static const MemoryRegionOps efuse_ctrl_ops = { + .read = register_read_memory, + .write = efuse_ctrl_reg_write, + .endianness = DEVICE_LITTLE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, +}; + +static void efuse_ctrl_realize(DeviceState *dev, Error **errp) +{ + XlnxVersalEFuseCtrl *s = XLNX_VERSAL_EFUSE_CTRL(dev); + const uint32_t lks_sz = sizeof(XlnxEFuseLkSpec) / 2; + + if (!s->efuse) { + error_setg(errp, "%s.efuse: link property not connected to XLNX-EFUSE", + object_get_canonical_path(OBJECT(dev))); + return; + } + + /* Sort property-defined pgm-locks for bsearch lookup */ + if ((s->extra_pg0_lock_n16 % lks_sz) != 0) { + error_setg(errp, + "%s.pg0-lock: array property item-count not multiple of %u", + object_get_canonical_path(OBJECT(dev)), lks_sz); + return; + } + + efuse_lk_spec_sort(s); +} + +static void efuse_ctrl_init(Object *obj) +{ + XlnxVersalEFuseCtrl *s = XLNX_VERSAL_EFUSE_CTRL(obj); + SysBusDevice *sbd = SYS_BUS_DEVICE(obj); + RegisterInfoArray *reg_array; + + reg_array = + register_init_block32(DEVICE(obj), efuse_ctrl_regs_info, + ARRAY_SIZE(efuse_ctrl_regs_info), + s->regs_info, s->regs, + &efuse_ctrl_ops, + XLNX_VERSAL_EFUSE_CTRL_ERR_DEBUG, + R_MAX * 4); + + sysbus_init_mmio(sbd, &reg_array->mem); + sysbus_init_irq(sbd, &s->irq_efuse_imr); +} + +static const VMStateDescription vmstate_efuse_ctrl = { + .name = TYPE_XLNX_VERSAL_EFUSE_CTRL, + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32_ARRAY(regs, XlnxVersalEFuseCtrl, R_MAX), + VMSTATE_END_OF_LIST(), + } +}; + +static Property efuse_ctrl_props[] = { + DEFINE_PROP_LINK("efuse", + XlnxVersalEFuseCtrl, efuse, + TYPE_XLNX_EFUSE, XlnxEFuse *), + DEFINE_PROP_ARRAY("pg0-lock", + XlnxVersalEFuseCtrl, extra_pg0_lock_n16, + extra_pg0_lock_spec, qdev_prop_uint16, uint16_t), + + DEFINE_PROP_END_OF_LIST(), +}; + +static void efuse_ctrl_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->reset = efuse_ctrl_reset; + dc->realize = efuse_ctrl_realize; + dc->vmsd = &vmstate_efuse_ctrl; + device_class_set_props(dc, efuse_ctrl_props); +} + +static const TypeInfo efuse_ctrl_info = { + .name = TYPE_XLNX_VERSAL_EFUSE_CTRL, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(XlnxVersalEFuseCtrl), + .class_init = efuse_ctrl_class_init, + .instance_init = efuse_ctrl_init, +}; + +static void efuse_ctrl_register_types(void) +{ + type_register_static(&efuse_ctrl_info); +} + +type_init(efuse_ctrl_register_types) + +/* + * Retrieve a row, with unreadable bits returned as 0. + */ +uint32_t xlnx_versal_efuse_read_row(XlnxEFuse *efuse, + uint32_t bit, bool *denied) +{ + bool dummy; + + if (!denied) { + denied = &dummy; + } + + if (bit >= EFUSE_RD_BLOCKED_START && bit <= EFUSE_RD_BLOCKED_END) { + *denied = true; + return 0; + } + + *denied = false; + return xlnx_efuse_get_row(efuse, bit); +} diff --git a/hw/nvram/Kconfig b/hw/nvram/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/nvram/Kconfig +++ b/hw/nvram/Kconfig @@ -XXX,XX +XXX,XX @@ config XLNX_EFUSE_CRC config XLNX_EFUSE bool select XLNX_EFUSE_CRC + +config XLNX_EFUSE_VERSAL + bool + select XLNX_EFUSE diff --git a/hw/nvram/meson.build b/hw/nvram/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/nvram/meson.build +++ b/hw/nvram/meson.build @@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_NPCM7XX', if_true: files('npcm7xx_otp.c')) softmmu_ss.add(when: 'CONFIG_NRF51_SOC', if_true: files('nrf51_nvm.c')) softmmu_ss.add(when: 'CONFIG_XLNX_EFUSE_CRC', if_true: files('xlnx-efuse-crc.c')) softmmu_ss.add(when: 'CONFIG_XLNX_EFUSE', if_true: files('xlnx-efuse.c')) +softmmu_ss.add(when: 'CONFIG_XLNX_EFUSE_VERSAL', if_true: files( + 'xlnx-versal-efuse-cache.c', + 'xlnx-versal-efuse-ctrl.c')) specific_ss.add(when: 'CONFIG_PSERIES', if_true: files('spapr_nvram.c')) -- 2.20.1 From: Tong Ho <tong.ho@xilinx.com> This implements the Xilinx ZynqMP eFuse, an one-time field-programmable non-volatile storage device. There is only one such device in the Xilinx ZynqMP product family. Co-authored-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Co-authored-by: Sai Pavan Boddu <sai.pavan.boddu@xilinx.com> Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Signed-off-by: Sai Pavan Boddu <sai.pavan.boddu@xilinx.com> Signed-off-by: Tong Ho <tong.ho@xilinx.com> Message-id: 20210917052400.1249094-4-tong.ho@xilinx.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/nvram/xlnx-zynqmp-efuse.h | 44 ++ hw/nvram/xlnx-zynqmp-efuse.c | 855 +++++++++++++++++++++++++++ hw/nvram/Kconfig | 4 + hw/nvram/meson.build | 2 + 4 files changed, 905 insertions(+) create mode 100644 include/hw/nvram/xlnx-zynqmp-efuse.h create mode 100644 hw/nvram/xlnx-zynqmp-efuse.c diff --git a/include/hw/nvram/xlnx-zynqmp-efuse.h b/include/hw/nvram/xlnx-zynqmp-efuse.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/nvram/xlnx-zynqmp-efuse.h @@ -XXX,XX +XXX,XX @@ +/* + * Copyright (c) 2021 Xilinx Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ +#ifndef XLNX_ZYNQMP_EFUSE_H +#define XLNX_ZYNQMP_EFUSE_H + +#include "hw/irq.h" +#include "hw/sysbus.h" +#include "hw/register.h" +#include "hw/nvram/xlnx-efuse.h" + +#define XLNX_ZYNQMP_EFUSE_R_MAX ((0x10fc / 4) + 1) + +#define TYPE_XLNX_ZYNQMP_EFUSE "xlnx,zynqmp-efuse" +OBJECT_DECLARE_SIMPLE_TYPE(XlnxZynqMPEFuse, XLNX_ZYNQMP_EFUSE); + +struct XlnxZynqMPEFuse { + SysBusDevice parent_obj; + qemu_irq irq; + + XlnxEFuse *efuse; + uint32_t regs[XLNX_ZYNQMP_EFUSE_R_MAX]; + RegisterInfo regs_info[XLNX_ZYNQMP_EFUSE_R_MAX]; +}; + +#endif diff --git a/hw/nvram/xlnx-zynqmp-efuse.c b/hw/nvram/xlnx-zynqmp-efuse.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/nvram/xlnx-zynqmp-efuse.c @@ -XXX,XX +XXX,XX @@ +/* + * QEMU model of the ZynqMP eFuse + * + * Copyright (c) 2015 Xilinx Inc. + * + * Written by Edgar E. Iglesias <edgari@xilinx.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "hw/nvram/xlnx-zynqmp-efuse.h" + +#include "qemu/log.h" +#include "qapi/error.h" +#include "migration/vmstate.h" +#include "hw/qdev-properties.h" + +#ifndef ZYNQMP_EFUSE_ERR_DEBUG +#define ZYNQMP_EFUSE_ERR_DEBUG 0 +#endif + +REG32(WR_LOCK, 0x0) + FIELD(WR_LOCK, LOCK, 0, 16) +REG32(CFG, 0x4) + FIELD(CFG, SLVERR_ENABLE, 5, 1) + FIELD(CFG, MARGIN_RD, 2, 2) + FIELD(CFG, PGM_EN, 1, 1) + FIELD(CFG, EFUSE_CLK_SEL, 0, 1) +REG32(STATUS, 0x8) + FIELD(STATUS, AES_CRC_PASS, 7, 1) + FIELD(STATUS, AES_CRC_DONE, 6, 1) + FIELD(STATUS, CACHE_DONE, 5, 1) + FIELD(STATUS, CACHE_LOAD, 4, 1) + FIELD(STATUS, EFUSE_3_TBIT, 2, 1) + FIELD(STATUS, EFUSE_2_TBIT, 1, 1) + FIELD(STATUS, EFUSE_0_TBIT, 0, 1) +REG32(EFUSE_PGM_ADDR, 0xc) + FIELD(EFUSE_PGM_ADDR, EFUSE, 11, 2) + FIELD(EFUSE_PGM_ADDR, ROW, 5, 6) + FIELD(EFUSE_PGM_ADDR, COLUMN, 0, 5) +REG32(EFUSE_RD_ADDR, 0x10) + FIELD(EFUSE_RD_ADDR, EFUSE, 11, 2) + FIELD(EFUSE_RD_ADDR, ROW, 5, 6) +REG32(EFUSE_RD_DATA, 0x14) +REG32(TPGM, 0x18) + FIELD(TPGM, VALUE, 0, 16) +REG32(TRD, 0x1c) + FIELD(TRD, VALUE, 0, 8) +REG32(TSU_H_PS, 0x20) + FIELD(TSU_H_PS, VALUE, 0, 8) +REG32(TSU_H_PS_CS, 0x24) + FIELD(TSU_H_PS_CS, VALUE, 0, 8) +REG32(TSU_H_CS, 0x2c) + FIELD(TSU_H_CS, VALUE, 0, 4) +REG32(EFUSE_ISR, 0x30) + FIELD(EFUSE_ISR, APB_SLVERR, 31, 1) + FIELD(EFUSE_ISR, CACHE_ERROR, 4, 1) + FIELD(EFUSE_ISR, RD_ERROR, 3, 1) + FIELD(EFUSE_ISR, RD_DONE, 2, 1) + FIELD(EFUSE_ISR, PGM_ERROR, 1, 1) + FIELD(EFUSE_ISR, PGM_DONE, 0, 1) +REG32(EFUSE_IMR, 0x34) + FIELD(EFUSE_IMR, APB_SLVERR, 31, 1) + FIELD(EFUSE_IMR, CACHE_ERROR, 4, 1) + FIELD(EFUSE_IMR, RD_ERROR, 3, 1) + FIELD(EFUSE_IMR, RD_DONE, 2, 1) + FIELD(EFUSE_IMR, PGM_ERROR, 1, 1) + FIELD(EFUSE_IMR, PGM_DONE, 0, 1) +REG32(EFUSE_IER, 0x38) + FIELD(EFUSE_IER, APB_SLVERR, 31, 1) + FIELD(EFUSE_IER, CACHE_ERROR, 4, 1) + FIELD(EFUSE_IER, RD_ERROR, 3, 1) + FIELD(EFUSE_IER, RD_DONE, 2, 1) + FIELD(EFUSE_IER, PGM_ERROR, 1, 1) + FIELD(EFUSE_IER, PGM_DONE, 0, 1) +REG32(EFUSE_IDR, 0x3c) + FIELD(EFUSE_IDR, APB_SLVERR, 31, 1) + FIELD(EFUSE_IDR, CACHE_ERROR, 4, 1) + FIELD(EFUSE_IDR, RD_ERROR, 3, 1) + FIELD(EFUSE_IDR, RD_DONE, 2, 1) + FIELD(EFUSE_IDR, PGM_ERROR, 1, 1) + FIELD(EFUSE_IDR, PGM_DONE, 0, 1) +REG32(EFUSE_CACHE_LOAD, 0x40) + FIELD(EFUSE_CACHE_LOAD, LOAD, 0, 1) +REG32(EFUSE_PGM_LOCK, 0x44) + FIELD(EFUSE_PGM_LOCK, SPK_ID_LOCK, 0, 1) +REG32(EFUSE_AES_CRC, 0x48) +REG32(EFUSE_TBITS_PRGRMG_EN, 0x100) + FIELD(EFUSE_TBITS_PRGRMG_EN, TBITS_PRGRMG_EN, 3, 1) +REG32(DNA_0, 0x100c) +REG32(DNA_1, 0x1010) +REG32(DNA_2, 0x1014) +REG32(IPDISABLE, 0x1018) + FIELD(IPDISABLE, VCU_DIS, 8, 1) + FIELD(IPDISABLE, GPU_DIS, 5, 1) + FIELD(IPDISABLE, APU3_DIS, 3, 1) + FIELD(IPDISABLE, APU2_DIS, 2, 1) + FIELD(IPDISABLE, APU1_DIS, 1, 1) + FIELD(IPDISABLE, APU0_DIS, 0, 1) +REG32(SYSOSC_CTRL, 0x101c) + FIELD(SYSOSC_CTRL, SYSOSC_EN, 0, 1) +REG32(USER_0, 0x1020) +REG32(USER_1, 0x1024) +REG32(USER_2, 0x1028) +REG32(USER_3, 0x102c) +REG32(USER_4, 0x1030) +REG32(USER_5, 0x1034) +REG32(USER_6, 0x1038) +REG32(USER_7, 0x103c) +REG32(MISC_USER_CTRL, 0x1040) + FIELD(MISC_USER_CTRL, FPD_SC_EN_0, 14, 1) + FIELD(MISC_USER_CTRL, LPD_SC_EN_0, 11, 1) + FIELD(MISC_USER_CTRL, LBIST_EN, 10, 1) + FIELD(MISC_USER_CTRL, USR_WRLK_7, 7, 1) + FIELD(MISC_USER_CTRL, USR_WRLK_6, 6, 1) + FIELD(MISC_USER_CTRL, USR_WRLK_5, 5, 1) + FIELD(MISC_USER_CTRL, USR_WRLK_4, 4, 1) + FIELD(MISC_USER_CTRL, USR_WRLK_3, 3, 1) + FIELD(MISC_USER_CTRL, USR_WRLK_2, 2, 1) + FIELD(MISC_USER_CTRL, USR_WRLK_1, 1, 1) + FIELD(MISC_USER_CTRL, USR_WRLK_0, 0, 1) +REG32(ROM_RSVD, 0x1044) + FIELD(ROM_RSVD, PBR_BOOT_ERROR, 0, 3) +REG32(PUF_CHASH, 0x1050) +REG32(PUF_MISC, 0x1054) + FIELD(PUF_MISC, REGISTER_DIS, 31, 1) + FIELD(PUF_MISC, SYN_WRLK, 30, 1) + FIELD(PUF_MISC, SYN_INVLD, 29, 1) + FIELD(PUF_MISC, TEST2_DIS, 28, 1) + FIELD(PUF_MISC, UNUSED27, 27, 1) + FIELD(PUF_MISC, UNUSED26, 26, 1) + FIELD(PUF_MISC, UNUSED25, 25, 1) + FIELD(PUF_MISC, UNUSED24, 24, 1) + FIELD(PUF_MISC, AUX, 0, 24) +REG32(SEC_CTRL, 0x1058) + FIELD(SEC_CTRL, PPK1_INVLD, 30, 2) + FIELD(SEC_CTRL, PPK1_WRLK, 29, 1) + FIELD(SEC_CTRL, PPK0_INVLD, 27, 2) + FIELD(SEC_CTRL, PPK0_WRLK, 26, 1) + FIELD(SEC_CTRL, RSA_EN, 11, 15) + FIELD(SEC_CTRL, SEC_LOCK, 10, 1) + FIELD(SEC_CTRL, PROG_GATE_2, 9, 1) + FIELD(SEC_CTRL, PROG_GATE_1, 8, 1) + FIELD(SEC_CTRL, PROG_GATE_0, 7, 1) + FIELD(SEC_CTRL, DFT_DIS, 6, 1) + FIELD(SEC_CTRL, JTAG_DIS, 5, 1) + FIELD(SEC_CTRL, ERROR_DIS, 4, 1) + FIELD(SEC_CTRL, BBRAM_DIS, 3, 1) + FIELD(SEC_CTRL, ENC_ONLY, 2, 1) + FIELD(SEC_CTRL, AES_WRLK, 1, 1) + FIELD(SEC_CTRL, AES_RDLK, 0, 1) +REG32(SPK_ID, 0x105c) +REG32(PPK0_0, 0x10a0) +REG32(PPK0_1, 0x10a4) +REG32(PPK0_2, 0x10a8) +REG32(PPK0_3, 0x10ac) +REG32(PPK0_4, 0x10b0) +REG32(PPK0_5, 0x10b4) +REG32(PPK0_6, 0x10b8) +REG32(PPK0_7, 0x10bc) +REG32(PPK0_8, 0x10c0) +REG32(PPK0_9, 0x10c4) +REG32(PPK0_10, 0x10c8) +REG32(PPK0_11, 0x10cc) +REG32(PPK1_0, 0x10d0) +REG32(PPK1_1, 0x10d4) +REG32(PPK1_2, 0x10d8) +REG32(PPK1_3, 0x10dc) +REG32(PPK1_4, 0x10e0) +REG32(PPK1_5, 0x10e4) +REG32(PPK1_6, 0x10e8) +REG32(PPK1_7, 0x10ec) +REG32(PPK1_8, 0x10f0) +REG32(PPK1_9, 0x10f4) +REG32(PPK1_10, 0x10f8) +REG32(PPK1_11, 0x10fc) + +#define BIT_POS(ROW, COLUMN) (ROW * 32 + COLUMN) +#define R_MAX (R_PPK1_11 + 1) + +/* #define EFUSE_XOSC 26 */ + +/* + * eFUSE layout references: + * ZynqMP: UG1085 (v2.1) August 21, 2019, p.277, Table 12-13 + */ +#define EFUSE_AES_RDLK BIT_POS(22, 0) +#define EFUSE_AES_WRLK BIT_POS(22, 1) +#define EFUSE_ENC_ONLY BIT_POS(22, 2) +#define EFUSE_BBRAM_DIS BIT_POS(22, 3) +#define EFUSE_ERROR_DIS BIT_POS(22, 4) +#define EFUSE_JTAG_DIS BIT_POS(22, 5) +#define EFUSE_DFT_DIS BIT_POS(22, 6) +#define EFUSE_PROG_GATE_0 BIT_POS(22, 7) +#define EFUSE_PROG_GATE_1 BIT_POS(22, 7) +#define EFUSE_PROG_GATE_2 BIT_POS(22, 9) +#define EFUSE_SEC_LOCK BIT_POS(22, 10) +#define EFUSE_RSA_EN BIT_POS(22, 11) +#define EFUSE_RSA_EN14 BIT_POS(22, 25) +#define EFUSE_PPK0_WRLK BIT_POS(22, 26) +#define EFUSE_PPK0_INVLD BIT_POS(22, 27) +#define EFUSE_PPK0_INVLD_1 BIT_POS(22, 28) +#define EFUSE_PPK1_WRLK BIT_POS(22, 29) +#define EFUSE_PPK1_INVLD BIT_POS(22, 30) +#define EFUSE_PPK1_INVLD_1 BIT_POS(22, 31) + +/* Areas. */ +#define EFUSE_TRIM_START BIT_POS(1, 0) +#define EFUSE_TRIM_END BIT_POS(1, 30) +#define EFUSE_DNA_START BIT_POS(3, 0) +#define EFUSE_DNA_END BIT_POS(5, 31) +#define EFUSE_AES_START BIT_POS(24, 0) +#define EFUSE_AES_END BIT_POS(31, 31) +#define EFUSE_ROM_START BIT_POS(17, 0) +#define EFUSE_ROM_END BIT_POS(17, 31) +#define EFUSE_IPDIS_START BIT_POS(6, 0) +#define EFUSE_IPDIS_END BIT_POS(6, 31) +#define EFUSE_USER_START BIT_POS(8, 0) +#define EFUSE_USER_END BIT_POS(15, 31) +#define EFUSE_BISR_START BIT_POS(32, 0) +#define EFUSE_BISR_END BIT_POS(39, 31) + +#define EFUSE_USER_CTRL_START BIT_POS(16, 0) +#define EFUSE_USER_CTRL_END BIT_POS(16, 16) +#define EFUSE_USER_CTRL_MASK ((uint32_t)MAKE_64BIT_MASK(0, 17)) + +#define EFUSE_PUF_CHASH_START BIT_POS(20, 0) +#define EFUSE_PUF_CHASH_END BIT_POS(20, 31) +#define EFUSE_PUF_MISC_START BIT_POS(21, 0) +#define EFUSE_PUF_MISC_END BIT_POS(21, 31) +#define EFUSE_PUF_SYN_WRLK BIT_POS(21, 30) + +#define EFUSE_SPK_START BIT_POS(23, 0) +#define EFUSE_SPK_END BIT_POS(23, 31) + +#define EFUSE_PPK0_START BIT_POS(40, 0) +#define EFUSE_PPK0_END BIT_POS(51, 31) +#define EFUSE_PPK1_START BIT_POS(52, 0) +#define EFUSE_PPK1_END BIT_POS(63, 31) + +#define EFUSE_CACHE_FLD(s, reg, field) \ + ARRAY_FIELD_DP32((s)->regs, reg, field, \ + (xlnx_efuse_get_row((s->efuse), EFUSE_ ## field) \ + >> (EFUSE_ ## field % 32))) + +#define EFUSE_CACHE_BIT(s, reg, field) \ + ARRAY_FIELD_DP32((s)->regs, reg, field, xlnx_efuse_get_bit((s->efuse), \ + EFUSE_ ## field)) + +#define FBIT_UNKNOWN (~0) + +QEMU_BUILD_BUG_ON(R_MAX != ARRAY_SIZE(((XlnxZynqMPEFuse *)0)->regs)); + +static void update_tbit_status(XlnxZynqMPEFuse *s) +{ + unsigned int check = xlnx_efuse_tbits_check(s->efuse); + uint32_t val = s->regs[R_STATUS]; + + val = FIELD_DP32(val, STATUS, EFUSE_0_TBIT, !!(check & (1 << 0))); + val = FIELD_DP32(val, STATUS, EFUSE_2_TBIT, !!(check & (1 << 1))); + val = FIELD_DP32(val, STATUS, EFUSE_3_TBIT, !!(check & (1 << 2))); + + s->regs[R_STATUS] = val; +} + +/* Update the u32 array from efuse bits. Slow but simple approach. */ +static void cache_sync_u32(XlnxZynqMPEFuse *s, unsigned int r_start, + unsigned int f_start, unsigned int f_end, + unsigned int f_written) +{ + uint32_t *u32 = &s->regs[r_start]; + unsigned int fbit, wbits = 0, u32_off = 0; + + /* Avoid working on bits that are not relevant. */ + if (f_written != FBIT_UNKNOWN + && (f_written < f_start || f_written > f_end)) { + return; + } + + for (fbit = f_start; fbit <= f_end; fbit++, wbits++) { + if (wbits == 32) { + /* Update the key offset. */ + u32_off += 1; + wbits = 0; + } + u32[u32_off] |= xlnx_efuse_get_bit(s->efuse, fbit) << wbits; + } +} + +/* + * Keep the syncs in bit order so we can bail out for the + * slower ones. + */ +static void zynqmp_efuse_sync_cache(XlnxZynqMPEFuse *s, unsigned int bit) +{ + EFUSE_CACHE_BIT(s, SEC_CTRL, AES_RDLK); + EFUSE_CACHE_BIT(s, SEC_CTRL, AES_WRLK); + EFUSE_CACHE_BIT(s, SEC_CTRL, ENC_ONLY); + EFUSE_CACHE_BIT(s, SEC_CTRL, BBRAM_DIS); + EFUSE_CACHE_BIT(s, SEC_CTRL, ERROR_DIS); + EFUSE_CACHE_BIT(s, SEC_CTRL, JTAG_DIS); + EFUSE_CACHE_BIT(s, SEC_CTRL, DFT_DIS); + EFUSE_CACHE_BIT(s, SEC_CTRL, PROG_GATE_0); + EFUSE_CACHE_BIT(s, SEC_CTRL, PROG_GATE_1); + EFUSE_CACHE_BIT(s, SEC_CTRL, PROG_GATE_2); + EFUSE_CACHE_BIT(s, SEC_CTRL, SEC_LOCK); + EFUSE_CACHE_BIT(s, SEC_CTRL, PPK0_WRLK); + EFUSE_CACHE_BIT(s, SEC_CTRL, PPK1_WRLK); + + EFUSE_CACHE_FLD(s, SEC_CTRL, RSA_EN); + EFUSE_CACHE_FLD(s, SEC_CTRL, PPK0_INVLD); + EFUSE_CACHE_FLD(s, SEC_CTRL, PPK1_INVLD); + + /* Update the tbits. */ + update_tbit_status(s); + + /* Sync the various areas. */ + s->regs[R_MISC_USER_CTRL] = xlnx_efuse_get_row(s->efuse, + EFUSE_USER_CTRL_START) + & EFUSE_USER_CTRL_MASK; + s->regs[R_PUF_CHASH] = xlnx_efuse_get_row(s->efuse, EFUSE_PUF_CHASH_START); + s->regs[R_PUF_MISC] = xlnx_efuse_get_row(s->efuse, EFUSE_PUF_MISC_START); + + cache_sync_u32(s, R_DNA_0, EFUSE_DNA_START, EFUSE_DNA_END, bit); + + if (bit < EFUSE_AES_START) { + return; + } + + cache_sync_u32(s, R_ROM_RSVD, EFUSE_ROM_START, EFUSE_ROM_END, bit); + cache_sync_u32(s, R_IPDISABLE, EFUSE_IPDIS_START, EFUSE_IPDIS_END, bit); + cache_sync_u32(s, R_USER_0, EFUSE_USER_START, EFUSE_USER_END, bit); + cache_sync_u32(s, R_SPK_ID, EFUSE_SPK_START, EFUSE_SPK_END, bit); + cache_sync_u32(s, R_PPK0_0, EFUSE_PPK0_START, EFUSE_PPK0_END, bit); + cache_sync_u32(s, R_PPK1_0, EFUSE_PPK1_START, EFUSE_PPK1_END, bit); +} + +static void zynqmp_efuse_update_irq(XlnxZynqMPEFuse *s) +{ + bool pending = s->regs[R_EFUSE_ISR] & s->regs[R_EFUSE_IMR]; + qemu_set_irq(s->irq, pending); +} + +static void zynqmp_efuse_isr_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxZynqMPEFuse *s = XLNX_ZYNQMP_EFUSE(reg->opaque); + zynqmp_efuse_update_irq(s); +} + +static uint64_t zynqmp_efuse_ier_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxZynqMPEFuse *s = XLNX_ZYNQMP_EFUSE(reg->opaque); + uint32_t val = val64; + + s->regs[R_EFUSE_IMR] |= val; + zynqmp_efuse_update_irq(s); + return 0; +} + +static uint64_t zynqmp_efuse_idr_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxZynqMPEFuse *s = XLNX_ZYNQMP_EFUSE(reg->opaque); + uint32_t val = val64; + + s->regs[R_EFUSE_IMR] &= ~val; + zynqmp_efuse_update_irq(s); + return 0; +} + +static void zynqmp_efuse_pgm_addr_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxZynqMPEFuse *s = XLNX_ZYNQMP_EFUSE(reg->opaque); + unsigned bit = val64; + unsigned page = FIELD_EX32(bit, EFUSE_PGM_ADDR, EFUSE); + bool puf_prot = false; + const char *errmsg = NULL; + + /* Allow only valid array, and adjust for skipped array 1 */ + switch (page) { + case 0: + break; + case 2 ... 3: + bit = FIELD_DP32(bit, EFUSE_PGM_ADDR, EFUSE, page - 1); + puf_prot = xlnx_efuse_get_bit(s->efuse, EFUSE_PUF_SYN_WRLK); + break; + default: + errmsg = "Invalid address"; + goto pgm_done; + } + + if (ARRAY_FIELD_EX32(s->regs, WR_LOCK, LOCK)) { + errmsg = "Array write-locked"; + goto pgm_done; + } + + if (!ARRAY_FIELD_EX32(s->regs, CFG, PGM_EN)) { + errmsg = "Array pgm-disabled"; + goto pgm_done; + } + + if (puf_prot) { + errmsg = "PUF_HD-store write-locked"; + goto pgm_done; + } + + if (ARRAY_FIELD_EX32(s->regs, SEC_CTRL, AES_WRLK) + && bit >= EFUSE_AES_START && bit <= EFUSE_AES_END) { + errmsg = "AES key-store Write-locked"; + goto pgm_done; + } + + if (!xlnx_efuse_set_bit(s->efuse, bit)) { + errmsg = "Write failed"; + } + + pgm_done: + if (!errmsg) { + ARRAY_FIELD_DP32(s->regs, EFUSE_ISR, PGM_ERROR, 0); + } else { + ARRAY_FIELD_DP32(s->regs, EFUSE_ISR, PGM_ERROR, 1); + qemu_log_mask(LOG_GUEST_ERROR, + "%s - eFuse write error: %s; addr=0x%x\n", + object_get_canonical_path(OBJECT(s)), + errmsg, (unsigned)val64); + } + + ARRAY_FIELD_DP32(s->regs, EFUSE_ISR, PGM_DONE, 1); + zynqmp_efuse_update_irq(s); +} + +static void zynqmp_efuse_rd_addr_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxZynqMPEFuse *s = XLNX_ZYNQMP_EFUSE(reg->opaque); + + /* + * Grant reads only to allowed bits; reference sources: + * 1/ XilSKey - XilSKey_ZynqMp_EfusePs_ReadRow() + * 2/ UG1085, v2.0, table 12-13 + * (note: enumerates the masks as <first, last> per described in + * references to avoid mental translation). + */ +#define COL_MASK(L_, H_) \ + ((uint32_t)MAKE_64BIT_MASK((L_), (1 + (H_) - (L_)))) + + static const uint32_t ary0_col_mask[] = { + /* XilSKey - XSK_ZYNQMP_EFUSEPS_TBITS_ROW */ + [0] = COL_MASK(28, 31), + + /* XilSKey - XSK_ZYNQMP_EFUSEPS_USR{0:7}_FUSE_ROW */ + [8] = COL_MASK(0, 31), [9] = COL_MASK(0, 31), + [10] = COL_MASK(0, 31), [11] = COL_MASK(0, 31), + [12] = COL_MASK(0, 31), [13] = COL_MASK(0, 31), + [14] = COL_MASK(0, 31), [15] = COL_MASK(0, 31), + + /* XilSKey - XSK_ZYNQMP_EFUSEPS_MISC_USR_CTRL_ROW */ + [16] = COL_MASK(0, 7) | COL_MASK(10, 16), + + /* XilSKey - XSK_ZYNQMP_EFUSEPS_PBR_BOOT_ERR_ROW */ + [17] = COL_MASK(0, 2), + + /* XilSKey - XSK_ZYNQMP_EFUSEPS_PUF_CHASH_ROW */ + [20] = COL_MASK(0, 31), + + /* XilSKey - XSK_ZYNQMP_EFUSEPS_PUF_AUX_ROW */ + [21] = COL_MASK(0, 23) | COL_MASK(29, 31), + + /* XilSKey - XSK_ZYNQMP_EFUSEPS_SEC_CTRL_ROW */ + [22] = COL_MASK(0, 31), + + /* XilSKey - XSK_ZYNQMP_EFUSEPS_SPK_ID_ROW */ + [23] = COL_MASK(0, 31), + + /* XilSKey - XSK_ZYNQMP_EFUSEPS_PPK0_START_ROW */ + [40] = COL_MASK(0, 31), [41] = COL_MASK(0, 31), + [42] = COL_MASK(0, 31), [43] = COL_MASK(0, 31), + [44] = COL_MASK(0, 31), [45] = COL_MASK(0, 31), + [46] = COL_MASK(0, 31), [47] = COL_MASK(0, 31), + [48] = COL_MASK(0, 31), [49] = COL_MASK(0, 31), + [50] = COL_MASK(0, 31), [51] = COL_MASK(0, 31), + + /* XilSKey - XSK_ZYNQMP_EFUSEPS_PPK1_START_ROW */ + [52] = COL_MASK(0, 31), [53] = COL_MASK(0, 31), + [54] = COL_MASK(0, 31), [55] = COL_MASK(0, 31), + [56] = COL_MASK(0, 31), [57] = COL_MASK(0, 31), + [58] = COL_MASK(0, 31), [59] = COL_MASK(0, 31), + [60] = COL_MASK(0, 31), [61] = COL_MASK(0, 31), + [62] = COL_MASK(0, 31), [63] = COL_MASK(0, 31), + }; + + uint32_t col_mask = COL_MASK(0, 31); +#undef COL_MASK + + uint32_t efuse_idx = s->regs[R_EFUSE_RD_ADDR]; + uint32_t efuse_ary = FIELD_EX32(efuse_idx, EFUSE_RD_ADDR, EFUSE); + uint32_t efuse_row = FIELD_EX32(efuse_idx, EFUSE_RD_ADDR, ROW); + + switch (efuse_ary) { + case 0: /* Various */ + if (efuse_row >= ARRAY_SIZE(ary0_col_mask)) { + goto denied; + } + + col_mask = ary0_col_mask[efuse_row]; + if (!col_mask) { + goto denied; + } + break; + case 2: /* PUF helper data, adjust for skipped array 1 */ + case 3: + val64 = FIELD_DP32(efuse_idx, EFUSE_RD_ADDR, EFUSE, efuse_ary - 1); + break; + default: + goto denied; + } + + s->regs[R_EFUSE_RD_DATA] = xlnx_efuse_get_row(s->efuse, val64) & col_mask; + + ARRAY_FIELD_DP32(s->regs, EFUSE_ISR, RD_ERROR, 0); + ARRAY_FIELD_DP32(s->regs, EFUSE_ISR, RD_DONE, 1); + zynqmp_efuse_update_irq(s); + return; + + denied: + qemu_log_mask(LOG_GUEST_ERROR, + "%s: Denied efuse read from array %u, row %u\n", + object_get_canonical_path(OBJECT(s)), + efuse_ary, efuse_row); + + s->regs[R_EFUSE_RD_DATA] = 0; + + ARRAY_FIELD_DP32(s->regs, EFUSE_ISR, RD_ERROR, 1); + ARRAY_FIELD_DP32(s->regs, EFUSE_ISR, RD_DONE, 0); + zynqmp_efuse_update_irq(s); +} + +static void zynqmp_efuse_aes_crc_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxZynqMPEFuse *s = XLNX_ZYNQMP_EFUSE(reg->opaque); + bool ok; + + ok = xlnx_efuse_k256_check(s->efuse, (uint32_t)val64, EFUSE_AES_START); + + ARRAY_FIELD_DP32(s->regs, STATUS, AES_CRC_PASS, (ok ? 1 : 0)); + ARRAY_FIELD_DP32(s->regs, STATUS, AES_CRC_DONE, 1); + + s->regs[R_EFUSE_AES_CRC] = 0; /* crc value is write-only */ +} + +static uint64_t zynqmp_efuse_cache_load_prew(RegisterInfo *reg, + uint64_t valu64) +{ + XlnxZynqMPEFuse *s = XLNX_ZYNQMP_EFUSE(reg->opaque); + + if (valu64 & R_EFUSE_CACHE_LOAD_LOAD_MASK) { + zynqmp_efuse_sync_cache(s, FBIT_UNKNOWN); + ARRAY_FIELD_DP32(s->regs, STATUS, CACHE_DONE, 1); + zynqmp_efuse_update_irq(s); + } + + return 0; +} + +static uint64_t zynqmp_efuse_wr_lock_prew(RegisterInfo *reg, uint64_t val) +{ + return val == 0xDF0D ? 0 : 1; +} + +static RegisterAccessInfo zynqmp_efuse_regs_info[] = { + { .name = "WR_LOCK", .addr = A_WR_LOCK, + .reset = 0x1, + .pre_write = zynqmp_efuse_wr_lock_prew, + },{ .name = "CFG", .addr = A_CFG, + },{ .name = "STATUS", .addr = A_STATUS, + .rsvd = 0x8, + .ro = 0xff, + },{ .name = "EFUSE_PGM_ADDR", .addr = A_EFUSE_PGM_ADDR, + .post_write = zynqmp_efuse_pgm_addr_postw + },{ .name = "EFUSE_RD_ADDR", .addr = A_EFUSE_RD_ADDR, + .rsvd = 0x1f, + .post_write = zynqmp_efuse_rd_addr_postw, + },{ .name = "EFUSE_RD_DATA", .addr = A_EFUSE_RD_DATA, + .ro = 0xffffffff, + },{ .name = "TPGM", .addr = A_TPGM, + },{ .name = "TRD", .addr = A_TRD, + .reset = 0x1b, + },{ .name = "TSU_H_PS", .addr = A_TSU_H_PS, + .reset = 0xff, + },{ .name = "TSU_H_PS_CS", .addr = A_TSU_H_PS_CS, + .reset = 0xb, + },{ .name = "TSU_H_CS", .addr = A_TSU_H_CS, + .reset = 0x7, + },{ .name = "EFUSE_ISR", .addr = A_EFUSE_ISR, + .rsvd = 0x7fffffe0, + .w1c = 0x8000001f, + .post_write = zynqmp_efuse_isr_postw, + },{ .name = "EFUSE_IMR", .addr = A_EFUSE_IMR, + .reset = 0x8000001f, + .rsvd = 0x7fffffe0, + .ro = 0xffffffff, + },{ .name = "EFUSE_IER", .addr = A_EFUSE_IER, + .rsvd = 0x7fffffe0, + .pre_write = zynqmp_efuse_ier_prew, + },{ .name = "EFUSE_IDR", .addr = A_EFUSE_IDR, + .rsvd = 0x7fffffe0, + .pre_write = zynqmp_efuse_idr_prew, + },{ .name = "EFUSE_CACHE_LOAD", .addr = A_EFUSE_CACHE_LOAD, + .pre_write = zynqmp_efuse_cache_load_prew, + },{ .name = "EFUSE_PGM_LOCK", .addr = A_EFUSE_PGM_LOCK, + },{ .name = "EFUSE_AES_CRC", .addr = A_EFUSE_AES_CRC, + .post_write = zynqmp_efuse_aes_crc_postw, + },{ .name = "EFUSE_TBITS_PRGRMG_EN", .addr = A_EFUSE_TBITS_PRGRMG_EN, + .reset = R_EFUSE_TBITS_PRGRMG_EN_TBITS_PRGRMG_EN_MASK, + },{ .name = "DNA_0", .addr = A_DNA_0, + .ro = 0xffffffff, + },{ .name = "DNA_1", .addr = A_DNA_1, + .ro = 0xffffffff, + },{ .name = "DNA_2", .addr = A_DNA_2, + .ro = 0xffffffff, + },{ .name = "IPDISABLE", .addr = A_IPDISABLE, + .ro = 0xffffffff, + },{ .name = "SYSOSC_CTRL", .addr = A_SYSOSC_CTRL, + .ro = 0xffffffff, + },{ .name = "USER_0", .addr = A_USER_0, + .ro = 0xffffffff, + },{ .name = "USER_1", .addr = A_USER_1, + .ro = 0xffffffff, + },{ .name = "USER_2", .addr = A_USER_2, + .ro = 0xffffffff, + },{ .name = "USER_3", .addr = A_USER_3, + .ro = 0xffffffff, + },{ .name = "USER_4", .addr = A_USER_4, + .ro = 0xffffffff, + },{ .name = "USER_5", .addr = A_USER_5, + .ro = 0xffffffff, + },{ .name = "USER_6", .addr = A_USER_6, + .ro = 0xffffffff, + },{ .name = "USER_7", .addr = A_USER_7, + .ro = 0xffffffff, + },{ .name = "MISC_USER_CTRL", .addr = A_MISC_USER_CTRL, + .ro = 0xffffffff, + },{ .name = "ROM_RSVD", .addr = A_ROM_RSVD, + .ro = 0xffffffff, + },{ .name = "PUF_CHASH", .addr = A_PUF_CHASH, + .ro = 0xffffffff, + },{ .name = "PUF_MISC", .addr = A_PUF_MISC, + .ro = 0xffffffff, + },{ .name = "SEC_CTRL", .addr = A_SEC_CTRL, + .ro = 0xffffffff, + },{ .name = "SPK_ID", .addr = A_SPK_ID, + .ro = 0xffffffff, + },{ .name = "PPK0_0", .addr = A_PPK0_0, + .ro = 0xffffffff, + },{ .name = "PPK0_1", .addr = A_PPK0_1, + .ro = 0xffffffff, + },{ .name = "PPK0_2", .addr = A_PPK0_2, + .ro = 0xffffffff, + },{ .name = "PPK0_3", .addr = A_PPK0_3, + .ro = 0xffffffff, + },{ .name = "PPK0_4", .addr = A_PPK0_4, + .ro = 0xffffffff, + },{ .name = "PPK0_5", .addr = A_PPK0_5, + .ro = 0xffffffff, + },{ .name = "PPK0_6", .addr = A_PPK0_6, + .ro = 0xffffffff, + },{ .name = "PPK0_7", .addr = A_PPK0_7, + .ro = 0xffffffff, + },{ .name = "PPK0_8", .addr = A_PPK0_8, + .ro = 0xffffffff, + },{ .name = "PPK0_9", .addr = A_PPK0_9, + .ro = 0xffffffff, + },{ .name = "PPK0_10", .addr = A_PPK0_10, + .ro = 0xffffffff, + },{ .name = "PPK0_11", .addr = A_PPK0_11, + .ro = 0xffffffff, + },{ .name = "PPK1_0", .addr = A_PPK1_0, + .ro = 0xffffffff, + },{ .name = "PPK1_1", .addr = A_PPK1_1, + .ro = 0xffffffff, + },{ .name = "PPK1_2", .addr = A_PPK1_2, + .ro = 0xffffffff, + },{ .name = "PPK1_3", .addr = A_PPK1_3, + .ro = 0xffffffff, + },{ .name = "PPK1_4", .addr = A_PPK1_4, + .ro = 0xffffffff, + },{ .name = "PPK1_5", .addr = A_PPK1_5, + .ro = 0xffffffff, + },{ .name = "PPK1_6", .addr = A_PPK1_6, + .ro = 0xffffffff, + },{ .name = "PPK1_7", .addr = A_PPK1_7, + .ro = 0xffffffff, + },{ .name = "PPK1_8", .addr = A_PPK1_8, + .ro = 0xffffffff, + },{ .name = "PPK1_9", .addr = A_PPK1_9, + .ro = 0xffffffff, + },{ .name = "PPK1_10", .addr = A_PPK1_10, + .ro = 0xffffffff, + },{ .name = "PPK1_11", .addr = A_PPK1_11, + .ro = 0xffffffff, + } +}; + +static void zynqmp_efuse_reg_write(void *opaque, hwaddr addr, + uint64_t data, unsigned size) +{ + RegisterInfoArray *reg_array = opaque; + XlnxZynqMPEFuse *s; + Object *dev; + + assert(reg_array != NULL); + + dev = reg_array->mem.owner; + assert(dev); + + s = XLNX_ZYNQMP_EFUSE(dev); + + if (addr != A_WR_LOCK && s->regs[R_WR_LOCK]) { + qemu_log_mask(LOG_GUEST_ERROR, + "%s[reg_0x%02lx]: Attempt to write locked register.\n", + object_get_canonical_path(OBJECT(s)), (long)addr); + } else { + register_write_memory(opaque, addr, data, size); + } +} + +static const MemoryRegionOps zynqmp_efuse_ops = { + .read = register_read_memory, + .write = zynqmp_efuse_reg_write, + .endianness = DEVICE_LITTLE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, +}; + +static void zynqmp_efuse_register_reset(RegisterInfo *reg) +{ + if (!reg->data || !reg->access) { + return; + } + + /* Reset must not trigger some registers' writers */ + switch (reg->access->addr) { + case A_EFUSE_AES_CRC: + *(uint32_t *)reg->data = reg->access->reset; + return; + } + + register_reset(reg); +} + +static void zynqmp_efuse_reset(DeviceState *dev) +{ + XlnxZynqMPEFuse *s = XLNX_ZYNQMP_EFUSE(dev); + unsigned int i; + + for (i = 0; i < ARRAY_SIZE(s->regs_info); ++i) { + zynqmp_efuse_register_reset(&s->regs_info[i]); + } + + zynqmp_efuse_sync_cache(s, FBIT_UNKNOWN); + ARRAY_FIELD_DP32(s->regs, STATUS, CACHE_DONE, 1); + zynqmp_efuse_update_irq(s); +} + +static void zynqmp_efuse_realize(DeviceState *dev, Error **errp) +{ + XlnxZynqMPEFuse *s = XLNX_ZYNQMP_EFUSE(dev); + + if (!s->efuse) { + error_setg(errp, "%s.efuse: link property not connected to XLNX-EFUSE", + object_get_canonical_path(OBJECT(dev))); + return; + } + + s->efuse->dev = dev; +} + +static void zynqmp_efuse_init(Object *obj) +{ + XlnxZynqMPEFuse *s = XLNX_ZYNQMP_EFUSE(obj); + SysBusDevice *sbd = SYS_BUS_DEVICE(obj); + RegisterInfoArray *reg_array; + + reg_array = + register_init_block32(DEVICE(obj), zynqmp_efuse_regs_info, + ARRAY_SIZE(zynqmp_efuse_regs_info), + s->regs_info, s->regs, + &zynqmp_efuse_ops, + ZYNQMP_EFUSE_ERR_DEBUG, + R_MAX * 4); + + sysbus_init_mmio(sbd, &reg_array->mem); + sysbus_init_irq(sbd, &s->irq); +} + +static const VMStateDescription vmstate_efuse = { + .name = TYPE_XLNX_ZYNQMP_EFUSE, + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32_ARRAY(regs, XlnxZynqMPEFuse, R_MAX), + VMSTATE_END_OF_LIST(), + } +}; + +static Property zynqmp_efuse_props[] = { + DEFINE_PROP_LINK("efuse", + XlnxZynqMPEFuse, efuse, + TYPE_XLNX_EFUSE, XlnxEFuse *), + + DEFINE_PROP_END_OF_LIST(), +}; + +static void zynqmp_efuse_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->reset = zynqmp_efuse_reset; + dc->realize = zynqmp_efuse_realize; + dc->vmsd = &vmstate_efuse; + device_class_set_props(dc, zynqmp_efuse_props); +} + + +static const TypeInfo efuse_info = { + .name = TYPE_XLNX_ZYNQMP_EFUSE, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(XlnxZynqMPEFuse), + .class_init = zynqmp_efuse_class_init, + .instance_init = zynqmp_efuse_init, +}; + +static void efuse_register_types(void) +{ + type_register_static(&efuse_info); +} + +type_init(efuse_register_types) diff --git a/hw/nvram/Kconfig b/hw/nvram/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/nvram/Kconfig +++ b/hw/nvram/Kconfig @@ -XXX,XX +XXX,XX @@ config XLNX_EFUSE config XLNX_EFUSE_VERSAL bool select XLNX_EFUSE + +config XLNX_EFUSE_ZYNQMP + bool + select XLNX_EFUSE diff --git a/hw/nvram/meson.build b/hw/nvram/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/nvram/meson.build +++ b/hw/nvram/meson.build @@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_XLNX_EFUSE', if_true: files('xlnx-efuse.c')) softmmu_ss.add(when: 'CONFIG_XLNX_EFUSE_VERSAL', if_true: files( 'xlnx-versal-efuse-cache.c', 'xlnx-versal-efuse-ctrl.c')) +softmmu_ss.add(when: 'CONFIG_XLNX_EFUSE_ZYNQMP', if_true: files( + 'xlnx-zynqmp-efuse.c')) specific_ss.add(when: 'CONFIG_PSERIES', if_true: files('spapr_nvram.c')) -- 2.20.1 From: Tong Ho <tong.ho@xilinx.com> This device is present in Versal and ZynqMP product families to store a 256-bit encryption key. Co-authored-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Co-authored-by: Sai Pavan Boddu <sai.pavan.boddu@xilinx.com> Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Signed-off-by: Sai Pavan Boddu <sai.pavan.boddu@xilinx.com> Signed-off-by: Tong Ho <tong.ho@xilinx.com> Message-id: 20210917052400.1249094-5-tong.ho@xilinx.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/nvram/xlnx-bbram.h | 54 ++++ hw/nvram/xlnx-bbram.c | 545 ++++++++++++++++++++++++++++++++++ hw/nvram/Kconfig | 4 + hw/nvram/meson.build | 1 + 4 files changed, 604 insertions(+) create mode 100644 include/hw/nvram/xlnx-bbram.h create mode 100644 hw/nvram/xlnx-bbram.c diff --git a/include/hw/nvram/xlnx-bbram.h b/include/hw/nvram/xlnx-bbram.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/nvram/xlnx-bbram.h @@ -XXX,XX +XXX,XX @@ +/* + * QEMU model of the Xilinx BBRAM Battery Backed RAM + * + * Copyright (c) 2015-2021 Xilinx Inc. + * + * Written by Edgar E. Iglesias <edgari@xilinx.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ +#ifndef XLNX_BBRAM_H +#define XLNX_BBRAM_H + +#include "sysemu/block-backend.h" +#include "hw/qdev-core.h" +#include "hw/irq.h" +#include "hw/sysbus.h" +#include "hw/register.h" + +#define RMAX_XLNX_BBRAM ((0x4c / 4) + 1) + +#define TYPE_XLNX_BBRAM "xlnx,bbram-ctrl" +OBJECT_DECLARE_SIMPLE_TYPE(XlnxBBRam, XLNX_BBRAM); + +struct XlnxBBRam { + SysBusDevice parent_obj; + qemu_irq irq_bbram; + + BlockBackend *blk; + + uint32_t crc_zpads; + bool bbram8_wo; + bool blk_ro; + + uint32_t regs[RMAX_XLNX_BBRAM]; + RegisterInfo regs_info[RMAX_XLNX_BBRAM]; +}; + +#endif diff --git a/hw/nvram/xlnx-bbram.c b/hw/nvram/xlnx-bbram.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/nvram/xlnx-bbram.c @@ -XXX,XX +XXX,XX @@ +/* + * QEMU model of the Xilinx BBRAM Battery Backed RAM + * + * Copyright (c) 2014-2021 Xilinx Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "hw/nvram/xlnx-bbram.h" + +#include "qemu/error-report.h" +#include "qemu/log.h" +#include "qapi/error.h" +#include "sysemu/blockdev.h" +#include "migration/vmstate.h" +#include "hw/qdev-properties.h" +#include "hw/qdev-properties-system.h" +#include "hw/nvram/xlnx-efuse.h" + +#ifndef XLNX_BBRAM_ERR_DEBUG +#define XLNX_BBRAM_ERR_DEBUG 0 +#endif + +REG32(BBRAM_STATUS, 0x0) + FIELD(BBRAM_STATUS, AES_CRC_PASS, 9, 1) + FIELD(BBRAM_STATUS, AES_CRC_DONE, 8, 1) + FIELD(BBRAM_STATUS, BBRAM_ZEROIZED, 4, 1) + FIELD(BBRAM_STATUS, PGM_MODE, 0, 1) +REG32(BBRAM_CTRL, 0x4) + FIELD(BBRAM_CTRL, ZEROIZE, 0, 1) +REG32(PGM_MODE, 0x8) +REG32(BBRAM_AES_CRC, 0xc) +REG32(BBRAM_0, 0x10) +REG32(BBRAM_1, 0x14) +REG32(BBRAM_2, 0x18) +REG32(BBRAM_3, 0x1c) +REG32(BBRAM_4, 0x20) +REG32(BBRAM_5, 0x24) +REG32(BBRAM_6, 0x28) +REG32(BBRAM_7, 0x2c) +REG32(BBRAM_8, 0x30) +REG32(BBRAM_SLVERR, 0x34) + FIELD(BBRAM_SLVERR, ENABLE, 0, 1) +REG32(BBRAM_ISR, 0x38) + FIELD(BBRAM_ISR, APB_SLVERR, 0, 1) +REG32(BBRAM_IMR, 0x3c) + FIELD(BBRAM_IMR, APB_SLVERR, 0, 1) +REG32(BBRAM_IER, 0x40) + FIELD(BBRAM_IER, APB_SLVERR, 0, 1) +REG32(BBRAM_IDR, 0x44) + FIELD(BBRAM_IDR, APB_SLVERR, 0, 1) +REG32(BBRAM_MSW_LOCK, 0x4c) + FIELD(BBRAM_MSW_LOCK, VAL, 0, 1) + +#define R_MAX (R_BBRAM_MSW_LOCK + 1) + +#define RAM_MAX (A_BBRAM_8 + 4 - A_BBRAM_0) + +#define BBRAM_PGM_MAGIC 0x757bdf0d + +QEMU_BUILD_BUG_ON(R_MAX != ARRAY_SIZE(((XlnxBBRam *)0)->regs)); + +static bool bbram_msw_locked(XlnxBBRam *s) +{ + return ARRAY_FIELD_EX32(s->regs, BBRAM_MSW_LOCK, VAL) != 0; +} + +static bool bbram_pgm_enabled(XlnxBBRam *s) +{ + return ARRAY_FIELD_EX32(s->regs, BBRAM_STATUS, PGM_MODE) != 0; +} + +static void bbram_bdrv_error(XlnxBBRam *s, int rc, gchar *detail) +{ + Error *errp; + + error_setg_errno(&errp, -rc, "%s: BBRAM backstore %s failed.", + blk_name(s->blk), detail); + error_report("%s", error_get_pretty(errp)); + error_free(errp); + + g_free(detail); +} + +static void bbram_bdrv_read(XlnxBBRam *s, Error **errp) +{ + uint32_t *ram = &s->regs[R_BBRAM_0]; + int nr = RAM_MAX; + + if (!s->blk) { + return; + } + + s->blk_ro = !blk_supports_write_perm(s->blk); + if (!s->blk_ro) { + int rc; + + rc = blk_set_perm(s->blk, + (BLK_PERM_CONSISTENT_READ | BLK_PERM_WRITE), + BLK_PERM_ALL, NULL); + if (rc) { + s->blk_ro = true; + } + } + if (s->blk_ro) { + warn_report("%s: Skip saving updates to read-only BBRAM backstore.", + blk_name(s->blk)); + } + + if (blk_pread(s->blk, 0, ram, nr) < 0) { + error_setg(errp, + "%s: Failed to read %u bytes from BBRAM backstore.", + blk_name(s->blk), nr); + return; + } + + /* Convert from little-endian backstore for each 32-bit word */ + nr /= 4; + while (nr--) { + ram[nr] = le32_to_cpu(ram[nr]); + } +} + +static void bbram_bdrv_sync(XlnxBBRam *s, uint64_t hwaddr) +{ + uint32_t le32; + unsigned offset; + int rc; + + assert(A_BBRAM_0 <= hwaddr && hwaddr <= A_BBRAM_8); + + /* Backstore is always in little-endian */ + le32 = cpu_to_le32(s->regs[hwaddr / 4]); + + /* Update zeroized flag */ + if (le32 && (hwaddr != A_BBRAM_8 || s->bbram8_wo)) { + ARRAY_FIELD_DP32(s->regs, BBRAM_STATUS, BBRAM_ZEROIZED, 0); + } + + if (!s->blk || s->blk_ro) { + return; + } + + offset = hwaddr - A_BBRAM_0; + rc = blk_pwrite(s->blk, offset, &le32, 4, 0); + if (rc < 0) { + bbram_bdrv_error(s, rc, g_strdup_printf("write to offset %u", offset)); + } +} + +static void bbram_bdrv_zero(XlnxBBRam *s) +{ + int rc; + + ARRAY_FIELD_DP32(s->regs, BBRAM_STATUS, BBRAM_ZEROIZED, 1); + + if (!s->blk || s->blk_ro) { + return; + } + + rc = blk_make_zero(s->blk, 0); + if (rc < 0) { + bbram_bdrv_error(s, rc, g_strdup("zeroizing")); + } + + /* Restore bbram8 if it is non-zero */ + if (s->regs[R_BBRAM_8]) { + bbram_bdrv_sync(s, A_BBRAM_8); + } +} + +static void bbram_zeroize(XlnxBBRam *s) +{ + int nr = RAM_MAX - (s->bbram8_wo ? 0 : 4); /* only wo bbram8 is cleared */ + + memset(&s->regs[R_BBRAM_0], 0, nr); + bbram_bdrv_zero(s); +} + +static void bbram_update_irq(XlnxBBRam *s) +{ + bool pending = s->regs[R_BBRAM_ISR] & ~s->regs[R_BBRAM_IMR]; + + qemu_set_irq(s->irq_bbram, pending); +} + +static void bbram_ctrl_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxBBRam *s = XLNX_BBRAM(reg->opaque); + uint32_t val = val64; + + if (val & R_BBRAM_CTRL_ZEROIZE_MASK) { + bbram_zeroize(s); + /* The bit is self clearing */ + s->regs[R_BBRAM_CTRL] &= ~R_BBRAM_CTRL_ZEROIZE_MASK; + } +} + +static void bbram_pgm_mode_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxBBRam *s = XLNX_BBRAM(reg->opaque); + uint32_t val = val64; + + if (val == BBRAM_PGM_MAGIC) { + bbram_zeroize(s); + + /* The status bit is cleared only by POR */ + ARRAY_FIELD_DP32(s->regs, BBRAM_STATUS, PGM_MODE, 1); + } +} + +static void bbram_aes_crc_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxBBRam *s = XLNX_BBRAM(reg->opaque); + uint32_t calc_crc; + + if (!bbram_pgm_enabled(s)) { + /* We are not in programming mode, don't do anything */ + return; + } + + /* Perform the AES integrity check */ + s->regs[R_BBRAM_STATUS] |= R_BBRAM_STATUS_AES_CRC_DONE_MASK; + + /* + * Set check status. + * + * ZynqMP BBRAM check has a zero-u32 prepended; see: + * https://github.com/Xilinx/embeddedsw/blob/release-2019.2/lib/sw_services/xilskey/src/xilskey_bbramps_zynqmp.c#L311 + */ + calc_crc = xlnx_efuse_calc_crc(&s->regs[R_BBRAM_0], + (R_BBRAM_8 - R_BBRAM_0), s->crc_zpads); + + ARRAY_FIELD_DP32(s->regs, BBRAM_STATUS, AES_CRC_PASS, + (s->regs[R_BBRAM_AES_CRC] == calc_crc)); +} + +static uint64_t bbram_key_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxBBRam *s = XLNX_BBRAM(reg->opaque); + uint32_t original_data = *(uint32_t *) reg->data; + + if (bbram_pgm_enabled(s)) { + return val64; + } else { + /* We are not in programming mode, don't do anything */ + qemu_log_mask(LOG_GUEST_ERROR, + "Not in programming mode, dropping the write\n"); + return original_data; + } +} + +static void bbram_key_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxBBRam *s = XLNX_BBRAM(reg->opaque); + + bbram_bdrv_sync(s, reg->access->addr); +} + +static uint64_t bbram_wo_postr(RegisterInfo *reg, uint64_t val) +{ + return 0; +} + +static uint64_t bbram_r8_postr(RegisterInfo *reg, uint64_t val) +{ + XlnxBBRam *s = XLNX_BBRAM(reg->opaque); + + return s->bbram8_wo ? bbram_wo_postr(reg, val) : val; +} + +static bool bbram_r8_readonly(XlnxBBRam *s) +{ + return !bbram_pgm_enabled(s) || bbram_msw_locked(s); +} + +static uint64_t bbram_r8_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxBBRam *s = XLNX_BBRAM(reg->opaque); + + if (bbram_r8_readonly(s)) { + val64 = *(uint32_t *)reg->data; + } + + return val64; +} + +static void bbram_r8_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxBBRam *s = XLNX_BBRAM(reg->opaque); + + if (!bbram_r8_readonly(s)) { + bbram_bdrv_sync(s, A_BBRAM_8); + } +} + +static uint64_t bbram_msw_lock_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxBBRam *s = XLNX_BBRAM(reg->opaque); + + /* Never lock if bbram8 is wo; and, only POR can clear the lock */ + if (s->bbram8_wo) { + val64 = 0; + } else { + val64 |= s->regs[R_BBRAM_MSW_LOCK]; + } + + return val64; +} + +static void bbram_isr_postw(RegisterInfo *reg, uint64_t val64) +{ + XlnxBBRam *s = XLNX_BBRAM(reg->opaque); + + bbram_update_irq(s); +} + +static uint64_t bbram_ier_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxBBRam *s = XLNX_BBRAM(reg->opaque); + uint32_t val = val64; + + s->regs[R_BBRAM_IMR] &= ~val; + bbram_update_irq(s); + return 0; +} + +static uint64_t bbram_idr_prew(RegisterInfo *reg, uint64_t val64) +{ + XlnxBBRam *s = XLNX_BBRAM(reg->opaque); + uint32_t val = val64; + + s->regs[R_BBRAM_IMR] |= val; + bbram_update_irq(s); + return 0; +} + +static RegisterAccessInfo bbram_ctrl_regs_info[] = { + { .name = "BBRAM_STATUS", .addr = A_BBRAM_STATUS, + .rsvd = 0xee, + .ro = 0x3ff, + },{ .name = "BBRAM_CTRL", .addr = A_BBRAM_CTRL, + .post_write = bbram_ctrl_postw, + },{ .name = "PGM_MODE", .addr = A_PGM_MODE, + .post_write = bbram_pgm_mode_postw, + },{ .name = "BBRAM_AES_CRC", .addr = A_BBRAM_AES_CRC, + .post_write = bbram_aes_crc_postw, + .post_read = bbram_wo_postr, + },{ .name = "BBRAM_0", .addr = A_BBRAM_0, + .pre_write = bbram_key_prew, + .post_write = bbram_key_postw, + .post_read = bbram_wo_postr, + },{ .name = "BBRAM_1", .addr = A_BBRAM_1, + .pre_write = bbram_key_prew, + .post_write = bbram_key_postw, + .post_read = bbram_wo_postr, + },{ .name = "BBRAM_2", .addr = A_BBRAM_2, + .pre_write = bbram_key_prew, + .post_write = bbram_key_postw, + .post_read = bbram_wo_postr, + },{ .name = "BBRAM_3", .addr = A_BBRAM_3, + .pre_write = bbram_key_prew, + .post_write = bbram_key_postw, + .post_read = bbram_wo_postr, + },{ .name = "BBRAM_4", .addr = A_BBRAM_4, + .pre_write = bbram_key_prew, + .post_write = bbram_key_postw, + .post_read = bbram_wo_postr, + },{ .name = "BBRAM_5", .addr = A_BBRAM_5, + .pre_write = bbram_key_prew, + .post_write = bbram_key_postw, + .post_read = bbram_wo_postr, + },{ .name = "BBRAM_6", .addr = A_BBRAM_6, + .pre_write = bbram_key_prew, + .post_write = bbram_key_postw, + .post_read = bbram_wo_postr, + },{ .name = "BBRAM_7", .addr = A_BBRAM_7, + .pre_write = bbram_key_prew, + .post_write = bbram_key_postw, + .post_read = bbram_wo_postr, + },{ .name = "BBRAM_8", .addr = A_BBRAM_8, + .pre_write = bbram_r8_prew, + .post_write = bbram_r8_postw, + .post_read = bbram_r8_postr, + },{ .name = "BBRAM_SLVERR", .addr = A_BBRAM_SLVERR, + .rsvd = ~1, + },{ .name = "BBRAM_ISR", .addr = A_BBRAM_ISR, + .w1c = 0x1, + .post_write = bbram_isr_postw, + },{ .name = "BBRAM_IMR", .addr = A_BBRAM_IMR, + .ro = 0x1, + },{ .name = "BBRAM_IER", .addr = A_BBRAM_IER, + .pre_write = bbram_ier_prew, + },{ .name = "BBRAM_IDR", .addr = A_BBRAM_IDR, + .pre_write = bbram_idr_prew, + },{ .name = "BBRAM_MSW_LOCK", .addr = A_BBRAM_MSW_LOCK, + .pre_write = bbram_msw_lock_prew, + .ro = ~R_BBRAM_MSW_LOCK_VAL_MASK, + } +}; + +static void bbram_ctrl_reset(DeviceState *dev) +{ + XlnxBBRam *s = XLNX_BBRAM(dev); + unsigned int i; + + for (i = 0; i < ARRAY_SIZE(s->regs_info); ++i) { + if (i < R_BBRAM_0 || i > R_BBRAM_8) { + register_reset(&s->regs_info[i]); + } + } + + bbram_update_irq(s); +} + +static const MemoryRegionOps bbram_ctrl_ops = { + .read = register_read_memory, + .write = register_write_memory, + .endianness = DEVICE_LITTLE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4, + }, +}; + +static void bbram_ctrl_realize(DeviceState *dev, Error **errp) +{ + XlnxBBRam *s = XLNX_BBRAM(dev); + + if (s->crc_zpads) { + s->bbram8_wo = true; + } + + bbram_bdrv_read(s, errp); +} + +static void bbram_ctrl_init(Object *obj) +{ + XlnxBBRam *s = XLNX_BBRAM(obj); + SysBusDevice *sbd = SYS_BUS_DEVICE(obj); + RegisterInfoArray *reg_array; + + reg_array = + register_init_block32(DEVICE(obj), bbram_ctrl_regs_info, + ARRAY_SIZE(bbram_ctrl_regs_info), + s->regs_info, s->regs, + &bbram_ctrl_ops, + XLNX_BBRAM_ERR_DEBUG, + R_MAX * 4); + + sysbus_init_mmio(sbd, &reg_array->mem); + sysbus_init_irq(sbd, &s->irq_bbram); +} + +static void bbram_prop_set_drive(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + DeviceState *dev = DEVICE(obj); + + qdev_prop_drive.set(obj, v, name, opaque, errp); + + /* Fill initial data if backend is attached after realized */ + if (dev->realized) { + bbram_bdrv_read(XLNX_BBRAM(obj), errp); + } +} + +static void bbram_prop_get_drive(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + qdev_prop_drive.get(obj, v, name, opaque, errp); +} + +static void bbram_prop_release_drive(Object *obj, const char *name, + void *opaque) +{ + qdev_prop_drive.release(obj, name, opaque); +} + +static const PropertyInfo bbram_prop_drive = { + .name = "str", + .description = "Node name or ID of a block device to use as BBRAM backend", + .realized_set_allowed = true, + .get = bbram_prop_get_drive, + .set = bbram_prop_set_drive, + .release = bbram_prop_release_drive, +}; + +static const VMStateDescription vmstate_bbram_ctrl = { + .name = TYPE_XLNX_BBRAM, + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32_ARRAY(regs, XlnxBBRam, R_MAX), + VMSTATE_END_OF_LIST(), + } +}; + +static Property bbram_ctrl_props[] = { + DEFINE_PROP("drive", XlnxBBRam, blk, bbram_prop_drive, BlockBackend *), + DEFINE_PROP_UINT32("crc-zpads", XlnxBBRam, crc_zpads, 1), + DEFINE_PROP_END_OF_LIST(), +}; + +static void bbram_ctrl_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->reset = bbram_ctrl_reset; + dc->realize = bbram_ctrl_realize; + dc->vmsd = &vmstate_bbram_ctrl; + device_class_set_props(dc, bbram_ctrl_props); +} + +static const TypeInfo bbram_ctrl_info = { + .name = TYPE_XLNX_BBRAM, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(XlnxBBRam), + .class_init = bbram_ctrl_class_init, + .instance_init = bbram_ctrl_init, +}; + +static void bbram_ctrl_register_types(void) +{ + type_register_static(&bbram_ctrl_info); +} + +type_init(bbram_ctrl_register_types) diff --git a/hw/nvram/Kconfig b/hw/nvram/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/nvram/Kconfig +++ b/hw/nvram/Kconfig @@ -XXX,XX +XXX,XX @@ config XLNX_EFUSE_VERSAL config XLNX_EFUSE_ZYNQMP bool select XLNX_EFUSE + +config XLNX_BBRAM + bool + select XLNX_EFUSE_CRC diff --git a/hw/nvram/meson.build b/hw/nvram/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/nvram/meson.build +++ b/hw/nvram/meson.build @@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_XLNX_EFUSE_VERSAL', if_true: files( 'xlnx-versal-efuse-ctrl.c')) softmmu_ss.add(when: 'CONFIG_XLNX_EFUSE_ZYNQMP', if_true: files( 'xlnx-zynqmp-efuse.c')) +softmmu_ss.add(when: 'CONFIG_XLNX_BBRAM', if_true: files('xlnx-bbram.c')) specific_ss.add(when: 'CONFIG_PSERIES', if_true: files('spapr_nvram.c')) -- 2.20.1 From: Tong Ho <tong.ho@xilinx.com> Connect the support for Versal Battery-Backed RAM (BBRAM) The command argument: -drive if=pflash,index=0,... Can be used to optionally connect the bbram to a backend storage, such that field-programmed values in one invocation can be made available to next invocation. The backend storage must be a seekable binary file, and its size must be 36 bytes or larger. A file with all binary 0's is a 'blank'. Signed-off-by: Tong Ho <tong.ho@xilinx.com> Message-id: 20210917052400.1249094-6-tong.ho@xilinx.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 5 +++++ hw/arm/xlnx-versal-virt.c | 36 ++++++++++++++++++++++++++++++++++++ hw/arm/xlnx-versal.c | 18 ++++++++++++++++++ hw/arm/Kconfig | 1 + 4 files changed, 60 insertions(+) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "qom/object.h" #include "hw/usb/xlnx-usb-subsystem.h" #include "hw/misc/xlnx-versal-xramc.h" +#include "hw/nvram/xlnx-bbram.h" #define TYPE_XLNX_VERSAL "xlnx-versal" OBJECT_DECLARE_SIMPLE_TYPE(Versal, XLNX_VERSAL) @@ -XXX,XX +XXX,XX @@ struct Versal { } iou; XlnxZynqMPRTC rtc; + XlnxBBRam bbram; } pmc; struct { @@ -XXX,XX +XXX,XX @@ struct Versal { #define VERSAL_GEM1_WAKE_IRQ_0 59 #define VERSAL_ADMA_IRQ_0 60 #define VERSAL_XRAM_IRQ_0 79 +#define VERSAL_BBRAM_APB_IRQ_0 121 #define VERSAL_RTC_APB_ERR_IRQ 121 #define VERSAL_SD0_IRQ_0 126 #define VERSAL_RTC_ALARM_IRQ 142 @@ -XXX,XX +XXX,XX @@ struct Versal { #define MM_PMC_SD0 0xf1040000U #define MM_PMC_SD0_SIZE 0x10000 +#define MM_PMC_BBRAM_CTRL 0xf11f0000 +#define MM_PMC_BBRAM_CTRL_SIZE 0x00050 #define MM_PMC_CRP 0xf1260000U #define MM_PMC_CRP_SIZE 0x10000 #define MM_PMC_RTC 0xf12a0000 diff --git a/hw/arm/xlnx-versal-virt.c b/hw/arm/xlnx-versal-virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal-virt.c +++ b/hw/arm/xlnx-versal-virt.c @@ -XXX,XX +XXX,XX @@ static void fdt_add_rtc_node(VersalVirt *s) g_free(name); } +static void fdt_add_bbram_node(VersalVirt *s) +{ + const char compat[] = TYPE_XLNX_BBRAM; + const char interrupt_names[] = "bbram-error"; + char *name = g_strdup_printf("/bbram@%x", MM_PMC_BBRAM_CTRL); + + qemu_fdt_add_subnode(s->fdt, name); + + qemu_fdt_setprop_cells(s->fdt, name, "interrupts", + GIC_FDT_IRQ_TYPE_SPI, VERSAL_BBRAM_APB_IRQ_0, + GIC_FDT_IRQ_FLAGS_LEVEL_HI); + qemu_fdt_setprop(s->fdt, name, "interrupt-names", + interrupt_names, sizeof(interrupt_names)); + qemu_fdt_setprop_sized_cells(s->fdt, name, "reg", + 2, MM_PMC_BBRAM_CTRL, + 2, MM_PMC_BBRAM_CTRL_SIZE); + qemu_fdt_setprop(s->fdt, name, "compatible", compat, sizeof(compat)); + g_free(name); +} + static void fdt_nop_memory_nodes(void *fdt, Error **errp) { Error *err = NULL; @@ -XXX,XX +XXX,XX @@ static void create_virtio_regions(VersalVirt *s) } } +static void bbram_attach_drive(XlnxBBRam *dev) +{ + DriveInfo *dinfo; + BlockBackend *blk; + + dinfo = drive_get_by_index(IF_PFLASH, 0); + blk = dinfo ? blk_by_legacy_dinfo(dinfo) : NULL; + if (blk) { + qdev_prop_set_drive(DEVICE(dev), "drive", blk); + } +} + static void sd_plugin_card(SDHCIState *sd, DriveInfo *di) { BlockBackend *blk = di ? blk_by_legacy_dinfo(di) : NULL; @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) fdt_add_usb_xhci_nodes(s); fdt_add_sd_nodes(s); fdt_add_rtc_node(s); + fdt_add_bbram_node(s); fdt_add_cpu_nodes(s, psci_conduit); fdt_add_clk_node(s, "/clk125", 125000000, s->phandle.clk_125Mhz); fdt_add_clk_node(s, "/clk25", 25000000, s->phandle.clk_25Mhz); @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) memory_region_add_subregion_overlap(get_system_memory(), 0, &s->soc.fpd.apu.mr, 0); + /* Attach bbram backend, if given */ + bbram_attach_drive(&s->soc.pmc.bbram); + /* Plugin SD cards. */ for (i = 0; i < ARRAY_SIZE(s->soc.pmc.iou.sd); i++) { sd_plugin_card(&s->soc.pmc.iou.sd[i], drive_get_next(IF_SD)); diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_xrams(Versal *s, qemu_irq *pic) } } +static void versal_create_bbram(Versal *s, qemu_irq *pic) +{ + SysBusDevice *sbd; + + object_initialize_child_with_props(OBJECT(s), "bbram", &s->pmc.bbram, + sizeof(s->pmc.bbram), TYPE_XLNX_BBRAM, + &error_fatal, + "crc-zpads", "0", + NULL); + sbd = SYS_BUS_DEVICE(&s->pmc.bbram); + + sysbus_realize(sbd, &error_fatal); + memory_region_add_subregion(&s->mr_ps, MM_PMC_BBRAM_CTRL, + sysbus_mmio_get_region(sbd, 0)); + sysbus_connect_irq(sbd, 0, pic[VERSAL_BBRAM_APB_IRQ_0]); +} + /* This takes the board allocated linear DDR memory and creates aliases * for each split DDR range/aperture on the Versal address map. */ @@ -XXX,XX +XXX,XX @@ static void versal_realize(DeviceState *dev, Error **errp) versal_create_sds(s, pic); versal_create_rtc(s, pic); versal_create_xrams(s, pic); + versal_create_bbram(s, pic); versal_map_ddr(s); versal_unimp(s); diff --git a/hw/arm/Kconfig b/hw/arm/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Kconfig +++ b/hw/arm/Kconfig @@ -XXX,XX +XXX,XX @@ config XLNX_VERSAL select XLNX_ZDMA select XLNX_ZYNQMP select OR_IRQ + select XLNX_BBRAM config NPCM7XX bool -- 2.20.1 From: Tong Ho <tong.ho@xilinx.com> Connect the support for Versal eFUSE one-time field-programmable bit array. The command argument: -drive if=pflash,index=1,... Can be used to optionally connect the bit array to a backend storage, such that field-programmed values in one invocation can be made available to next invocation. The backend storage must be a seekable binary file, and its size must be 3072 bytes or larger. A file with all binary 0's is a 'blank'. Signed-off-by: Tong Ho <tong.ho@xilinx.com> Message-id: 20210917052400.1249094-7-tong.ho@xilinx.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 10 +++++++ hw/arm/xlnx-versal-virt.c | 52 ++++++++++++++++++++++++++++++++++++ hw/arm/xlnx-versal.c | 39 +++++++++++++++++++++++++++ hw/arm/Kconfig | 1 + 4 files changed, 102 insertions(+) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "hw/usb/xlnx-usb-subsystem.h" #include "hw/misc/xlnx-versal-xramc.h" #include "hw/nvram/xlnx-bbram.h" +#include "hw/nvram/xlnx-versal-efuse.h" #define TYPE_XLNX_VERSAL "xlnx-versal" OBJECT_DECLARE_SIMPLE_TYPE(Versal, XLNX_VERSAL) @@ -XXX,XX +XXX,XX @@ struct Versal { XlnxZynqMPRTC rtc; XlnxBBRam bbram; + XlnxEFuse efuse; + XlnxVersalEFuseCtrl efuse_ctrl; + XlnxVersalEFuseCache efuse_cache; } pmc; struct { @@ -XXX,XX +XXX,XX @@ struct Versal { #define VERSAL_BBRAM_APB_IRQ_0 121 #define VERSAL_RTC_APB_ERR_IRQ 121 #define VERSAL_SD0_IRQ_0 126 +#define VERSAL_EFUSE_IRQ 139 #define VERSAL_RTC_ALARM_IRQ 142 #define VERSAL_RTC_SECONDS_IRQ 143 @@ -XXX,XX +XXX,XX @@ struct Versal { #define MM_PMC_SD0_SIZE 0x10000 #define MM_PMC_BBRAM_CTRL 0xf11f0000 #define MM_PMC_BBRAM_CTRL_SIZE 0x00050 +#define MM_PMC_EFUSE_CTRL 0xf1240000 +#define MM_PMC_EFUSE_CTRL_SIZE 0x00104 +#define MM_PMC_EFUSE_CACHE 0xf1250000 +#define MM_PMC_EFUSE_CACHE_SIZE 0x00C00 + #define MM_PMC_CRP 0xf1260000U #define MM_PMC_CRP_SIZE 0x10000 #define MM_PMC_RTC 0xf12a0000 diff --git a/hw/arm/xlnx-versal-virt.c b/hw/arm/xlnx-versal-virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal-virt.c +++ b/hw/arm/xlnx-versal-virt.c @@ -XXX,XX +XXX,XX @@ static void fdt_add_bbram_node(VersalVirt *s) g_free(name); } +static void fdt_add_efuse_ctrl_node(VersalVirt *s) +{ + const char compat[] = TYPE_XLNX_VERSAL_EFUSE_CTRL; + const char interrupt_names[] = "pmc_efuse"; + char *name = g_strdup_printf("/pmc_efuse@%x", MM_PMC_EFUSE_CTRL); + + qemu_fdt_add_subnode(s->fdt, name); + + qemu_fdt_setprop_cells(s->fdt, name, "interrupts", + GIC_FDT_IRQ_TYPE_SPI, VERSAL_EFUSE_IRQ, + GIC_FDT_IRQ_FLAGS_LEVEL_HI); + qemu_fdt_setprop(s->fdt, name, "interrupt-names", + interrupt_names, sizeof(interrupt_names)); + qemu_fdt_setprop_sized_cells(s->fdt, name, "reg", + 2, MM_PMC_EFUSE_CTRL, + 2, MM_PMC_EFUSE_CTRL_SIZE); + qemu_fdt_setprop(s->fdt, name, "compatible", compat, sizeof(compat)); + g_free(name); +} + +static void fdt_add_efuse_cache_node(VersalVirt *s) +{ + const char compat[] = TYPE_XLNX_VERSAL_EFUSE_CACHE; + char *name = g_strdup_printf("/xlnx_pmc_efuse_cache@%x", + MM_PMC_EFUSE_CACHE); + + qemu_fdt_add_subnode(s->fdt, name); + + qemu_fdt_setprop_sized_cells(s->fdt, name, "reg", + 2, MM_PMC_EFUSE_CACHE, + 2, MM_PMC_EFUSE_CACHE_SIZE); + qemu_fdt_setprop(s->fdt, name, "compatible", compat, sizeof(compat)); + g_free(name); +} + static void fdt_nop_memory_nodes(void *fdt, Error **errp) { Error *err = NULL; @@ -XXX,XX +XXX,XX @@ static void bbram_attach_drive(XlnxBBRam *dev) } } +static void efuse_attach_drive(XlnxEFuse *dev) +{ + DriveInfo *dinfo; + BlockBackend *blk; + + dinfo = drive_get_by_index(IF_PFLASH, 1); + blk = dinfo ? blk_by_legacy_dinfo(dinfo) : NULL; + if (blk) { + qdev_prop_set_drive(DEVICE(dev), "drive", blk); + } +} + static void sd_plugin_card(SDHCIState *sd, DriveInfo *di) { BlockBackend *blk = di ? blk_by_legacy_dinfo(di) : NULL; @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) fdt_add_sd_nodes(s); fdt_add_rtc_node(s); fdt_add_bbram_node(s); + fdt_add_efuse_ctrl_node(s); + fdt_add_efuse_cache_node(s); fdt_add_cpu_nodes(s, psci_conduit); fdt_add_clk_node(s, "/clk125", 125000000, s->phandle.clk_125Mhz); fdt_add_clk_node(s, "/clk25", 25000000, s->phandle.clk_25Mhz); @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) /* Attach bbram backend, if given */ bbram_attach_drive(&s->soc.pmc.bbram); + /* Attach efuse backend, if given */ + efuse_attach_drive(&s->soc.pmc.efuse); + /* Plugin SD cards. */ for (i = 0; i < ARRAY_SIZE(s->soc.pmc.iou.sd); i++) { sd_plugin_card(&s->soc.pmc.iou.sd[i], drive_get_next(IF_SD)); diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_bbram(Versal *s, qemu_irq *pic) sysbus_connect_irq(sbd, 0, pic[VERSAL_BBRAM_APB_IRQ_0]); } +static void versal_realize_efuse_part(Versal *s, Object *dev, hwaddr base) +{ + SysBusDevice *part = SYS_BUS_DEVICE(dev); + + object_property_set_link(OBJECT(part), "efuse", + OBJECT(&s->pmc.efuse), &error_abort); + + sysbus_realize(part, &error_abort); + memory_region_add_subregion(&s->mr_ps, base, + sysbus_mmio_get_region(part, 0)); +} + +static void versal_create_efuse(Versal *s, qemu_irq *pic) +{ + Object *bits = OBJECT(&s->pmc.efuse); + Object *ctrl = OBJECT(&s->pmc.efuse_ctrl); + Object *cache = OBJECT(&s->pmc.efuse_cache); + + object_initialize_child(OBJECT(s), "efuse-ctrl", &s->pmc.efuse_ctrl, + TYPE_XLNX_VERSAL_EFUSE_CTRL); + + object_initialize_child(OBJECT(s), "efuse-cache", &s->pmc.efuse_cache, + TYPE_XLNX_VERSAL_EFUSE_CACHE); + + object_initialize_child_with_props(ctrl, "xlnx-efuse@0", bits, + sizeof(s->pmc.efuse), + TYPE_XLNX_EFUSE, &error_abort, + "efuse-nr", "3", + "efuse-size", "8192", + NULL); + + qdev_realize(DEVICE(bits), NULL, &error_abort); + versal_realize_efuse_part(s, ctrl, MM_PMC_EFUSE_CTRL); + versal_realize_efuse_part(s, cache, MM_PMC_EFUSE_CACHE); + + sysbus_connect_irq(SYS_BUS_DEVICE(ctrl), 0, pic[VERSAL_EFUSE_IRQ]); +} + /* This takes the board allocated linear DDR memory and creates aliases * for each split DDR range/aperture on the Versal address map. */ @@ -XXX,XX +XXX,XX @@ static void versal_realize(DeviceState *dev, Error **errp) versal_create_rtc(s, pic); versal_create_xrams(s, pic); versal_create_bbram(s, pic); + versal_create_efuse(s, pic); versal_map_ddr(s); versal_unimp(s); diff --git a/hw/arm/Kconfig b/hw/arm/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Kconfig +++ b/hw/arm/Kconfig @@ -XXX,XX +XXX,XX @@ config XLNX_VERSAL select XLNX_ZYNQMP select OR_IRQ select XLNX_BBRAM + select XLNX_EFUSE_VERSAL config NPCM7XX bool -- 2.20.1 From: Tong Ho <tong.ho@xilinx.com> Connect the support for Xilinx ZynqMP Battery-Backed RAM (BBRAM) The command argument: -drive if=pflash,index=2,... Can be used to optionally connect the bbram to a backend storage, such that field-programmed values in one invocation can be made available to next invocation. The backend storage must be a seekable binary file, and its size must be 36 bytes or larger. A file with all binary 0's is a 'blank'. Signed-off-by: Tong Ho <tong.ho@xilinx.com> Message-id: 20210917052400.1249094-8-tong.ho@xilinx.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-zynqmp.h | 2 ++ hw/arm/xlnx-zcu102.c | 15 +++++++++++++++ hw/arm/xlnx-zynqmp.c | 20 ++++++++++++++++++++ hw/Kconfig | 1 + 4 files changed, 38 insertions(+) diff --git a/include/hw/arm/xlnx-zynqmp.h b/include/hw/arm/xlnx-zynqmp.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-zynqmp.h +++ b/include/hw/arm/xlnx-zynqmp.h @@ -XXX,XX +XXX,XX @@ #include "qom/object.h" #include "net/can_emu.h" #include "hw/dma/xlnx_csu_dma.h" +#include "hw/nvram/xlnx-bbram.h" #define TYPE_XLNX_ZYNQMP "xlnx-zynqmp" OBJECT_DECLARE_SIMPLE_TYPE(XlnxZynqMPState, XLNX_ZYNQMP) @@ -XXX,XX +XXX,XX @@ struct XlnxZynqMPState { MemoryRegion *ddr_ram; MemoryRegion ddr_ram_low, ddr_ram_high; + XlnxBBRam bbram; MemoryRegion mr_unimp[XLNX_ZYNQMP_NUM_UNIMP_AREAS]; diff --git a/hw/arm/xlnx-zcu102.c b/hw/arm/xlnx-zcu102.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-zcu102.c +++ b/hw/arm/xlnx-zcu102.c @@ -XXX,XX +XXX,XX @@ static void zcu102_modify_dtb(const struct arm_boot_info *binfo, void *fdt) } } +static void bbram_attach_drive(XlnxBBRam *dev) +{ + DriveInfo *dinfo; + BlockBackend *blk; + + dinfo = drive_get_by_index(IF_PFLASH, 2); + blk = dinfo ? blk_by_legacy_dinfo(dinfo) : NULL; + if (blk) { + qdev_prop_set_drive(DEVICE(dev), "drive", blk); + } +} + static void xlnx_zcu102_init(MachineState *machine) { XlnxZCU102 *s = ZCU102_MACHINE(machine); @@ -XXX,XX +XXX,XX @@ static void xlnx_zcu102_init(MachineState *machine) qdev_realize(DEVICE(&s->soc), NULL, &error_fatal); + /* Attach bbram backend, if given */ + bbram_attach_drive(&s->soc.bbram); + /* Create and plug in the SD cards */ for (i = 0; i < XLNX_ZYNQMP_NUM_SDHCI; i++) { BusState *bus; diff --git a/hw/arm/xlnx-zynqmp.c b/hw/arm/xlnx-zynqmp.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-zynqmp.c +++ b/hw/arm/xlnx-zynqmp.c @@ -XXX,XX +XXX,XX @@ #define RTC_ADDR 0xffa60000 #define RTC_IRQ 26 +#define BBRAM_ADDR 0xffcd0000 +#define BBRAM_IRQ 11 + #define SDHCI_CAPABILITIES 0x280737ec6481 /* Datasheet: UG1085 (v1.7) */ static const uint64_t gem_addr[XLNX_ZYNQMP_NUM_GEMS] = { @@ -XXX,XX +XXX,XX @@ static void xlnx_zynqmp_create_rpu(MachineState *ms, XlnxZynqMPState *s, qdev_realize(DEVICE(&s->rpu_cluster), NULL, &error_fatal); } +static void xlnx_zynqmp_create_bbram(XlnxZynqMPState *s, qemu_irq *gic) +{ + SysBusDevice *sbd; + + object_initialize_child_with_props(OBJECT(s), "bbram", &s->bbram, + sizeof(s->bbram), TYPE_XLNX_BBRAM, + &error_fatal, + "crc-zpads", "1", + NULL); + sbd = SYS_BUS_DEVICE(&s->bbram); + + sysbus_realize(sbd, &error_fatal); + sysbus_mmio_map(sbd, 0, BBRAM_ADDR); + sysbus_connect_irq(sbd, 0, gic[BBRAM_IRQ]); +} + static void xlnx_zynqmp_create_unimp_mmio(XlnxZynqMPState *s) { static const struct UnimpInfo { @@ -XXX,XX +XXX,XX @@ static void xlnx_zynqmp_realize(DeviceState *dev, Error **errp) sysbus_mmio_map(SYS_BUS_DEVICE(&s->rtc), 0, RTC_ADDR); sysbus_connect_irq(SYS_BUS_DEVICE(&s->rtc), 0, gic_spi[RTC_IRQ]); + xlnx_zynqmp_create_bbram(s, gic_spi); xlnx_zynqmp_create_unimp_mmio(s); for (i = 0; i < XLNX_ZYNQMP_NUM_GDMA_CH; i++) { diff --git a/hw/Kconfig b/hw/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/Kconfig +++ b/hw/Kconfig @@ -XXX,XX +XXX,XX @@ config XLNX_ZYNQMP select REGISTER select CAN_BUS select PTIMER + select XLNX_BBRAM -- 2.20.1 From: Tong Ho <tong.ho@xilinx.com> Connect the support for ZynqMP eFUSE one-time field-programmable bit array. The command argument: -drive if=pflash,index=3,... Can be used to optionally connect the bit array to a backend storage, such that field-programmed values in one invocation can be made available to next invocation. The backend storage must be a seekable binary file, and its size must be 768 bytes or larger. A file with all binary 0's is a 'blank'. Signed-off-by: Tong Ho <tong.ho@xilinx.com> Message-id: 20210917052400.1249094-9-tong.ho@xilinx.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-zynqmp.h | 3 +++ hw/arm/xlnx-zcu102.c | 15 +++++++++++++++ hw/arm/xlnx-zynqmp.c | 29 +++++++++++++++++++++++++++++ hw/Kconfig | 1 + 4 files changed, 48 insertions(+) diff --git a/include/hw/arm/xlnx-zynqmp.h b/include/hw/arm/xlnx-zynqmp.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-zynqmp.h +++ b/include/hw/arm/xlnx-zynqmp.h @@ -XXX,XX +XXX,XX @@ #include "net/can_emu.h" #include "hw/dma/xlnx_csu_dma.h" #include "hw/nvram/xlnx-bbram.h" +#include "hw/nvram/xlnx-zynqmp-efuse.h" #define TYPE_XLNX_ZYNQMP "xlnx-zynqmp" OBJECT_DECLARE_SIMPLE_TYPE(XlnxZynqMPState, XLNX_ZYNQMP) @@ -XXX,XX +XXX,XX @@ struct XlnxZynqMPState { MemoryRegion *ddr_ram; MemoryRegion ddr_ram_low, ddr_ram_high; XlnxBBRam bbram; + XlnxEFuse efuse; + XlnxZynqMPEFuse efuse_ctrl; MemoryRegion mr_unimp[XLNX_ZYNQMP_NUM_UNIMP_AREAS]; diff --git a/hw/arm/xlnx-zcu102.c b/hw/arm/xlnx-zcu102.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-zcu102.c +++ b/hw/arm/xlnx-zcu102.c @@ -XXX,XX +XXX,XX @@ static void bbram_attach_drive(XlnxBBRam *dev) } } +static void efuse_attach_drive(XlnxEFuse *dev) +{ + DriveInfo *dinfo; + BlockBackend *blk; + + dinfo = drive_get_by_index(IF_PFLASH, 3); + blk = dinfo ? blk_by_legacy_dinfo(dinfo) : NULL; + if (blk) { + qdev_prop_set_drive(DEVICE(dev), "drive", blk); + } +} + static void xlnx_zcu102_init(MachineState *machine) { XlnxZCU102 *s = ZCU102_MACHINE(machine); @@ -XXX,XX +XXX,XX @@ static void xlnx_zcu102_init(MachineState *machine) /* Attach bbram backend, if given */ bbram_attach_drive(&s->soc.bbram); + /* Attach efuse backend, if given */ + efuse_attach_drive(&s->soc.efuse); + /* Create and plug in the SD cards */ for (i = 0; i < XLNX_ZYNQMP_NUM_SDHCI; i++) { BusState *bus; diff --git a/hw/arm/xlnx-zynqmp.c b/hw/arm/xlnx-zynqmp.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-zynqmp.c +++ b/hw/arm/xlnx-zynqmp.c @@ -XXX,XX +XXX,XX @@ #define BBRAM_ADDR 0xffcd0000 #define BBRAM_IRQ 11 +#define EFUSE_ADDR 0xffcc0000 +#define EFUSE_IRQ 87 + #define SDHCI_CAPABILITIES 0x280737ec6481 /* Datasheet: UG1085 (v1.7) */ static const uint64_t gem_addr[XLNX_ZYNQMP_NUM_GEMS] = { @@ -XXX,XX +XXX,XX @@ static void xlnx_zynqmp_create_bbram(XlnxZynqMPState *s, qemu_irq *gic) sysbus_connect_irq(sbd, 0, gic[BBRAM_IRQ]); } +static void xlnx_zynqmp_create_efuse(XlnxZynqMPState *s, qemu_irq *gic) +{ + Object *bits = OBJECT(&s->efuse); + Object *ctrl = OBJECT(&s->efuse_ctrl); + SysBusDevice *sbd; + + object_initialize_child(OBJECT(s), "efuse-ctrl", &s->efuse_ctrl, + TYPE_XLNX_ZYNQMP_EFUSE); + + object_initialize_child_with_props(ctrl, "xlnx-efuse@0", bits, + sizeof(s->efuse), + TYPE_XLNX_EFUSE, &error_abort, + "efuse-nr", "3", + "efuse-size", "2048", + NULL); + + qdev_realize(DEVICE(bits), NULL, &error_abort); + object_property_set_link(ctrl, "efuse", bits, &error_abort); + + sbd = SYS_BUS_DEVICE(ctrl); + sysbus_realize(sbd, &error_abort); + sysbus_mmio_map(sbd, 0, EFUSE_ADDR); + sysbus_connect_irq(sbd, 0, gic[EFUSE_IRQ]); +} + static void xlnx_zynqmp_create_unimp_mmio(XlnxZynqMPState *s) { static const struct UnimpInfo { @@ -XXX,XX +XXX,XX @@ static void xlnx_zynqmp_realize(DeviceState *dev, Error **errp) sysbus_connect_irq(SYS_BUS_DEVICE(&s->rtc), 0, gic_spi[RTC_IRQ]); xlnx_zynqmp_create_bbram(s, gic_spi); + xlnx_zynqmp_create_efuse(s, gic_spi); xlnx_zynqmp_create_unimp_mmio(s); for (i = 0; i < XLNX_ZYNQMP_NUM_GDMA_CH; i++) { diff --git a/hw/Kconfig b/hw/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/Kconfig +++ b/hw/Kconfig @@ -XXX,XX +XXX,XX @@ config XLNX_ZYNQMP select CAN_BUS select PTIMER select XLNX_BBRAM + select XLNX_EFUSE_ZYNQMP -- 2.20.1 From: Tong Ho <tong.ho@xilinx.com> Add BBRAM and eFUSE usage to the Xilinx Versal Virt board document. Signed-off-by: Tong Ho <tong.ho@xilinx.com> Message-id: 20210917052400.1249094-10-tong.ho@xilinx.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/xlnx-versal-virt.rst | 49 ++++++++++++++++++++++++++++ 1 file changed, 49 insertions(+) diff --git a/docs/system/arm/xlnx-versal-virt.rst b/docs/system/arm/xlnx-versal-virt.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/arm/xlnx-versal-virt.rst +++ b/docs/system/arm/xlnx-versal-virt.rst @@ -XXX,XX +XXX,XX @@ Implemented devices: - OCM (256KB of On Chip Memory) - XRAM (4MB of on chip Accelerator RAM) - DDR memory +- BBRAM (36 bytes of Battery-backed RAM) +- eFUSE (3072 bytes of one-time field-programmable bit array) QEMU does not yet model any other devices, including the PL and the AI Engine. @@ -XXX,XX +XXX,XX @@ Run the following at the U-Boot prompt: fdt set /chosen/dom0 reg <0x00000000 0x40000000 0x0 0x03100000> booti 30000000 - 20000000 +BBRAM File Backend +"""""""""""""""""" +BBRAM can have an optional file backend, which must be a seekable +binary file with a size of 36 bytes or larger. A file with all +binary 0s is a 'blank'. + +To add a file-backend for the BBRAM: + +.. code-block:: bash + + -drive if=pflash,index=0,file=versal-bbram.bin,format=raw + +To use a different index value, N, from default of 0, add: + +.. code-block:: bash + + -global xlnx,bbram-ctrl.drive-index=N + +eFUSE File Backend +"""""""""""""""""" +eFUSE can have an optional file backend, which must be a seekable +binary file with a size of 3072 bytes or larger. A file with all +binary 0s is a 'blank'. + +To add a file-backend for the eFUSE: + +.. code-block:: bash + + -drive if=pflash,index=1,file=versal-efuse.bin,format=raw + +To use a different index value, N, from default of 1, add: + +.. code-block:: bash + + -global xlnx,efuse.drive-index=N + +.. warning:: + In actual physical Versal, BBRAM and eFUSE contain sensitive data. + The QEMU device models do **not** encrypt nor obfuscate any data + when holding them in models' memory or when writing them to their + file backends. + + Thus, a file backend should be used with caution, and 'format=luks' + is highly recommended (albeit with usage complexity). + + Better yet, do not use actual product data when running guest image + on this Xilinx Versal Virt board. -- 2.20.1 The aarch64-linux QEMU usermode binaries can never run 32-bit code, so they do not need to include the GDB XML for it. (arm_cpu_register_gdb_regs_for_features() will not use these XML files if the CPU has ARM_FEATURE_AARCH64, so we will not advertise to gdb that we have them.) Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210921162901.17508-2-peter.maydell@linaro.org --- configs/targets/aarch64-linux-user.mak | 2 +- configs/targets/aarch64_be-linux-user.mak | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/configs/targets/aarch64-linux-user.mak b/configs/targets/aarch64-linux-user.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/aarch64-linux-user.mak +++ b/configs/targets/aarch64-linux-user.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=aarch64 TARGET_BASE_ARCH=arm -TARGET_XML_FILES= gdb-xml/aarch64-core.xml gdb-xml/aarch64-fpu.xml gdb-xml/arm-core.xml gdb-xml/arm-vfp.xml gdb-xml/arm-vfp3.xml gdb-xml/arm-neon.xml gdb-xml/arm-m-profile.xml +TARGET_XML_FILES= gdb-xml/aarch64-core.xml gdb-xml/aarch64-fpu.xml TARGET_HAS_BFLT=y CONFIG_ARM_COMPATIBLE_SEMIHOSTING=y diff --git a/configs/targets/aarch64_be-linux-user.mak b/configs/targets/aarch64_be-linux-user.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/aarch64_be-linux-user.mak +++ b/configs/targets/aarch64_be-linux-user.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=aarch64 TARGET_BASE_ARCH=arm TARGET_WORDS_BIGENDIAN=y -TARGET_XML_FILES= gdb-xml/aarch64-core.xml gdb-xml/aarch64-fpu.xml gdb-xml/arm-core.xml gdb-xml/arm-vfp.xml gdb-xml/arm-vfp3.xml gdb-xml/arm-neon.xml gdb-xml/arm-m-profile.xml +TARGET_XML_FILES= gdb-xml/aarch64-core.xml gdb-xml/aarch64-fpu.xml TARGET_HAS_BFLT=y CONFIG_ARM_COMPATIBLE_SEMIHOSTING=y -- 2.20.1 We're going to move this code to a different file; fix the coding style first so checkpatch doesn't complain. This includes deleting the spurious 'break' statements after returns in the vfp_gdb_get_reg() function. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210921162901.17508-3-peter.maydell@linaro.org --- target/arm/helper.c | 23 ++++++++++++++++------- 1 file changed, 16 insertions(+), 7 deletions(-) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ static int vfp_gdb_get_reg(CPUARMState *env, GByteArray *buf, int reg) } } switch (reg - nregs) { - case 0: return gdb_get_reg32(buf, env->vfp.xregs[ARM_VFP_FPSID]); break; - case 1: return gdb_get_reg32(buf, vfp_get_fpscr(env)); break; - case 2: return gdb_get_reg32(buf, env->vfp.xregs[ARM_VFP_FPEXC]); break; + case 0: + return gdb_get_reg32(buf, env->vfp.xregs[ARM_VFP_FPSID]); + case 1: + return gdb_get_reg32(buf, vfp_get_fpscr(env)); + case 2: + return gdb_get_reg32(buf, env->vfp.xregs[ARM_VFP_FPEXC]); } return 0; } @@ -XXX,XX +XXX,XX @@ static int vfp_gdb_set_reg(CPUARMState *env, uint8_t *buf, int reg) } } switch (reg - nregs) { - case 0: env->vfp.xregs[ARM_VFP_FPSID] = ldl_p(buf); return 4; - case 1: vfp_set_fpscr(env, ldl_p(buf)); return 4; - case 2: env->vfp.xregs[ARM_VFP_FPEXC] = ldl_p(buf) & (1 << 30); return 4; + case 0: + env->vfp.xregs[ARM_VFP_FPSID] = ldl_p(buf); + return 4; + case 1: + vfp_set_fpscr(env, ldl_p(buf)); + return 4; + case 2: + env->vfp.xregs[ARM_VFP_FPEXC] = ldl_p(buf) & (1 << 30); + return 4; } return 0; } @@ -XXX,XX +XXX,XX @@ static int aarch64_fpu_gdb_get_reg(CPUARMState *env, GByteArray *buf, int reg) return gdb_get_reg32(buf, vfp_get_fpsr(env)); case 33: /* FPCR */ - return gdb_get_reg32(buf,vfp_get_fpcr(env)); + return gdb_get_reg32(buf, vfp_get_fpcr(env)); default: return 0; } -- 2.20.1 Currently helper.c includes some code which is part of the arm target's gdbstub support. This code has a better home: in gdbstub.c and gdbstub64.c. Move it there. Because aarch64_fpu_gdb_get_reg() and aarch64_fpu_gdb_set_reg() move into gdbstub64.c, this means that they're now compiled only for TARGET_AARCH64 rather than always. That is the only case when they would ever be used, but it does mean that the ifdef in arm_cpu_register_gdb_regs_for_features() needs to be adjusted to match. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210921162901.17508-4-peter.maydell@linaro.org --- target/arm/internals.h | 7 ++ target/arm/gdbstub.c | 130 ++++++++++++++++++++ target/arm/gdbstub64.c | 140 +++++++++++++++++++++ target/arm/helper.c | 271 ----------------------------------------- 4 files changed, 277 insertions(+), 271 deletions(-) diff --git a/target/arm/internals.h b/target/arm/internals.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/internals.h +++ b/target/arm/internals.h @@ -XXX,XX +XXX,XX @@ static inline uint64_t pmu_counter_mask(CPUARMState *env) return (1 << 31) | ((1 << pmu_num_counters(env)) - 1); } +#ifdef TARGET_AARCH64 +int arm_gdb_get_svereg(CPUARMState *env, GByteArray *buf, int reg); +int arm_gdb_set_svereg(CPUARMState *env, uint8_t *buf, int reg); +int aarch64_fpu_gdb_get_reg(CPUARMState *env, GByteArray *buf, int reg); +int aarch64_fpu_gdb_set_reg(CPUARMState *env, uint8_t *buf, int reg); +#endif + #endif diff --git a/target/arm/gdbstub.c b/target/arm/gdbstub.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/gdbstub.c +++ b/target/arm/gdbstub.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" #include "cpu.h" +#include "internals.h" #include "exec/gdbstub.h" typedef struct RegisterSysregXmlParam { @@ -XXX,XX +XXX,XX @@ int arm_cpu_gdb_write_register(CPUState *cs, uint8_t *mem_buf, int n) return 0; } +static int vfp_gdb_get_reg(CPUARMState *env, GByteArray *buf, int reg) +{ + ARMCPU *cpu = env_archcpu(env); + int nregs = cpu_isar_feature(aa32_simd_r32, cpu) ? 32 : 16; + + /* VFP data registers are always little-endian. */ + if (reg < nregs) { + return gdb_get_reg64(buf, *aa32_vfp_dreg(env, reg)); + } + if (arm_feature(env, ARM_FEATURE_NEON)) { + /* Aliases for Q regs. */ + nregs += 16; + if (reg < nregs) { + uint64_t *q = aa32_vfp_qreg(env, reg - 32); + return gdb_get_reg128(buf, q[0], q[1]); + } + } + switch (reg - nregs) { + case 0: + return gdb_get_reg32(buf, env->vfp.xregs[ARM_VFP_FPSID]); + case 1: + return gdb_get_reg32(buf, vfp_get_fpscr(env)); + case 2: + return gdb_get_reg32(buf, env->vfp.xregs[ARM_VFP_FPEXC]); + } + return 0; +} + +static int vfp_gdb_set_reg(CPUARMState *env, uint8_t *buf, int reg) +{ + ARMCPU *cpu = env_archcpu(env); + int nregs = cpu_isar_feature(aa32_simd_r32, cpu) ? 32 : 16; + + if (reg < nregs) { + *aa32_vfp_dreg(env, reg) = ldq_le_p(buf); + return 8; + } + if (arm_feature(env, ARM_FEATURE_NEON)) { + nregs += 16; + if (reg < nregs) { + uint64_t *q = aa32_vfp_qreg(env, reg - 32); + q[0] = ldq_le_p(buf); + q[1] = ldq_le_p(buf + 8); + return 16; + } + } + switch (reg - nregs) { + case 0: + env->vfp.xregs[ARM_VFP_FPSID] = ldl_p(buf); + return 4; + case 1: + vfp_set_fpscr(env, ldl_p(buf)); + return 4; + case 2: + env->vfp.xregs[ARM_VFP_FPEXC] = ldl_p(buf) & (1 << 30); + return 4; + } + return 0; +} + +/** + * arm_get/set_gdb_*: get/set a gdb register + * @env: the CPU state + * @buf: a buffer to copy to/from + * @reg: register number (offset from start of group) + * + * We return the number of bytes copied + */ + +static int arm_gdb_get_sysreg(CPUARMState *env, GByteArray *buf, int reg) +{ + ARMCPU *cpu = env_archcpu(env); + const ARMCPRegInfo *ri; + uint32_t key; + + key = cpu->dyn_sysreg_xml.data.cpregs.keys[reg]; + ri = get_arm_cp_reginfo(cpu->cp_regs, key); + if (ri) { + if (cpreg_field_is_64bit(ri)) { + return gdb_get_reg64(buf, (uint64_t)read_raw_cp_reg(env, ri)); + } else { + return gdb_get_reg32(buf, (uint32_t)read_raw_cp_reg(env, ri)); + } + } + return 0; +} + +static int arm_gdb_set_sysreg(CPUARMState *env, uint8_t *buf, int reg) +{ + return 0; +} + static void arm_gen_one_xml_sysreg_tag(GString *s, DynamicGDBXMLInfo *dyn_xml, ARMCPRegInfo *ri, uint32_t ri_key, int bitsize, int regnum) @@ -XXX,XX +XXX,XX @@ const char *arm_gdb_get_dynamic_xml(CPUState *cs, const char *xmlname) } return NULL; } + +void arm_cpu_register_gdb_regs_for_features(ARMCPU *cpu) +{ + CPUState *cs = CPU(cpu); + CPUARMState *env = &cpu->env; + + if (arm_feature(env, ARM_FEATURE_AARCH64)) { + /* + * The lower part of each SVE register aliases to the FPU + * registers so we don't need to include both. + */ +#ifdef TARGET_AARCH64 + if (isar_feature_aa64_sve(&cpu->isar)) { + gdb_register_coprocessor(cs, arm_gdb_get_svereg, arm_gdb_set_svereg, + arm_gen_dynamic_svereg_xml(cs, cs->gdb_num_regs), + "sve-registers.xml", 0); + } else { + gdb_register_coprocessor(cs, aarch64_fpu_gdb_get_reg, + aarch64_fpu_gdb_set_reg, + 34, "aarch64-fpu.xml", 0); + } +#endif + } else if (arm_feature(env, ARM_FEATURE_NEON)) { + gdb_register_coprocessor(cs, vfp_gdb_get_reg, vfp_gdb_set_reg, + 51, "arm-neon.xml", 0); + } else if (cpu_isar_feature(aa32_simd_r32, cpu)) { + gdb_register_coprocessor(cs, vfp_gdb_get_reg, vfp_gdb_set_reg, + 35, "arm-vfp3.xml", 0); + } else if (cpu_isar_feature(aa32_vfp_simd, cpu)) { + gdb_register_coprocessor(cs, vfp_gdb_get_reg, vfp_gdb_set_reg, + 19, "arm-vfp.xml", 0); + } + gdb_register_coprocessor(cs, arm_gdb_get_sysreg, arm_gdb_set_sysreg, + arm_gen_dynamic_sysreg_xml(cs, cs->gdb_num_regs), + "system-registers.xml", 0); + +} diff --git a/target/arm/gdbstub64.c b/target/arm/gdbstub64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/gdbstub64.c +++ b/target/arm/gdbstub64.c @@ -XXX,XX +XXX,XX @@ * License along with this library; if not, see <http://www.gnu.org/licenses/>. */ #include "qemu/osdep.h" +#include "qemu/log.h" #include "cpu.h" +#include "internals.h" #include "exec/gdbstub.h" int aarch64_cpu_gdb_read_register(CPUState *cs, GByteArray *mem_buf, int n) @@ -XXX,XX +XXX,XX @@ int aarch64_cpu_gdb_write_register(CPUState *cs, uint8_t *mem_buf, int n) /* Unknown register. */ return 0; } + +int aarch64_fpu_gdb_get_reg(CPUARMState *env, GByteArray *buf, int reg) +{ + switch (reg) { + case 0 ... 31: + { + /* 128 bit FP register - quads are in LE order */ + uint64_t *q = aa64_vfp_qreg(env, reg); + return gdb_get_reg128(buf, q[1], q[0]); + } + case 32: + /* FPSR */ + return gdb_get_reg32(buf, vfp_get_fpsr(env)); + case 33: + /* FPCR */ + return gdb_get_reg32(buf, vfp_get_fpcr(env)); + default: + return 0; + } +} + +int aarch64_fpu_gdb_set_reg(CPUARMState *env, uint8_t *buf, int reg) +{ + switch (reg) { + case 0 ... 31: + /* 128 bit FP register */ + { + uint64_t *q = aa64_vfp_qreg(env, reg); + q[0] = ldq_le_p(buf); + q[1] = ldq_le_p(buf + 8); + return 16; + } + case 32: + /* FPSR */ + vfp_set_fpsr(env, ldl_p(buf)); + return 4; + case 33: + /* FPCR */ + vfp_set_fpcr(env, ldl_p(buf)); + return 4; + default: + return 0; + } +} + +int arm_gdb_get_svereg(CPUARMState *env, GByteArray *buf, int reg) +{ + ARMCPU *cpu = env_archcpu(env); + + switch (reg) { + /* The first 32 registers are the zregs */ + case 0 ... 31: + { + int vq, len = 0; + for (vq = 0; vq < cpu->sve_max_vq; vq++) { + len += gdb_get_reg128(buf, + env->vfp.zregs[reg].d[vq * 2 + 1], + env->vfp.zregs[reg].d[vq * 2]); + } + return len; + } + case 32: + return gdb_get_reg32(buf, vfp_get_fpsr(env)); + case 33: + return gdb_get_reg32(buf, vfp_get_fpcr(env)); + /* then 16 predicates and the ffr */ + case 34 ... 50: + { + int preg = reg - 34; + int vq, len = 0; + for (vq = 0; vq < cpu->sve_max_vq; vq = vq + 4) { + len += gdb_get_reg64(buf, env->vfp.pregs[preg].p[vq / 4]); + } + return len; + } + case 51: + { + /* + * We report in Vector Granules (VG) which is 64bit in a Z reg + * while the ZCR works in Vector Quads (VQ) which is 128bit chunks. + */ + int vq = sve_zcr_len_for_el(env, arm_current_el(env)) + 1; + return gdb_get_reg64(buf, vq * 2); + } + default: + /* gdbstub asked for something out our range */ + qemu_log_mask(LOG_UNIMP, "%s: out of range register %d", __func__, reg); + break; + } + + return 0; +} + +int arm_gdb_set_svereg(CPUARMState *env, uint8_t *buf, int reg) +{ + ARMCPU *cpu = env_archcpu(env); + + /* The first 32 registers are the zregs */ + switch (reg) { + /* The first 32 registers are the zregs */ + case 0 ... 31: + { + int vq, len = 0; + uint64_t *p = (uint64_t *) buf; + for (vq = 0; vq < cpu->sve_max_vq; vq++) { + env->vfp.zregs[reg].d[vq * 2 + 1] = *p++; + env->vfp.zregs[reg].d[vq * 2] = *p++; + len += 16; + } + return len; + } + case 32: + vfp_set_fpsr(env, *(uint32_t *)buf); + return 4; + case 33: + vfp_set_fpcr(env, *(uint32_t *)buf); + return 4; + case 34 ... 50: + { + int preg = reg - 34; + int vq, len = 0; + uint64_t *p = (uint64_t *) buf; + for (vq = 0; vq < cpu->sve_max_vq; vq = vq + 4) { + env->vfp.pregs[preg].p[vq / 4] = *p++; + len += 8; + } + return len; + } + case 51: + /* cannot set vg via gdbstub */ + return 0; + default: + /* gdbstub asked for something out our range */ + break; + } + + return 0; +} diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ #include "trace.h" #include "cpu.h" #include "internals.h" -#include "exec/gdbstub.h" #include "exec/helper-proto.h" #include "qemu/host-utils.h" #include "qemu/main-loop.h" @@ -XXX,XX +XXX,XX @@ static bool get_phys_addr_lpae(CPUARMState *env, uint64_t address, static void switch_mode(CPUARMState *env, int mode); static int aa64_va_parameter_tbi(uint64_t tcr, ARMMMUIdx mmu_idx); -static int vfp_gdb_get_reg(CPUARMState *env, GByteArray *buf, int reg) -{ - ARMCPU *cpu = env_archcpu(env); - int nregs = cpu_isar_feature(aa32_simd_r32, cpu) ? 32 : 16; - - /* VFP data registers are always little-endian. */ - if (reg < nregs) { - return gdb_get_reg64(buf, *aa32_vfp_dreg(env, reg)); - } - if (arm_feature(env, ARM_FEATURE_NEON)) { - /* Aliases for Q regs. */ - nregs += 16; - if (reg < nregs) { - uint64_t *q = aa32_vfp_qreg(env, reg - 32); - return gdb_get_reg128(buf, q[0], q[1]); - } - } - switch (reg - nregs) { - case 0: - return gdb_get_reg32(buf, env->vfp.xregs[ARM_VFP_FPSID]); - case 1: - return gdb_get_reg32(buf, vfp_get_fpscr(env)); - case 2: - return gdb_get_reg32(buf, env->vfp.xregs[ARM_VFP_FPEXC]); - } - return 0; -} - -static int vfp_gdb_set_reg(CPUARMState *env, uint8_t *buf, int reg) -{ - ARMCPU *cpu = env_archcpu(env); - int nregs = cpu_isar_feature(aa32_simd_r32, cpu) ? 32 : 16; - - if (reg < nregs) { - *aa32_vfp_dreg(env, reg) = ldq_le_p(buf); - return 8; - } - if (arm_feature(env, ARM_FEATURE_NEON)) { - nregs += 16; - if (reg < nregs) { - uint64_t *q = aa32_vfp_qreg(env, reg - 32); - q[0] = ldq_le_p(buf); - q[1] = ldq_le_p(buf + 8); - return 16; - } - } - switch (reg - nregs) { - case 0: - env->vfp.xregs[ARM_VFP_FPSID] = ldl_p(buf); - return 4; - case 1: - vfp_set_fpscr(env, ldl_p(buf)); - return 4; - case 2: - env->vfp.xregs[ARM_VFP_FPEXC] = ldl_p(buf) & (1 << 30); - return 4; - } - return 0; -} - -static int aarch64_fpu_gdb_get_reg(CPUARMState *env, GByteArray *buf, int reg) -{ - switch (reg) { - case 0 ... 31: - { - /* 128 bit FP register - quads are in LE order */ - uint64_t *q = aa64_vfp_qreg(env, reg); - return gdb_get_reg128(buf, q[1], q[0]); - } - case 32: - /* FPSR */ - return gdb_get_reg32(buf, vfp_get_fpsr(env)); - case 33: - /* FPCR */ - return gdb_get_reg32(buf, vfp_get_fpcr(env)); - default: - return 0; - } -} - -static int aarch64_fpu_gdb_set_reg(CPUARMState *env, uint8_t *buf, int reg) -{ - switch (reg) { - case 0 ... 31: - /* 128 bit FP register */ - { - uint64_t *q = aa64_vfp_qreg(env, reg); - q[0] = ldq_le_p(buf); - q[1] = ldq_le_p(buf + 8); - return 16; - } - case 32: - /* FPSR */ - vfp_set_fpsr(env, ldl_p(buf)); - return 4; - case 33: - /* FPCR */ - vfp_set_fpcr(env, ldl_p(buf)); - return 4; - default: - return 0; - } -} - static uint64_t raw_read(CPUARMState *env, const ARMCPRegInfo *ri) { assert(ri->fieldoffset); @@ -XXX,XX +XXX,XX @@ static void write_raw_cp_reg(CPUARMState *env, const ARMCPRegInfo *ri, } } -/** - * arm_get/set_gdb_*: get/set a gdb register - * @env: the CPU state - * @buf: a buffer to copy to/from - * @reg: register number (offset from start of group) - * - * We return the number of bytes copied - */ - -static int arm_gdb_get_sysreg(CPUARMState *env, GByteArray *buf, int reg) -{ - ARMCPU *cpu = env_archcpu(env); - const ARMCPRegInfo *ri; - uint32_t key; - - key = cpu->dyn_sysreg_xml.data.cpregs.keys[reg]; - ri = get_arm_cp_reginfo(cpu->cp_regs, key); - if (ri) { - if (cpreg_field_is_64bit(ri)) { - return gdb_get_reg64(buf, (uint64_t)read_raw_cp_reg(env, ri)); - } else { - return gdb_get_reg32(buf, (uint32_t)read_raw_cp_reg(env, ri)); - } - } - return 0; -} - -static int arm_gdb_set_sysreg(CPUARMState *env, uint8_t *buf, int reg) -{ - return 0; -} - -#ifdef TARGET_AARCH64 -static int arm_gdb_get_svereg(CPUARMState *env, GByteArray *buf, int reg) -{ - ARMCPU *cpu = env_archcpu(env); - - switch (reg) { - /* The first 32 registers are the zregs */ - case 0 ... 31: - { - int vq, len = 0; - for (vq = 0; vq < cpu->sve_max_vq; vq++) { - len += gdb_get_reg128(buf, - env->vfp.zregs[reg].d[vq * 2 + 1], - env->vfp.zregs[reg].d[vq * 2]); - } - return len; - } - case 32: - return gdb_get_reg32(buf, vfp_get_fpsr(env)); - case 33: - return gdb_get_reg32(buf, vfp_get_fpcr(env)); - /* then 16 predicates and the ffr */ - case 34 ... 50: - { - int preg = reg - 34; - int vq, len = 0; - for (vq = 0; vq < cpu->sve_max_vq; vq = vq + 4) { - len += gdb_get_reg64(buf, env->vfp.pregs[preg].p[vq / 4]); - } - return len; - } - case 51: - { - /* - * We report in Vector Granules (VG) which is 64bit in a Z reg - * while the ZCR works in Vector Quads (VQ) which is 128bit chunks. - */ - int vq = sve_zcr_len_for_el(env, arm_current_el(env)) + 1; - return gdb_get_reg64(buf, vq * 2); - } - default: - /* gdbstub asked for something out our range */ - qemu_log_mask(LOG_UNIMP, "%s: out of range register %d", __func__, reg); - break; - } - - return 0; -} - -static int arm_gdb_set_svereg(CPUARMState *env, uint8_t *buf, int reg) -{ - ARMCPU *cpu = env_archcpu(env); - - /* The first 32 registers are the zregs */ - switch (reg) { - /* The first 32 registers are the zregs */ - case 0 ... 31: - { - int vq, len = 0; - uint64_t *p = (uint64_t *) buf; - for (vq = 0; vq < cpu->sve_max_vq; vq++) { - env->vfp.zregs[reg].d[vq * 2 + 1] = *p++; - env->vfp.zregs[reg].d[vq * 2] = *p++; - len += 16; - } - return len; - } - case 32: - vfp_set_fpsr(env, *(uint32_t *)buf); - return 4; - case 33: - vfp_set_fpcr(env, *(uint32_t *)buf); - return 4; - case 34 ... 50: - { - int preg = reg - 34; - int vq, len = 0; - uint64_t *p = (uint64_t *) buf; - for (vq = 0; vq < cpu->sve_max_vq; vq = vq + 4) { - env->vfp.pregs[preg].p[vq / 4] = *p++; - len += 8; - } - return len; - } - case 51: - /* cannot set vg via gdbstub */ - return 0; - default: - /* gdbstub asked for something out our range */ - break; - } - - return 0; -} -#endif /* TARGET_AARCH64 */ - static bool raw_accessors_invalid(const ARMCPRegInfo *ri) { /* Return true if the regdef would cause an assertion if you called @@ -XXX,XX +XXX,XX @@ void register_cp_regs_for_features(ARMCPU *cpu) #endif } -void arm_cpu_register_gdb_regs_for_features(ARMCPU *cpu) -{ - CPUState *cs = CPU(cpu); - CPUARMState *env = &cpu->env; - - if (arm_feature(env, ARM_FEATURE_AARCH64)) { - /* - * The lower part of each SVE register aliases to the FPU - * registers so we don't need to include both. - */ -#ifdef TARGET_AARCH64 - if (isar_feature_aa64_sve(&cpu->isar)) { - gdb_register_coprocessor(cs, arm_gdb_get_svereg, arm_gdb_set_svereg, - arm_gen_dynamic_svereg_xml(cs, cs->gdb_num_regs), - "sve-registers.xml", 0); - } else -#endif - { - gdb_register_coprocessor(cs, aarch64_fpu_gdb_get_reg, - aarch64_fpu_gdb_set_reg, - 34, "aarch64-fpu.xml", 0); - } - } else if (arm_feature(env, ARM_FEATURE_NEON)) { - gdb_register_coprocessor(cs, vfp_gdb_get_reg, vfp_gdb_set_reg, - 51, "arm-neon.xml", 0); - } else if (cpu_isar_feature(aa32_simd_r32, cpu)) { - gdb_register_coprocessor(cs, vfp_gdb_get_reg, vfp_gdb_set_reg, - 35, "arm-vfp3.xml", 0); - } else if (cpu_isar_feature(aa32_vfp_simd, cpu)) { - gdb_register_coprocessor(cs, vfp_gdb_get_reg, vfp_gdb_set_reg, - 19, "arm-vfp.xml", 0); - } - gdb_register_coprocessor(cs, arm_gdb_get_sysreg, arm_gdb_set_sysreg, - arm_gen_dynamic_sysreg_xml(cs, cs->gdb_num_regs), - "system-registers.xml", 0); - -} - /* Sort alphabetically by type name, except for "any". */ static gint arm_cpu_list_compare(gconstpointer a, gconstpointer b) { -- 2.20.1 Currently we send VFP XML which includes D0..D15 or D0..D31, plus FPSID, FPSCR and FPEXC. The upstream GDB tolerates this, but its definition of this XML feature does not include FPSID or FPEXC. In particular, for M-profile cores there are no FPSID or FPEXC registers, so advertising those is wrong. Move FPSID and FPEXC into their own bit of XML which we only send for A and R profile cores. This brings our definition of the XML org.gnu.gdb.arm.vfp feature into line with GDB's own (at least for non-Neon cores...) and means we don't claim to have FPSID and FPEXC on M-profile. (It seems unlikely to me that any gdbstub users really care about being able to look at FPEXC and FPSID; but we've supplied them to gdb for a decade and it's not hard to keep doing so.) Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210921162901.17508-5-peter.maydell@linaro.org --- configs/targets/aarch64-softmmu.mak | 2 +- configs/targets/arm-linux-user.mak | 2 +- configs/targets/arm-softmmu.mak | 2 +- configs/targets/armeb-linux-user.mak | 2 +- target/arm/gdbstub.c | 56 ++++++++++++++++++++-------- gdb-xml/arm-neon.xml | 2 - gdb-xml/arm-vfp-sysregs.xml | 17 +++++++++ gdb-xml/arm-vfp.xml | 2 - gdb-xml/arm-vfp3.xml | 2 - 9 files changed, 61 insertions(+), 26 deletions(-) create mode 100644 gdb-xml/arm-vfp-sysregs.xml diff --git a/configs/targets/aarch64-softmmu.mak b/configs/targets/aarch64-softmmu.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/aarch64-softmmu.mak +++ b/configs/targets/aarch64-softmmu.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=aarch64 TARGET_BASE_ARCH=arm TARGET_SUPPORTS_MTTCG=y -TARGET_XML_FILES= gdb-xml/aarch64-core.xml gdb-xml/aarch64-fpu.xml gdb-xml/arm-core.xml gdb-xml/arm-vfp.xml gdb-xml/arm-vfp3.xml gdb-xml/arm-neon.xml gdb-xml/arm-m-profile.xml +TARGET_XML_FILES= gdb-xml/aarch64-core.xml gdb-xml/aarch64-fpu.xml gdb-xml/arm-core.xml gdb-xml/arm-vfp.xml gdb-xml/arm-vfp3.xml gdb-xml/arm-vfp-sysregs.xml gdb-xml/arm-neon.xml gdb-xml/arm-m-profile.xml TARGET_NEED_FDT=y diff --git a/configs/targets/arm-linux-user.mak b/configs/targets/arm-linux-user.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/arm-linux-user.mak +++ b/configs/targets/arm-linux-user.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=arm TARGET_SYSTBL_ABI=common,oabi TARGET_SYSTBL=syscall.tbl -TARGET_XML_FILES= gdb-xml/arm-core.xml gdb-xml/arm-vfp.xml gdb-xml/arm-vfp3.xml gdb-xml/arm-neon.xml gdb-xml/arm-m-profile.xml +TARGET_XML_FILES= gdb-xml/arm-core.xml gdb-xml/arm-vfp.xml gdb-xml/arm-vfp3.xml gdb-xml/arm-vfp-sysregs.xml gdb-xml/arm-neon.xml gdb-xml/arm-m-profile.xml TARGET_HAS_BFLT=y CONFIG_ARM_COMPATIBLE_SEMIHOSTING=y diff --git a/configs/targets/arm-softmmu.mak b/configs/targets/arm-softmmu.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/arm-softmmu.mak +++ b/configs/targets/arm-softmmu.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=arm TARGET_SUPPORTS_MTTCG=y -TARGET_XML_FILES= gdb-xml/arm-core.xml gdb-xml/arm-vfp.xml gdb-xml/arm-vfp3.xml gdb-xml/arm-neon.xml gdb-xml/arm-m-profile.xml +TARGET_XML_FILES= gdb-xml/arm-core.xml gdb-xml/arm-vfp.xml gdb-xml/arm-vfp3.xml gdb-xml/arm-vfp-sysregs.xml gdb-xml/arm-neon.xml gdb-xml/arm-m-profile.xml TARGET_NEED_FDT=y diff --git a/configs/targets/armeb-linux-user.mak b/configs/targets/armeb-linux-user.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/armeb-linux-user.mak +++ b/configs/targets/armeb-linux-user.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=arm TARGET_SYSTBL_ABI=common,oabi TARGET_SYSTBL=syscall.tbl TARGET_WORDS_BIGENDIAN=y -TARGET_XML_FILES= gdb-xml/arm-core.xml gdb-xml/arm-vfp.xml gdb-xml/arm-vfp3.xml gdb-xml/arm-neon.xml gdb-xml/arm-m-profile.xml +TARGET_XML_FILES= gdb-xml/arm-core.xml gdb-xml/arm-vfp.xml gdb-xml/arm-vfp3.xml gdb-xml/arm-vfp-sysregs.xml gdb-xml/arm-neon.xml gdb-xml/arm-m-profile.xml TARGET_HAS_BFLT=y CONFIG_ARM_COMPATIBLE_SEMIHOSTING=y diff --git a/target/arm/gdbstub.c b/target/arm/gdbstub.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/gdbstub.c +++ b/target/arm/gdbstub.c @@ -XXX,XX +XXX,XX @@ static int vfp_gdb_get_reg(CPUARMState *env, GByteArray *buf, int reg) } switch (reg - nregs) { case 0: - return gdb_get_reg32(buf, env->vfp.xregs[ARM_VFP_FPSID]); - case 1: return gdb_get_reg32(buf, vfp_get_fpscr(env)); - case 2: - return gdb_get_reg32(buf, env->vfp.xregs[ARM_VFP_FPEXC]); } return 0; } @@ -XXX,XX +XXX,XX @@ static int vfp_gdb_set_reg(CPUARMState *env, uint8_t *buf, int reg) } } switch (reg - nregs) { + case 0: + vfp_set_fpscr(env, ldl_p(buf)); + return 4; + } + return 0; +} + +static int vfp_gdb_get_sysreg(CPUARMState *env, GByteArray *buf, int reg) +{ + switch (reg) { + case 0: + return gdb_get_reg32(buf, env->vfp.xregs[ARM_VFP_FPSID]); + case 1: + return gdb_get_reg32(buf, env->vfp.xregs[ARM_VFP_FPEXC]); + } + return 0; +} + +static int vfp_gdb_set_sysreg(CPUARMState *env, uint8_t *buf, int reg) +{ + switch (reg) { case 0: env->vfp.xregs[ARM_VFP_FPSID] = ldl_p(buf); return 4; case 1: - vfp_set_fpscr(env, ldl_p(buf)); - return 4; - case 2: env->vfp.xregs[ARM_VFP_FPEXC] = ldl_p(buf) & (1 << 30); return 4; } @@ -XXX,XX +XXX,XX @@ void arm_cpu_register_gdb_regs_for_features(ARMCPU *cpu) 34, "aarch64-fpu.xml", 0); } #endif - } else if (arm_feature(env, ARM_FEATURE_NEON)) { - gdb_register_coprocessor(cs, vfp_gdb_get_reg, vfp_gdb_set_reg, - 51, "arm-neon.xml", 0); - } else if (cpu_isar_feature(aa32_simd_r32, cpu)) { - gdb_register_coprocessor(cs, vfp_gdb_get_reg, vfp_gdb_set_reg, - 35, "arm-vfp3.xml", 0); - } else if (cpu_isar_feature(aa32_vfp_simd, cpu)) { - gdb_register_coprocessor(cs, vfp_gdb_get_reg, vfp_gdb_set_reg, - 19, "arm-vfp.xml", 0); + } else { + if (arm_feature(env, ARM_FEATURE_NEON)) { + gdb_register_coprocessor(cs, vfp_gdb_get_reg, vfp_gdb_set_reg, + 49, "arm-neon.xml", 0); + } else if (cpu_isar_feature(aa32_simd_r32, cpu)) { + gdb_register_coprocessor(cs, vfp_gdb_get_reg, vfp_gdb_set_reg, + 33, "arm-vfp3.xml", 0); + } else if (cpu_isar_feature(aa32_vfp_simd, cpu)) { + gdb_register_coprocessor(cs, vfp_gdb_get_reg, vfp_gdb_set_reg, + 17, "arm-vfp.xml", 0); + } + if (!arm_feature(env, ARM_FEATURE_M)) { + /* + * A and R profile have FP sysregs FPEXC and FPSID that we + * expose to gdb. + */ + gdb_register_coprocessor(cs, vfp_gdb_get_sysreg, vfp_gdb_set_sysreg, + 2, "arm-vfp-sysregs.xml", 0); + } } gdb_register_coprocessor(cs, arm_gdb_get_sysreg, arm_gdb_set_sysreg, arm_gen_dynamic_sysreg_xml(cs, cs->gdb_num_regs), diff --git a/gdb-xml/arm-neon.xml b/gdb-xml/arm-neon.xml index XXXXXXX..XXXXXXX 100644 --- a/gdb-xml/arm-neon.xml +++ b/gdb-xml/arm-neon.xml @@ -XXX,XX +XXX,XX @@ <reg name="q14" bitsize="128" type="neon_q"/> <reg name="q15" bitsize="128" type="neon_q"/> - <reg name="fpsid" bitsize="32" type="int" group="float"/> <reg name="fpscr" bitsize="32" type="int" group="float"/> - <reg name="fpexc" bitsize="32" type="int" group="float"/> </feature> diff --git a/gdb-xml/arm-vfp-sysregs.xml b/gdb-xml/arm-vfp-sysregs.xml new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/gdb-xml/arm-vfp-sysregs.xml @@ -XXX,XX +XXX,XX @@ +<?xml version="1.0"?> +<!-- Copyright (C) 2021 Linaro Ltd. + + Copying and distribution of this file, with or without modification, + are permitted in any medium without royalty provided the copyright + notice and this notice are preserved. + + These are A/R profile VFP system registers. Debugger users probably + don't really care about these, but because we used to (incorrectly) + provide them to gdb in the org.gnu.gdb.arm.vfp XML we continue + to do so via this separate XML. + --> +<!DOCTYPE feature SYSTEM "gdb-target.dtd"> +<feature name="org.qemu.gdb.arm.vfp-sysregs"> + <reg name="fpsid" bitsize="32" type="int" group="float"/> + <reg name="fpexc" bitsize="32" type="int" group="float"/> +</feature> diff --git a/gdb-xml/arm-vfp.xml b/gdb-xml/arm-vfp.xml index XXXXXXX..XXXXXXX 100644 --- a/gdb-xml/arm-vfp.xml +++ b/gdb-xml/arm-vfp.xml @@ -XXX,XX +XXX,XX @@ <reg name="d14" bitsize="64" type="float"/> <reg name="d15" bitsize="64" type="float"/> - <reg name="fpsid" bitsize="32" type="int" group="float"/> <reg name="fpscr" bitsize="32" type="int" group="float"/> - <reg name="fpexc" bitsize="32" type="int" group="float"/> </feature> diff --git a/gdb-xml/arm-vfp3.xml b/gdb-xml/arm-vfp3.xml index XXXXXXX..XXXXXXX 100644 --- a/gdb-xml/arm-vfp3.xml +++ b/gdb-xml/arm-vfp3.xml @@ -XXX,XX +XXX,XX @@ <reg name="d30" bitsize="64" type="float"/> <reg name="d31" bitsize="64" type="float"/> - <reg name="fpsid" bitsize="32" type="int" group="float"/> <reg name="fpscr" bitsize="32" type="int" group="float"/> - <reg name="fpexc" bitsize="32" type="int" group="float"/> </feature> -- 2.20.1 The function scsi_bus_new() creates a new SCSI bus; callers can either pass in a name argument to specify the name of the new bus, or they can pass in NULL to allow the bus to be given an automatically generated unique name. Almost all callers want to use the autogenerated name; the only exception is the virtio-scsi device. Taking a name argument that should almost always be NULL is an easy-to-misuse API design -- it encourages callers to think perhaps they should pass in some standard name like "scsi" or "scsi-bus". We don't do this anywhere for SCSI, but we do (incorrectly) do it for other bus types such as i2c. The function name also implies that it will return a newly allocated object, when it in fact does in-place allocation. We more commonly name such functions foo_init(), with foo_new() being the allocate-and-return variant. Replace all the scsi_bus_new() callsites with either: * scsi_bus_init() for the usual case where the caller wants an autogenerated bus name * scsi_bus_init_named() for the rare case where the caller needs to specify the bus name and document that for the _named() version it's then the caller's responsibility to think about uniqueness of bus names. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: Paolo Bonzini <pbonzini@redhat.com> Message-id: 20210923121153.23754-2-peter.maydell@linaro.org --- include/hw/scsi/scsi.h | 30 ++++++++++++++++++++++++++++-- hw/scsi/esp-pci.c | 2 +- hw/scsi/esp.c | 2 +- hw/scsi/lsi53c895a.c | 2 +- hw/scsi/megasas.c | 3 +-- hw/scsi/mptsas.c | 2 +- hw/scsi/scsi-bus.c | 4 ++-- hw/scsi/spapr_vscsi.c | 3 +-- hw/scsi/virtio-scsi.c | 4 ++-- hw/scsi/vmw_pvscsi.c | 3 +-- hw/usb/dev-storage-bot.c | 3 +-- hw/usb/dev-storage-classic.c | 4 ++-- hw/usb/dev-uas.c | 3 +-- 13 files changed, 43 insertions(+), 22 deletions(-) diff --git a/include/hw/scsi/scsi.h b/include/hw/scsi/scsi.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/scsi/scsi.h +++ b/include/hw/scsi/scsi.h @@ -XXX,XX +XXX,XX @@ struct SCSIBus { const SCSIBusInfo *info; }; -void scsi_bus_new(SCSIBus *bus, size_t bus_size, DeviceState *host, - const SCSIBusInfo *info, const char *bus_name); +/** + * scsi_bus_init_named: Initialize a SCSI bus with the specified name + * @bus: SCSIBus object to initialize + * @bus_size: size of @bus object + * @host: Device which owns the bus (generally the SCSI controller) + * @info: structure defining callbacks etc for the controller + * @bus_name: Name to use for this bus + * + * This in-place initializes @bus as a new SCSI bus with a name + * provided by the caller. It is the caller's responsibility to make + * sure that name does not clash with the name of any other bus in the + * system. Unless you need the new bus to have a specific name, you + * should use scsi_bus_new() instead. + */ +void scsi_bus_init_named(SCSIBus *bus, size_t bus_size, DeviceState *host, + const SCSIBusInfo *info, const char *bus_name); + +/** + * scsi_bus_init: Initialize a SCSI bus + * + * This in-place-initializes @bus as a new SCSI bus and gives it + * an automatically generated unique name. + */ +static inline void scsi_bus_init(SCSIBus *bus, size_t bus_size, + DeviceState *host, const SCSIBusInfo *info) +{ + scsi_bus_init_named(bus, bus_size, host, info, NULL); +} static inline SCSIBus *scsi_bus_from_device(SCSIDevice *d) { diff --git a/hw/scsi/esp-pci.c b/hw/scsi/esp-pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/scsi/esp-pci.c +++ b/hw/scsi/esp-pci.c @@ -XXX,XX +XXX,XX @@ static void esp_pci_scsi_realize(PCIDevice *dev, Error **errp) pci_register_bar(dev, 0, PCI_BASE_ADDRESS_SPACE_IO, &pci->io); s->irq = pci_allocate_irq(dev); - scsi_bus_new(&s->bus, sizeof(s->bus), d, &esp_pci_scsi_info, NULL); + scsi_bus_init(&s->bus, sizeof(s->bus), d, &esp_pci_scsi_info); } static void esp_pci_scsi_exit(PCIDevice *d) diff --git a/hw/scsi/esp.c b/hw/scsi/esp.c index XXXXXXX..XXXXXXX 100644 --- a/hw/scsi/esp.c +++ b/hw/scsi/esp.c @@ -XXX,XX +XXX,XX @@ static void sysbus_esp_realize(DeviceState *dev, Error **errp) qdev_init_gpio_in(dev, sysbus_esp_gpio_demux, 2); - scsi_bus_new(&s->bus, sizeof(s->bus), dev, &esp_scsi_info, NULL); + scsi_bus_init(&s->bus, sizeof(s->bus), dev, &esp_scsi_info); } static void sysbus_esp_hard_reset(DeviceState *dev) diff --git a/hw/scsi/lsi53c895a.c b/hw/scsi/lsi53c895a.c index XXXXXXX..XXXXXXX 100644 --- a/hw/scsi/lsi53c895a.c +++ b/hw/scsi/lsi53c895a.c @@ -XXX,XX +XXX,XX @@ static void lsi_scsi_realize(PCIDevice *dev, Error **errp) pci_register_bar(dev, 2, PCI_BASE_ADDRESS_SPACE_MEMORY, &s->ram_io); QTAILQ_INIT(&s->queue); - scsi_bus_new(&s->bus, sizeof(s->bus), d, &lsi_scsi_info, NULL); + scsi_bus_init(&s->bus, sizeof(s->bus), d, &lsi_scsi_info); } static void lsi_scsi_exit(PCIDevice *dev) diff --git a/hw/scsi/megasas.c b/hw/scsi/megasas.c index XXXXXXX..XXXXXXX 100644 --- a/hw/scsi/megasas.c +++ b/hw/scsi/megasas.c @@ -XXX,XX +XXX,XX @@ static void megasas_scsi_realize(PCIDevice *dev, Error **errp) s->frames[i].state = s; } - scsi_bus_new(&s->bus, sizeof(s->bus), DEVICE(dev), - &megasas_scsi_info, NULL); + scsi_bus_init(&s->bus, sizeof(s->bus), DEVICE(dev), &megasas_scsi_info); } static Property megasas_properties_gen1[] = { diff --git a/hw/scsi/mptsas.c b/hw/scsi/mptsas.c index XXXXXXX..XXXXXXX 100644 --- a/hw/scsi/mptsas.c +++ b/hw/scsi/mptsas.c @@ -XXX,XX +XXX,XX @@ static void mptsas_scsi_realize(PCIDevice *dev, Error **errp) s->request_bh = qemu_bh_new(mptsas_fetch_requests, s); - scsi_bus_new(&s->bus, sizeof(s->bus), &dev->qdev, &mptsas_scsi_info, NULL); + scsi_bus_init(&s->bus, sizeof(s->bus), &dev->qdev, &mptsas_scsi_info); } static void mptsas_scsi_uninit(PCIDevice *dev) diff --git a/hw/scsi/scsi-bus.c b/hw/scsi/scsi-bus.c index XXXXXXX..XXXXXXX 100644 --- a/hw/scsi/scsi-bus.c +++ b/hw/scsi/scsi-bus.c @@ -XXX,XX +XXX,XX @@ void scsi_device_unit_attention_reported(SCSIDevice *s) } /* Create a scsi bus, and attach devices to it. */ -void scsi_bus_new(SCSIBus *bus, size_t bus_size, DeviceState *host, - const SCSIBusInfo *info, const char *bus_name) +void scsi_bus_init_named(SCSIBus *bus, size_t bus_size, DeviceState *host, + const SCSIBusInfo *info, const char *bus_name) { qbus_create_inplace(bus, bus_size, TYPE_SCSI_BUS, host, bus_name); bus->busnr = next_scsi_bus++; diff --git a/hw/scsi/spapr_vscsi.c b/hw/scsi/spapr_vscsi.c index XXXXXXX..XXXXXXX 100644 --- a/hw/scsi/spapr_vscsi.c +++ b/hw/scsi/spapr_vscsi.c @@ -XXX,XX +XXX,XX @@ static void spapr_vscsi_realize(SpaprVioDevice *dev, Error **errp) dev->crq.SendFunc = vscsi_do_crq; - scsi_bus_new(&s->bus, sizeof(s->bus), DEVICE(dev), - &vscsi_scsi_info, NULL); + scsi_bus_init(&s->bus, sizeof(s->bus), DEVICE(dev), &vscsi_scsi_info); /* ibmvscsi SCSI bus does not allow hotplug. */ qbus_set_hotplug_handler(BUS(&s->bus), NULL); diff --git a/hw/scsi/virtio-scsi.c b/hw/scsi/virtio-scsi.c index XXXXXXX..XXXXXXX 100644 --- a/hw/scsi/virtio-scsi.c +++ b/hw/scsi/virtio-scsi.c @@ -XXX,XX +XXX,XX @@ static void virtio_scsi_device_realize(DeviceState *dev, Error **errp) return; } - scsi_bus_new(&s->bus, sizeof(s->bus), dev, - &virtio_scsi_scsi_info, vdev->bus_name); + scsi_bus_init_named(&s->bus, sizeof(s->bus), dev, + &virtio_scsi_scsi_info, vdev->bus_name); /* override default SCSI bus hotplug-handler, with virtio-scsi's one */ qbus_set_hotplug_handler(BUS(&s->bus), OBJECT(dev)); diff --git a/hw/scsi/vmw_pvscsi.c b/hw/scsi/vmw_pvscsi.c index XXXXXXX..XXXXXXX 100644 --- a/hw/scsi/vmw_pvscsi.c +++ b/hw/scsi/vmw_pvscsi.c @@ -XXX,XX +XXX,XX @@ pvscsi_realizefn(PCIDevice *pci_dev, Error **errp) s->completion_worker = qemu_bh_new(pvscsi_process_completion_queue, s); - scsi_bus_new(&s->bus, sizeof(s->bus), DEVICE(pci_dev), - &pvscsi_scsi_info, NULL); + scsi_bus_init(&s->bus, sizeof(s->bus), DEVICE(pci_dev), &pvscsi_scsi_info); /* override default SCSI bus hotplug-handler, with pvscsi's one */ qbus_set_hotplug_handler(BUS(&s->bus), OBJECT(s)); pvscsi_reset_state(s); diff --git a/hw/usb/dev-storage-bot.c b/hw/usb/dev-storage-bot.c index XXXXXXX..XXXXXXX 100644 --- a/hw/usb/dev-storage-bot.c +++ b/hw/usb/dev-storage-bot.c @@ -XXX,XX +XXX,XX @@ static void usb_msd_bot_realize(USBDevice *dev, Error **errp) s->dev.auto_attach = 0; } - scsi_bus_new(&s->bus, sizeof(s->bus), DEVICE(dev), - &usb_msd_scsi_info_bot, NULL); + scsi_bus_init(&s->bus, sizeof(s->bus), DEVICE(dev), &usb_msd_scsi_info_bot); usb_msd_handle_reset(dev); } diff --git a/hw/usb/dev-storage-classic.c b/hw/usb/dev-storage-classic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/usb/dev-storage-classic.c +++ b/hw/usb/dev-storage-classic.c @@ -XXX,XX +XXX,XX @@ static void usb_msd_storage_realize(USBDevice *dev, Error **errp) usb_desc_create_serial(dev); usb_desc_init(dev); dev->flags |= (1 << USB_DEV_FLAG_IS_SCSI_STORAGE); - scsi_bus_new(&s->bus, sizeof(s->bus), DEVICE(dev), - &usb_msd_scsi_info_storage, NULL); + scsi_bus_init(&s->bus, sizeof(s->bus), DEVICE(dev), + &usb_msd_scsi_info_storage); scsi_dev = scsi_bus_legacy_add_drive(&s->bus, blk, 0, !!s->removable, s->conf.bootindex, s->conf.share_rw, s->conf.rerror, s->conf.werror, diff --git a/hw/usb/dev-uas.c b/hw/usb/dev-uas.c index XXXXXXX..XXXXXXX 100644 --- a/hw/usb/dev-uas.c +++ b/hw/usb/dev-uas.c @@ -XXX,XX +XXX,XX @@ static void usb_uas_realize(USBDevice *dev, Error **errp) uas->status_bh = qemu_bh_new(usb_uas_send_status_bh, uas); dev->flags |= (1 << USB_DEV_FLAG_IS_SCSI_STORAGE); - scsi_bus_new(&uas->bus, sizeof(uas->bus), DEVICE(dev), - &usb_uas_scsi_info, NULL); + scsi_bus_init(&uas->bus, sizeof(uas->bus), DEVICE(dev), &usb_uas_scsi_info); } static const VMStateDescription vmstate_usb_uas = { -- 2.20.1 Rename ipack_bus_new_inplace() to ipack_bus_init(), to bring it in to line with a "_init for in-place init, _new for allocate-and-return" convention. Drop the 'name' argument, because the only caller does not pass in a name. If a future caller does need to specify the bus name, we should create an ipack_bus_init_named() function at that point. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Message-id: 20210923121153.23754-3-peter.maydell@linaro.org --- include/hw/ipack/ipack.h | 8 ++++---- hw/ipack/ipack.c | 10 +++++----- hw/ipack/tpci200.c | 4 ++-- 3 files changed, 11 insertions(+), 11 deletions(-) diff --git a/include/hw/ipack/ipack.h b/include/hw/ipack/ipack.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/ipack/ipack.h +++ b/include/hw/ipack/ipack.h @@ -XXX,XX +XXX,XX @@ extern const VMStateDescription vmstate_ipack_device; VMSTATE_STRUCT(_field, _state, 1, vmstate_ipack_device, IPackDevice) IPackDevice *ipack_device_find(IPackBus *bus, int32_t slot); -void ipack_bus_new_inplace(IPackBus *bus, size_t bus_size, - DeviceState *parent, - const char *name, uint8_t n_slots, - qemu_irq_handler handler); +void ipack_bus_init(IPackBus *bus, size_t bus_size, + DeviceState *parent, + uint8_t n_slots, + qemu_irq_handler handler); #endif diff --git a/hw/ipack/ipack.c b/hw/ipack/ipack.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ipack/ipack.c +++ b/hw/ipack/ipack.c @@ -XXX,XX +XXX,XX @@ IPackDevice *ipack_device_find(IPackBus *bus, int32_t slot) return NULL; } -void ipack_bus_new_inplace(IPackBus *bus, size_t bus_size, - DeviceState *parent, - const char *name, uint8_t n_slots, - qemu_irq_handler handler) +void ipack_bus_init(IPackBus *bus, size_t bus_size, + DeviceState *parent, + uint8_t n_slots, + qemu_irq_handler handler) { - qbus_create_inplace(bus, bus_size, TYPE_IPACK_BUS, parent, name); + qbus_create_inplace(bus, bus_size, TYPE_IPACK_BUS, parent, NULL); bus->n_slots = n_slots; bus->set_irq = handler; } diff --git a/hw/ipack/tpci200.c b/hw/ipack/tpci200.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ipack/tpci200.c +++ b/hw/ipack/tpci200.c @@ -XXX,XX +XXX,XX @@ static void tpci200_realize(PCIDevice *pci_dev, Error **errp) pci_register_bar(&s->dev, 4, PCI_BASE_ADDRESS_SPACE_MEMORY, &s->las2); pci_register_bar(&s->dev, 5, PCI_BASE_ADDRESS_SPACE_MEMORY, &s->las3); - ipack_bus_new_inplace(&s->bus, sizeof(s->bus), DEVICE(pci_dev), NULL, - N_MODULES, tpci200_set_irq); + ipack_bus_init(&s->bus, sizeof(s->bus), DEVICE(pci_dev), + N_MODULES, tpci200_set_irq); } static const VMStateDescription vmstate_tpci200 = { -- 2.20.1 Rename the pci_root_bus_new_inplace() function to pci_root_bus_init(); this brings the bus type in to line with a "_init for in-place init, _new for allocate-and-return" convention. To do this we need to rename the implementation-internal function that was using the pci_root_bus_init() name to pci_root_bus_internal_init(). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Message-id: 20210923121153.23754-4-peter.maydell@linaro.org --- include/hw/pci/pci.h | 10 +++++----- hw/pci-host/raven.c | 4 ++-- hw/pci-host/versatile.c | 6 +++--- hw/pci/pci.c | 26 +++++++++++++------------- 4 files changed, 23 insertions(+), 23 deletions(-) diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/pci/pci.h +++ b/include/hw/pci/pci.h @@ -XXX,XX +XXX,XX @@ OBJECT_DECLARE_TYPE(PCIBus, PCIBusClass, PCI_BUS) bool pci_bus_is_express(PCIBus *bus); -void pci_root_bus_new_inplace(PCIBus *bus, size_t bus_size, DeviceState *parent, - const char *name, - MemoryRegion *address_space_mem, - MemoryRegion *address_space_io, - uint8_t devfn_min, const char *typename); +void pci_root_bus_init(PCIBus *bus, size_t bus_size, DeviceState *parent, + const char *name, + MemoryRegion *address_space_mem, + MemoryRegion *address_space_io, + uint8_t devfn_min, const char *typename); PCIBus *pci_root_bus_new(DeviceState *parent, const char *name, MemoryRegion *address_space_mem, MemoryRegion *address_space_io, diff --git a/hw/pci-host/raven.c b/hw/pci-host/raven.c index XXXXXXX..XXXXXXX 100644 --- a/hw/pci-host/raven.c +++ b/hw/pci-host/raven.c @@ -XXX,XX +XXX,XX @@ static void raven_pcihost_initfn(Object *obj) memory_region_add_subregion_overlap(address_space_mem, PCI_IO_BASE_ADDR, &s->pci_io_non_contiguous, 1); memory_region_add_subregion(address_space_mem, 0xc0000000, &s->pci_memory); - pci_root_bus_new_inplace(&s->pci_bus, sizeof(s->pci_bus), DEVICE(obj), NULL, - &s->pci_memory, &s->pci_io, 0, TYPE_PCI_BUS); + pci_root_bus_init(&s->pci_bus, sizeof(s->pci_bus), DEVICE(obj), NULL, + &s->pci_memory, &s->pci_io, 0, TYPE_PCI_BUS); /* Bus master address space */ memory_region_init(&s->bm, obj, "bm-raven", 4 * GiB); diff --git a/hw/pci-host/versatile.c b/hw/pci-host/versatile.c index XXXXXXX..XXXXXXX 100644 --- a/hw/pci-host/versatile.c +++ b/hw/pci-host/versatile.c @@ -XXX,XX +XXX,XX @@ static void pci_vpb_realize(DeviceState *dev, Error **errp) memory_region_init(&s->pci_io_space, OBJECT(s), "pci_io", 4 * GiB); memory_region_init(&s->pci_mem_space, OBJECT(s), "pci_mem", 4 * GiB); - pci_root_bus_new_inplace(&s->pci_bus, sizeof(s->pci_bus), dev, "pci", - &s->pci_mem_space, &s->pci_io_space, - PCI_DEVFN(11, 0), TYPE_PCI_BUS); + pci_root_bus_init(&s->pci_bus, sizeof(s->pci_bus), dev, "pci", + &s->pci_mem_space, &s->pci_io_space, + PCI_DEVFN(11, 0), TYPE_PCI_BUS); h->bus = &s->pci_bus; object_initialize(&s->pci_dev, sizeof(s->pci_dev), TYPE_VERSATILE_PCI_HOST); diff --git a/hw/pci/pci.c b/hw/pci/pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/pci/pci.c +++ b/hw/pci/pci.c @@ -XXX,XX +XXX,XX @@ bool pci_bus_bypass_iommu(PCIBus *bus) return host_bridge->bypass_iommu; } -static void pci_root_bus_init(PCIBus *bus, DeviceState *parent, - MemoryRegion *address_space_mem, - MemoryRegion *address_space_io, - uint8_t devfn_min) +static void pci_root_bus_internal_init(PCIBus *bus, DeviceState *parent, + MemoryRegion *address_space_mem, + MemoryRegion *address_space_io, + uint8_t devfn_min) { assert(PCI_FUNC(devfn_min) == 0); bus->devfn_min = devfn_min; @@ -XXX,XX +XXX,XX @@ bool pci_bus_is_express(PCIBus *bus) return object_dynamic_cast(OBJECT(bus), TYPE_PCIE_BUS); } -void pci_root_bus_new_inplace(PCIBus *bus, size_t bus_size, DeviceState *parent, - const char *name, - MemoryRegion *address_space_mem, - MemoryRegion *address_space_io, - uint8_t devfn_min, const char *typename) +void pci_root_bus_init(PCIBus *bus, size_t bus_size, DeviceState *parent, + const char *name, + MemoryRegion *address_space_mem, + MemoryRegion *address_space_io, + uint8_t devfn_min, const char *typename) { qbus_create_inplace(bus, bus_size, typename, parent, name); - pci_root_bus_init(bus, parent, address_space_mem, address_space_io, - devfn_min); + pci_root_bus_internal_init(bus, parent, address_space_mem, + address_space_io, devfn_min); } PCIBus *pci_root_bus_new(DeviceState *parent, const char *name, @@ -XXX,XX +XXX,XX @@ PCIBus *pci_root_bus_new(DeviceState *parent, const char *name, PCIBus *bus; bus = PCI_BUS(qbus_create(typename, parent, name)); - pci_root_bus_init(bus, parent, address_space_mem, address_space_io, - devfn_min); + pci_root_bus_internal_init(bus, parent, address_space_mem, + address_space_io, devfn_min); return bus; } -- 2.20.1 Rename qbus_create_inplace() to qbus_init(); this is more in line with our usual naming convention for functions that in-place initialize objects. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Message-id: 20210923121153.23754-5-peter.maydell@linaro.org --- include/hw/qdev-core.h | 4 ++-- hw/audio/intel-hda.c | 2 +- hw/block/fdc.c | 2 +- hw/block/swim.c | 3 +-- hw/char/virtio-serial-bus.c | 4 ++-- hw/core/bus.c | 11 ++++++----- hw/core/sysbus.c | 10 ++++++---- hw/gpio/bcm2835_gpio.c | 3 +-- hw/ide/qdev.c | 2 +- hw/ipack/ipack.c | 2 +- hw/misc/mac_via.c | 4 ++-- hw/misc/macio/cuda.c | 4 ++-- hw/misc/macio/macio.c | 4 ++-- hw/misc/macio/pmu.c | 4 ++-- hw/nubus/nubus-bridge.c | 2 +- hw/nvme/ctrl.c | 4 ++-- hw/nvme/subsys.c | 3 +-- hw/pci/pci.c | 2 +- hw/pci/pci_bridge.c | 4 ++-- hw/s390x/event-facility.c | 4 ++-- hw/s390x/virtio-ccw.c | 3 +-- hw/scsi/scsi-bus.c | 2 +- hw/sd/allwinner-sdhost.c | 4 ++-- hw/sd/bcm2835_sdhost.c | 4 ++-- hw/sd/pl181.c | 3 +-- hw/sd/pxa2xx_mmci.c | 4 ++-- hw/sd/sdhci.c | 3 +-- hw/sd/ssi-sd.c | 3 +-- hw/usb/bus.c | 2 +- hw/usb/dev-smartcard-reader.c | 3 +-- hw/virtio/virtio-mmio.c | 3 +-- hw/virtio/virtio-pci.c | 3 +-- 32 files changed, 54 insertions(+), 61 deletions(-) diff --git a/include/hw/qdev-core.h b/include/hw/qdev-core.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/qdev-core.h +++ b/include/hw/qdev-core.h @@ -XXX,XX +XXX,XX @@ DeviceState *qdev_find_recursive(BusState *bus, const char *id); typedef int (qbus_walkerfn)(BusState *bus, void *opaque); typedef int (qdev_walkerfn)(DeviceState *dev, void *opaque); -void qbus_create_inplace(void *bus, size_t size, const char *typename, - DeviceState *parent, const char *name); +void qbus_init(void *bus, size_t size, const char *typename, + DeviceState *parent, const char *name); BusState *qbus_create(const char *typename, DeviceState *parent, const char *name); bool qbus_realize(BusState *bus, Error **errp); void qbus_unrealize(BusState *bus); diff --git a/hw/audio/intel-hda.c b/hw/audio/intel-hda.c index XXXXXXX..XXXXXXX 100644 --- a/hw/audio/intel-hda.c +++ b/hw/audio/intel-hda.c @@ -XXX,XX +XXX,XX @@ void hda_codec_bus_init(DeviceState *dev, HDACodecBus *bus, size_t bus_size, hda_codec_response_func response, hda_codec_xfer_func xfer) { - qbus_create_inplace(bus, bus_size, TYPE_HDA_BUS, dev, NULL); + qbus_init(bus, bus_size, TYPE_HDA_BUS, dev, NULL); bus->response = response; bus->xfer = xfer; } diff --git a/hw/block/fdc.c b/hw/block/fdc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/block/fdc.c +++ b/hw/block/fdc.c @@ -XXX,XX +XXX,XX @@ static const TypeInfo floppy_bus_info = { static void floppy_bus_create(FDCtrl *fdc, FloppyBus *bus, DeviceState *dev) { - qbus_create_inplace(bus, sizeof(FloppyBus), TYPE_FLOPPY_BUS, dev, NULL); + qbus_init(bus, sizeof(FloppyBus), TYPE_FLOPPY_BUS, dev, NULL); bus->fdc = fdc; } diff --git a/hw/block/swim.c b/hw/block/swim.c index XXXXXXX..XXXXXXX 100644 --- a/hw/block/swim.c +++ b/hw/block/swim.c @@ -XXX,XX +XXX,XX @@ static void sysbus_swim_realize(DeviceState *dev, Error **errp) Swim *sys = SWIM(dev); SWIMCtrl *swimctrl = &sys->ctrl; - qbus_create_inplace(&swimctrl->bus, sizeof(SWIMBus), TYPE_SWIM_BUS, dev, - NULL); + qbus_init(&swimctrl->bus, sizeof(SWIMBus), TYPE_SWIM_BUS, dev, NULL); swimctrl->bus.ctrl = swimctrl; } diff --git a/hw/char/virtio-serial-bus.c b/hw/char/virtio-serial-bus.c index XXXXXXX..XXXXXXX 100644 --- a/hw/char/virtio-serial-bus.c +++ b/hw/char/virtio-serial-bus.c @@ -XXX,XX +XXX,XX @@ static void virtio_serial_device_realize(DeviceState *dev, Error **errp) config_size); /* Spawn a new virtio-serial bus on which the ports will ride as devices */ - qbus_create_inplace(&vser->bus, sizeof(vser->bus), TYPE_VIRTIO_SERIAL_BUS, - dev, vdev->bus_name); + qbus_init(&vser->bus, sizeof(vser->bus), TYPE_VIRTIO_SERIAL_BUS, + dev, vdev->bus_name); qbus_set_hotplug_handler(BUS(&vser->bus), OBJECT(vser)); vser->bus.vser = vser; QTAILQ_INIT(&vser->ports); diff --git a/hw/core/bus.c b/hw/core/bus.c index XXXXXXX..XXXXXXX 100644 --- a/hw/core/bus.c +++ b/hw/core/bus.c @@ -XXX,XX +XXX,XX @@ static void bus_reset_child_foreach(Object *obj, ResettableChildCallback cb, } } -static void qbus_init(BusState *bus, DeviceState *parent, const char *name) +static void qbus_init_internal(BusState *bus, DeviceState *parent, + const char *name) { const char *typename = object_get_typename(OBJECT(bus)); BusClass *bc; @@ -XXX,XX +XXX,XX @@ static void bus_unparent(Object *obj) bus->parent = NULL; } -void qbus_create_inplace(void *bus, size_t size, const char *typename, - DeviceState *parent, const char *name) +void qbus_init(void *bus, size_t size, const char *typename, + DeviceState *parent, const char *name) { object_initialize(bus, size, typename); - qbus_init(bus, parent, name); + qbus_init_internal(bus, parent, name); } BusState *qbus_create(const char *typename, DeviceState *parent, const char *name) @@ -XXX,XX +XXX,XX @@ BusState *qbus_create(const char *typename, DeviceState *parent, const char *nam BusState *bus; bus = BUS(object_new(typename)); - qbus_init(bus, parent, name); + qbus_init_internal(bus, parent, name); return bus; } diff --git a/hw/core/sysbus.c b/hw/core/sysbus.c index XXXXXXX..XXXXXXX 100644 --- a/hw/core/sysbus.c +++ b/hw/core/sysbus.c @@ -XXX,XX +XXX,XX @@ static BusState *main_system_bus; static void main_system_bus_create(void) { - /* assign main_system_bus before qbus_create_inplace() - * in order to make "if (bus != sysbus_get_default())" work */ + /* + * assign main_system_bus before qbus_init() + * in order to make "if (bus != sysbus_get_default())" work + */ main_system_bus = g_malloc0(system_bus_info.instance_size); - qbus_create_inplace(main_system_bus, system_bus_info.instance_size, - TYPE_SYSTEM_BUS, NULL, "main-system-bus"); + qbus_init(main_system_bus, system_bus_info.instance_size, + TYPE_SYSTEM_BUS, NULL, "main-system-bus"); OBJECT(main_system_bus)->free = g_free; } diff --git a/hw/gpio/bcm2835_gpio.c b/hw/gpio/bcm2835_gpio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/gpio/bcm2835_gpio.c +++ b/hw/gpio/bcm2835_gpio.c @@ -XXX,XX +XXX,XX @@ static void bcm2835_gpio_init(Object *obj) DeviceState *dev = DEVICE(obj); SysBusDevice *sbd = SYS_BUS_DEVICE(obj); - qbus_create_inplace(&s->sdbus, sizeof(s->sdbus), - TYPE_SD_BUS, DEVICE(s), "sd-bus"); + qbus_init(&s->sdbus, sizeof(s->sdbus), TYPE_SD_BUS, DEVICE(s), "sd-bus"); memory_region_init_io(&s->iomem, obj, &bcm2835_gpio_ops, s, "bcm2835_gpio", 0x1000); diff --git a/hw/ide/qdev.c b/hw/ide/qdev.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ide/qdev.c +++ b/hw/ide/qdev.c @@ -XXX,XX +XXX,XX @@ static const TypeInfo ide_bus_info = { void ide_bus_new(IDEBus *idebus, size_t idebus_size, DeviceState *dev, int bus_id, int max_units) { - qbus_create_inplace(idebus, idebus_size, TYPE_IDE_BUS, dev, NULL); + qbus_init(idebus, idebus_size, TYPE_IDE_BUS, dev, NULL); idebus->bus_id = bus_id; idebus->max_units = max_units; } diff --git a/hw/ipack/ipack.c b/hw/ipack/ipack.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ipack/ipack.c +++ b/hw/ipack/ipack.c @@ -XXX,XX +XXX,XX @@ void ipack_bus_init(IPackBus *bus, size_t bus_size, uint8_t n_slots, qemu_irq_handler handler) { - qbus_create_inplace(bus, bus_size, TYPE_IPACK_BUS, parent, NULL); + qbus_init(bus, bus_size, TYPE_IPACK_BUS, parent, NULL); bus->n_slots = n_slots; bus->set_irq = handler; } diff --git a/hw/misc/mac_via.c b/hw/misc/mac_via.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/mac_via.c +++ b/hw/misc/mac_via.c @@ -XXX,XX +XXX,XX @@ static void mos6522_q800_via1_init(Object *obj) sysbus_init_mmio(sbd, &v1s->via_mem); /* ADB */ - qbus_create_inplace((BusState *)&v1s->adb_bus, sizeof(v1s->adb_bus), - TYPE_ADB_BUS, DEVICE(v1s), "adb.0"); + qbus_init((BusState *)&v1s->adb_bus, sizeof(v1s->adb_bus), + TYPE_ADB_BUS, DEVICE(v1s), "adb.0"); qdev_init_gpio_in(DEVICE(obj), via1_irq_request, VIA1_IRQ_NB); } diff --git a/hw/misc/macio/cuda.c b/hw/misc/macio/cuda.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/macio/cuda.c +++ b/hw/misc/macio/cuda.c @@ -XXX,XX +XXX,XX @@ static void cuda_init(Object *obj) memory_region_init_io(&s->mem, obj, &mos6522_cuda_ops, s, "cuda", 0x2000); sysbus_init_mmio(sbd, &s->mem); - qbus_create_inplace(&s->adb_bus, sizeof(s->adb_bus), TYPE_ADB_BUS, - DEVICE(obj), "adb.0"); + qbus_init(&s->adb_bus, sizeof(s->adb_bus), TYPE_ADB_BUS, + DEVICE(obj), "adb.0"); } static Property cuda_properties[] = { diff --git a/hw/misc/macio/macio.c b/hw/misc/macio/macio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/macio/macio.c +++ b/hw/misc/macio/macio.c @@ -XXX,XX +XXX,XX @@ static void macio_instance_init(Object *obj) memory_region_init(&s->bar, obj, "macio", 0x80000); - qbus_create_inplace(&s->macio_bus, sizeof(s->macio_bus), TYPE_MACIO_BUS, - DEVICE(obj), "macio.0"); + qbus_init(&s->macio_bus, sizeof(s->macio_bus), TYPE_MACIO_BUS, + DEVICE(obj), "macio.0"); object_initialize_child(OBJECT(s), "dbdma", &s->dbdma, TYPE_MAC_DBDMA); diff --git a/hw/misc/macio/pmu.c b/hw/misc/macio/pmu.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/macio/pmu.c +++ b/hw/misc/macio/pmu.c @@ -XXX,XX +XXX,XX @@ static void pmu_realize(DeviceState *dev, Error **errp) timer_mod(s->one_sec_timer, s->one_sec_target); if (s->has_adb) { - qbus_create_inplace(&s->adb_bus, sizeof(s->adb_bus), TYPE_ADB_BUS, - dev, "adb.0"); + qbus_init(&s->adb_bus, sizeof(s->adb_bus), TYPE_ADB_BUS, + dev, "adb.0"); adb_register_autopoll_callback(adb_bus, pmu_adb_poll, s); } } diff --git a/hw/nubus/nubus-bridge.c b/hw/nubus/nubus-bridge.c index XXXXXXX..XXXXXXX 100644 --- a/hw/nubus/nubus-bridge.c +++ b/hw/nubus/nubus-bridge.c @@ -XXX,XX +XXX,XX @@ static void nubus_bridge_init(Object *obj) NubusBridge *s = NUBUS_BRIDGE(obj); NubusBus *bus = &s->bus; - qbus_create_inplace(bus, sizeof(s->bus), TYPE_NUBUS_BUS, DEVICE(s), NULL); + qbus_init(bus, sizeof(s->bus), TYPE_NUBUS_BUS, DEVICE(s), NULL); qdev_init_gpio_out(DEVICE(s), bus->irqs, NUBUS_IRQS); } diff --git a/hw/nvme/ctrl.c b/hw/nvme/ctrl.c index XXXXXXX..XXXXXXX 100644 --- a/hw/nvme/ctrl.c +++ b/hw/nvme/ctrl.c @@ -XXX,XX +XXX,XX @@ static void nvme_realize(PCIDevice *pci_dev, Error **errp) return; } - qbus_create_inplace(&n->bus, sizeof(NvmeBus), TYPE_NVME_BUS, - &pci_dev->qdev, n->parent_obj.qdev.id); + qbus_init(&n->bus, sizeof(NvmeBus), TYPE_NVME_BUS, + &pci_dev->qdev, n->parent_obj.qdev.id); nvme_init_state(n); if (nvme_init_pci(n, pci_dev, errp)) { diff --git a/hw/nvme/subsys.c b/hw/nvme/subsys.c index XXXXXXX..XXXXXXX 100644 --- a/hw/nvme/subsys.c +++ b/hw/nvme/subsys.c @@ -XXX,XX +XXX,XX @@ static void nvme_subsys_realize(DeviceState *dev, Error **errp) { NvmeSubsystem *subsys = NVME_SUBSYS(dev); - qbus_create_inplace(&subsys->bus, sizeof(NvmeBus), TYPE_NVME_BUS, dev, - dev->id); + qbus_init(&subsys->bus, sizeof(NvmeBus), TYPE_NVME_BUS, dev, dev->id); nvme_subsys_setup(subsys); } diff --git a/hw/pci/pci.c b/hw/pci/pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/pci/pci.c +++ b/hw/pci/pci.c @@ -XXX,XX +XXX,XX @@ void pci_root_bus_init(PCIBus *bus, size_t bus_size, DeviceState *parent, MemoryRegion *address_space_io, uint8_t devfn_min, const char *typename) { - qbus_create_inplace(bus, bus_size, typename, parent, name); + qbus_init(bus, bus_size, typename, parent, name); pci_root_bus_internal_init(bus, parent, address_space_mem, address_space_io, devfn_min); } diff --git a/hw/pci/pci_bridge.c b/hw/pci/pci_bridge.c index XXXXXXX..XXXXXXX 100644 --- a/hw/pci/pci_bridge.c +++ b/hw/pci/pci_bridge.c @@ -XXX,XX +XXX,XX @@ void pci_bridge_initfn(PCIDevice *dev, const char *typename) br->bus_name = dev->qdev.id; } - qbus_create_inplace(sec_bus, sizeof(br->sec_bus), typename, DEVICE(dev), - br->bus_name); + qbus_init(sec_bus, sizeof(br->sec_bus), typename, DEVICE(dev), + br->bus_name); sec_bus->parent_dev = dev; sec_bus->map_irq = br->map_irq ? br->map_irq : pci_swizzle_map_irq_fn; sec_bus->address_space_mem = &br->address_space_mem; diff --git a/hw/s390x/event-facility.c b/hw/s390x/event-facility.c index XXXXXXX..XXXXXXX 100644 --- a/hw/s390x/event-facility.c +++ b/hw/s390x/event-facility.c @@ -XXX,XX +XXX,XX @@ static void init_event_facility(Object *obj) sclp_event_set_allow_all_mask_sizes); /* Spawn a new bus for SCLP events */ - qbus_create_inplace(&event_facility->sbus, sizeof(event_facility->sbus), - TYPE_SCLP_EVENTS_BUS, sdev, NULL); + qbus_init(&event_facility->sbus, sizeof(event_facility->sbus), + TYPE_SCLP_EVENTS_BUS, sdev, NULL); object_initialize_child(obj, TYPE_SCLP_QUIESCE, &event_facility->quiesce, diff --git a/hw/s390x/virtio-ccw.c b/hw/s390x/virtio-ccw.c index XXXXXXX..XXXXXXX 100644 --- a/hw/s390x/virtio-ccw.c +++ b/hw/s390x/virtio-ccw.c @@ -XXX,XX +XXX,XX @@ static void virtio_ccw_bus_new(VirtioBusState *bus, size_t bus_size, DeviceState *qdev = DEVICE(dev); char virtio_bus_name[] = "virtio-bus"; - qbus_create_inplace(bus, bus_size, TYPE_VIRTIO_CCW_BUS, - qdev, virtio_bus_name); + qbus_init(bus, bus_size, TYPE_VIRTIO_CCW_BUS, qdev, virtio_bus_name); } static void virtio_ccw_bus_class_init(ObjectClass *klass, void *data) diff --git a/hw/scsi/scsi-bus.c b/hw/scsi/scsi-bus.c index XXXXXXX..XXXXXXX 100644 --- a/hw/scsi/scsi-bus.c +++ b/hw/scsi/scsi-bus.c @@ -XXX,XX +XXX,XX @@ void scsi_device_unit_attention_reported(SCSIDevice *s) void scsi_bus_init_named(SCSIBus *bus, size_t bus_size, DeviceState *host, const SCSIBusInfo *info, const char *bus_name) { - qbus_create_inplace(bus, bus_size, TYPE_SCSI_BUS, host, bus_name); + qbus_init(bus, bus_size, TYPE_SCSI_BUS, host, bus_name); bus->busnr = next_scsi_bus++; bus->info = info; qbus_set_bus_hotplug_handler(BUS(bus)); diff --git a/hw/sd/allwinner-sdhost.c b/hw/sd/allwinner-sdhost.c index XXXXXXX..XXXXXXX 100644 --- a/hw/sd/allwinner-sdhost.c +++ b/hw/sd/allwinner-sdhost.c @@ -XXX,XX +XXX,XX @@ static void allwinner_sdhost_init(Object *obj) { AwSdHostState *s = AW_SDHOST(obj); - qbus_create_inplace(&s->sdbus, sizeof(s->sdbus), - TYPE_AW_SDHOST_BUS, DEVICE(s), "sd-bus"); + qbus_init(&s->sdbus, sizeof(s->sdbus), + TYPE_AW_SDHOST_BUS, DEVICE(s), "sd-bus"); memory_region_init_io(&s->iomem, obj, &allwinner_sdhost_ops, s, TYPE_AW_SDHOST, 4 * KiB); diff --git a/hw/sd/bcm2835_sdhost.c b/hw/sd/bcm2835_sdhost.c index XXXXXXX..XXXXXXX 100644 --- a/hw/sd/bcm2835_sdhost.c +++ b/hw/sd/bcm2835_sdhost.c @@ -XXX,XX +XXX,XX @@ static void bcm2835_sdhost_init(Object *obj) { BCM2835SDHostState *s = BCM2835_SDHOST(obj); - qbus_create_inplace(&s->sdbus, sizeof(s->sdbus), - TYPE_BCM2835_SDHOST_BUS, DEVICE(s), "sd-bus"); + qbus_init(&s->sdbus, sizeof(s->sdbus), + TYPE_BCM2835_SDHOST_BUS, DEVICE(s), "sd-bus"); memory_region_init_io(&s->iomem, obj, &bcm2835_sdhost_ops, s, TYPE_BCM2835_SDHOST, 0x1000); diff --git a/hw/sd/pl181.c b/hw/sd/pl181.c index XXXXXXX..XXXXXXX 100644 --- a/hw/sd/pl181.c +++ b/hw/sd/pl181.c @@ -XXX,XX +XXX,XX @@ static void pl181_init(Object *obj) qdev_init_gpio_out_named(dev, &s->card_readonly, "card-read-only", 1); qdev_init_gpio_out_named(dev, &s->card_inserted, "card-inserted", 1); - qbus_create_inplace(&s->sdbus, sizeof(s->sdbus), - TYPE_PL181_BUS, dev, "sd-bus"); + qbus_init(&s->sdbus, sizeof(s->sdbus), TYPE_PL181_BUS, dev, "sd-bus"); } static void pl181_class_init(ObjectClass *klass, void *data) diff --git a/hw/sd/pxa2xx_mmci.c b/hw/sd/pxa2xx_mmci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/sd/pxa2xx_mmci.c +++ b/hw/sd/pxa2xx_mmci.c @@ -XXX,XX +XXX,XX @@ static void pxa2xx_mmci_instance_init(Object *obj) qdev_init_gpio_out_named(dev, &s->rx_dma, "rx-dma", 1); qdev_init_gpio_out_named(dev, &s->tx_dma, "tx-dma", 1); - qbus_create_inplace(&s->sdbus, sizeof(s->sdbus), - TYPE_PXA2XX_MMCI_BUS, DEVICE(obj), "sd-bus"); + qbus_init(&s->sdbus, sizeof(s->sdbus), + TYPE_PXA2XX_MMCI_BUS, DEVICE(obj), "sd-bus"); } static void pxa2xx_mmci_class_init(ObjectClass *klass, void *data) diff --git a/hw/sd/sdhci.c b/hw/sd/sdhci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/sd/sdhci.c +++ b/hw/sd/sdhci.c @@ -XXX,XX +XXX,XX @@ static void sdhci_init_readonly_registers(SDHCIState *s, Error **errp) void sdhci_initfn(SDHCIState *s) { - qbus_create_inplace(&s->sdbus, sizeof(s->sdbus), - TYPE_SDHCI_BUS, DEVICE(s), "sd-bus"); + qbus_init(&s->sdbus, sizeof(s->sdbus), TYPE_SDHCI_BUS, DEVICE(s), "sd-bus"); s->insert_timer = timer_new_ns(QEMU_CLOCK_VIRTUAL, sdhci_raise_insertion_irq, s); s->transfer_timer = timer_new_ns(QEMU_CLOCK_VIRTUAL, sdhci_data_transfer, s); diff --git a/hw/sd/ssi-sd.c b/hw/sd/ssi-sd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/sd/ssi-sd.c +++ b/hw/sd/ssi-sd.c @@ -XXX,XX +XXX,XX @@ static void ssi_sd_realize(SSIPeripheral *d, Error **errp) DeviceState *carddev; DriveInfo *dinfo; - qbus_create_inplace(&s->sdbus, sizeof(s->sdbus), TYPE_SD_BUS, - DEVICE(d), "sd-bus"); + qbus_init(&s->sdbus, sizeof(s->sdbus), TYPE_SD_BUS, DEVICE(d), "sd-bus"); /* Create and plug in the sd card */ /* FIXME use a qdev drive property instead of drive_get_next() */ diff --git a/hw/usb/bus.c b/hw/usb/bus.c index XXXXXXX..XXXXXXX 100644 --- a/hw/usb/bus.c +++ b/hw/usb/bus.c @@ -XXX,XX +XXX,XX @@ const VMStateDescription vmstate_usb_device = { void usb_bus_new(USBBus *bus, size_t bus_size, USBBusOps *ops, DeviceState *host) { - qbus_create_inplace(bus, bus_size, TYPE_USB_BUS, host, NULL); + qbus_init(bus, bus_size, TYPE_USB_BUS, host, NULL); qbus_set_bus_hotplug_handler(BUS(bus)); bus->ops = ops; bus->busnr = next_usb_bus++; diff --git a/hw/usb/dev-smartcard-reader.c b/hw/usb/dev-smartcard-reader.c index XXXXXXX..XXXXXXX 100644 --- a/hw/usb/dev-smartcard-reader.c +++ b/hw/usb/dev-smartcard-reader.c @@ -XXX,XX +XXX,XX @@ static void ccid_realize(USBDevice *dev, Error **errp) usb_desc_create_serial(dev); usb_desc_init(dev); - qbus_create_inplace(&s->bus, sizeof(s->bus), TYPE_CCID_BUS, DEVICE(dev), - NULL); + qbus_init(&s->bus, sizeof(s->bus), TYPE_CCID_BUS, DEVICE(dev), NULL); qbus_set_hotplug_handler(BUS(&s->bus), OBJECT(dev)); s->intr = usb_ep_get(dev, USB_TOKEN_IN, CCID_INT_IN_EP); s->bulk = usb_ep_get(dev, USB_TOKEN_IN, CCID_BULK_IN_EP); diff --git a/hw/virtio/virtio-mmio.c b/hw/virtio/virtio-mmio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/virtio/virtio-mmio.c +++ b/hw/virtio/virtio-mmio.c @@ -XXX,XX +XXX,XX @@ static void virtio_mmio_realizefn(DeviceState *d, Error **errp) VirtIOMMIOProxy *proxy = VIRTIO_MMIO(d); SysBusDevice *sbd = SYS_BUS_DEVICE(d); - qbus_create_inplace(&proxy->bus, sizeof(proxy->bus), TYPE_VIRTIO_MMIO_BUS, - d, NULL); + qbus_init(&proxy->bus, sizeof(proxy->bus), TYPE_VIRTIO_MMIO_BUS, d, NULL); sysbus_init_irq(sbd, &proxy->irq); if (!kvm_eventfds_enabled()) { diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/virtio/virtio-pci.c +++ b/hw/virtio/virtio-pci.c @@ -XXX,XX +XXX,XX @@ static void virtio_pci_bus_new(VirtioBusState *bus, size_t bus_size, DeviceState *qdev = DEVICE(dev); char virtio_bus_name[] = "virtio-bus"; - qbus_create_inplace(bus, bus_size, TYPE_VIRTIO_PCI_BUS, qdev, - virtio_bus_name); + qbus_init(bus, bus_size, TYPE_VIRTIO_PCI_BUS, qdev, virtio_bus_name); } static void virtio_pci_bus_class_init(ObjectClass *klass, void *data) -- 2.20.1 Rename the "allocate and return" qbus creation function to qbus_new(), to bring it into line with our _init vs _new convention. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Corey Minyard <cminyard@mvista.com> Message-id: 20210923121153.23754-6-peter.maydell@linaro.org --- include/hw/qdev-core.h | 2 +- hw/core/bus.c | 2 +- hw/hyperv/vmbus.c | 2 +- hw/i2c/core.c | 2 +- hw/isa/isa-bus.c | 2 +- hw/misc/auxbus.c | 2 +- hw/pci/pci.c | 2 +- hw/ppc/spapr_vio.c | 2 +- hw/s390x/ap-bridge.c | 2 +- hw/s390x/css-bridge.c | 2 +- hw/s390x/s390-pci-bus.c | 2 +- hw/ssi/ssi.c | 2 +- hw/xen/xen-bus.c | 2 +- hw/xen/xen-legacy-backend.c | 2 +- 14 files changed, 14 insertions(+), 14 deletions(-) diff --git a/include/hw/qdev-core.h b/include/hw/qdev-core.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/qdev-core.h +++ b/include/hw/qdev-core.h @@ -XXX,XX +XXX,XX @@ typedef int (qdev_walkerfn)(DeviceState *dev, void *opaque); void qbus_init(void *bus, size_t size, const char *typename, DeviceState *parent, const char *name); -BusState *qbus_create(const char *typename, DeviceState *parent, const char *name); +BusState *qbus_new(const char *typename, DeviceState *parent, const char *name); bool qbus_realize(BusState *bus, Error **errp); void qbus_unrealize(BusState *bus); diff --git a/hw/core/bus.c b/hw/core/bus.c index XXXXXXX..XXXXXXX 100644 --- a/hw/core/bus.c +++ b/hw/core/bus.c @@ -XXX,XX +XXX,XX @@ void qbus_init(void *bus, size_t size, const char *typename, qbus_init_internal(bus, parent, name); } -BusState *qbus_create(const char *typename, DeviceState *parent, const char *name) +BusState *qbus_new(const char *typename, DeviceState *parent, const char *name) { BusState *bus; diff --git a/hw/hyperv/vmbus.c b/hw/hyperv/vmbus.c index XXXXXXX..XXXXXXX 100644 --- a/hw/hyperv/vmbus.c +++ b/hw/hyperv/vmbus.c @@ -XXX,XX +XXX,XX @@ static void vmbus_bridge_realize(DeviceState *dev, Error **errp) return; } - bridge->bus = VMBUS(qbus_create(TYPE_VMBUS, dev, "vmbus")); + bridge->bus = VMBUS(qbus_new(TYPE_VMBUS, dev, "vmbus")); } static char *vmbus_bridge_ofw_unit_address(const SysBusDevice *dev) diff --git a/hw/i2c/core.c b/hw/i2c/core.c index XXXXXXX..XXXXXXX 100644 --- a/hw/i2c/core.c +++ b/hw/i2c/core.c @@ -XXX,XX +XXX,XX @@ I2CBus *i2c_init_bus(DeviceState *parent, const char *name) { I2CBus *bus; - bus = I2C_BUS(qbus_create(TYPE_I2C_BUS, parent, name)); + bus = I2C_BUS(qbus_new(TYPE_I2C_BUS, parent, name)); QLIST_INIT(&bus->current_devs); vmstate_register(NULL, VMSTATE_INSTANCE_ID_ANY, &vmstate_i2c_bus, bus); return bus; diff --git a/hw/isa/isa-bus.c b/hw/isa/isa-bus.c index XXXXXXX..XXXXXXX 100644 --- a/hw/isa/isa-bus.c +++ b/hw/isa/isa-bus.c @@ -XXX,XX +XXX,XX @@ ISABus *isa_bus_new(DeviceState *dev, MemoryRegion* address_space, sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal); } - isabus = ISA_BUS(qbus_create(TYPE_ISA_BUS, dev, NULL)); + isabus = ISA_BUS(qbus_new(TYPE_ISA_BUS, dev, NULL)); isabus->address_space = address_space; isabus->address_space_io = address_space_io; return isabus; diff --git a/hw/misc/auxbus.c b/hw/misc/auxbus.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/auxbus.c +++ b/hw/misc/auxbus.c @@ -XXX,XX +XXX,XX @@ AUXBus *aux_bus_init(DeviceState *parent, const char *name) AUXBus *bus; Object *auxtoi2c; - bus = AUX_BUS(qbus_create(TYPE_AUX_BUS, parent, name)); + bus = AUX_BUS(qbus_new(TYPE_AUX_BUS, parent, name)); auxtoi2c = object_new_with_props(TYPE_AUXTOI2C, OBJECT(bus), "i2c", &error_abort, NULL); diff --git a/hw/pci/pci.c b/hw/pci/pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/pci/pci.c +++ b/hw/pci/pci.c @@ -XXX,XX +XXX,XX @@ PCIBus *pci_root_bus_new(DeviceState *parent, const char *name, { PCIBus *bus; - bus = PCI_BUS(qbus_create(typename, parent, name)); + bus = PCI_BUS(qbus_new(typename, parent, name)); pci_root_bus_internal_init(bus, parent, address_space_mem, address_space_io, devfn_min); return bus; diff --git a/hw/ppc/spapr_vio.c b/hw/ppc/spapr_vio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ppc/spapr_vio.c +++ b/hw/ppc/spapr_vio.c @@ -XXX,XX +XXX,XX @@ SpaprVioBus *spapr_vio_bus_init(void) sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal); /* Create bus on bridge device */ - qbus = qbus_create(TYPE_SPAPR_VIO_BUS, dev, "spapr-vio"); + qbus = qbus_new(TYPE_SPAPR_VIO_BUS, dev, "spapr-vio"); bus = SPAPR_VIO_BUS(qbus); bus->next_reg = SPAPR_VIO_REG_BASE; diff --git a/hw/s390x/ap-bridge.c b/hw/s390x/ap-bridge.c index XXXXXXX..XXXXXXX 100644 --- a/hw/s390x/ap-bridge.c +++ b/hw/s390x/ap-bridge.c @@ -XXX,XX +XXX,XX @@ void s390_init_ap(void) sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal); /* Create bus on bridge device */ - bus = qbus_create(TYPE_AP_BUS, dev, TYPE_AP_BUS); + bus = qbus_new(TYPE_AP_BUS, dev, TYPE_AP_BUS); /* Enable hotplugging */ qbus_set_hotplug_handler(bus, OBJECT(dev)); diff --git a/hw/s390x/css-bridge.c b/hw/s390x/css-bridge.c index XXXXXXX..XXXXXXX 100644 --- a/hw/s390x/css-bridge.c +++ b/hw/s390x/css-bridge.c @@ -XXX,XX +XXX,XX @@ VirtualCssBus *virtual_css_bus_init(void) sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal); /* Create bus on bridge device */ - bus = qbus_create(TYPE_VIRTUAL_CSS_BUS, dev, "virtual-css"); + bus = qbus_new(TYPE_VIRTUAL_CSS_BUS, dev, "virtual-css"); cbus = VIRTUAL_CSS_BUS(bus); /* Enable hotplugging */ diff --git a/hw/s390x/s390-pci-bus.c b/hw/s390x/s390-pci-bus.c index XXXXXXX..XXXXXXX 100644 --- a/hw/s390x/s390-pci-bus.c +++ b/hw/s390x/s390-pci-bus.c @@ -XXX,XX +XXX,XX @@ static void s390_pcihost_realize(DeviceState *dev, Error **errp) qbus_set_hotplug_handler(bus, OBJECT(dev)); phb->bus = b; - s->bus = S390_PCI_BUS(qbus_create(TYPE_S390_PCI_BUS, dev, NULL)); + s->bus = S390_PCI_BUS(qbus_new(TYPE_S390_PCI_BUS, dev, NULL)); qbus_set_hotplug_handler(BUS(s->bus), OBJECT(dev)); s->iommu_table = g_hash_table_new_full(g_int64_hash, g_int64_equal, diff --git a/hw/ssi/ssi.c b/hw/ssi/ssi.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/ssi.c +++ b/hw/ssi/ssi.c @@ -XXX,XX +XXX,XX @@ DeviceState *ssi_create_peripheral(SSIBus *bus, const char *name) SSIBus *ssi_create_bus(DeviceState *parent, const char *name) { BusState *bus; - bus = qbus_create(TYPE_SSI_BUS, parent, name); + bus = qbus_new(TYPE_SSI_BUS, parent, name); return SSI_BUS(bus); } diff --git a/hw/xen/xen-bus.c b/hw/xen/xen-bus.c index XXXXXXX..XXXXXXX 100644 --- a/hw/xen/xen-bus.c +++ b/hw/xen/xen-bus.c @@ -XXX,XX +XXX,XX @@ type_init(xen_register_types) void xen_bus_init(void) { DeviceState *dev = qdev_new(TYPE_XEN_BRIDGE); - BusState *bus = qbus_create(TYPE_XEN_BUS, dev, NULL); + BusState *bus = qbus_new(TYPE_XEN_BUS, dev, NULL); sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal); qbus_set_bus_hotplug_handler(bus); diff --git a/hw/xen/xen-legacy-backend.c b/hw/xen/xen-legacy-backend.c index XXXXXXX..XXXXXXX 100644 --- a/hw/xen/xen-legacy-backend.c +++ b/hw/xen/xen-legacy-backend.c @@ -XXX,XX +XXX,XX @@ int xen_be_init(void) xen_sysdev = qdev_new(TYPE_XENSYSDEV); sysbus_realize_and_unref(SYS_BUS_DEVICE(xen_sysdev), &error_fatal); - xen_sysbus = qbus_create(TYPE_XENSYSBUS, xen_sysdev, "xen-sysbus"); + xen_sysbus = qbus_new(TYPE_XENSYSBUS, xen_sysdev, "xen-sysbus"); qbus_set_bus_hotplug_handler(xen_sysbus); return 0; -- 2.20.1 The function ide_bus_new() does an in-place initialization. Rename it to ide_bus_init() to follow our _init vs _new convention. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Corey Minyard <cminyard@mvista.com> Reviewed-by: John Snow <jsnow@redhat.com> Acked-by: John Snow <jsnow@redhat.com> (Feel free to merge.) Message-id: 20210923121153.23754-7-peter.maydell@linaro.org --- include/hw/ide/internal.h | 4 ++-- hw/ide/ahci.c | 2 +- hw/ide/cmd646.c | 2 +- hw/ide/isa.c | 2 +- hw/ide/macio.c | 2 +- hw/ide/microdrive.c | 2 +- hw/ide/mmio.c | 2 +- hw/ide/piix.c | 2 +- hw/ide/qdev.c | 2 +- hw/ide/sii3112.c | 2 +- hw/ide/via.c | 2 +- 11 files changed, 12 insertions(+), 12 deletions(-) diff --git a/include/hw/ide/internal.h b/include/hw/ide/internal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/ide/internal.h +++ b/include/hw/ide/internal.h @@ -XXX,XX +XXX,XX @@ void ide_atapi_cmd(IDEState *s); void ide_atapi_cmd_reply_end(IDEState *s); /* hw/ide/qdev.c */ -void ide_bus_new(IDEBus *idebus, size_t idebus_size, DeviceState *dev, - int bus_id, int max_units); +void ide_bus_init(IDEBus *idebus, size_t idebus_size, DeviceState *dev, + int bus_id, int max_units); IDEDevice *ide_create_drive(IDEBus *bus, int unit, DriveInfo *drive); int ide_handle_rw_error(IDEState *s, int error, int op); diff --git a/hw/ide/ahci.c b/hw/ide/ahci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ide/ahci.c +++ b/hw/ide/ahci.c @@ -XXX,XX +XXX,XX @@ void ahci_realize(AHCIState *s, DeviceState *qdev, AddressSpace *as, int ports) for (i = 0; i < s->ports; i++) { AHCIDevice *ad = &s->dev[i]; - ide_bus_new(&ad->port, sizeof(ad->port), qdev, i, 1); + ide_bus_init(&ad->port, sizeof(ad->port), qdev, i, 1); ide_init2(&ad->port, irqs[i]); ad->hba = s; diff --git a/hw/ide/cmd646.c b/hw/ide/cmd646.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ide/cmd646.c +++ b/hw/ide/cmd646.c @@ -XXX,XX +XXX,XX @@ static void pci_cmd646_ide_realize(PCIDevice *dev, Error **errp) qdev_init_gpio_in(ds, cmd646_set_irq, 2); for (i = 0; i < 2; i++) { - ide_bus_new(&d->bus[i], sizeof(d->bus[i]), ds, i, 2); + ide_bus_init(&d->bus[i], sizeof(d->bus[i]), ds, i, 2); ide_init2(&d->bus[i], qdev_get_gpio_in(ds, i)); bmdma_init(&d->bus[i], &d->bmdma[i], d); diff --git a/hw/ide/isa.c b/hw/ide/isa.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ide/isa.c +++ b/hw/ide/isa.c @@ -XXX,XX +XXX,XX @@ static void isa_ide_realizefn(DeviceState *dev, Error **errp) ISADevice *isadev = ISA_DEVICE(dev); ISAIDEState *s = ISA_IDE(dev); - ide_bus_new(&s->bus, sizeof(s->bus), dev, 0, 2); + ide_bus_init(&s->bus, sizeof(s->bus), dev, 0, 2); ide_init_ioport(&s->bus, isadev, s->iobase, s->iobase2); isa_init_irq(isadev, &s->irq, s->isairq); ide_init2(&s->bus, s->irq); diff --git a/hw/ide/macio.c b/hw/ide/macio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ide/macio.c +++ b/hw/ide/macio.c @@ -XXX,XX +XXX,XX @@ static void macio_ide_initfn(Object *obj) SysBusDevice *d = SYS_BUS_DEVICE(obj); MACIOIDEState *s = MACIO_IDE(obj); - ide_bus_new(&s->bus, sizeof(s->bus), DEVICE(obj), 0, 2); + ide_bus_init(&s->bus, sizeof(s->bus), DEVICE(obj), 0, 2); memory_region_init_io(&s->mem, obj, &pmac_ide_ops, s, "pmac-ide", 0x1000); sysbus_init_mmio(d, &s->mem); sysbus_init_irq(d, &s->real_ide_irq); diff --git a/hw/ide/microdrive.c b/hw/ide/microdrive.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ide/microdrive.c +++ b/hw/ide/microdrive.c @@ -XXX,XX +XXX,XX @@ static void microdrive_init(Object *obj) { MicroDriveState *md = MICRODRIVE(obj); - ide_bus_new(&md->bus, sizeof(md->bus), DEVICE(obj), 0, 1); + ide_bus_init(&md->bus, sizeof(md->bus), DEVICE(obj), 0, 1); } static void microdrive_class_init(ObjectClass *oc, void *data) diff --git a/hw/ide/mmio.c b/hw/ide/mmio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ide/mmio.c +++ b/hw/ide/mmio.c @@ -XXX,XX +XXX,XX @@ static void mmio_ide_initfn(Object *obj) SysBusDevice *d = SYS_BUS_DEVICE(obj); MMIOState *s = MMIO_IDE(obj); - ide_bus_new(&s->bus, sizeof(s->bus), DEVICE(obj), 0, 2); + ide_bus_init(&s->bus, sizeof(s->bus), DEVICE(obj), 0, 2); sysbus_init_irq(d, &s->irq); } diff --git a/hw/ide/piix.c b/hw/ide/piix.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ide/piix.c +++ b/hw/ide/piix.c @@ -XXX,XX +XXX,XX @@ static int pci_piix_init_ports(PCIIDEState *d) int i, ret; for (i = 0; i < 2; i++) { - ide_bus_new(&d->bus[i], sizeof(d->bus[i]), DEVICE(d), i, 2); + ide_bus_init(&d->bus[i], sizeof(d->bus[i]), DEVICE(d), i, 2); ret = ide_init_ioport(&d->bus[i], NULL, port_info[i].iobase, port_info[i].iobase2); if (ret) { diff --git a/hw/ide/qdev.c b/hw/ide/qdev.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ide/qdev.c +++ b/hw/ide/qdev.c @@ -XXX,XX +XXX,XX @@ static const TypeInfo ide_bus_info = { .class_init = ide_bus_class_init, }; -void ide_bus_new(IDEBus *idebus, size_t idebus_size, DeviceState *dev, +void ide_bus_init(IDEBus *idebus, size_t idebus_size, DeviceState *dev, int bus_id, int max_units) { qbus_init(idebus, idebus_size, TYPE_IDE_BUS, dev, NULL); diff --git a/hw/ide/sii3112.c b/hw/ide/sii3112.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ide/sii3112.c +++ b/hw/ide/sii3112.c @@ -XXX,XX +XXX,XX @@ static void sii3112_pci_realize(PCIDevice *dev, Error **errp) qdev_init_gpio_in(ds, sii3112_set_irq, 2); for (i = 0; i < 2; i++) { - ide_bus_new(&s->bus[i], sizeof(s->bus[i]), ds, i, 1); + ide_bus_init(&s->bus[i], sizeof(s->bus[i]), ds, i, 1); ide_init2(&s->bus[i], qdev_get_gpio_in(ds, i)); bmdma_init(&s->bus[i], &s->bmdma[i], s); diff --git a/hw/ide/via.c b/hw/ide/via.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ide/via.c +++ b/hw/ide/via.c @@ -XXX,XX +XXX,XX @@ static void via_ide_realize(PCIDevice *dev, Error **errp) qdev_init_gpio_in(ds, via_ide_set_irq, 2); for (i = 0; i < 2; i++) { - ide_bus_new(&d->bus[i], sizeof(d->bus[i]), ds, i, 2); + ide_bus_init(&d->bus[i], sizeof(d->bus[i]), ds, i, 2); ide_init2(&d->bus[i], qdev_get_gpio_in(ds, i)); bmdma_init(&d->bus[i], &d->bmdma[i], d); -- 2.20.1 From: Xuzhou Cheng <xuzhou.cheng@windriver.com> The Linux spi-imx driver does not work on QEMU. The reason is that the state of m25p80 loops in STATE_READING_DATA state after receiving RDSR command, the new command is ignored. Before sending a new command, CS line should be pulled high to make the state of m25p80 back to IDLE. Currently the SPI flash CS line is connected to the SPI controller, but on the real board, it's connected to GPIO3_19. This matches the ecspi1 device node in the board dts. ecspi1 node in imx6qdl-sabrelite.dtsi: &ecspi1 { cs-gpios = <&gpio3 19 GPIO_ACTIVE_LOW>; pinctrl-names = "default"; pinctrl-0 = <&pinctrl_ecspi1>; status = "okay"; flash: m25p80@0 { compatible = "sst,sst25vf016b", "jedec,spi-nor"; spi-max-frequency = <20000000>; reg = <0>; }; }; Should connect the SSI_GPIO_CS to GPIO3_19 when adding a spi-nor to spi1 on sabrelite machine. Verified this patch on Linux v5.14. Logs: # echo "01234567899876543210" > test # mtd_debug erase /dev/mtd0 0x0 0x1000 Erased 4096 bytes from address 0x00000000 in flash # mtd_debug write /dev/mtdblock0 0x0 20 test Copied 20 bytes from test to address 0x00000000 in flash # mtd_debug read /dev/mtdblock0 0x0 20 test_out Copied 20 bytes from address 0x00000000 in flash to test_out # cat test_out 01234567899876543210# Signed-off-by: Xuzhou Cheng <xuzhou.cheng@windriver.com> Reported-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Bin Meng <bin.meng@windriver.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210927142825.491-1-xchengl.cn@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/sabrelite.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/arm/sabrelite.c b/hw/arm/sabrelite.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/sabrelite.c +++ b/hw/arm/sabrelite.c @@ -XXX,XX +XXX,XX @@ static void sabrelite_init(MachineState *machine) qdev_realize_and_unref(flash_dev, BUS(spi_bus), &error_fatal); cs_line = qdev_get_gpio_in_named(flash_dev, SSI_GPIO_CS, 0); - sysbus_connect_irq(SYS_BUS_DEVICE(spi_dev), 1, cs_line); + qdev_connect_gpio_out(DEVICE(&s->gpio[2]), 19, cs_line); } } } -- 2.20.1 I don't have anything else queued up at the moment, so this is just Richard's SME patches. -- PMM The following changes since commit 63b38f6c85acd312c2cab68554abf33adf4ee2b3: Merge tag 'pull-target-arm-20220707' of https://git.linaro.org/people/pmaydell/qemu-arm into staging (2022-07-08 06:17:11 +0530) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20220711 for you to fetch changes up to f9982ceaf26df27d15547a3a7990a95019e9e3a8: linux-user/aarch64: Add SME related hwcap entries (2022-07-11 13:43:52 +0100) ---------------------------------------------------------------- target-arm: * Implement SME emulation, for both system and linux-user ---------------------------------------------------------------- Richard Henderson (45): target/arm: Handle SME in aarch64_cpu_dump_state target/arm: Add infrastructure for disas_sme target/arm: Trap non-streaming usage when Streaming SVE is active target/arm: Mark ADR as non-streaming target/arm: Mark RDFFR, WRFFR, SETFFR as non-streaming target/arm: Mark BDEP, BEXT, BGRP, COMPACT, FEXPA, FTSSEL as non-streaming target/arm: Mark PMULL, FMMLA as non-streaming target/arm: Mark FTSMUL, FTMAD, FADDA as non-streaming target/arm: Mark SMMLA, UMMLA, USMMLA as non-streaming target/arm: Mark string/histo/crypto as non-streaming target/arm: Mark gather/scatter load/store as non-streaming target/arm: Mark gather prefetch as non-streaming target/arm: Mark LDFF1 and LDNF1 as non-streaming target/arm: Mark LD1RO as non-streaming target/arm: Add SME enablement checks target/arm: Handle SME in sve_access_check target/arm: Implement SME RDSVL, ADDSVL, ADDSPL target/arm: Implement SME ZERO target/arm: Implement SME MOVA target/arm: Implement SME LD1, ST1 target/arm: Export unpredicated ld/st from translate-sve.c target/arm: Implement SME LDR, STR target/arm: Implement SME ADDHA, ADDVA target/arm: Implement FMOPA, FMOPS (non-widening) target/arm: Implement BFMOPA, BFMOPS target/arm: Implement FMOPA, FMOPS (widening) target/arm: Implement SME integer outer product target/arm: Implement PSEL target/arm: Implement REVD target/arm: Implement SCLAMP, UCLAMP target/arm: Reset streaming sve state on exception boundaries target/arm: Enable SME for -cpu max linux-user/aarch64: Clear tpidr2_el0 if CLONE_SETTLS linux-user/aarch64: Reset PSTATE.SM on syscalls linux-user/aarch64: Add SM bit to SVE signal context linux-user/aarch64: Tidy target_restore_sigframe error return linux-user/aarch64: Do not allow duplicate or short sve records linux-user/aarch64: Verify extra record lock succeeded linux-user/aarch64: Move sve record checks into restore linux-user/aarch64: Implement SME signal handling linux-user: Rename sve prctls linux-user/aarch64: Implement PR_SME_GET_VL, PR_SME_SET_VL target/arm: Only set ZEN in reset if SVE present target/arm: Enable SME for user-only linux-user/aarch64: Add SME related hwcap entries docs/system/arm/emulation.rst | 4 + linux-user/aarch64/target_cpu.h | 5 +- linux-user/aarch64/target_prctl.h | 62 +- target/arm/cpu.h | 7 + target/arm/helper-sme.h | 126 ++++ target/arm/helper-sve.h | 4 + target/arm/helper.h | 18 + target/arm/translate-a64.h | 45 ++ target/arm/translate.h | 16 + target/arm/sme-fa64.decode | 60 ++ target/arm/sme.decode | 88 +++ target/arm/sve.decode | 41 +- linux-user/aarch64/cpu_loop.c | 9 + linux-user/aarch64/signal.c | 243 ++++++-- linux-user/elfload.c | 20 + linux-user/syscall.c | 28 +- target/arm/cpu.c | 35 +- target/arm/cpu64.c | 11 + target/arm/helper.c | 56 +- target/arm/sme_helper.c | 1140 +++++++++++++++++++++++++++++++++++++ target/arm/sve_helper.c | 28 + target/arm/translate-a64.c | 103 +++- target/arm/translate-sme.c | 373 ++++++++++++ target/arm/translate-sve.c | 393 ++++++++++--- target/arm/translate-vfp.c | 12 + target/arm/translate.c | 2 + target/arm/vec_helper.c | 24 + target/arm/meson.build | 3 + 28 files changed, 2821 insertions(+), 135 deletions(-) create mode 100644 target/arm/sme-fa64.decode create mode 100644 target/arm/sme.decode create mode 100644 target/arm/translate-sme.c From: Richard Henderson <richard.henderson@linaro.org> Dump SVCR, plus use the correct access check for Streaming Mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-2-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-) diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) int i; int el = arm_current_el(env); const char *ns_status; + bool sve; qemu_fprintf(f, " PC=%016" PRIx64 " ", env->pc); for (i = 0; i < 32; i++) { @@ -XXX,XX +XXX,XX @@ static void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) el, psr & PSTATE_SP ? 'h' : 't'); + if (cpu_isar_feature(aa64_sme, cpu)) { + qemu_fprintf(f, " SVCR=%08" PRIx64 " %c%c", + env->svcr, + (FIELD_EX64(env->svcr, SVCR, ZA) ? 'Z' : '-'), + (FIELD_EX64(env->svcr, SVCR, SM) ? 'S' : '-')); + } if (cpu_isar_feature(aa64_bti, cpu)) { qemu_fprintf(f, " BTYPE=%d", (psr & PSTATE_BTYPE) >> 10); } @@ -XXX,XX +XXX,XX @@ static void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) qemu_fprintf(f, " FPCR=%08x FPSR=%08x\n", vfp_get_fpcr(env), vfp_get_fpsr(env)); - if (cpu_isar_feature(aa64_sve, cpu) && sve_exception_el(env, el) == 0) { + if (cpu_isar_feature(aa64_sme, cpu) && FIELD_EX64(env->svcr, SVCR, SM)) { + sve = sme_exception_el(env, el) == 0; + } else if (cpu_isar_feature(aa64_sve, cpu)) { + sve = sve_exception_el(env, el) == 0; + } else { + sve = false; + } + + if (sve) { int j, zcr_len = sve_vqm1_for_el(env, el); for (i = 0; i <= FFR_PRED_NUM; i++) { -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This includes the build rules for the decoder, and the new file for translation, but excludes any instructions. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-3-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.h | 1 + target/arm/sme.decode | 20 ++++++++++++++++++++ target/arm/translate-a64.c | 7 ++++++- target/arm/translate-sme.c | 35 +++++++++++++++++++++++++++++++++++ target/arm/meson.build | 2 ++ 5 files changed, 64 insertions(+), 1 deletion(-) create mode 100644 target/arm/sme.decode create mode 100644 target/arm/translate-sme.c diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ static inline int pred_gvec_reg_size(DisasContext *s) } bool disas_sve(DisasContext *, uint32_t); +bool disas_sme(DisasContext *, uint32_t); void gen_gvec_rax1(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz); diff --git a/target/arm/sme.decode b/target/arm/sme.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ +# AArch64 SME instruction descriptions +# +# Copyright (c) 2022 Linaro, Ltd +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. + +# +# This file is processed by scripts/decodetree.py +# diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void aarch64_tr_translate_insn(DisasContextBase *dcbase, CPUState *cpu) } switch (extract32(insn, 25, 4)) { - case 0x0: case 0x1: case 0x3: /* UNALLOCATED */ + case 0x0: + if (!extract32(insn, 31, 1) || !disas_sme(s, insn)) { + unallocated_encoding(s); + } + break; + case 0x1: case 0x3: /* UNALLOCATED */ unallocated_encoding(s); break; case 0x2: diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ +/* + * AArch64 SME translation + * + * Copyright (c) 2022 Linaro, Ltd + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "cpu.h" +#include "tcg/tcg-op.h" +#include "tcg/tcg-op-gvec.h" +#include "tcg/tcg-gvec-desc.h" +#include "translate.h" +#include "exec/helper-gen.h" +#include "translate-a64.h" +#include "fpu/softfloat.h" + + +/* + * Include the generated decoder. + */ + +#include "decode-sme.c.inc" diff --git a/target/arm/meson.build b/target/arm/meson.build index XXXXXXX..XXXXXXX 100644 --- a/target/arm/meson.build +++ b/target/arm/meson.build @@ -XXX,XX +XXX,XX @@ gen = [ decodetree.process('sve.decode', extra_args: '--decode=disas_sve'), + decodetree.process('sme.decode', extra_args: '--decode=disas_sme'), decodetree.process('neon-shared.decode', extra_args: '--decode=disas_neon_shared'), decodetree.process('neon-dp.decode', extra_args: '--decode=disas_neon_dp'), decodetree.process('neon-ls.decode', extra_args: '--decode=disas_neon_ls'), @@ -XXX,XX +XXX,XX @@ arm_ss.add(when: 'TARGET_AARCH64', if_true: files( 'sme_helper.c', 'translate-a64.c', 'translate-sve.c', + 'translate-sme.c', )) arm_softmmu_ss = ss.source_set() -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This new behaviour is in the ARM pseudocode function AArch64.CheckFPAdvSIMDEnabled, which applies to AArch32 via AArch32.CheckAdvSIMDOrFPEnabled when the EL to which the trap would be delivered is in AArch64 mode. Given that ARMv9 drops support for AArch32 outside EL0, the trap EL detection ought to be trivially true, but the pseudocode still contains a number of conditions, and QEMU has not yet committed to dropping A32 support for EL[12] when v9 features are present. Since the computation of SME_TRAP_NONSTREAMING is necessarily different for the two modes, we might as well preserve bits within TBFLAG_ANY and allocate separate bits within TBFLAG_A32 and TBFLAG_A64 instead. Note that DDI0616A.a has typos for bits [22:21] of LD1RO in the table of instructions illegal in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-4-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 7 +++ target/arm/translate.h | 4 ++ target/arm/sme-fa64.decode | 90 ++++++++++++++++++++++++++++++++++++++ target/arm/helper.c | 41 +++++++++++++++++ target/arm/translate-a64.c | 40 ++++++++++++++++- target/arm/translate-vfp.c | 12 +++++ target/arm/translate.c | 2 + target/arm/meson.build | 1 + 8 files changed, 195 insertions(+), 2 deletions(-) create mode 100644 target/arm/sme-fa64.decode diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ FIELD(TBFLAG_A32, HSTR_ACTIVE, 9, 1) * the same thing as the current security state of the processor! */ FIELD(TBFLAG_A32, NS, 10, 1) +/* + * Indicates that SME Streaming mode is active, and SMCR_ELx.FA64 is not. + * This requires an SME trap from AArch32 mode when using NEON. + */ +FIELD(TBFLAG_A32, SME_TRAP_NONSTREAMING, 11, 1) /* * Bit usage when in AArch32 state, for M-profile only. @@ -XXX,XX +XXX,XX @@ FIELD(TBFLAG_A64, SMEEXC_EL, 20, 2) FIELD(TBFLAG_A64, PSTATE_SM, 22, 1) FIELD(TBFLAG_A64, PSTATE_ZA, 23, 1) FIELD(TBFLAG_A64, SVL, 24, 4) +/* Indicates that SME Streaming mode is active, and SMCR_ELx.FA64 is not. */ +FIELD(TBFLAG_A64, SME_TRAP_NONSTREAMING, 28, 1) /* * Helpers for using the above. diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ typedef struct DisasContext { bool pstate_sm; /* True if PSTATE.ZA is set. */ bool pstate_za; + /* True if non-streaming insns should raise an SME Streaming exception. */ + bool sme_trap_nonstreaming; + /* True if the current instruction is non-streaming. */ + bool is_nonstreaming; /* True if MVE insns are definitely not predicated by VPR or LTPSIZE */ bool mve_no_pred; /* diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ +# AArch64 SME allowed instruction decoding +# +# Copyright (c) 2022 Linaro, Ltd +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. + +# +# This file is processed by scripts/decodetree.py +# + +# These patterns are taken from Appendix E1.1 of DDI0616 A.a, +# Arm Architecture Reference Manual Supplement, +# The Scalable Matrix Extension (SME), for Armv9-A + +{ + [ + OK 0-00 1110 0000 0001 0010 11-- ---- ---- # SMOV W|Xd,Vn.B[0] + OK 0-00 1110 0000 0010 0010 11-- ---- ---- # SMOV W|Xd,Vn.H[0] + OK 0100 1110 0000 0100 0010 11-- ---- ---- # SMOV Xd,Vn.S[0] + OK 0000 1110 0000 0001 0011 11-- ---- ---- # UMOV Wd,Vn.B[0] + OK 0000 1110 0000 0010 0011 11-- ---- ---- # UMOV Wd,Vn.H[0] + OK 0000 1110 0000 0100 0011 11-- ---- ---- # UMOV Wd,Vn.S[0] + OK 0100 1110 0000 1000 0011 11-- ---- ---- # UMOV Xd,Vn.D[0] + ] + FAIL 0--0 111- ---- ---- ---- ---- ---- ---- # Advanced SIMD vector operations +} + +{ + [ + OK 0101 1110 --1- ---- 11-1 11-- ---- ---- # FMULX/FRECPS/FRSQRTS (scalar) + OK 0101 1110 -10- ---- 00-1 11-- ---- ---- # FMULX/FRECPS/FRSQRTS (scalar, FP16) + OK 01-1 1110 1-10 0001 11-1 10-- ---- ---- # FRECPE/FRSQRTE/FRECPX (scalar) + OK 01-1 1110 1111 1001 11-1 10-- ---- ---- # FRECPE/FRSQRTE/FRECPX (scalar, FP16) + ] + FAIL 01-1 111- ---- ---- ---- ---- ---- ---- # Advanced SIMD single-element operations +} + +FAIL 0-00 110- ---- ---- ---- ---- ---- ---- # Advanced SIMD structure load/store +FAIL 1100 1110 ---- ---- ---- ---- ---- ---- # Advanced SIMD cryptography extensions +FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS + +# These are the "avoidance of doubt" final table of Illegal Advanced SIMD instructions +# We don't actually need to include these, as the default is OK. +# -001 111- ---- ---- ---- ---- ---- ---- # Scalar floating-point operations +# --10 110- ---- ---- ---- ---- ---- ---- # Load/store pair of FP registers +# --01 1100 ---- ---- ---- ---- ---- ---- # Load FP register (PC-relative literal) +# --11 1100 --0- ---- ---- ---- ---- ---- # Load/store FP register (unscaled imm) +# --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) +# --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) + +FAIL 0000 0100 --1- ---- 1010 ---- ---- ---- # ADR +FAIL 0000 0100 --1- ---- 1011 -0-- ---- ---- # FTSSEL, FEXPA +FAIL 0000 0101 --10 0001 100- ---- ---- ---- # COMPACT +FAIL 0010 0101 --01 100- 1111 000- ---0 ---- # RDFFR, RDFFRS +FAIL 0010 0101 --10 1--- 1001 ---- ---- ---- # WRFFR, SETFFR +FAIL 0100 0101 --0- ---- 1011 ---- ---- ---- # BDEP, BEXT, BGRP +FAIL 0100 0101 000- ---- 0110 1--- ---- ---- # PMULLB, PMULLT (128b result) +FAIL 0110 0100 --1- ---- 1110 01-- ---- ---- # FMMLA, BFMMLA +FAIL 0110 0101 --0- ---- 0000 11-- ---- ---- # FTSMUL +FAIL 0110 0101 --01 0--- 100- ---- ---- ---- # FTMAD +FAIL 0110 0101 --01 1--- 001- ---- ---- ---- # FADDA +FAIL 0100 0101 --0- ---- 1001 10-- ---- ---- # SMMLA, UMMLA, USMMLA +FAIL 0100 0101 --1- ---- 1--- ---- ---- ---- # SVE2 string/histo/crypto instructions +FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) +FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) +FAIL 1000 0100 0-1- ---- 0--- ---- ---- ---- # SVE 32-bit gather prefetch (scalar+vector) +FAIL 1000 010- -01- ---- 1--- ---- ---- ---- # SVE 32-bit gather load (vector+imm) +FAIL 1000 0100 0-0- ---- 0--- ---- ---- ---- # SVE 32-bit gather load byte (scalar+vector) +FAIL 1000 0100 1--- ---- 0--- ---- ---- ---- # SVE 32-bit gather load half (scalar+vector) +FAIL 1000 0101 0--- ---- 0--- ---- ---- ---- # SVE 32-bit gather load word (scalar+vector) +FAIL 1010 010- ---- ---- 011- ---- ---- ---- # SVE contiguous FF load (scalar+scalar) +FAIL 1010 010- ---1 ---- 101- ---- ---- ---- # SVE contiguous NF load (scalar+imm) +FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) +FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) +FAIL 1100 010- ---- ---- ---- ---- ---- ---- # SVE 64-bit gather load/prefetch +FAIL 1110 010- -00- ---- 001- ---- ---- ---- # SVE2 64-bit scatter NT store (vector+scalar) +FAIL 1110 010- -10- ---- 001- ---- ---- ---- # SVE2 32-bit scatter NT store (vector+scalar) +FAIL 1110 010- ---- ---- 1-0- ---- ---- ---- # SVE scatter store (scalar+32-bit vector) +FAIL 1110 010- ---- ---- 101- ---- ---- ---- # SVE scatter store (misc) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ int sme_exception_el(CPUARMState *env, int el) return 0; } +/* This corresponds to the ARM pseudocode function IsFullA64Enabled(). */ +static bool sme_fa64(CPUARMState *env, int el) +{ + if (!cpu_isar_feature(aa64_sme_fa64, env_archcpu(env))) { + return false; + } + + if (el <= 1 && !el_is_in_host(env, el)) { + if (!FIELD_EX64(env->vfp.smcr_el[1], SMCR, FA64)) { + return false; + } + } + if (el <= 2 && arm_is_el2_enabled(env)) { + if (!FIELD_EX64(env->vfp.smcr_el[2], SMCR, FA64)) { + return false; + } + } + if (arm_feature(env, ARM_FEATURE_EL3)) { + if (!FIELD_EX64(env->vfp.smcr_el[3], SMCR, FA64)) { + return false; + } + } + + return true; +} + /* * Given that SVE is enabled, return the vector length for EL. */ @@ -XXX,XX +XXX,XX @@ static CPUARMTBFlags rebuild_hflags_a32(CPUARMState *env, int fp_el, DP_TBFLAG_ANY(flags, PSTATE__IL, 1); } + /* + * The SME exception we are testing for is raised via + * AArch64.CheckFPAdvSIMDEnabled(), as called from + * AArch32.CheckAdvSIMDOrFPEnabled(). + */ + if (el == 0 + && FIELD_EX64(env->svcr, SVCR, SM) + && (!arm_is_el2_enabled(env) + || (arm_el_is_aa64(env, 2) && !(env->cp15.hcr_el2 & HCR_TGE))) + && arm_el_is_aa64(env, 1) + && !sme_fa64(env, el)) { + DP_TBFLAG_A32(flags, SME_TRAP_NONSTREAMING, 1); + } + return rebuild_hflags_common_32(env, fp_el, mmu_idx, flags); } @@ -XXX,XX +XXX,XX @@ static CPUARMTBFlags rebuild_hflags_a64(CPUARMState *env, int el, int fp_el, } if (FIELD_EX64(env->svcr, SVCR, SM)) { DP_TBFLAG_A64(flags, PSTATE_SM, 1); + DP_TBFLAG_A64(flags, SME_TRAP_NONSTREAMING, !sme_fa64(env, el)); } DP_TBFLAG_A64(flags, PSTATE_ZA, FIELD_EX64(env->svcr, SVCR, ZA)); } diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void do_vec_ld(DisasContext *s, int destidx, int element, * unallocated-encoding checks (otherwise the syndrome information * for the resulting exception will be incorrect). */ -static bool fp_access_check(DisasContext *s) +static bool fp_access_check_only(DisasContext *s) { if (s->fp_excp_el) { assert(!s->fp_access_checked); @@ -XXX,XX +XXX,XX @@ static bool fp_access_check(DisasContext *s) return true; } +static bool fp_access_check(DisasContext *s) +{ + if (!fp_access_check_only(s)) { + return false; + } + if (s->sme_trap_nonstreaming && s->is_nonstreaming) { + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_smetrap(SME_ET_Streaming, false)); + return false; + } + return true; +} + /* Check that SVE access is enabled. If it is, return true. * If not, emit code to generate an appropriate exception and return false. */ @@ -XXX,XX +XXX,XX @@ static void handle_sys(DisasContext *s, uint32_t insn, bool isread, default: g_assert_not_reached(); } - if ((ri->type & ARM_CP_FPU) && !fp_access_check(s)) { + if ((ri->type & ARM_CP_FPU) && !fp_access_check_only(s)) { return; } else if ((ri->type & ARM_CP_SVE) && !sve_access_check(s)) { return; @@ -XXX,XX +XXX,XX @@ static void disas_data_proc_simd_fp(DisasContext *s, uint32_t insn) } } +/* + * Include the generated SME FA64 decoder. + */ + +#include "decode-sme-fa64.c.inc" + +static bool trans_OK(DisasContext *s, arg_OK *a) +{ + return true; +} + +static bool trans_FAIL(DisasContext *s, arg_OK *a) +{ + s->is_nonstreaming = true; + return true; +} + /** * is_guarded_page: * @env: The cpu environment @@ -XXX,XX +XXX,XX @@ static void aarch64_tr_init_disas_context(DisasContextBase *dcbase, dc->mte_active[1] = EX_TBFLAG_A64(tb_flags, MTE0_ACTIVE); dc->pstate_sm = EX_TBFLAG_A64(tb_flags, PSTATE_SM); dc->pstate_za = EX_TBFLAG_A64(tb_flags, PSTATE_ZA); + dc->sme_trap_nonstreaming = EX_TBFLAG_A64(tb_flags, SME_TRAP_NONSTREAMING); dc->vec_len = 0; dc->vec_stride = 0; dc->cp_regs = arm_cpu->cp_regs; @@ -XXX,XX +XXX,XX @@ static void aarch64_tr_translate_insn(DisasContextBase *dcbase, CPUState *cpu) } } + s->is_nonstreaming = false; + if (s->sme_trap_nonstreaming) { + disas_sme_fa64(s, insn); + } + switch (extract32(insn, 25, 4)) { case 0x0: if (!extract32(insn, 31, 1) || !disas_sme(s, insn)) { diff --git a/target/arm/translate-vfp.c b/target/arm/translate-vfp.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-vfp.c +++ b/target/arm/translate-vfp.c @@ -XXX,XX +XXX,XX @@ static bool vfp_access_check_a(DisasContext *s, bool ignore_vfp_enabled) return false; } + /* + * Note that rebuild_hflags_a32 has already accounted for being in EL0 + * and the higher EL in A64 mode, etc. Unlike A64 mode, there do not + * appear to be any insns which touch VFP which are allowed. + */ + if (s->sme_trap_nonstreaming) { + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_smetrap(SME_ET_Streaming, + s->base.pc_next - s->pc_curr == 2)); + return false; + } + if (!s->vfp_enabled && !ignore_vfp_enabled) { assert(!arm_dc_feature(s, ARM_FEATURE_M)); unallocated_encoding(s); diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static void arm_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs) dc->vec_len = EX_TBFLAG_A32(tb_flags, VECLEN); dc->vec_stride = EX_TBFLAG_A32(tb_flags, VECSTRIDE); } + dc->sme_trap_nonstreaming = + EX_TBFLAG_A32(tb_flags, SME_TRAP_NONSTREAMING); } dc->cp_regs = cpu->cp_regs; dc->features = env->features; diff --git a/target/arm/meson.build b/target/arm/meson.build index XXXXXXX..XXXXXXX 100644 --- a/target/arm/meson.build +++ b/target/arm/meson.build @@ -XXX,XX +XXX,XX @@ gen = [ decodetree.process('sve.decode', extra_args: '--decode=disas_sve'), decodetree.process('sme.decode', extra_args: '--decode=disas_sme'), + decodetree.process('sme-fa64.decode', extra_args: '--static-decode=disas_sme_fa64'), decodetree.process('neon-shared.decode', extra_args: '--decode=disas_neon_shared'), decodetree.process('neon-dp.decode', extra_args: '--decode=disas_neon_dp'), decodetree.process('neon-ls.decode', extra_args: '--decode=disas_neon_ls'), -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark ADR as a non-streaming instruction, which should trap if full a64 support is not enabled in streaming mode. Removing entries from sme-fa64.decode is an easy way to see what remains to be done. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-5-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate.h | 7 +++++++ target/arm/sme-fa64.decode | 1 - target/arm/translate-sve.c | 8 ++++---- 3 files changed, 11 insertions(+), 5 deletions(-) diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ uint64_t asimd_imm_const(uint32_t imm, int cmode, int op); static bool trans_##NAME(DisasContext *s, arg_##NAME *a) \ { return dc_isar_feature(FEAT, s) && FUNC(s, __VA_ARGS__); } +#define TRANS_FEAT_NONSTREAMING(NAME, FEAT, FUNC, ...) \ + static bool trans_##NAME(DisasContext *s, arg_##NAME *a) \ + { \ + s->is_nonstreaming = true; \ + return dc_isar_feature(FEAT, s) && FUNC(s, __VA_ARGS__); \ + } + #endif /* TARGET_ARM_TRANSLATE_H */ diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0000 0100 --1- ---- 1010 ---- ---- ---- # ADR FAIL 0000 0100 --1- ---- 1011 -0-- ---- ---- # FTSSEL, FEXPA FAIL 0000 0101 --10 0001 100- ---- ---- ---- # COMPACT FAIL 0010 0101 --01 100- 1111 000- ---0 ---- # RDFFR, RDFFRS diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool do_adr(DisasContext *s, arg_rrri *a, gen_helper_gvec_3 *fn) return gen_gvec_ool_zzz(s, fn, a->rd, a->rn, a->rm, a->imm); } -TRANS_FEAT(ADR_p32, aa64_sve, do_adr, a, gen_helper_sve_adr_p32) -TRANS_FEAT(ADR_p64, aa64_sve, do_adr, a, gen_helper_sve_adr_p64) -TRANS_FEAT(ADR_s32, aa64_sve, do_adr, a, gen_helper_sve_adr_s32) -TRANS_FEAT(ADR_u32, aa64_sve, do_adr, a, gen_helper_sve_adr_u32) +TRANS_FEAT_NONSTREAMING(ADR_p32, aa64_sve, do_adr, a, gen_helper_sve_adr_p32) +TRANS_FEAT_NONSTREAMING(ADR_p64, aa64_sve, do_adr, a, gen_helper_sve_adr_p64) +TRANS_FEAT_NONSTREAMING(ADR_s32, aa64_sve, do_adr, a, gen_helper_sve_adr_s32) +TRANS_FEAT_NONSTREAMING(ADR_u32, aa64_sve, do_adr, a, gen_helper_sve_adr_u32) /* *** SVE Integer Misc - Unpredicated Group -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-6-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 2 -- target/arm/translate-sve.c | 9 ++++++--- 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS FAIL 0000 0100 --1- ---- 1011 -0-- ---- ---- # FTSSEL, FEXPA FAIL 0000 0101 --10 0001 100- ---- ---- ---- # COMPACT -FAIL 0010 0101 --01 100- 1111 000- ---0 ---- # RDFFR, RDFFRS -FAIL 0010 0101 --10 1--- 1001 ---- ---- ---- # WRFFR, SETFFR FAIL 0100 0101 --0- ---- 1011 ---- ---- ---- # BDEP, BEXT, BGRP FAIL 0100 0101 000- ---- 0110 1--- ---- ---- # PMULLB, PMULLT (128b result) FAIL 0110 0100 --1- ---- 1110 01-- ---- ---- # FMMLA, BFMMLA diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool do_predset(DisasContext *s, int esz, int rd, int pat, bool setflag) TRANS_FEAT(PTRUE, aa64_sve, do_predset, a->esz, a->rd, a->pat, a->s) /* Note pat == 31 is #all, to set all elements. */ -TRANS_FEAT(SETFFR, aa64_sve, do_predset, 0, FFR_PRED_NUM, 31, false) +TRANS_FEAT_NONSTREAMING(SETFFR, aa64_sve, + do_predset, 0, FFR_PRED_NUM, 31, false) /* Note pat == 32 is #unimp, to set no elements. */ TRANS_FEAT(PFALSE, aa64_sve, do_predset, 0, a->rd, 32, false) @@ -XXX,XX +XXX,XX @@ static bool trans_RDFFR_p(DisasContext *s, arg_RDFFR_p *a) .rd = a->rd, .pg = a->pg, .s = a->s, .rn = FFR_PRED_NUM, .rm = FFR_PRED_NUM, }; + + s->is_nonstreaming = true; return trans_AND_pppp(s, &alt_a); } -TRANS_FEAT(RDFFR, aa64_sve, do_mov_p, a->rd, FFR_PRED_NUM) -TRANS_FEAT(WRFFR, aa64_sve, do_mov_p, FFR_PRED_NUM, a->rn) +TRANS_FEAT_NONSTREAMING(RDFFR, aa64_sve, do_mov_p, a->rd, FFR_PRED_NUM) +TRANS_FEAT_NONSTREAMING(WRFFR, aa64_sve, do_mov_p, FFR_PRED_NUM, a->rn) static bool do_pfirst_pnext(DisasContext *s, arg_rr_esz *a, void (*gen_fn)(TCGv_i32, TCGv_ptr, -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-7-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 3 --- target/arm/translate-sve.c | 22 ++++++++++++---------- 2 files changed, 12 insertions(+), 13 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0000 0100 --1- ---- 1011 -0-- ---- ---- # FTSSEL, FEXPA -FAIL 0000 0101 --10 0001 100- ---- ---- ---- # COMPACT -FAIL 0100 0101 --0- ---- 1011 ---- ---- ---- # BDEP, BEXT, BGRP FAIL 0100 0101 000- ---- 0110 1--- ---- ---- # PMULLB, PMULLT (128b result) FAIL 0110 0100 --1- ---- 1110 01-- ---- ---- # FMMLA, BFMMLA FAIL 0110 0101 --0- ---- 0000 11-- ---- ---- # FTSMUL diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static gen_helper_gvec_2 * const fexpa_fns[4] = { NULL, gen_helper_sve_fexpa_h, gen_helper_sve_fexpa_s, gen_helper_sve_fexpa_d, }; -TRANS_FEAT(FEXPA, aa64_sve, gen_gvec_ool_zz, - fexpa_fns[a->esz], a->rd, a->rn, 0) +TRANS_FEAT_NONSTREAMING(FEXPA, aa64_sve, gen_gvec_ool_zz, + fexpa_fns[a->esz], a->rd, a->rn, 0) static gen_helper_gvec_3 * const ftssel_fns[4] = { NULL, gen_helper_sve_ftssel_h, gen_helper_sve_ftssel_s, gen_helper_sve_ftssel_d, }; -TRANS_FEAT(FTSSEL, aa64_sve, gen_gvec_ool_arg_zzz, ftssel_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(FTSSEL, aa64_sve, gen_gvec_ool_arg_zzz, + ftssel_fns[a->esz], a, 0) /* *** SVE Predicate Logical Operations Group @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(TRN2_q, aa64_sve_f64mm, gen_gvec_ool_arg_zzz, static gen_helper_gvec_3 * const compact_fns[4] = { NULL, NULL, gen_helper_sve_compact_s, gen_helper_sve_compact_d }; -TRANS_FEAT(COMPACT, aa64_sve, gen_gvec_ool_arg_zpz, compact_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(COMPACT, aa64_sve, gen_gvec_ool_arg_zpz, + compact_fns[a->esz], a, 0) /* Call the helper that computes the ARM LastActiveElement pseudocode * function, scaled by the element size. This includes the not found @@ -XXX,XX +XXX,XX @@ static gen_helper_gvec_3 * const bext_fns[4] = { gen_helper_sve2_bext_b, gen_helper_sve2_bext_h, gen_helper_sve2_bext_s, gen_helper_sve2_bext_d, }; -TRANS_FEAT(BEXT, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, - bext_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(BEXT, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, + bext_fns[a->esz], a, 0) static gen_helper_gvec_3 * const bdep_fns[4] = { gen_helper_sve2_bdep_b, gen_helper_sve2_bdep_h, gen_helper_sve2_bdep_s, gen_helper_sve2_bdep_d, }; -TRANS_FEAT(BDEP, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, - bdep_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(BDEP, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, + bdep_fns[a->esz], a, 0) static gen_helper_gvec_3 * const bgrp_fns[4] = { gen_helper_sve2_bgrp_b, gen_helper_sve2_bgrp_h, gen_helper_sve2_bgrp_s, gen_helper_sve2_bgrp_d, }; -TRANS_FEAT(BGRP, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, - bgrp_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(BGRP, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, + bgrp_fns[a->esz], a, 0) static gen_helper_gvec_3 * const cadd_fns[4] = { gen_helper_sve2_cadd_b, gen_helper_sve2_cadd_h, -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-8-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 2 -- target/arm/translate-sve.c | 24 +++++++++++++++--------- 2 files changed, 15 insertions(+), 11 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0100 0101 000- ---- 0110 1--- ---- ---- # PMULLB, PMULLT (128b result) -FAIL 0110 0100 --1- ---- 1110 01-- ---- ---- # FMMLA, BFMMLA FAIL 0110 0101 --0- ---- 0000 11-- ---- ---- # FTSMUL FAIL 0110 0101 --01 0--- 100- ---- ---- ---- # FTMAD FAIL 0110 0101 --01 1--- 001- ---- ---- ---- # FADDA diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool do_trans_pmull(DisasContext *s, arg_rrr_esz *a, bool sel) gen_helper_gvec_pmull_q, gen_helper_sve2_pmull_h, NULL, gen_helper_sve2_pmull_d, }; - if (a->esz == 0 - ? !dc_isar_feature(aa64_sve2_pmull128, s) - : !dc_isar_feature(aa64_sve, s)) { + + if (a->esz == 0) { + if (!dc_isar_feature(aa64_sve2_pmull128, s)) { + return false; + } + s->is_nonstreaming = true; + } else if (!dc_isar_feature(aa64_sve, s)) { return false; } return gen_gvec_ool_arg_zzz(s, fns[a->esz], a, sel); @@ -XXX,XX +XXX,XX @@ DO_ZPZZ_FP(FMINP, aa64_sve2, sve2_fminp_zpzz) * SVE Integer Multiply-Add (unpredicated) */ -TRANS_FEAT(FMMLA_s, aa64_sve_f32mm, gen_gvec_fpst_zzzz, gen_helper_fmmla_s, - a->rd, a->rn, a->rm, a->ra, 0, FPST_FPCR) -TRANS_FEAT(FMMLA_d, aa64_sve_f64mm, gen_gvec_fpst_zzzz, gen_helper_fmmla_d, - a->rd, a->rn, a->rm, a->ra, 0, FPST_FPCR) +TRANS_FEAT_NONSTREAMING(FMMLA_s, aa64_sve_f32mm, gen_gvec_fpst_zzzz, + gen_helper_fmmla_s, a->rd, a->rn, a->rm, a->ra, + 0, FPST_FPCR) +TRANS_FEAT_NONSTREAMING(FMMLA_d, aa64_sve_f64mm, gen_gvec_fpst_zzzz, + gen_helper_fmmla_d, a->rd, a->rn, a->rm, a->ra, + 0, FPST_FPCR) static gen_helper_gvec_4 * const sqdmlal_zzzw_fns[] = { NULL, gen_helper_sve2_sqdmlal_zzzw_h, @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(BFDOT_zzzz, aa64_sve_bf16, gen_gvec_ool_arg_zzzz, TRANS_FEAT(BFDOT_zzxz, aa64_sve_bf16, gen_gvec_ool_arg_zzxz, gen_helper_gvec_bfdot_idx, a) -TRANS_FEAT(BFMMLA, aa64_sve_bf16, gen_gvec_ool_arg_zzzz, - gen_helper_gvec_bfmmla, a, 0) +TRANS_FEAT_NONSTREAMING(BFMMLA, aa64_sve_bf16, gen_gvec_ool_arg_zzzz, + gen_helper_gvec_bfmmla, a, 0) static bool do_BFMLAL_zzzw(DisasContext *s, arg_rrrr_esz *a, bool sel) { -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-9-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 3 --- target/arm/translate-sve.c | 15 +++++++++++---- 2 files changed, 11 insertions(+), 7 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0110 0101 --0- ---- 0000 11-- ---- ---- # FTSMUL -FAIL 0110 0101 --01 0--- 100- ---- ---- ---- # FTMAD -FAIL 0110 0101 --01 1--- 001- ---- ---- ---- # FADDA FAIL 0100 0101 --0- ---- 1001 10-- ---- ---- # SMMLA, UMMLA, USMMLA FAIL 0100 0101 --1- ---- 1--- ---- ---- ---- # SVE2 string/histo/crypto instructions FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static gen_helper_gvec_3_ptr * const ftmad_fns[4] = { NULL, gen_helper_sve_ftmad_h, gen_helper_sve_ftmad_s, gen_helper_sve_ftmad_d, }; -TRANS_FEAT(FTMAD, aa64_sve, gen_gvec_fpst_zzz, - ftmad_fns[a->esz], a->rd, a->rn, a->rm, a->imm, - a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR) +TRANS_FEAT_NONSTREAMING(FTMAD, aa64_sve, gen_gvec_fpst_zzz, + ftmad_fns[a->esz], a->rd, a->rn, a->rm, a->imm, + a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR) /* *** SVE Floating Point Accumulating Reduction Group @@ -XXX,XX +XXX,XX @@ static bool trans_FADDA(DisasContext *s, arg_rprr_esz *a) if (a->esz == 0 || !dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_FADDA(DisasContext *s, arg_rprr_esz *a) DO_FP3(FADD_zzz, fadd) DO_FP3(FSUB_zzz, fsub) DO_FP3(FMUL_zzz, fmul) -DO_FP3(FTSMUL, ftsmul) DO_FP3(FRECPS, recps) DO_FP3(FRSQRTS, rsqrts) #undef DO_FP3 +static gen_helper_gvec_3_ptr * const ftsmul_fns[4] = { + NULL, gen_helper_gvec_ftsmul_h, + gen_helper_gvec_ftsmul_s, gen_helper_gvec_ftsmul_d +}; +TRANS_FEAT_NONSTREAMING(FTSMUL, aa64_sve, gen_gvec_fpst_arg_zzz, + ftsmul_fns[a->esz], a, 0) + /* *** SVE Floating Point Arithmetic - Predicated Group */ -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-10-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 1 - target/arm/translate-sve.c | 12 ++++++------ 2 files changed, 6 insertions(+), 7 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0100 0101 --0- ---- 1001 10-- ---- ---- # SMMLA, UMMLA, USMMLA FAIL 0100 0101 --1- ---- 1--- ---- ---- ---- # SVE2 string/histo/crypto instructions FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(FMLALT_zzxw, aa64_sve2, do_FMLAL_zzxw, a, false, true) TRANS_FEAT(FMLSLB_zzxw, aa64_sve2, do_FMLAL_zzxw, a, true, false) TRANS_FEAT(FMLSLT_zzxw, aa64_sve2, do_FMLAL_zzxw, a, true, true) -TRANS_FEAT(SMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, - gen_helper_gvec_smmla_b, a, 0) -TRANS_FEAT(USMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, - gen_helper_gvec_usmmla_b, a, 0) -TRANS_FEAT(UMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, - gen_helper_gvec_ummla_b, a, 0) +TRANS_FEAT_NONSTREAMING(SMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, + gen_helper_gvec_smmla_b, a, 0) +TRANS_FEAT_NONSTREAMING(USMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, + gen_helper_gvec_usmmla_b, a, 0) +TRANS_FEAT_NONSTREAMING(UMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, + gen_helper_gvec_ummla_b, a, 0) TRANS_FEAT(BFDOT_zzzz, aa64_sve_bf16, gen_gvec_ool_arg_zzzz, gen_helper_gvec_bfdot, a, 0) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-11-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 1 - target/arm/translate-sve.c | 35 ++++++++++++++++++----------------- 2 files changed, 18 insertions(+), 18 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0100 0101 --1- ---- 1--- ---- ---- ---- # SVE2 string/histo/crypto instructions FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) FAIL 1000 0100 0-1- ---- 0--- ---- ---- ---- # SVE 32-bit gather prefetch (scalar+vector) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ DO_SVE2_ZZZ_NARROW(RSUBHNT, rsubhnt) static gen_helper_gvec_flags_4 * const match_fns[4] = { gen_helper_sve2_match_ppzz_b, gen_helper_sve2_match_ppzz_h, NULL, NULL }; -TRANS_FEAT(MATCH, aa64_sve2, do_ppzz_flags, a, match_fns[a->esz]) +TRANS_FEAT_NONSTREAMING(MATCH, aa64_sve2, do_ppzz_flags, a, match_fns[a->esz]) static gen_helper_gvec_flags_4 * const nmatch_fns[4] = { gen_helper_sve2_nmatch_ppzz_b, gen_helper_sve2_nmatch_ppzz_h, NULL, NULL }; -TRANS_FEAT(NMATCH, aa64_sve2, do_ppzz_flags, a, nmatch_fns[a->esz]) +TRANS_FEAT_NONSTREAMING(NMATCH, aa64_sve2, do_ppzz_flags, a, nmatch_fns[a->esz]) static gen_helper_gvec_4 * const histcnt_fns[4] = { NULL, NULL, gen_helper_sve2_histcnt_s, gen_helper_sve2_histcnt_d }; -TRANS_FEAT(HISTCNT, aa64_sve2, gen_gvec_ool_arg_zpzz, - histcnt_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(HISTCNT, aa64_sve2, gen_gvec_ool_arg_zpzz, + histcnt_fns[a->esz], a, 0) -TRANS_FEAT(HISTSEG, aa64_sve2, gen_gvec_ool_arg_zzz, - a->esz == 0 ? gen_helper_sve2_histseg : NULL, a, 0) +TRANS_FEAT_NONSTREAMING(HISTSEG, aa64_sve2, gen_gvec_ool_arg_zzz, + a->esz == 0 ? gen_helper_sve2_histseg : NULL, a, 0) DO_ZPZZ_FP(FADDP, aa64_sve2, sve2_faddp_zpzz) DO_ZPZZ_FP(FMAXNMP, aa64_sve2, sve2_fmaxnmp_zpzz) @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(SQRDCMLAH_zzzz, aa64_sve2, gen_gvec_ool_zzzz, TRANS_FEAT(USDOT_zzzz, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, a->esz == 2 ? gen_helper_gvec_usdot_b : NULL, a, 0) -TRANS_FEAT(AESMC, aa64_sve2_aes, gen_gvec_ool_zz, - gen_helper_crypto_aesmc, a->rd, a->rd, a->decrypt) +TRANS_FEAT_NONSTREAMING(AESMC, aa64_sve2_aes, gen_gvec_ool_zz, + gen_helper_crypto_aesmc, a->rd, a->rd, a->decrypt) -TRANS_FEAT(AESE, aa64_sve2_aes, gen_gvec_ool_arg_zzz, - gen_helper_crypto_aese, a, false) -TRANS_FEAT(AESD, aa64_sve2_aes, gen_gvec_ool_arg_zzz, - gen_helper_crypto_aese, a, true) +TRANS_FEAT_NONSTREAMING(AESE, aa64_sve2_aes, gen_gvec_ool_arg_zzz, + gen_helper_crypto_aese, a, false) +TRANS_FEAT_NONSTREAMING(AESD, aa64_sve2_aes, gen_gvec_ool_arg_zzz, + gen_helper_crypto_aese, a, true) -TRANS_FEAT(SM4E, aa64_sve2_sm4, gen_gvec_ool_arg_zzz, - gen_helper_crypto_sm4e, a, 0) -TRANS_FEAT(SM4EKEY, aa64_sve2_sm4, gen_gvec_ool_arg_zzz, - gen_helper_crypto_sm4ekey, a, 0) +TRANS_FEAT_NONSTREAMING(SM4E, aa64_sve2_sm4, gen_gvec_ool_arg_zzz, + gen_helper_crypto_sm4e, a, 0) +TRANS_FEAT_NONSTREAMING(SM4EKEY, aa64_sve2_sm4, gen_gvec_ool_arg_zzz, + gen_helper_crypto_sm4ekey, a, 0) -TRANS_FEAT(RAX1, aa64_sve2_sha3, gen_gvec_fn_arg_zzz, gen_gvec_rax1, a) +TRANS_FEAT_NONSTREAMING(RAX1, aa64_sve2_sha3, gen_gvec_fn_arg_zzz, + gen_gvec_rax1, a) TRANS_FEAT(FCVTNT_sh, aa64_sve2, gen_gvec_fpst_arg_zpz, gen_helper_sve2_fcvtnt_sh, a, 0, FPST_FPCR) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-12-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 9 --------- target/arm/translate-sve.c | 6 ++++++ 2 files changed, 6 insertions(+), 9 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) FAIL 1000 0100 0-1- ---- 0--- ---- ---- ---- # SVE 32-bit gather prefetch (scalar+vector) -FAIL 1000 010- -01- ---- 1--- ---- ---- ---- # SVE 32-bit gather load (vector+imm) -FAIL 1000 0100 0-0- ---- 0--- ---- ---- ---- # SVE 32-bit gather load byte (scalar+vector) -FAIL 1000 0100 1--- ---- 0--- ---- ---- ---- # SVE 32-bit gather load half (scalar+vector) -FAIL 1000 0101 0--- ---- 0--- ---- ---- ---- # SVE 32-bit gather load word (scalar+vector) FAIL 1010 010- ---- ---- 011- ---- ---- ---- # SVE contiguous FF load (scalar+scalar) FAIL 1010 010- ---1 ---- 101- ---- ---- ---- # SVE contiguous NF load (scalar+imm) FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) FAIL 1100 010- ---- ---- ---- ---- ---- ---- # SVE 64-bit gather load/prefetch -FAIL 1110 010- -00- ---- 001- ---- ---- ---- # SVE2 64-bit scatter NT store (vector+scalar) -FAIL 1110 010- -10- ---- 001- ---- ---- ---- # SVE2 32-bit scatter NT store (vector+scalar) -FAIL 1110 010- ---- ---- 1-0- ---- ---- ---- # SVE scatter store (scalar+32-bit vector) -FAIL 1110 010- ---- ---- 101- ---- ---- ---- # SVE scatter store (misc) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_LD1_zprz(DisasContext *s, arg_LD1_zprz *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_LD1_zpiz(DisasContext *s, arg_LD1_zpiz *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_LDNT1_zprz(DisasContext *s, arg_LD1_zprz *a) if (!dc_isar_feature(aa64_sve2, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_ST1_zprz(DisasContext *s, arg_ST1_zprz *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_ST1_zpiz(DisasContext *s, arg_ST1_zpiz *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_STNT1_zprz(DisasContext *s, arg_ST1_zprz *a) if (!dc_isar_feature(aa64_sve2, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. In this case, introduce PRF_ns (prefetch non-streaming) to handle the checks. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-13-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 3 --- target/arm/sve.decode | 10 +++++----- target/arm/translate-sve.c | 11 +++++++++++ 3 files changed, 16 insertions(+), 8 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) -FAIL 1000 0100 0-1- ---- 0--- ---- ---- ---- # SVE 32-bit gather prefetch (scalar+vector) FAIL 1010 010- ---- ---- 011- ---- ---- ---- # SVE contiguous FF load (scalar+scalar) FAIL 1010 010- ---1 ---- 101- ---- ---- ---- # SVE contiguous NF load (scalar+imm) FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) -FAIL 1100 010- ---- ---- ---- ---- ---- ---- # SVE 64-bit gather load/prefetch diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ LD1RO_zpri 1010010 .. 01 0.... 001 ... ..... ..... \ @rpri_load_msz nreg=0 # SVE 32-bit gather prefetch (scalar plus 32-bit scaled offsets) -PRF 1000010 00 -1 ----- 0-- --- ----- 0 ---- +PRF_ns 1000010 00 -1 ----- 0-- --- ----- 0 ---- # SVE 32-bit gather prefetch (vector plus immediate) -PRF 1000010 -- 00 ----- 111 --- ----- 0 ---- +PRF_ns 1000010 -- 00 ----- 111 --- ----- 0 ---- # SVE contiguous prefetch (scalar plus immediate) PRF 1000010 11 1- ----- 0-- --- ----- 0 ---- @@ -XXX,XX +XXX,XX @@ LD1_zpiz 1100010 .. 01 ..... 1.. ... ..... ..... \ @rpri_g_load esz=3 # SVE 64-bit gather prefetch (scalar plus 64-bit scaled offsets) -PRF 1100010 00 11 ----- 1-- --- ----- 0 ---- +PRF_ns 1100010 00 11 ----- 1-- --- ----- 0 ---- # SVE 64-bit gather prefetch (scalar plus unpacked 32-bit scaled offsets) -PRF 1100010 00 -1 ----- 0-- --- ----- 0 ---- +PRF_ns 1100010 00 -1 ----- 0-- --- ----- 0 ---- # SVE 64-bit gather prefetch (vector plus immediate) -PRF 1100010 -- 00 ----- 111 --- ----- 0 ---- +PRF_ns 1100010 -- 00 ----- 111 --- ----- 0 ---- ### SVE Memory Store Group diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_PRF_rr(DisasContext *s, arg_PRF_rr *a) return true; } +static bool trans_PRF_ns(DisasContext *s, arg_PRF_ns *a) +{ + if (!dc_isar_feature(aa64_sve, s)) { + return false; + } + /* Prefetch is a nop within QEMU. */ + s->is_nonstreaming = true; + (void)sve_access_check(s); + return true; +} + /* * Move Prefix * -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-14-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 2 -- target/arm/translate-sve.c | 2 ++ 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 1010 010- ---- ---- 011- ---- ---- ---- # SVE contiguous FF load (scalar+scalar) -FAIL 1010 010- ---1 ---- 101- ---- ---- ---- # SVE contiguous NF load (scalar+imm) FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_LDFF1_zprr(DisasContext *s, arg_rprr_load *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (sve_access_check(s)) { TCGv_i64 addr = new_tmp_a64(s); tcg_gen_shli_i64(addr, cpu_reg(s, a->rm), dtype_msz(a->dtype)); @@ -XXX,XX +XXX,XX @@ static bool trans_LDNF1_zpri(DisasContext *s, arg_rpri_load *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (sve_access_check(s)) { int vsz = vec_full_reg_size(s); int elements = vsz >> dtype_esz[a->dtype]; -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-15-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 3 --- target/arm/translate-sve.c | 2 ++ 2 files changed, 2 insertions(+), 3 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --0- ---- ---- ---- ---- ---- # Load/store FP register (unscaled imm) # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) - -FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) -FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_LD1RO_zprr(DisasContext *s, arg_rprr_load *a) if (a->rm == 31) { return false; } + s->is_nonstreaming = true; if (sve_access_check(s)) { TCGv_i64 addr = new_tmp_a64(s); tcg_gen_shli_i64(addr, cpu_reg(s, a->rm), dtype_msz(a->dtype)); @@ -XXX,XX +XXX,XX @@ static bool trans_LD1RO_zpri(DisasContext *s, arg_rpri_load *a) if (!dc_isar_feature(aa64_sve_f64mm, s)) { return false; } + s->is_nonstreaming = true; if (sve_access_check(s)) { TCGv_i64 addr = new_tmp_a64(s); tcg_gen_addi_i64(addr, cpu_reg_sp(s, a->rn), a->imm * 32); -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> These functions will be used to verify that the cpu is in the correct state for a given instruction. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-16-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.h | 21 +++++++++++++++++++++ target/arm/translate-a64.c | 34 ++++++++++++++++++++++++++++++++++ 2 files changed, 55 insertions(+) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ void write_fp_dreg(DisasContext *s, int reg, TCGv_i64 v); bool logic_imm_decode_wmask(uint64_t *result, unsigned int immn, unsigned int imms, unsigned int immr); bool sve_access_check(DisasContext *s); +bool sme_enabled_check(DisasContext *s); +bool sme_enabled_check_with_svcr(DisasContext *s, unsigned); + +/* This function corresponds to CheckStreamingSVEEnabled. */ +static inline bool sme_sm_enabled_check(DisasContext *s) +{ + return sme_enabled_check_with_svcr(s, R_SVCR_SM_MASK); +} + +/* This function corresponds to CheckSMEAndZAEnabled. */ +static inline bool sme_za_enabled_check(DisasContext *s) +{ + return sme_enabled_check_with_svcr(s, R_SVCR_ZA_MASK); +} + +/* Note that this function corresponds to CheckStreamingSVEAndZAEnabled. */ +static inline bool sme_smza_enabled_check(DisasContext *s) +{ + return sme_enabled_check_with_svcr(s, R_SVCR_SM_MASK | R_SVCR_ZA_MASK); +} + TCGv_i64 clean_data_tbi(DisasContext *s, TCGv_i64 addr); TCGv_i64 gen_mte_check1(DisasContext *s, TCGv_i64 addr, bool is_write, bool tag_checked, int log2_size); diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static bool sme_access_check(DisasContext *s) return true; } +/* This function corresponds to CheckSMEEnabled. */ +bool sme_enabled_check(DisasContext *s) +{ + /* + * Note that unlike sve_excp_el, we have not constrained sme_excp_el + * to be zero when fp_excp_el has priority. This is because we need + * sme_excp_el by itself for cpregs access checks. + */ + if (!s->fp_excp_el || s->sme_excp_el < s->fp_excp_el) { + s->fp_access_checked = true; + return sme_access_check(s); + } + return fp_access_check_only(s); +} + +/* Common subroutine for CheckSMEAnd*Enabled. */ +bool sme_enabled_check_with_svcr(DisasContext *s, unsigned req) +{ + if (!sme_enabled_check(s)) { + return false; + } + if (FIELD_EX64(req, SVCR, SM) && !s->pstate_sm) { + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_smetrap(SME_ET_NotStreaming, false)); + return false; + } + if (FIELD_EX64(req, SVCR, ZA) && !s->pstate_za) { + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_smetrap(SME_ET_InactiveZA, false)); + return false; + } + return true; +} + /* * This utility function is for doing register extension with an * optional shift. You will likely want to pass a temporary for the -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> The pseudocode for CheckSVEEnabled gains a check for Streaming SVE mode, and for SME present but SVE absent. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-17-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.c | 22 ++++++++++++++++------ 1 file changed, 16 insertions(+), 6 deletions(-) diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static bool fp_access_check(DisasContext *s) return true; } -/* Check that SVE access is enabled. If it is, return true. +/* + * Check that SVE access is enabled. If it is, return true. * If not, emit code to generate an appropriate exception and return false. + * This function corresponds to CheckSVEEnabled(). */ bool sve_access_check(DisasContext *s) { - if (s->sve_excp_el) { - assert(!s->sve_access_checked); - s->sve_access_checked = true; - + if (s->pstate_sm || !dc_isar_feature(aa64_sve, s)) { + assert(dc_isar_feature(aa64_sme, s)); + if (!sme_sm_enabled_check(s)) { + goto fail_exit; + } + } else if (s->sve_excp_el) { gen_exception_insn_el(s, s->pc_curr, EXCP_UDEF, syn_sve_access_trap(), s->sve_excp_el); - return false; + goto fail_exit; } s->sve_access_checked = true; return fp_access_check(s); + + fail_exit: + /* Assert that we only raise one exception per instruction. */ + assert(!s->sve_access_checked); + s->sve_access_checked = true; + return false; } /* -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> These SME instructions are nominally within the SVE decode space, so we add them to sve.decode and translate-sve.c. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-18-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.h | 12 ++++++++++++ target/arm/sve.decode | 5 ++++- target/arm/translate-sve.c | 38 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 54 insertions(+), 1 deletion(-) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ static inline int vec_full_reg_size(DisasContext *s) return s->vl; } +/* Return the byte size of the vector register, SVL / 8. */ +static inline int streaming_vec_reg_size(DisasContext *s) +{ + return s->svl; +} + /* * Return the offset info CPUARMState of the predicate vector register Pn. * Note for this purpose, FFR is P16. @@ -XXX,XX +XXX,XX @@ static inline int pred_full_reg_size(DisasContext *s) return s->vl >> 3; } +/* Return the byte size of the predicate register, SVL / 64. */ +static inline int streaming_pred_reg_size(DisasContext *s) +{ + return s->svl >> 3; +} + /* * Round up the size of a register to a size allowed by * the tcg vector infrastructure. Any operation which uses this diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ INDEX_ri 00000100 esz:2 1 imm:s5 010001 rn:5 rd:5 # SVE index generation (register start, register increment) INDEX_rr 00000100 .. 1 ..... 010011 ..... ..... @rd_rn_rm -### SVE Stack Allocation Group +### SVE / Streaming SVE Stack Allocation Group # SVE stack frame adjustment ADDVL 00000100 001 ..... 01010 ...... ..... @rd_rn_i6 +ADDSVL 00000100 001 ..... 01011 ...... ..... @rd_rn_i6 ADDPL 00000100 011 ..... 01010 ...... ..... @rd_rn_i6 +ADDSPL 00000100 011 ..... 01011 ...... ..... @rd_rn_i6 # SVE stack frame size RDVL 00000100 101 11111 01010 imm:s6 rd:5 +RDSVL 00000100 101 11111 01011 imm:s6 rd:5 ### SVE Bitwise Shift - Unpredicated Group diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_ADDVL(DisasContext *s, arg_ADDVL *a) return true; } +static bool trans_ADDSVL(DisasContext *s, arg_ADDSVL *a) +{ + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (sme_enabled_check(s)) { + TCGv_i64 rd = cpu_reg_sp(s, a->rd); + TCGv_i64 rn = cpu_reg_sp(s, a->rn); + tcg_gen_addi_i64(rd, rn, a->imm * streaming_vec_reg_size(s)); + } + return true; +} + static bool trans_ADDPL(DisasContext *s, arg_ADDPL *a) { if (!dc_isar_feature(aa64_sve, s)) { @@ -XXX,XX +XXX,XX @@ static bool trans_ADDPL(DisasContext *s, arg_ADDPL *a) return true; } +static bool trans_ADDSPL(DisasContext *s, arg_ADDSPL *a) +{ + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (sme_enabled_check(s)) { + TCGv_i64 rd = cpu_reg_sp(s, a->rd); + TCGv_i64 rn = cpu_reg_sp(s, a->rn); + tcg_gen_addi_i64(rd, rn, a->imm * streaming_pred_reg_size(s)); + } + return true; +} + static bool trans_RDVL(DisasContext *s, arg_RDVL *a) { if (!dc_isar_feature(aa64_sve, s)) { @@ -XXX,XX +XXX,XX @@ static bool trans_RDVL(DisasContext *s, arg_RDVL *a) return true; } +static bool trans_RDSVL(DisasContext *s, arg_RDSVL *a) +{ + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (sme_enabled_check(s)) { + TCGv_i64 reg = cpu_reg(s, a->rd); + tcg_gen_movi_i64(reg, a->imm * streaming_vec_reg_size(s)); + } + return true; +} + /* *** SVE Compute Vector Address Group */ -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-19-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 2 ++ target/arm/sme.decode | 4 ++++ target/arm/sme_helper.c | 25 +++++++++++++++++++++++++ target/arm/translate-sme.c | 13 +++++++++++++ 4 files changed, 44 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_2(set_pstate_sm, TCG_CALL_NO_RWG, void, env, i32) DEF_HELPER_FLAGS_2(set_pstate_za, TCG_CALL_NO_RWG, void, env, i32) + +DEF_HELPER_FLAGS_3(sme_zero, TCG_CALL_NO_RWG, void, env, i32, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ # # This file is processed by scripts/decodetree.py # + +### SME Misc + +ZERO 11000000 00 001 00000000000 imm:8 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ void helper_set_pstate_za(CPUARMState *env, uint32_t i) memset(env->zarray, 0, sizeof(env->zarray)); } } + +void helper_sme_zero(CPUARMState *env, uint32_t imm, uint32_t svl) +{ + uint32_t i; + + /* + * Special case clearing the entire ZA space. + * This falls into the CONSTRAINED UNPREDICTABLE zeroing of any + * parts of the ZA storage outside of SVL. + */ + if (imm == 0xff) { + memset(env->zarray, 0, sizeof(env->zarray)); + return; + } + + /* + * Recall that ZAnH.D[m] is spread across ZA[n+8*m], + * so each row is discontiguous within ZA[]. + */ + for (i = 0; i < svl; i++) { + if (imm & (1 << (i % 8))) { + memset(&env->zarray[i], 0, svl); + } + } +} diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ */ #include "decode-sme.c.inc" + + +static bool trans_ZERO(DisasContext *s, arg_ZERO *a) +{ + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (sme_za_enabled_check(s)) { + gen_helper_sme_zero(cpu_env, tcg_constant_i32(a->imm), + tcg_constant_i32(streaming_vec_reg_size(s))); + } + return true; +} -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> We can reuse the SVE functions for implementing moves to/from horizontal tile slices, but we need new ones for moves to/from vertical tile slices. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-20-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 12 +++ target/arm/helper-sve.h | 2 + target/arm/translate-a64.h | 8 ++ target/arm/translate.h | 5 ++ target/arm/sme.decode | 15 ++++ target/arm/sme_helper.c | 151 ++++++++++++++++++++++++++++++++++++- target/arm/sve_helper.c | 12 +++ target/arm/translate-sme.c | 127 +++++++++++++++++++++++++++++++ 8 files changed, 331 insertions(+), 1 deletion(-) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_2(set_pstate_sm, TCG_CALL_NO_RWG, void, env, i32) DEF_HELPER_FLAGS_2(set_pstate_za, TCG_CALL_NO_RWG, void, env, i32) DEF_HELPER_FLAGS_3(sme_zero, TCG_CALL_NO_RWG, void, env, i32, i32) + +/* Move to/from vertical array slices, i.e. columns, so 'c'. */ +DEF_HELPER_FLAGS_4(sme_mova_cz_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_cz_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_cz_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_cz_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_cz_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) diff --git a/target/arm/helper-sve.h b/target/arm/helper-sve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sve.h +++ b/target/arm/helper-sve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_5(sve_sel_zpzz_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sve_sel_zpzz_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(sve_sel_zpzz_q, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sve2_addp_zpzz_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ static inline int pred_gvec_reg_size(DisasContext *s) return size_for_gvec(pred_full_reg_size(s)); } +/* Return a newly allocated pointer to the predicate register. */ +static inline TCGv_ptr pred_full_reg_ptr(DisasContext *s, int regno) +{ + TCGv_ptr ret = tcg_temp_new_ptr(); + tcg_gen_addi_ptr(ret, cpu_env, pred_full_reg_offset(s, regno)); + return ret; +} + bool disas_sve(DisasContext *, uint32_t); bool disas_sme(DisasContext *, uint32_t); diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ static inline int plus_2(DisasContext *s, int x) return x + 2; } +static inline int plus_12(DisasContext *s, int x) +{ + return x + 12; +} + static inline int times_2(DisasContext *s, int x) { return x * 2; diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ ### SME Misc ZERO 11000000 00 001 00000000000 imm:8 + +### SME Move into/from Array + +%mova_rs 13:2 !function=plus_12 +&mova esz rs pg zr za_imm v:bool to_vec:bool + +MOVA 11000000 esz:2 00000 0 v:1 .. pg:3 zr:5 0 za_imm:4 \ + &mova to_vec=0 rs=%mova_rs +MOVA 11000000 11 00000 1 v:1 .. pg:3 zr:5 0 za_imm:4 \ + &mova to_vec=0 rs=%mova_rs esz=4 + +MOVA 11000000 esz:2 00001 0 v:1 .. pg:3 0 za_imm:4 zr:5 \ + &mova to_vec=1 rs=%mova_rs +MOVA 11000000 11 00001 1 v:1 .. pg:3 0 za_imm:4 zr:5 \ + &mova to_vec=1 rs=%mova_rs esz=4 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "cpu.h" -#include "internals.h" +#include "tcg/tcg-gvec-desc.h" #include "exec/helper-proto.h" +#include "qemu/int128.h" +#include "vec_internal.h" /* ResetSVEState */ void arm_reset_sve_state(CPUARMState *env) @@ -XXX,XX +XXX,XX @@ void helper_sme_zero(CPUARMState *env, uint32_t imm, uint32_t svl) } } } + + +/* + * When considering the ZA storage as an array of elements of + * type T, the index within that array of the Nth element of + * a vertical slice of a tile can be calculated like this, + * regardless of the size of type T. This is because the tiles + * are interleaved, so if type T is size N bytes then row 1 of + * the tile is N rows away from row 0. The division by N to + * convert a byte offset into an array index and the multiplication + * by N to convert from vslice-index-within-the-tile to + * the index within the ZA storage cancel out. + */ +#define tile_vslice_index(i) ((i) * sizeof(ARMVectorReg)) + +/* + * When doing byte arithmetic on the ZA storage, the element + * byteoff bytes away in a tile vertical slice is always this + * many bytes away in the ZA storage, regardless of the + * size of the tile element, assuming that byteoff is a multiple + * of the element size. Again this is because of the interleaving + * of the tiles. For instance if we have 1 byte per element then + * each row of the ZA storage has one byte of the vslice data, + * and (counting from 0) byte 8 goes in row 8 of the storage + * at offset (8 * row-size-in-bytes). + * If we have 8 bytes per element then each row of the ZA storage + * has 8 bytes of the data, but there are 8 interleaved tiles and + * so byte 8 of the data goes into row 1 of the tile, + * which is again row 8 of the storage, so the offset is still + * (8 * row-size-in-bytes). Similarly for other element sizes. + */ +#define tile_vslice_offset(byteoff) ((byteoff) * sizeof(ARMVectorReg)) + + +/* + * Move Zreg vector to ZArray column. + */ +#define DO_MOVA_C(NAME, TYPE, H) \ +void HELPER(NAME)(void *za, void *vn, void *vg, uint32_t desc) \ +{ \ + int i, oprsz = simd_oprsz(desc); \ + for (i = 0; i < oprsz; ) { \ + uint16_t pg = *(uint16_t *)(vg + H1_2(i >> 3)); \ + do { \ + if (pg & 1) { \ + *(TYPE *)(za + tile_vslice_offset(i)) = *(TYPE *)(vn + H(i)); \ + } \ + i += sizeof(TYPE); \ + pg >>= sizeof(TYPE); \ + } while (i & 15); \ + } \ +} + +DO_MOVA_C(sme_mova_cz_b, uint8_t, H1) +DO_MOVA_C(sme_mova_cz_h, uint16_t, H1_2) +DO_MOVA_C(sme_mova_cz_s, uint32_t, H1_4) + +void HELPER(sme_mova_cz_d)(void *za, void *vn, void *vg, uint32_t desc) +{ + int i, oprsz = simd_oprsz(desc) / 8; + uint8_t *pg = vg; + uint64_t *n = vn; + uint64_t *a = za; + + for (i = 0; i < oprsz; i++) { + if (pg[H1(i)] & 1) { + a[tile_vslice_index(i)] = n[i]; + } + } +} + +void HELPER(sme_mova_cz_q)(void *za, void *vn, void *vg, uint32_t desc) +{ + int i, oprsz = simd_oprsz(desc) / 16; + uint16_t *pg = vg; + Int128 *n = vn; + Int128 *a = za; + + /* + * Int128 is used here simply to copy 16 bytes, and to simplify + * the address arithmetic. + */ + for (i = 0; i < oprsz; i++) { + if (pg[H2(i)] & 1) { + a[tile_vslice_index(i)] = n[i]; + } + } +} + +#undef DO_MOVA_C + +/* + * Move ZArray column to Zreg vector. + */ +#define DO_MOVA_Z(NAME, TYPE, H) \ +void HELPER(NAME)(void *vd, void *za, void *vg, uint32_t desc) \ +{ \ + int i, oprsz = simd_oprsz(desc); \ + for (i = 0; i < oprsz; ) { \ + uint16_t pg = *(uint16_t *)(vg + H1_2(i >> 3)); \ + do { \ + if (pg & 1) { \ + *(TYPE *)(vd + H(i)) = *(TYPE *)(za + tile_vslice_offset(i)); \ + } \ + i += sizeof(TYPE); \ + pg >>= sizeof(TYPE); \ + } while (i & 15); \ + } \ +} + +DO_MOVA_Z(sme_mova_zc_b, uint8_t, H1) +DO_MOVA_Z(sme_mova_zc_h, uint16_t, H1_2) +DO_MOVA_Z(sme_mova_zc_s, uint32_t, H1_4) + +void HELPER(sme_mova_zc_d)(void *vd, void *za, void *vg, uint32_t desc) +{ + int i, oprsz = simd_oprsz(desc) / 8; + uint8_t *pg = vg; + uint64_t *d = vd; + uint64_t *a = za; + + for (i = 0; i < oprsz; i++) { + if (pg[H1(i)] & 1) { + d[i] = a[tile_vslice_index(i)]; + } + } +} + +void HELPER(sme_mova_zc_q)(void *vd, void *za, void *vg, uint32_t desc) +{ + int i, oprsz = simd_oprsz(desc) / 16; + uint16_t *pg = vg; + Int128 *d = vd; + Int128 *a = za; + + /* + * Int128 is used here simply to copy 16 bytes, and to simplify + * the address arithmetic. + */ + for (i = 0; i < oprsz; i++, za += sizeof(ARMVectorReg)) { + if (pg[H2(i)] & 1) { + d[i] = a[tile_vslice_index(i)]; + } + } +} + +#undef DO_MOVA_Z diff --git a/target/arm/sve_helper.c b/target/arm/sve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve_helper.c +++ b/target/arm/sve_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(sve_sel_zpzz_d)(void *vd, void *vn, void *vm, } } +void HELPER(sve_sel_zpzz_q)(void *vd, void *vn, void *vm, + void *vg, uint32_t desc) +{ + intptr_t i, opr_sz = simd_oprsz(desc) / 16; + Int128 *d = vd, *n = vn, *m = vm; + uint16_t *pg = vg; + + for (i = 0; i < opr_sz; i += 1) { + d[i] = (pg[H2(i)] & 1 ? n : m)[i]; + } +} + /* Two operand comparison controlled by a predicate. * ??? It is very tempting to want to be able to expand this inline * with x86 instructions, e.g. diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ #include "decode-sme.c.inc" +/* + * Resolve tile.size[index] to a host pointer, where tile and index + * are always decoded together, dependent on the element size. + */ +static TCGv_ptr get_tile_rowcol(DisasContext *s, int esz, int rs, + int tile_index, bool vertical) +{ + int tile = tile_index >> (4 - esz); + int index = esz == MO_128 ? 0 : extract32(tile_index, 0, 4 - esz); + int pos, len, offset; + TCGv_i32 tmp; + TCGv_ptr addr; + + /* Compute the final index, which is Rs+imm. */ + tmp = tcg_temp_new_i32(); + tcg_gen_trunc_tl_i32(tmp, cpu_reg(s, rs)); + tcg_gen_addi_i32(tmp, tmp, index); + + /* Prepare a power-of-two modulo via extraction of @len bits. */ + len = ctz32(streaming_vec_reg_size(s)) - esz; + + if (vertical) { + /* + * Compute the byte offset of the index within the tile: + * (index % (svl / size)) * size + * = (index % (svl >> esz)) << esz + * Perform the power-of-two modulo via extraction of the low @len bits. + * Perform the multiply by shifting left by @pos bits. + * Perform these operations simultaneously via deposit into zero. + */ + pos = esz; + tcg_gen_deposit_z_i32(tmp, tmp, pos, len); + + /* + * For big-endian, adjust the indexed column byte offset within + * the uint64_t host words that make up env->zarray[]. + */ + if (HOST_BIG_ENDIAN && esz < MO_64) { + tcg_gen_xori_i32(tmp, tmp, 8 - (1 << esz)); + } + } else { + /* + * Compute the byte offset of the index within the tile: + * (index % (svl / size)) * (size * sizeof(row)) + * = (index % (svl >> esz)) << (esz + log2(sizeof(row))) + */ + pos = esz + ctz32(sizeof(ARMVectorReg)); + tcg_gen_deposit_z_i32(tmp, tmp, pos, len); + + /* Row slices are always aligned and need no endian adjustment. */ + } + + /* The tile byte offset within env->zarray is the row. */ + offset = tile * sizeof(ARMVectorReg); + + /* Include the byte offset of zarray to make this relative to env. */ + offset += offsetof(CPUARMState, zarray); + tcg_gen_addi_i32(tmp, tmp, offset); + + /* Add the byte offset to env to produce the final pointer. */ + addr = tcg_temp_new_ptr(); + tcg_gen_ext_i32_ptr(addr, tmp); + tcg_temp_free_i32(tmp); + tcg_gen_add_ptr(addr, addr, cpu_env); + + return addr; +} + static bool trans_ZERO(DisasContext *s, arg_ZERO *a) { if (!dc_isar_feature(aa64_sme, s)) { @@ -XXX,XX +XXX,XX @@ static bool trans_ZERO(DisasContext *s, arg_ZERO *a) } return true; } + +static bool trans_MOVA(DisasContext *s, arg_MOVA *a) +{ + static gen_helper_gvec_4 * const h_fns[5] = { + gen_helper_sve_sel_zpzz_b, gen_helper_sve_sel_zpzz_h, + gen_helper_sve_sel_zpzz_s, gen_helper_sve_sel_zpzz_d, + gen_helper_sve_sel_zpzz_q + }; + static gen_helper_gvec_3 * const cz_fns[5] = { + gen_helper_sme_mova_cz_b, gen_helper_sme_mova_cz_h, + gen_helper_sme_mova_cz_s, gen_helper_sme_mova_cz_d, + gen_helper_sme_mova_cz_q, + }; + static gen_helper_gvec_3 * const zc_fns[5] = { + gen_helper_sme_mova_zc_b, gen_helper_sme_mova_zc_h, + gen_helper_sme_mova_zc_s, gen_helper_sme_mova_zc_d, + gen_helper_sme_mova_zc_q, + }; + + TCGv_ptr t_za, t_zr, t_pg; + TCGv_i32 t_desc; + int svl; + + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (!sme_smza_enabled_check(s)) { + return true; + } + + t_za = get_tile_rowcol(s, a->esz, a->rs, a->za_imm, a->v); + t_zr = vec_full_reg_ptr(s, a->zr); + t_pg = pred_full_reg_ptr(s, a->pg); + + svl = streaming_vec_reg_size(s); + t_desc = tcg_constant_i32(simd_desc(svl, svl, 0)); + + if (a->v) { + /* Vertical slice -- use sme mova helpers. */ + if (a->to_vec) { + zc_fns[a->esz](t_zr, t_za, t_pg, t_desc); + } else { + cz_fns[a->esz](t_za, t_zr, t_pg, t_desc); + } + } else { + /* Horizontal slice -- reuse sve sel helpers. */ + if (a->to_vec) { + h_fns[a->esz](t_zr, t_za, t_zr, t_pg, t_desc); + } else { + h_fns[a->esz](t_za, t_zr, t_za, t_pg, t_desc); + } + } + + tcg_temp_free_ptr(t_za); + tcg_temp_free_ptr(t_zr); + tcg_temp_free_ptr(t_pg); + + return true; +} -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> We cannot reuse the SVE functions for LD[1-4] and ST[1-4], because those functions accept only a Zreg register number. For SME, we want to pass a pointer into ZA storage. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-21-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 82 +++++ target/arm/sme.decode | 9 + target/arm/sme_helper.c | 595 +++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 70 +++++ 4 files changed, 756 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(sme_mova_cz_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sme_mova_zc_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sme_mova_cz_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sme_mova_zc_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) + +DEF_HELPER_FLAGS_5(sme_ld1b_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1b_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1b_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1b_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_ld1h_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_ld1s_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_ld1d_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_ld1q_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1b_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1b_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1b_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1b_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1h_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1s_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1d_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1q_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ MOVA 11000000 esz:2 00001 0 v:1 .. pg:3 0 za_imm:4 zr:5 \ &mova to_vec=1 rs=%mova_rs MOVA 11000000 11 00001 1 v:1 .. pg:3 0 za_imm:4 zr:5 \ &mova to_vec=1 rs=%mova_rs esz=4 + +### SME Memory + +&ldst esz rs pg rn rm za_imm v:bool st:bool + +LDST1 1110000 0 esz:2 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ + &ldst rs=%mova_rs +LDST1 1110000 111 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ + &ldst esz=4 rs=%mova_rs diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "cpu.h" +#include "internals.h" #include "tcg/tcg-gvec-desc.h" #include "exec/helper-proto.h" +#include "exec/cpu_ldst.h" +#include "exec/exec-all.h" #include "qemu/int128.h" #include "vec_internal.h" +#include "sve_ldst_internal.h" /* ResetSVEState */ void arm_reset_sve_state(CPUARMState *env) @@ -XXX,XX +XXX,XX @@ void HELPER(sme_mova_zc_q)(void *vd, void *za, void *vg, uint32_t desc) } #undef DO_MOVA_Z + +/* + * Clear elements in a tile slice comprising len bytes. + */ + +typedef void ClearFn(void *ptr, size_t off, size_t len); + +static void clear_horizontal(void *ptr, size_t off, size_t len) +{ + memset(ptr + off, 0, len); +} + +static void clear_vertical_b(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; ++i) { + *(uint8_t *)(vptr + tile_vslice_offset(i + off)) = 0; + } +} + +static void clear_vertical_h(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; i += 2) { + *(uint16_t *)(vptr + tile_vslice_offset(i + off)) = 0; + } +} + +static void clear_vertical_s(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; i += 4) { + *(uint32_t *)(vptr + tile_vslice_offset(i + off)) = 0; + } +} + +static void clear_vertical_d(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; i += 8) { + *(uint64_t *)(vptr + tile_vslice_offset(i + off)) = 0; + } +} + +static void clear_vertical_q(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; i += 16) { + memset(vptr + tile_vslice_offset(i + off), 0, 16); + } +} + +/* + * Copy elements from an array into a tile slice comprising len bytes. + */ + +typedef void CopyFn(void *dst, const void *src, size_t len); + +static void copy_horizontal(void *dst, const void *src, size_t len) +{ + memcpy(dst, src, len); +} + +static void copy_vertical_b(void *vdst, const void *vsrc, size_t len) +{ + const uint8_t *src = vsrc; + uint8_t *dst = vdst; + size_t i; + + for (i = 0; i < len; ++i) { + dst[tile_vslice_index(i)] = src[i]; + } +} + +static void copy_vertical_h(void *vdst, const void *vsrc, size_t len) +{ + const uint16_t *src = vsrc; + uint16_t *dst = vdst; + size_t i; + + for (i = 0; i < len / 2; ++i) { + dst[tile_vslice_index(i)] = src[i]; + } +} + +static void copy_vertical_s(void *vdst, const void *vsrc, size_t len) +{ + const uint32_t *src = vsrc; + uint32_t *dst = vdst; + size_t i; + + for (i = 0; i < len / 4; ++i) { + dst[tile_vslice_index(i)] = src[i]; + } +} + +static void copy_vertical_d(void *vdst, const void *vsrc, size_t len) +{ + const uint64_t *src = vsrc; + uint64_t *dst = vdst; + size_t i; + + for (i = 0; i < len / 8; ++i) { + dst[tile_vslice_index(i)] = src[i]; + } +} + +static void copy_vertical_q(void *vdst, const void *vsrc, size_t len) +{ + for (size_t i = 0; i < len; i += 16) { + memcpy(vdst + tile_vslice_offset(i), vsrc + i, 16); + } +} + +/* + * Host and TLB primitives for vertical tile slice addressing. + */ + +#define DO_LD(NAME, TYPE, HOST, TLB) \ +static inline void sme_##NAME##_v_host(void *za, intptr_t off, void *host) \ +{ \ + TYPE val = HOST(host); \ + *(TYPE *)(za + tile_vslice_offset(off)) = val; \ +} \ +static inline void sme_##NAME##_v_tlb(CPUARMState *env, void *za, \ + intptr_t off, target_ulong addr, uintptr_t ra) \ +{ \ + TYPE val = TLB(env, useronly_clean_ptr(addr), ra); \ + *(TYPE *)(za + tile_vslice_offset(off)) = val; \ +} + +#define DO_ST(NAME, TYPE, HOST, TLB) \ +static inline void sme_##NAME##_v_host(void *za, intptr_t off, void *host) \ +{ \ + TYPE val = *(TYPE *)(za + tile_vslice_offset(off)); \ + HOST(host, val); \ +} \ +static inline void sme_##NAME##_v_tlb(CPUARMState *env, void *za, \ + intptr_t off, target_ulong addr, uintptr_t ra) \ +{ \ + TYPE val = *(TYPE *)(za + tile_vslice_offset(off)); \ + TLB(env, useronly_clean_ptr(addr), val, ra); \ +} + +/* + * The ARMVectorReg elements are stored in host-endian 64-bit units. + * For 128-bit quantities, the sequence defined by the Elem[] pseudocode + * corresponds to storing the two 64-bit pieces in little-endian order. + */ +#define DO_LDQ(HNAME, VNAME, BE, HOST, TLB) \ +static inline void HNAME##_host(void *za, intptr_t off, void *host) \ +{ \ + uint64_t val0 = HOST(host), val1 = HOST(host + 8); \ + uint64_t *ptr = za + off; \ + ptr[0] = BE ? val1 : val0, ptr[1] = BE ? val0 : val1; \ +} \ +static inline void VNAME##_v_host(void *za, intptr_t off, void *host) \ +{ \ + HNAME##_host(za, tile_vslice_offset(off), host); \ +} \ +static inline void HNAME##_tlb(CPUARMState *env, void *za, intptr_t off, \ + target_ulong addr, uintptr_t ra) \ +{ \ + uint64_t val0 = TLB(env, useronly_clean_ptr(addr), ra); \ + uint64_t val1 = TLB(env, useronly_clean_ptr(addr + 8), ra); \ + uint64_t *ptr = za + off; \ + ptr[0] = BE ? val1 : val0, ptr[1] = BE ? val0 : val1; \ +} \ +static inline void VNAME##_v_tlb(CPUARMState *env, void *za, intptr_t off, \ + target_ulong addr, uintptr_t ra) \ +{ \ + HNAME##_tlb(env, za, tile_vslice_offset(off), addr, ra); \ +} + +#define DO_STQ(HNAME, VNAME, BE, HOST, TLB) \ +static inline void HNAME##_host(void *za, intptr_t off, void *host) \ +{ \ + uint64_t *ptr = za + off; \ + HOST(host, ptr[BE]); \ + HOST(host + 1, ptr[!BE]); \ +} \ +static inline void VNAME##_v_host(void *za, intptr_t off, void *host) \ +{ \ + HNAME##_host(za, tile_vslice_offset(off), host); \ +} \ +static inline void HNAME##_tlb(CPUARMState *env, void *za, intptr_t off, \ + target_ulong addr, uintptr_t ra) \ +{ \ + uint64_t *ptr = za + off; \ + TLB(env, useronly_clean_ptr(addr), ptr[BE], ra); \ + TLB(env, useronly_clean_ptr(addr + 8), ptr[!BE], ra); \ +} \ +static inline void VNAME##_v_tlb(CPUARMState *env, void *za, intptr_t off, \ + target_ulong addr, uintptr_t ra) \ +{ \ + HNAME##_tlb(env, za, tile_vslice_offset(off), addr, ra); \ +} + +DO_LD(ld1b, uint8_t, ldub_p, cpu_ldub_data_ra) +DO_LD(ld1h_be, uint16_t, lduw_be_p, cpu_lduw_be_data_ra) +DO_LD(ld1h_le, uint16_t, lduw_le_p, cpu_lduw_le_data_ra) +DO_LD(ld1s_be, uint32_t, ldl_be_p, cpu_ldl_be_data_ra) +DO_LD(ld1s_le, uint32_t, ldl_le_p, cpu_ldl_le_data_ra) +DO_LD(ld1d_be, uint64_t, ldq_be_p, cpu_ldq_be_data_ra) +DO_LD(ld1d_le, uint64_t, ldq_le_p, cpu_ldq_le_data_ra) + +DO_LDQ(sve_ld1qq_be, sme_ld1q_be, 1, ldq_be_p, cpu_ldq_be_data_ra) +DO_LDQ(sve_ld1qq_le, sme_ld1q_le, 0, ldq_le_p, cpu_ldq_le_data_ra) + +DO_ST(st1b, uint8_t, stb_p, cpu_stb_data_ra) +DO_ST(st1h_be, uint16_t, stw_be_p, cpu_stw_be_data_ra) +DO_ST(st1h_le, uint16_t, stw_le_p, cpu_stw_le_data_ra) +DO_ST(st1s_be, uint32_t, stl_be_p, cpu_stl_be_data_ra) +DO_ST(st1s_le, uint32_t, stl_le_p, cpu_stl_le_data_ra) +DO_ST(st1d_be, uint64_t, stq_be_p, cpu_stq_be_data_ra) +DO_ST(st1d_le, uint64_t, stq_le_p, cpu_stq_le_data_ra) + +DO_STQ(sve_st1qq_be, sme_st1q_be, 1, stq_be_p, cpu_stq_be_data_ra) +DO_STQ(sve_st1qq_le, sme_st1q_le, 0, stq_le_p, cpu_stq_le_data_ra) + +#undef DO_LD +#undef DO_ST +#undef DO_LDQ +#undef DO_STQ + +/* + * Common helper for all contiguous predicated loads. + */ + +static inline QEMU_ALWAYS_INLINE +void sme_ld1(CPUARMState *env, void *za, uint64_t *vg, + const target_ulong addr, uint32_t desc, const uintptr_t ra, + const int esz, uint32_t mtedesc, bool vertical, + sve_ldst1_host_fn *host_fn, + sve_ldst1_tlb_fn *tlb_fn, + ClearFn *clr_fn, + CopyFn *cpy_fn) +{ + const intptr_t reg_max = simd_oprsz(desc); + const intptr_t esize = 1 << esz; + intptr_t reg_off, reg_last; + SVEContLdSt info; + void *host; + int flags; + + /* Find the active elements. */ + if (!sve_cont_ldst_elements(&info, addr, vg, reg_max, esz, esize)) { + /* The entire predicate was false; no load occurs. */ + clr_fn(za, 0, reg_max); + return; + } + + /* Probe the page(s). Exit with exception for any invalid page. */ + sve_cont_ldst_pages(&info, FAULT_ALL, env, addr, MMU_DATA_LOAD, ra); + + /* Handle watchpoints for all active elements. */ + sve_cont_ldst_watchpoints(&info, env, vg, addr, esize, esize, + BP_MEM_READ, ra); + + /* + * Handle mte checks for all active elements. + * Since TBI must be set for MTE, !mtedesc => !mte_active. + */ + if (mtedesc) { + sve_cont_ldst_mte_check(&info, env, vg, addr, esize, esize, + mtedesc, ra); + } + + flags = info.page[0].flags | info.page[1].flags; + if (unlikely(flags != 0)) { +#ifdef CONFIG_USER_ONLY + g_assert_not_reached(); +#else + /* + * At least one page includes MMIO. + * Any bus operation can fail with cpu_transaction_failed, + * which for ARM will raise SyncExternal. Perform the load + * into scratch memory to preserve register state until the end. + */ + ARMVectorReg scratch = { }; + + reg_off = info.reg_off_first[0]; + reg_last = info.reg_off_last[1]; + if (reg_last < 0) { + reg_last = info.reg_off_split; + if (reg_last < 0) { + reg_last = info.reg_off_last[0]; + } + } + + do { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + tlb_fn(env, &scratch, reg_off, addr + reg_off, ra); + } + reg_off += esize; + } while (reg_off & 63); + } while (reg_off <= reg_last); + + cpy_fn(za, &scratch, reg_max); + return; +#endif + } + + /* The entire operation is in RAM, on valid pages. */ + + reg_off = info.reg_off_first[0]; + reg_last = info.reg_off_last[0]; + host = info.page[0].host; + + if (!vertical) { + memset(za, 0, reg_max); + } else if (reg_off) { + clr_fn(za, 0, reg_off); + } + + while (reg_off <= reg_last) { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + host_fn(za, reg_off, host + reg_off); + } else if (vertical) { + clr_fn(za, reg_off, esize); + } + reg_off += esize; + } while (reg_off <= reg_last && (reg_off & 63)); + } + + /* + * Use the slow path to manage the cross-page misalignment. + * But we know this is RAM and cannot trap. + */ + reg_off = info.reg_off_split; + if (unlikely(reg_off >= 0)) { + tlb_fn(env, za, reg_off, addr + reg_off, ra); + } + + reg_off = info.reg_off_first[1]; + if (unlikely(reg_off >= 0)) { + reg_last = info.reg_off_last[1]; + host = info.page[1].host; + + do { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + host_fn(za, reg_off, host + reg_off); + } else if (vertical) { + clr_fn(za, reg_off, esize); + } + reg_off += esize; + } while (reg_off & 63); + } while (reg_off <= reg_last); + } +} + +static inline QEMU_ALWAYS_INLINE +void sme_ld1_mte(CPUARMState *env, void *za, uint64_t *vg, + target_ulong addr, uint32_t desc, uintptr_t ra, + const int esz, bool vertical, + sve_ldst1_host_fn *host_fn, + sve_ldst1_tlb_fn *tlb_fn, + ClearFn *clr_fn, + CopyFn *cpy_fn) +{ + uint32_t mtedesc = desc >> (SIMD_DATA_SHIFT + SVE_MTEDESC_SHIFT); + int bit55 = extract64(addr, 55, 1); + + /* Remove mtedesc from the normal sve descriptor. */ + desc = extract32(desc, 0, SIMD_DATA_SHIFT + SVE_MTEDESC_SHIFT); + + /* Perform gross MTE suppression early. */ + if (!tbi_check(desc, bit55) || + tcma_check(desc, bit55, allocation_tag_from_addr(addr))) { + mtedesc = 0; + } + + sme_ld1(env, za, vg, addr, desc, ra, esz, mtedesc, vertical, + host_fn, tlb_fn, clr_fn, cpy_fn); +} + +#define DO_LD(L, END, ESZ) \ +void HELPER(sme_ld1##L##END##_h)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_ld1(env, za, vg, addr, desc, GETPC(), ESZ, 0, false, \ + sve_ld1##L##L##END##_host, sve_ld1##L##L##END##_tlb, \ + clear_horizontal, copy_horizontal); \ +} \ +void HELPER(sme_ld1##L##END##_v)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_ld1(env, za, vg, addr, desc, GETPC(), ESZ, 0, true, \ + sme_ld1##L##END##_v_host, sme_ld1##L##END##_v_tlb, \ + clear_vertical_##L, copy_vertical_##L); \ +} \ +void HELPER(sme_ld1##L##END##_h_mte)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_ld1_mte(env, za, vg, addr, desc, GETPC(), ESZ, false, \ + sve_ld1##L##L##END##_host, sve_ld1##L##L##END##_tlb, \ + clear_horizontal, copy_horizontal); \ +} \ +void HELPER(sme_ld1##L##END##_v_mte)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_ld1_mte(env, za, vg, addr, desc, GETPC(), ESZ, true, \ + sme_ld1##L##END##_v_host, sme_ld1##L##END##_v_tlb, \ + clear_vertical_##L, copy_vertical_##L); \ +} + +DO_LD(b, , MO_8) +DO_LD(h, _be, MO_16) +DO_LD(h, _le, MO_16) +DO_LD(s, _be, MO_32) +DO_LD(s, _le, MO_32) +DO_LD(d, _be, MO_64) +DO_LD(d, _le, MO_64) +DO_LD(q, _be, MO_128) +DO_LD(q, _le, MO_128) + +#undef DO_LD + +/* + * Common helper for all contiguous predicated stores. + */ + +static inline QEMU_ALWAYS_INLINE +void sme_st1(CPUARMState *env, void *za, uint64_t *vg, + const target_ulong addr, uint32_t desc, const uintptr_t ra, + const int esz, uint32_t mtedesc, bool vertical, + sve_ldst1_host_fn *host_fn, + sve_ldst1_tlb_fn *tlb_fn) +{ + const intptr_t reg_max = simd_oprsz(desc); + const intptr_t esize = 1 << esz; + intptr_t reg_off, reg_last; + SVEContLdSt info; + void *host; + int flags; + + /* Find the active elements. */ + if (!sve_cont_ldst_elements(&info, addr, vg, reg_max, esz, esize)) { + /* The entire predicate was false; no store occurs. */ + return; + } + + /* Probe the page(s). Exit with exception for any invalid page. */ + sve_cont_ldst_pages(&info, FAULT_ALL, env, addr, MMU_DATA_STORE, ra); + + /* Handle watchpoints for all active elements. */ + sve_cont_ldst_watchpoints(&info, env, vg, addr, esize, esize, + BP_MEM_WRITE, ra); + + /* + * Handle mte checks for all active elements. + * Since TBI must be set for MTE, !mtedesc => !mte_active. + */ + if (mtedesc) { + sve_cont_ldst_mte_check(&info, env, vg, addr, esize, esize, + mtedesc, ra); + } + + flags = info.page[0].flags | info.page[1].flags; + if (unlikely(flags != 0)) { +#ifdef CONFIG_USER_ONLY + g_assert_not_reached(); +#else + /* + * At least one page includes MMIO. + * Any bus operation can fail with cpu_transaction_failed, + * which for ARM will raise SyncExternal. We cannot avoid + * this fault and will leave with the store incomplete. + */ + reg_off = info.reg_off_first[0]; + reg_last = info.reg_off_last[1]; + if (reg_last < 0) { + reg_last = info.reg_off_split; + if (reg_last < 0) { + reg_last = info.reg_off_last[0]; + } + } + + do { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + tlb_fn(env, za, reg_off, addr + reg_off, ra); + } + reg_off += esize; + } while (reg_off & 63); + } while (reg_off <= reg_last); + return; +#endif + } + + reg_off = info.reg_off_first[0]; + reg_last = info.reg_off_last[0]; + host = info.page[0].host; + + while (reg_off <= reg_last) { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + host_fn(za, reg_off, host + reg_off); + } + reg_off += 1 << esz; + } while (reg_off <= reg_last && (reg_off & 63)); + } + + /* + * Use the slow path to manage the cross-page misalignment. + * But we know this is RAM and cannot trap. + */ + reg_off = info.reg_off_split; + if (unlikely(reg_off >= 0)) { + tlb_fn(env, za, reg_off, addr + reg_off, ra); + } + + reg_off = info.reg_off_first[1]; + if (unlikely(reg_off >= 0)) { + reg_last = info.reg_off_last[1]; + host = info.page[1].host; + + do { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + host_fn(za, reg_off, host + reg_off); + } + reg_off += 1 << esz; + } while (reg_off & 63); + } while (reg_off <= reg_last); + } +} + +static inline QEMU_ALWAYS_INLINE +void sme_st1_mte(CPUARMState *env, void *za, uint64_t *vg, target_ulong addr, + uint32_t desc, uintptr_t ra, int esz, bool vertical, + sve_ldst1_host_fn *host_fn, + sve_ldst1_tlb_fn *tlb_fn) +{ + uint32_t mtedesc = desc >> (SIMD_DATA_SHIFT + SVE_MTEDESC_SHIFT); + int bit55 = extract64(addr, 55, 1); + + /* Remove mtedesc from the normal sve descriptor. */ + desc = extract32(desc, 0, SIMD_DATA_SHIFT + SVE_MTEDESC_SHIFT); + + /* Perform gross MTE suppression early. */ + if (!tbi_check(desc, bit55) || + tcma_check(desc, bit55, allocation_tag_from_addr(addr))) { + mtedesc = 0; + } + + sme_st1(env, za, vg, addr, desc, ra, esz, mtedesc, + vertical, host_fn, tlb_fn); +} + +#define DO_ST(L, END, ESZ) \ +void HELPER(sme_st1##L##END##_h)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_st1(env, za, vg, addr, desc, GETPC(), ESZ, 0, false, \ + sve_st1##L##L##END##_host, sve_st1##L##L##END##_tlb); \ +} \ +void HELPER(sme_st1##L##END##_v)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_st1(env, za, vg, addr, desc, GETPC(), ESZ, 0, true, \ + sme_st1##L##END##_v_host, sme_st1##L##END##_v_tlb); \ +} \ +void HELPER(sme_st1##L##END##_h_mte)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_st1_mte(env, za, vg, addr, desc, GETPC(), ESZ, false, \ + sve_st1##L##L##END##_host, sve_st1##L##L##END##_tlb); \ +} \ +void HELPER(sme_st1##L##END##_v_mte)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_st1_mte(env, za, vg, addr, desc, GETPC(), ESZ, true, \ + sme_st1##L##END##_v_host, sme_st1##L##END##_v_tlb); \ +} + +DO_ST(b, , MO_8) +DO_ST(h, _be, MO_16) +DO_ST(h, _le, MO_16) +DO_ST(s, _be, MO_32) +DO_ST(s, _le, MO_32) +DO_ST(d, _be, MO_64) +DO_ST(d, _le, MO_64) +DO_ST(q, _be, MO_128) +DO_ST(q, _le, MO_128) + +#undef DO_ST diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ static bool trans_MOVA(DisasContext *s, arg_MOVA *a) return true; } + +static bool trans_LDST1(DisasContext *s, arg_LDST1 *a) +{ + typedef void GenLdSt1(TCGv_env, TCGv_ptr, TCGv_ptr, TCGv, TCGv_i32); + + /* + * Indexed by [esz][be][v][mte][st], which is (except for load/store) + * also the order in which the elements appear in the function names, + * and so how we must concatenate the pieces. + */ + +#define FN_LS(F) { gen_helper_sme_ld1##F, gen_helper_sme_st1##F } +#define FN_MTE(F) { FN_LS(F), FN_LS(F##_mte) } +#define FN_HV(F) { FN_MTE(F##_h), FN_MTE(F##_v) } +#define FN_END(L, B) { FN_HV(L), FN_HV(B) } + + static GenLdSt1 * const fns[5][2][2][2][2] = { + FN_END(b, b), + FN_END(h_le, h_be), + FN_END(s_le, s_be), + FN_END(d_le, d_be), + FN_END(q_le, q_be), + }; + +#undef FN_LS +#undef FN_MTE +#undef FN_HV +#undef FN_END + + TCGv_ptr t_za, t_pg; + TCGv_i64 addr; + int svl, desc = 0; + bool be = s->be_data == MO_BE; + bool mte = s->mte_active[0]; + + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (!sme_smza_enabled_check(s)) { + return true; + } + + t_za = get_tile_rowcol(s, a->esz, a->rs, a->za_imm, a->v); + t_pg = pred_full_reg_ptr(s, a->pg); + addr = tcg_temp_new_i64(); + + tcg_gen_shli_i64(addr, cpu_reg(s, a->rm), a->esz); + tcg_gen_add_i64(addr, addr, cpu_reg_sp(s, a->rn)); + + if (mte) { + desc = FIELD_DP32(desc, MTEDESC, MIDX, get_mem_index(s)); + desc = FIELD_DP32(desc, MTEDESC, TBI, s->tbid); + desc = FIELD_DP32(desc, MTEDESC, TCMA, s->tcma); + desc = FIELD_DP32(desc, MTEDESC, WRITE, a->st); + desc = FIELD_DP32(desc, MTEDESC, SIZEM1, (1 << a->esz) - 1); + desc <<= SVE_MTEDESC_SHIFT; + } else { + addr = clean_data_tbi(s, addr); + } + svl = streaming_vec_reg_size(s); + desc = simd_desc(svl, svl, desc); + + fns[a->esz][be][a->v][mte][a->st](cpu_env, t_za, t_pg, addr, + tcg_constant_i32(desc)); + + tcg_temp_free_ptr(t_za); + tcg_temp_free_ptr(t_pg); + tcg_temp_free_i64(addr); + return true; +} -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Add a TCGv_ptr base argument, which will be cpu_env for SVE. We will reuse this for SME save and restore array insns. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-22-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.h | 3 +++ target/arm/translate-sve.c | 48 ++++++++++++++++++++++++++++---------- 2 files changed, 39 insertions(+), 12 deletions(-) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ void gen_gvec_xar(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, uint32_t rm_ofs, int64_t shift, uint32_t opr_sz, uint32_t max_sz); +void gen_sve_ldr(DisasContext *s, TCGv_ptr, int vofs, int len, int rn, int imm); +void gen_sve_str(DisasContext *s, TCGv_ptr, int vofs, int len, int rn, int imm); + #endif /* TARGET_ARM_TRANSLATE_A64_H */ diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(UCVTF_dd, aa64_sve, gen_gvec_fpst_arg_zpz, * The load should begin at the address Rn + IMM. */ -static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) +void gen_sve_ldr(DisasContext *s, TCGv_ptr base, int vofs, + int len, int rn, int imm) { int len_align = QEMU_ALIGN_DOWN(len, 8); int len_remain = len % 8; @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) t0 = tcg_temp_new_i64(); for (i = 0; i < len_align; i += 8) { tcg_gen_qemu_ld_i64(t0, clean_addr, midx, MO_LEUQ); - tcg_gen_st_i64(t0, cpu_env, vofs + i); + tcg_gen_st_i64(t0, base, vofs + i); tcg_gen_addi_i64(clean_addr, clean_addr, 8); } tcg_temp_free_i64(t0); @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) clean_addr = new_tmp_a64_local(s); tcg_gen_mov_i64(clean_addr, t0); + if (base != cpu_env) { + TCGv_ptr b = tcg_temp_local_new_ptr(); + tcg_gen_mov_ptr(b, base); + base = b; + } + gen_set_label(loop); t0 = tcg_temp_new_i64(); @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) tcg_gen_addi_i64(clean_addr, clean_addr, 8); tp = tcg_temp_new_ptr(); - tcg_gen_add_ptr(tp, cpu_env, i); + tcg_gen_add_ptr(tp, base, i); tcg_gen_addi_ptr(i, i, 8); tcg_gen_st_i64(t0, tp, vofs); tcg_temp_free_ptr(tp); @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) tcg_gen_brcondi_ptr(TCG_COND_LTU, i, len_align, loop); tcg_temp_free_ptr(i); + + if (base != cpu_env) { + tcg_temp_free_ptr(base); + assert(len_remain == 0); + } } /* @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) default: g_assert_not_reached(); } - tcg_gen_st_i64(t0, cpu_env, vofs + len_align); + tcg_gen_st_i64(t0, base, vofs + len_align); tcg_temp_free_i64(t0); } } /* Similarly for stores. */ -static void do_str(DisasContext *s, uint32_t vofs, int len, int rn, int imm) +void gen_sve_str(DisasContext *s, TCGv_ptr base, int vofs, + int len, int rn, int imm) { int len_align = QEMU_ALIGN_DOWN(len, 8); int len_remain = len % 8; @@ -XXX,XX +XXX,XX @@ static void do_str(DisasContext *s, uint32_t vofs, int len, int rn, int imm) t0 = tcg_temp_new_i64(); for (i = 0; i < len_align; i += 8) { - tcg_gen_ld_i64(t0, cpu_env, vofs + i); + tcg_gen_ld_i64(t0, base, vofs + i); tcg_gen_qemu_st_i64(t0, clean_addr, midx, MO_LEUQ); tcg_gen_addi_i64(clean_addr, clean_addr, 8); } @@ -XXX,XX +XXX,XX @@ static void do_str(DisasContext *s, uint32_t vofs, int len, int rn, int imm) clean_addr = new_tmp_a64_local(s); tcg_gen_mov_i64(clean_addr, t0); + if (base != cpu_env) { + TCGv_ptr b = tcg_temp_local_new_ptr(); + tcg_gen_mov_ptr(b, base); + base = b; + } + gen_set_label(loop); t0 = tcg_temp_new_i64(); tp = tcg_temp_new_ptr(); - tcg_gen_add_ptr(tp, cpu_env, i); + tcg_gen_add_ptr(tp, base, i); tcg_gen_ld_i64(t0, tp, vofs); tcg_gen_addi_ptr(i, i, 8); tcg_temp_free_ptr(tp); @@ -XXX,XX +XXX,XX @@ static void do_str(DisasContext *s, uint32_t vofs, int len, int rn, int imm) tcg_gen_brcondi_ptr(TCG_COND_LTU, i, len_align, loop); tcg_temp_free_ptr(i); + + if (base != cpu_env) { + tcg_temp_free_ptr(base); + assert(len_remain == 0); + } } /* Predicate register stores can be any multiple of 2. */ if (len_remain) { t0 = tcg_temp_new_i64(); - tcg_gen_ld_i64(t0, cpu_env, vofs + len_align); + tcg_gen_ld_i64(t0, base, vofs + len_align); switch (len_remain) { case 2: @@ -XXX,XX +XXX,XX @@ static bool trans_LDR_zri(DisasContext *s, arg_rri *a) if (sve_access_check(s)) { int size = vec_full_reg_size(s); int off = vec_full_reg_offset(s, a->rd); - do_ldr(s, off, size, a->rn, a->imm * size); + gen_sve_ldr(s, cpu_env, off, size, a->rn, a->imm * size); } return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_LDR_pri(DisasContext *s, arg_rri *a) if (sve_access_check(s)) { int size = pred_full_reg_size(s); int off = pred_full_reg_offset(s, a->rd); - do_ldr(s, off, size, a->rn, a->imm * size); + gen_sve_ldr(s, cpu_env, off, size, a->rn, a->imm * size); } return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_STR_zri(DisasContext *s, arg_rri *a) if (sve_access_check(s)) { int size = vec_full_reg_size(s); int off = vec_full_reg_offset(s, a->rd); - do_str(s, off, size, a->rn, a->imm * size); + gen_sve_str(s, cpu_env, off, size, a->rn, a->imm * size); } return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_STR_pri(DisasContext *s, arg_rri *a) if (sve_access_check(s)) { int size = pred_full_reg_size(s); int off = pred_full_reg_offset(s, a->rd); - do_str(s, off, size, a->rn, a->imm * size); + gen_sve_str(s, cpu_env, off, size, a->rn, a->imm * size); } return true; } -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> We can reuse the SVE functions for LDR and STR, passing in the base of the ZA vector and a zero offset. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-23-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme.decode | 7 +++++++ target/arm/translate-sme.c | 24 ++++++++++++++++++++++++ 2 files changed, 31 insertions(+) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ LDST1 1110000 0 esz:2 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ &ldst rs=%mova_rs LDST1 1110000 111 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ &ldst esz=4 rs=%mova_rs + +&ldstr rv rn imm +@ldstr ....... ... . ...... .. ... rn:5 . imm:4 \ + &ldstr rv=%mova_rs + +LDR 1110000 100 0 000000 .. 000 ..... 0 .... @ldstr +STR 1110000 100 1 000000 .. 000 ..... 0 .... @ldstr diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ static bool trans_LDST1(DisasContext *s, arg_LDST1 *a) tcg_temp_free_i64(addr); return true; } + +typedef void GenLdStR(DisasContext *, TCGv_ptr, int, int, int, int); + +static bool do_ldst_r(DisasContext *s, arg_ldstr *a, GenLdStR *fn) +{ + int svl = streaming_vec_reg_size(s); + int imm = a->imm; + TCGv_ptr base; + + if (!sme_za_enabled_check(s)) { + return true; + } + + /* ZA[n] equates to ZA0H.B[n]. */ + base = get_tile_rowcol(s, MO_8, a->rv, imm, false); + + fn(s, base, 0, svl, a->rn, imm * svl); + + tcg_temp_free_ptr(base); + return true; +} + +TRANS_FEAT(LDR, aa64_sme, do_ldst_r, a, gen_sve_ldr) +TRANS_FEAT(STR, aa64_sme, do_ldst_r, a, gen_sve_str) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-24-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 5 +++ target/arm/sme.decode | 11 +++++ target/arm/sme_helper.c | 90 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 31 +++++++++++++ 4 files changed, 137 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_5(sme_st1q_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i DEF_HELPER_FLAGS_5(sme_st1q_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) DEF_HELPER_FLAGS_5(sme_st1q_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) DEF_HELPER_FLAGS_5(sme_st1q_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_addha_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(sme_addva_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(sme_addha_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(sme_addva_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ LDST1 1110000 111 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ LDR 1110000 100 0 000000 .. 000 ..... 0 .... @ldstr STR 1110000 100 1 000000 .. 000 ..... 0 .... @ldstr + +### SME Add Vector to Array + +&adda zad zn pm pn +@adda_32 ........ .. ..... . pm:3 pn:3 zn:5 ... zad:2 &adda +@adda_64 ........ .. ..... . pm:3 pn:3 zn:5 .. zad:3 &adda + +ADDHA_s 11000000 10 01000 0 ... ... ..... 000 .. @adda_32 +ADDVA_s 11000000 10 01000 1 ... ... ..... 000 .. @adda_32 +ADDHA_d 11000000 11 01000 0 ... ... ..... 00 ... @adda_64 +ADDVA_d 11000000 11 01000 1 ... ... ..... 00 ... @adda_64 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ DO_ST(q, _be, MO_128) DO_ST(q, _le, MO_128) #undef DO_ST + +void HELPER(sme_addha_s)(void *vzda, void *vzn, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 4; + uint64_t *pn = vpn, *pm = vpm; + uint32_t *zda = vzda, *zn = vzn; + + for (row = 0; row < oprsz; ) { + uint64_t pa = pn[row >> 4]; + do { + if (pa & 1) { + for (col = 0; col < oprsz; ) { + uint64_t pb = pm[col >> 4]; + do { + if (pb & 1) { + zda[tile_vslice_index(row) + H4(col)] += zn[H4(col)]; + } + pb >>= 4; + } while (++col & 15); + } + } + pa >>= 4; + } while (++row & 15); + } +} + +void HELPER(sme_addha_d)(void *vzda, void *vzn, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 8; + uint8_t *pn = vpn, *pm = vpm; + uint64_t *zda = vzda, *zn = vzn; + + for (row = 0; row < oprsz; ++row) { + if (pn[H1(row)] & 1) { + for (col = 0; col < oprsz; ++col) { + if (pm[H1(col)] & 1) { + zda[tile_vslice_index(row) + col] += zn[col]; + } + } + } + } +} + +void HELPER(sme_addva_s)(void *vzda, void *vzn, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 4; + uint64_t *pn = vpn, *pm = vpm; + uint32_t *zda = vzda, *zn = vzn; + + for (row = 0; row < oprsz; ) { + uint64_t pa = pn[row >> 4]; + do { + if (pa & 1) { + uint32_t zn_row = zn[H4(row)]; + for (col = 0; col < oprsz; ) { + uint64_t pb = pm[col >> 4]; + do { + if (pb & 1) { + zda[tile_vslice_index(row) + H4(col)] += zn_row; + } + pb >>= 4; + } while (++col & 15); + } + } + pa >>= 4; + } while (++row & 15); + } +} + +void HELPER(sme_addva_d)(void *vzda, void *vzn, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 8; + uint8_t *pn = vpn, *pm = vpm; + uint64_t *zda = vzda, *zn = vzn; + + for (row = 0; row < oprsz; ++row) { + if (pn[H1(row)] & 1) { + uint64_t zn_row = zn[row]; + for (col = 0; col < oprsz; ++col) { + if (pm[H1(col)] & 1) { + zda[tile_vslice_index(row) + col] += zn_row; + } + } + } + } +} diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ static bool do_ldst_r(DisasContext *s, arg_ldstr *a, GenLdStR *fn) TRANS_FEAT(LDR, aa64_sme, do_ldst_r, a, gen_sve_ldr) TRANS_FEAT(STR, aa64_sme, do_ldst_r, a, gen_sve_str) + +static bool do_adda(DisasContext *s, arg_adda *a, MemOp esz, + gen_helper_gvec_4 *fn) +{ + int svl = streaming_vec_reg_size(s); + uint32_t desc = simd_desc(svl, svl, 0); + TCGv_ptr za, zn, pn, pm; + + if (!sme_smza_enabled_check(s)) { + return true; + } + + /* Sum XZR+zad to find ZAd. */ + za = get_tile_rowcol(s, esz, 31, a->zad, false); + zn = vec_full_reg_ptr(s, a->zn); + pn = pred_full_reg_ptr(s, a->pn); + pm = pred_full_reg_ptr(s, a->pm); + + fn(za, zn, pn, pm, tcg_constant_i32(desc)); + + tcg_temp_free_ptr(za); + tcg_temp_free_ptr(zn); + tcg_temp_free_ptr(pn); + tcg_temp_free_ptr(pm); + return true; +} + +TRANS_FEAT(ADDHA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addha_s) +TRANS_FEAT(ADDVA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addva_s) +TRANS_FEAT(ADDHA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addha_d) +TRANS_FEAT(ADDVA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addva_d) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-25-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 5 +++ target/arm/sme.decode | 9 +++++ target/arm/sme_helper.c | 69 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 32 ++++++++++++++++++ 4 files changed, 115 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_5(sme_addha_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addva_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addha_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addva_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) + +DEF_HELPER_FLAGS_7(sme_fmopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_7(sme_fmopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ ADDHA_s 11000000 10 01000 0 ... ... ..... 000 .. @adda_32 ADDVA_s 11000000 10 01000 1 ... ... ..... 000 .. @adda_32 ADDHA_d 11000000 11 01000 0 ... ... ..... 00 ... @adda_64 ADDVA_d 11000000 11 01000 1 ... ... ..... 00 ... @adda_64 + +### SME Outer Product + +&op zad zn zm pm pn sub:bool +@op_32 ........ ... zm:5 pm:3 pn:3 zn:5 sub:1 .. zad:2 &op +@op_64 ........ ... zm:5 pm:3 pn:3 zn:5 sub:1 . zad:3 &op + +FMOPA_s 10000000 100 ..... ... ... ..... . 00 .. @op_32 +FMOPA_d 10000000 110 ..... ... ... ..... . 0 ... @op_64 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ #include "exec/cpu_ldst.h" #include "exec/exec-all.h" #include "qemu/int128.h" +#include "fpu/softfloat.h" #include "vec_internal.h" #include "sve_ldst_internal.h" @@ -XXX,XX +XXX,XX @@ void HELPER(sme_addva_d)(void *vzda, void *vzn, void *vpn, } } } + +void HELPER(sme_fmopa_s)(void *vza, void *vzn, void *vzm, void *vpn, + void *vpm, void *vst, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_maxsz(desc); + uint32_t neg = simd_data(desc) << 31; + uint16_t *pn = vpn, *pm = vpm; + float_status fpst; + + /* + * Make a copy of float_status because this operation does not + * update the cumulative fp exception status. It also produces + * default nans. + */ + fpst = *(float_status *)vst; + set_default_nan_mode(true, &fpst); + + for (row = 0; row < oprsz; ) { + uint16_t pa = pn[H2(row >> 4)]; + do { + if (pa & 1) { + void *vza_row = vza + tile_vslice_offset(row); + uint32_t n = *(uint32_t *)(vzn + H1_4(row)) ^ neg; + + for (col = 0; col < oprsz; ) { + uint16_t pb = pm[H2(col >> 4)]; + do { + if (pb & 1) { + uint32_t *a = vza_row + H1_4(col); + uint32_t *m = vzm + H1_4(col); + *a = float32_muladd(n, *m, *a, 0, vst); + } + col += 4; + pb >>= 4; + } while (col & 15); + } + } + row += 4; + pa >>= 4; + } while (row & 15); + } +} + +void HELPER(sme_fmopa_d)(void *vza, void *vzn, void *vzm, void *vpn, + void *vpm, void *vst, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 8; + uint64_t neg = (uint64_t)simd_data(desc) << 63; + uint64_t *za = vza, *zn = vzn, *zm = vzm; + uint8_t *pn = vpn, *pm = vpm; + float_status fpst = *(float_status *)vst; + + set_default_nan_mode(true, &fpst); + + for (row = 0; row < oprsz; ++row) { + if (pn[H1(row)] & 1) { + uint64_t *za_row = &za[tile_vslice_index(row)]; + uint64_t n = zn[row] ^ neg; + + for (col = 0; col < oprsz; ++col) { + if (pm[H1(col)] & 1) { + uint64_t *a = &za_row[col]; + *a = float64_muladd(n, zm[col], *a, 0, &fpst); + } + } + } + } +} diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(ADDHA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addha_s) TRANS_FEAT(ADDVA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addva_s) TRANS_FEAT(ADDHA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addha_d) TRANS_FEAT(ADDVA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addva_d) + +static bool do_outprod_fpst(DisasContext *s, arg_op *a, MemOp esz, + gen_helper_gvec_5_ptr *fn) +{ + int svl = streaming_vec_reg_size(s); + uint32_t desc = simd_desc(svl, svl, a->sub); + TCGv_ptr za, zn, zm, pn, pm, fpst; + + if (!sme_smza_enabled_check(s)) { + return true; + } + + /* Sum XZR+zad to find ZAd. */ + za = get_tile_rowcol(s, esz, 31, a->zad, false); + zn = vec_full_reg_ptr(s, a->zn); + zm = vec_full_reg_ptr(s, a->zm); + pn = pred_full_reg_ptr(s, a->pn); + pm = pred_full_reg_ptr(s, a->pm); + fpst = fpstatus_ptr(FPST_FPCR); + + fn(za, zn, zm, pn, pm, fpst, tcg_constant_i32(desc)); + + tcg_temp_free_ptr(za); + tcg_temp_free_ptr(zn); + tcg_temp_free_ptr(pn); + tcg_temp_free_ptr(pm); + tcg_temp_free_ptr(fpst); + return true; +} + +TRANS_FEAT(FMOPA_s, aa64_sme, do_outprod_fpst, a, MO_32, gen_helper_sme_fmopa_s) +TRANS_FEAT(FMOPA_d, aa64_sme_f64f64, do_outprod_fpst, a, MO_64, gen_helper_sme_fmopa_d) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-26-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 2 ++ target/arm/sme.decode | 2 ++ target/arm/sme_helper.c | 56 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 30 ++++++++++++++++++++ 4 files changed, 90 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_7(sme_fmopa_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_7(sme_fmopa_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_bfmopa, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ ADDVA_d 11000000 11 01000 1 ... ... ..... 00 ... @adda_64 FMOPA_s 10000000 100 ..... ... ... ..... . 00 .. @op_32 FMOPA_d 10000000 110 ..... ... ... ..... . 0 ... @op_64 + +BFMOPA 10000001 100 ..... ... ... ..... . 00 .. @op_32 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(sme_fmopa_d)(void *vza, void *vzn, void *vzm, void *vpn, } } } + +/* + * Alter PAIR as needed for controlling predicates being false, + * and for NEG on an enabled row element. + */ +static inline uint32_t f16mop_adj_pair(uint32_t pair, uint32_t pg, uint32_t neg) +{ + /* + * The pseudocode uses a conditional negate after the conditional zero. + * It is simpler here to unconditionally negate before conditional zero. + */ + pair ^= neg; + if (!(pg & 1)) { + pair &= 0xffff0000u; + } + if (!(pg & 4)) { + pair &= 0x0000ffffu; + } + return pair; +} + +void HELPER(sme_bfmopa)(void *vza, void *vzn, void *vzm, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_maxsz(desc); + uint32_t neg = simd_data(desc) * 0x80008000u; + uint16_t *pn = vpn, *pm = vpm; + + for (row = 0; row < oprsz; ) { + uint16_t prow = pn[H2(row >> 4)]; + do { + void *vza_row = vza + tile_vslice_offset(row); + uint32_t n = *(uint32_t *)(vzn + H1_4(row)); + + n = f16mop_adj_pair(n, prow, neg); + + for (col = 0; col < oprsz; ) { + uint16_t pcol = pm[H2(col >> 4)]; + do { + if (prow & pcol & 0b0101) { + uint32_t *a = vza_row + H1_4(col); + uint32_t m = *(uint32_t *)(vzm + H1_4(col)); + + m = f16mop_adj_pair(m, pcol, 0); + *a = bfdotadd(*a, n, m); + + col += 4; + pcol >>= 4; + } + } while (col & 15); + } + row += 4; + prow >>= 4; + } while (row & 15); + } +} diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(ADDVA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addva_s) TRANS_FEAT(ADDHA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addha_d) TRANS_FEAT(ADDVA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addva_d) +static bool do_outprod(DisasContext *s, arg_op *a, MemOp esz, + gen_helper_gvec_5 *fn) +{ + int svl = streaming_vec_reg_size(s); + uint32_t desc = simd_desc(svl, svl, a->sub); + TCGv_ptr za, zn, zm, pn, pm; + + if (!sme_smza_enabled_check(s)) { + return true; + } + + /* Sum XZR+zad to find ZAd. */ + za = get_tile_rowcol(s, esz, 31, a->zad, false); + zn = vec_full_reg_ptr(s, a->zn); + zm = vec_full_reg_ptr(s, a->zm); + pn = pred_full_reg_ptr(s, a->pn); + pm = pred_full_reg_ptr(s, a->pm); + + fn(za, zn, zm, pn, pm, tcg_constant_i32(desc)); + + tcg_temp_free_ptr(za); + tcg_temp_free_ptr(zn); + tcg_temp_free_ptr(pn); + tcg_temp_free_ptr(pm); + return true; +} + static bool do_outprod_fpst(DisasContext *s, arg_op *a, MemOp esz, gen_helper_gvec_5_ptr *fn) { @@ -XXX,XX +XXX,XX @@ static bool do_outprod_fpst(DisasContext *s, arg_op *a, MemOp esz, TRANS_FEAT(FMOPA_s, aa64_sme, do_outprod_fpst, a, MO_32, gen_helper_sme_fmopa_s) TRANS_FEAT(FMOPA_d, aa64_sme_f64f64, do_outprod_fpst, a, MO_64, gen_helper_sme_fmopa_d) + +/* TODO: FEAT_EBF16 */ +TRANS_FEAT(BFMOPA, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_bfmopa) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-27-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 2 ++ target/arm/sme.decode | 1 + target/arm/sme_helper.c | 74 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 1 + 4 files changed, 78 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_5(sme_addva_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addha_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addva_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_7(sme_fmopa_h, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_7(sme_fmopa_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_7(sme_fmopa_d, TCG_CALL_NO_RWG, diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ FMOPA_s 10000000 100 ..... ... ... ..... . 00 .. @op_32 FMOPA_d 10000000 110 ..... ... ... ..... . 0 ... @op_64 BFMOPA 10000001 100 ..... ... ... ..... . 00 .. @op_32 +FMOPA_h 10000001 101 ..... ... ... ..... . 00 .. @op_32 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ static inline uint32_t f16mop_adj_pair(uint32_t pair, uint32_t pg, uint32_t neg) return pair; } +static float32 f16_dotadd(float32 sum, uint32_t e1, uint32_t e2, + float_status *s_std, float_status *s_odd) +{ + float64 e1r = float16_to_float64(e1 & 0xffff, true, s_std); + float64 e1c = float16_to_float64(e1 >> 16, true, s_std); + float64 e2r = float16_to_float64(e2 & 0xffff, true, s_std); + float64 e2c = float16_to_float64(e2 >> 16, true, s_std); + float64 t64; + float32 t32; + + /* + * The ARM pseudocode function FPDot performs both multiplies + * and the add with a single rounding operation. Emulate this + * by performing the first multiply in round-to-odd, then doing + * the second multiply as fused multiply-add, and rounding to + * float32 all in one step. + */ + t64 = float64_mul(e1r, e2r, s_odd); + t64 = float64r32_muladd(e1c, e2c, t64, 0, s_std); + + /* This conversion is exact, because we've already rounded. */ + t32 = float64_to_float32(t64, s_std); + + /* The final accumulation step is not fused. */ + return float32_add(sum, t32, s_std); +} + +void HELPER(sme_fmopa_h)(void *vza, void *vzn, void *vzm, void *vpn, + void *vpm, void *vst, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_maxsz(desc); + uint32_t neg = simd_data(desc) * 0x80008000u; + uint16_t *pn = vpn, *pm = vpm; + float_status fpst_odd, fpst_std; + + /* + * Make a copy of float_status because this operation does not + * update the cumulative fp exception status. It also produces + * default nans. Make a second copy with round-to-odd -- see above. + */ + fpst_std = *(float_status *)vst; + set_default_nan_mode(true, &fpst_std); + fpst_odd = fpst_std; + set_float_rounding_mode(float_round_to_odd, &fpst_odd); + + for (row = 0; row < oprsz; ) { + uint16_t prow = pn[H2(row >> 4)]; + do { + void *vza_row = vza + tile_vslice_offset(row); + uint32_t n = *(uint32_t *)(vzn + H1_4(row)); + + n = f16mop_adj_pair(n, prow, neg); + + for (col = 0; col < oprsz; ) { + uint16_t pcol = pm[H2(col >> 4)]; + do { + if (prow & pcol & 0b0101) { + uint32_t *a = vza_row + H1_4(col); + uint32_t m = *(uint32_t *)(vzm + H1_4(col)); + + m = f16mop_adj_pair(m, pcol, 0); + *a = f16_dotadd(*a, n, m, &fpst_std, &fpst_odd); + + col += 4; + pcol >>= 4; + } + } while (col & 15); + } + row += 4; + prow >>= 4; + } while (row & 15); + } +} + void HELPER(sme_bfmopa)(void *vza, void *vzn, void *vzm, void *vpn, void *vpm, uint32_t desc) { diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ static bool do_outprod_fpst(DisasContext *s, arg_op *a, MemOp esz, return true; } +TRANS_FEAT(FMOPA_h, aa64_sme, do_outprod_fpst, a, MO_32, gen_helper_sme_fmopa_h) TRANS_FEAT(FMOPA_s, aa64_sme, do_outprod_fpst, a, MO_32, gen_helper_sme_fmopa_s) TRANS_FEAT(FMOPA_d, aa64_sme_f64f64, do_outprod_fpst, a, MO_64, gen_helper_sme_fmopa_d) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This is SMOPA, SUMOPA, USMOPA_s, UMOPA, for both Int8 and Int16. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-28-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 16 ++++++++ target/arm/sme.decode | 10 +++++ target/arm/sme_helper.c | 82 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 10 +++++ 4 files changed, 118 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_7(sme_fmopa_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_6(sme_bfmopa, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_smopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_umopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_sumopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_usmopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_smopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_umopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_sumopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_usmopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ FMOPA_d 10000000 110 ..... ... ... ..... . 0 ... @op_64 BFMOPA 10000001 100 ..... ... ... ..... . 00 .. @op_32 FMOPA_h 10000001 101 ..... ... ... ..... . 00 .. @op_32 + +SMOPA_s 1010000 0 10 0 ..... ... ... ..... . 00 .. @op_32 +SUMOPA_s 1010000 0 10 1 ..... ... ... ..... . 00 .. @op_32 +USMOPA_s 1010000 1 10 0 ..... ... ... ..... . 00 .. @op_32 +UMOPA_s 1010000 1 10 1 ..... ... ... ..... . 00 .. @op_32 + +SMOPA_d 1010000 0 11 0 ..... ... ... ..... . 0 ... @op_64 +SUMOPA_d 1010000 0 11 1 ..... ... ... ..... . 0 ... @op_64 +USMOPA_d 1010000 1 11 0 ..... ... ... ..... . 0 ... @op_64 +UMOPA_d 1010000 1 11 1 ..... ... ... ..... . 0 ... @op_64 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(sme_bfmopa)(void *vza, void *vzn, void *vzm, void *vpn, } while (row & 15); } } + +typedef uint64_t IMOPFn(uint64_t, uint64_t, uint64_t, uint8_t, bool); + +static inline void do_imopa(uint64_t *za, uint64_t *zn, uint64_t *zm, + uint8_t *pn, uint8_t *pm, + uint32_t desc, IMOPFn *fn) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 8; + bool neg = simd_data(desc); + + for (row = 0; row < oprsz; ++row) { + uint8_t pa = pn[H1(row)]; + uint64_t *za_row = &za[tile_vslice_index(row)]; + uint64_t n = zn[row]; + + for (col = 0; col < oprsz; ++col) { + uint8_t pb = pm[H1(col)]; + uint64_t *a = &za_row[col]; + + *a = fn(n, zm[col], *a, pa & pb, neg); + } + } +} + +#define DEF_IMOP_32(NAME, NTYPE, MTYPE) \ +static uint64_t NAME(uint64_t n, uint64_t m, uint64_t a, uint8_t p, bool neg) \ +{ \ + uint32_t sum0 = 0, sum1 = 0; \ + /* Apply P to N as a mask, making the inactive elements 0. */ \ + n &= expand_pred_b(p); \ + sum0 += (NTYPE)(n >> 0) * (MTYPE)(m >> 0); \ + sum0 += (NTYPE)(n >> 8) * (MTYPE)(m >> 8); \ + sum0 += (NTYPE)(n >> 16) * (MTYPE)(m >> 16); \ + sum0 += (NTYPE)(n >> 24) * (MTYPE)(m >> 24); \ + sum1 += (NTYPE)(n >> 32) * (MTYPE)(m >> 32); \ + sum1 += (NTYPE)(n >> 40) * (MTYPE)(m >> 40); \ + sum1 += (NTYPE)(n >> 48) * (MTYPE)(m >> 48); \ + sum1 += (NTYPE)(n >> 56) * (MTYPE)(m >> 56); \ + if (neg) { \ + sum0 = (uint32_t)a - sum0, sum1 = (uint32_t)(a >> 32) - sum1; \ + } else { \ + sum0 = (uint32_t)a + sum0, sum1 = (uint32_t)(a >> 32) + sum1; \ + } \ + return ((uint64_t)sum1 << 32) | sum0; \ +} + +#define DEF_IMOP_64(NAME, NTYPE, MTYPE) \ +static uint64_t NAME(uint64_t n, uint64_t m, uint64_t a, uint8_t p, bool neg) \ +{ \ + uint64_t sum = 0; \ + /* Apply P to N as a mask, making the inactive elements 0. */ \ + n &= expand_pred_h(p); \ + sum += (NTYPE)(n >> 0) * (MTYPE)(m >> 0); \ + sum += (NTYPE)(n >> 16) * (MTYPE)(m >> 16); \ + sum += (NTYPE)(n >> 32) * (MTYPE)(m >> 32); \ + sum += (NTYPE)(n >> 48) * (MTYPE)(m >> 48); \ + return neg ? a - sum : a + sum; \ +} + +DEF_IMOP_32(smopa_s, int8_t, int8_t) +DEF_IMOP_32(umopa_s, uint8_t, uint8_t) +DEF_IMOP_32(sumopa_s, int8_t, uint8_t) +DEF_IMOP_32(usmopa_s, uint8_t, int8_t) + +DEF_IMOP_64(smopa_d, int16_t, int16_t) +DEF_IMOP_64(umopa_d, uint16_t, uint16_t) +DEF_IMOP_64(sumopa_d, int16_t, uint16_t) +DEF_IMOP_64(usmopa_d, uint16_t, int16_t) + +#define DEF_IMOPH(NAME) \ + void HELPER(sme_##NAME)(void *vza, void *vzn, void *vzm, void *vpn, \ + void *vpm, uint32_t desc) \ + { do_imopa(vza, vzn, vzm, vpn, vpm, desc, NAME); } + +DEF_IMOPH(smopa_s) +DEF_IMOPH(umopa_s) +DEF_IMOPH(sumopa_s) +DEF_IMOPH(usmopa_s) +DEF_IMOPH(smopa_d) +DEF_IMOPH(umopa_d) +DEF_IMOPH(sumopa_d) +DEF_IMOPH(usmopa_d) diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(FMOPA_d, aa64_sme_f64f64, do_outprod_fpst, a, MO_64, gen_helper_sme_f /* TODO: FEAT_EBF16 */ TRANS_FEAT(BFMOPA, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_bfmopa) + +TRANS_FEAT(SMOPA_s, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_smopa_s) +TRANS_FEAT(UMOPA_s, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_umopa_s) +TRANS_FEAT(SUMOPA_s, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_sumopa_s) +TRANS_FEAT(USMOPA_s, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_usmopa_s) + +TRANS_FEAT(SMOPA_d, aa64_sme_i16i64, do_outprod, a, MO_64, gen_helper_sme_smopa_d) +TRANS_FEAT(UMOPA_d, aa64_sme_i16i64, do_outprod, a, MO_64, gen_helper_sme_umopa_d) +TRANS_FEAT(SUMOPA_d, aa64_sme_i16i64, do_outprod, a, MO_64, gen_helper_sme_sumopa_d) +TRANS_FEAT(USMOPA_d, aa64_sme_i16i64, do_outprod, a, MO_64, gen_helper_sme_usmopa_d) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This is an SVE instruction that operates using the SVE vector length but that it is present only if SME is implemented. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-29-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sve.decode | 20 +++++++++++++ target/arm/translate-sve.c | 57 ++++++++++++++++++++++++++++++++++++++ 2 files changed, 77 insertions(+) diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ BFMLALT_zzxw 01100100 11 1 ..... 0100.1 ..... ..... @rrxr_3a esz=2 ### SVE2 floating-point bfloat16 dot-product (indexed) BFDOT_zzxz 01100100 01 1 ..... 010000 ..... ..... @rrxr_2 esz=2 + +### SVE broadcast predicate element + +&psel esz pd pn pm rv imm +%psel_rv 16:2 !function=plus_12 +%psel_imm_b 22:2 19:2 +%psel_imm_h 22:2 20:1 +%psel_imm_s 22:2 +%psel_imm_d 23:1 +@psel ........ .. . ... .. .. pn:4 . pm:4 . pd:4 \ + &psel rv=%psel_rv + +PSEL 00100101 .. 1 ..1 .. 01 .... 0 .... 0 .... \ + @psel esz=0 imm=%psel_imm_b +PSEL 00100101 .. 1 .10 .. 01 .... 0 .... 0 .... \ + @psel esz=1 imm=%psel_imm_h +PSEL 00100101 .. 1 100 .. 01 .... 0 .... 0 .... \ + @psel esz=2 imm=%psel_imm_s +PSEL 00100101 .1 1 000 .. 01 .... 0 .... 0 .... \ + @psel esz=3 imm=%psel_imm_d diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool do_BFMLAL_zzxw(DisasContext *s, arg_rrxr_esz *a, bool sel) TRANS_FEAT(BFMLALB_zzxw, aa64_sve_bf16, do_BFMLAL_zzxw, a, false) TRANS_FEAT(BFMLALT_zzxw, aa64_sve_bf16, do_BFMLAL_zzxw, a, true) + +static bool trans_PSEL(DisasContext *s, arg_psel *a) +{ + int vl = vec_full_reg_size(s); + int pl = pred_gvec_reg_size(s); + int elements = vl >> a->esz; + TCGv_i64 tmp, didx, dbit; + TCGv_ptr ptr; + + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (!sve_access_check(s)) { + return true; + } + + tmp = tcg_temp_new_i64(); + dbit = tcg_temp_new_i64(); + didx = tcg_temp_new_i64(); + ptr = tcg_temp_new_ptr(); + + /* Compute the predicate element. */ + tcg_gen_addi_i64(tmp, cpu_reg(s, a->rv), a->imm); + if (is_power_of_2(elements)) { + tcg_gen_andi_i64(tmp, tmp, elements - 1); + } else { + tcg_gen_remu_i64(tmp, tmp, tcg_constant_i64(elements)); + } + + /* Extract the predicate byte and bit indices. */ + tcg_gen_shli_i64(tmp, tmp, a->esz); + tcg_gen_andi_i64(dbit, tmp, 7); + tcg_gen_shri_i64(didx, tmp, 3); + if (HOST_BIG_ENDIAN) { + tcg_gen_xori_i64(didx, didx, 7); + } + + /* Load the predicate word. */ + tcg_gen_trunc_i64_ptr(ptr, didx); + tcg_gen_add_ptr(ptr, ptr, cpu_env); + tcg_gen_ld8u_i64(tmp, ptr, pred_full_reg_offset(s, a->pm)); + + /* Extract the predicate bit and replicate to MO_64. */ + tcg_gen_shr_i64(tmp, tmp, dbit); + tcg_gen_andi_i64(tmp, tmp, 1); + tcg_gen_neg_i64(tmp, tmp); + + /* Apply to either copy the source, or write zeros. */ + tcg_gen_gvec_ands(MO_64, pred_full_reg_offset(s, a->pd), + pred_full_reg_offset(s, a->pn), tmp, pl, pl); + + tcg_temp_free_i64(tmp); + tcg_temp_free_i64(dbit); + tcg_temp_free_i64(didx); + tcg_temp_free_ptr(ptr); + return true; +} -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This is an SVE instruction that operates using the SVE vector length but that it is present only if SME is implemented. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-30-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sve.h | 2 ++ target/arm/sve.decode | 1 + target/arm/sve_helper.c | 16 ++++++++++++++++ target/arm/translate-sve.c | 2 ++ 4 files changed, 21 insertions(+) diff --git a/target/arm/helper-sve.h b/target/arm/helper-sve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sve.h +++ b/target/arm/helper-sve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(sve_revh_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sve_revw_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_revd_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) + DEF_HELPER_FLAGS_4(sve_rbit_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sve_rbit_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sve_rbit_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ REVB 00000101 .. 1001 00 100 ... ..... ..... @rd_pg_rn REVH 00000101 .. 1001 01 100 ... ..... ..... @rd_pg_rn REVW 00000101 .. 1001 10 100 ... ..... ..... @rd_pg_rn RBIT 00000101 .. 1001 11 100 ... ..... ..... @rd_pg_rn +REVD 00000101 00 1011 10 100 ... ..... ..... @rd_pg_rn_e0 # SVE vector splice (predicated, destructive) SPLICE 00000101 .. 101 100 100 ... ..... ..... @rdn_pg_rm diff --git a/target/arm/sve_helper.c b/target/arm/sve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve_helper.c +++ b/target/arm/sve_helper.c @@ -XXX,XX +XXX,XX @@ DO_ZPZ_D(sve_revh_d, uint64_t, hswap64) DO_ZPZ_D(sve_revw_d, uint64_t, wswap64) +void HELPER(sme_revd_q)(void *vd, void *vn, void *vg, uint32_t desc) +{ + intptr_t i, opr_sz = simd_oprsz(desc) / 8; + uint64_t *d = vd, *n = vn; + uint8_t *pg = vg; + + for (i = 0; i < opr_sz; i += 2) { + if (pg[H1(i)] & 1) { + uint64_t n0 = n[i + 0]; + uint64_t n1 = n[i + 1]; + d[i + 0] = n1; + d[i + 1] = n0; + } + } +} + DO_ZPZ(sve_rbit_b, uint8_t, H1, revbit8) DO_ZPZ(sve_rbit_h, uint16_t, H1_2, revbit16) DO_ZPZ(sve_rbit_s, uint32_t, H1_4, revbit32) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(REVH, aa64_sve, gen_gvec_ool_arg_zpz, revh_fns[a->esz], a, 0) TRANS_FEAT(REVW, aa64_sve, gen_gvec_ool_arg_zpz, a->esz == 3 ? gen_helper_sve_revw_d : NULL, a, 0) +TRANS_FEAT(REVD, aa64_sme, gen_gvec_ool_arg_zpz, gen_helper_sme_revd_q, a, 0) + TRANS_FEAT(SPLICE, aa64_sve, gen_gvec_ool_arg_zpzz, gen_helper_sve_splice, a, a->esz) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This is an SVE instruction that operates using the SVE vector length but that it is present only if SME is implemented. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-31-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.h | 18 +++++++ target/arm/sve.decode | 5 ++ target/arm/translate-sve.c | 102 +++++++++++++++++++++++++++++++++++++ target/arm/vec_helper.c | 24 +++++++++ 4 files changed, 149 insertions(+) diff --git a/target/arm/helper.h b/target/arm/helper.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.h +++ b/target/arm/helper.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_6(gvec_bfmlal, TCG_CALL_NO_RWG, DEF_HELPER_FLAGS_6(gvec_bfmlal_idx, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_sclamp_b, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_sclamp_h, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_sclamp_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_sclamp_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) + +DEF_HELPER_FLAGS_5(gvec_uclamp_b, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_uclamp_h, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_uclamp_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_uclamp_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) + #ifdef TARGET_AARCH64 #include "helper-a64.h" #include "helper-sve.h" diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ PSEL 00100101 .. 1 100 .. 01 .... 0 .... 0 .... \ @psel esz=2 imm=%psel_imm_s PSEL 00100101 .1 1 000 .. 01 .... 0 .... 0 .... \ @psel esz=3 imm=%psel_imm_d + +### SVE clamp + +SCLAMP 01000100 .. 0 ..... 110000 ..... ..... @rda_rn_rm +UCLAMP 01000100 .. 0 ..... 110001 ..... ..... @rda_rn_rm diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_PSEL(DisasContext *s, arg_psel *a) tcg_temp_free_ptr(ptr); return true; } + +static void gen_sclamp_i32(TCGv_i32 d, TCGv_i32 n, TCGv_i32 m, TCGv_i32 a) +{ + tcg_gen_smax_i32(d, a, n); + tcg_gen_smin_i32(d, d, m); +} + +static void gen_sclamp_i64(TCGv_i64 d, TCGv_i64 n, TCGv_i64 m, TCGv_i64 a) +{ + tcg_gen_smax_i64(d, a, n); + tcg_gen_smin_i64(d, d, m); +} + +static void gen_sclamp_vec(unsigned vece, TCGv_vec d, TCGv_vec n, + TCGv_vec m, TCGv_vec a) +{ + tcg_gen_smax_vec(vece, d, a, n); + tcg_gen_smin_vec(vece, d, d, m); +} + +static void gen_sclamp(unsigned vece, uint32_t d, uint32_t n, uint32_t m, + uint32_t a, uint32_t oprsz, uint32_t maxsz) +{ + static const TCGOpcode vecop[] = { + INDEX_op_smin_vec, INDEX_op_smax_vec, 0 + }; + static const GVecGen4 ops[4] = { + { .fniv = gen_sclamp_vec, + .fno = gen_helper_gvec_sclamp_b, + .opt_opc = vecop, + .vece = MO_8 }, + { .fniv = gen_sclamp_vec, + .fno = gen_helper_gvec_sclamp_h, + .opt_opc = vecop, + .vece = MO_16 }, + { .fni4 = gen_sclamp_i32, + .fniv = gen_sclamp_vec, + .fno = gen_helper_gvec_sclamp_s, + .opt_opc = vecop, + .vece = MO_32 }, + { .fni8 = gen_sclamp_i64, + .fniv = gen_sclamp_vec, + .fno = gen_helper_gvec_sclamp_d, + .opt_opc = vecop, + .vece = MO_64, + .prefer_i64 = TCG_TARGET_REG_BITS == 64 } + }; + tcg_gen_gvec_4(d, n, m, a, oprsz, maxsz, &ops[vece]); +} + +TRANS_FEAT(SCLAMP, aa64_sme, gen_gvec_fn_arg_zzzz, gen_sclamp, a) + +static void gen_uclamp_i32(TCGv_i32 d, TCGv_i32 n, TCGv_i32 m, TCGv_i32 a) +{ + tcg_gen_umax_i32(d, a, n); + tcg_gen_umin_i32(d, d, m); +} + +static void gen_uclamp_i64(TCGv_i64 d, TCGv_i64 n, TCGv_i64 m, TCGv_i64 a) +{ + tcg_gen_umax_i64(d, a, n); + tcg_gen_umin_i64(d, d, m); +} + +static void gen_uclamp_vec(unsigned vece, TCGv_vec d, TCGv_vec n, + TCGv_vec m, TCGv_vec a) +{ + tcg_gen_umax_vec(vece, d, a, n); + tcg_gen_umin_vec(vece, d, d, m); +} + +static void gen_uclamp(unsigned vece, uint32_t d, uint32_t n, uint32_t m, + uint32_t a, uint32_t oprsz, uint32_t maxsz) +{ + static const TCGOpcode vecop[] = { + INDEX_op_umin_vec, INDEX_op_umax_vec, 0 + }; + static const GVecGen4 ops[4] = { + { .fniv = gen_uclamp_vec, + .fno = gen_helper_gvec_uclamp_b, + .opt_opc = vecop, + .vece = MO_8 }, + { .fniv = gen_uclamp_vec, + .fno = gen_helper_gvec_uclamp_h, + .opt_opc = vecop, + .vece = MO_16 }, + { .fni4 = gen_uclamp_i32, + .fniv = gen_uclamp_vec, + .fno = gen_helper_gvec_uclamp_s, + .opt_opc = vecop, + .vece = MO_32 }, + { .fni8 = gen_uclamp_i64, + .fniv = gen_uclamp_vec, + .fno = gen_helper_gvec_uclamp_d, + .opt_opc = vecop, + .vece = MO_64, + .prefer_i64 = TCG_TARGET_REG_BITS == 64 } + }; + tcg_gen_gvec_4(d, n, m, a, oprsz, maxsz, &ops[vece]); +} + +TRANS_FEAT(UCLAMP, aa64_sme, gen_gvec_fn_arg_zzzz, gen_uclamp, a) diff --git a/target/arm/vec_helper.c b/target/arm/vec_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/vec_helper.c +++ b/target/arm/vec_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(gvec_bfmlal_idx)(void *vd, void *vn, void *vm, } clear_tail(d, opr_sz, simd_maxsz(desc)); } + +#define DO_CLAMP(NAME, TYPE) \ +void HELPER(NAME)(void *d, void *n, void *m, void *a, uint32_t desc) \ +{ \ + intptr_t i, opr_sz = simd_oprsz(desc); \ + for (i = 0; i < opr_sz; i += sizeof(TYPE)) { \ + TYPE aa = *(TYPE *)(a + i); \ + TYPE nn = *(TYPE *)(n + i); \ + TYPE mm = *(TYPE *)(m + i); \ + TYPE dd = MIN(MAX(aa, nn), mm); \ + *(TYPE *)(d + i) = dd; \ + } \ + clear_tail(d, opr_sz, simd_maxsz(desc)); \ +} + +DO_CLAMP(gvec_sclamp_b, int8_t) +DO_CLAMP(gvec_sclamp_h, int16_t) +DO_CLAMP(gvec_sclamp_s, int32_t) +DO_CLAMP(gvec_sclamp_d, int64_t) + +DO_CLAMP(gvec_uclamp_b, uint8_t) +DO_CLAMP(gvec_uclamp_h, uint16_t) +DO_CLAMP(gvec_uclamp_s, uint32_t) +DO_CLAMP(gvec_uclamp_d, uint64_t) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> We can handle both exception entry and exception return by hooking into aarch64_sve_change_el. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-32-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ void aarch64_sve_change_el(CPUARMState *env, int old_el, return; } + old_a64 = old_el ? arm_el_is_aa64(env, old_el) : el0_a64; + new_a64 = new_el ? arm_el_is_aa64(env, new_el) : el0_a64; + + /* + * Both AArch64.TakeException and AArch64.ExceptionReturn + * invoke ResetSVEState when taking an exception from, or + * returning to, AArch32 state when PSTATE.SM is enabled. + */ + if (old_a64 != new_a64 && FIELD_EX64(env->svcr, SVCR, SM)) { + arm_reset_sve_state(env); + return; + } + /* * DDI0584A.d sec 3.2: "If SVE instructions are disabled or trapped * at ELx, or not available because the EL is in AArch32 state, then @@ -XXX,XX +XXX,XX @@ void aarch64_sve_change_el(CPUARMState *env, int old_el, * we already have the correct register contents when encountering the * vq0->vq0 transition between EL0->EL1. */ - old_a64 = old_el ? arm_el_is_aa64(env, old_el) : el0_a64; old_len = (old_a64 && !sve_exception_el(env, old_el) ? sve_vqm1_for_el(env, old_el) : 0); - new_a64 = new_el ? arm_el_is_aa64(env, new_el) : el0_a64; new_len = (new_a64 && !sve_exception_el(env, new_el) ? sve_vqm1_for_el(env, new_el) : 0); -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Note that SME remains effectively disabled for user-only, because we do not yet set CPACR_EL1.SMEN. This needs to wait until the kernel ABI is implemented. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-33-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/emulation.rst | 4 ++++ target/arm/cpu64.c | 11 +++++++++++ 2 files changed, 15 insertions(+) diff --git a/docs/system/arm/emulation.rst b/docs/system/arm/emulation.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/arm/emulation.rst +++ b/docs/system/arm/emulation.rst @@ -XXX,XX +XXX,XX @@ the following architecture extensions: - FEAT_SHA512 (Advanced SIMD SHA512 instructions) - FEAT_SM3 (Advanced SIMD SM3 instructions) - FEAT_SM4 (Advanced SIMD SM4 instructions) +- FEAT_SME (Scalable Matrix Extension) +- FEAT_SME_FA64 (Full A64 instruction set in Streaming SVE mode) +- FEAT_SME_F64F64 (Double-precision floating-point outer product instructions) +- FEAT_SME_I16I64 (16-bit to 64-bit integer widening outer product instructions) - FEAT_SPECRES (Speculation restriction instructions) - FEAT_SSBS (Speculative Store Bypass Safe) - FEAT_TLBIOS (TLB invalidate instructions in Outer Shareable domain) diff --git a/target/arm/cpu64.c b/target/arm/cpu64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu64.c +++ b/target/arm/cpu64.c @@ -XXX,XX +XXX,XX @@ static void aarch64_max_initfn(Object *obj) */ t = FIELD_DP64(t, ID_AA64PFR1, MTE, 3); /* FEAT_MTE3 */ t = FIELD_DP64(t, ID_AA64PFR1, RAS_FRAC, 0); /* FEAT_RASv1p1 + FEAT_DoubleFault */ + t = FIELD_DP64(t, ID_AA64PFR1, SME, 1); /* FEAT_SME */ t = FIELD_DP64(t, ID_AA64PFR1, CSV2_FRAC, 0); /* FEAT_CSV2_2 */ cpu->isar.id_aa64pfr1 = t; @@ -XXX,XX +XXX,XX @@ static void aarch64_max_initfn(Object *obj) t = FIELD_DP64(t, ID_AA64DFR0, PMUVER, 5); /* FEAT_PMUv3p4 */ cpu->isar.id_aa64dfr0 = t; + t = cpu->isar.id_aa64smfr0; + t = FIELD_DP64(t, ID_AA64SMFR0, F32F32, 1); /* FEAT_SME */ + t = FIELD_DP64(t, ID_AA64SMFR0, B16F32, 1); /* FEAT_SME */ + t = FIELD_DP64(t, ID_AA64SMFR0, F16F32, 1); /* FEAT_SME */ + t = FIELD_DP64(t, ID_AA64SMFR0, I8I32, 0xf); /* FEAT_SME */ + t = FIELD_DP64(t, ID_AA64SMFR0, F64F64, 1); /* FEAT_SME_F64F64 */ + t = FIELD_DP64(t, ID_AA64SMFR0, I16I64, 0xf); /* FEAT_SME_I16I64 */ + t = FIELD_DP64(t, ID_AA64SMFR0, FA64, 1); /* FEAT_SME_FA64 */ + cpu->isar.id_aa64smfr0 = t; + /* Replicate the same data to the 32-bit id registers. */ aa32_max_features(cpu); -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-34-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/target_cpu.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/linux-user/aarch64/target_cpu.h b/linux-user/aarch64/target_cpu.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/target_cpu.h +++ b/linux-user/aarch64/target_cpu.h @@ -XXX,XX +XXX,XX @@ static inline void cpu_clone_regs_parent(CPUARMState *env, unsigned flags) static inline void cpu_set_tls(CPUARMState *env, target_ulong newtls) { - /* Note that AArch64 Linux keeps the TLS pointer in TPIDR; this is + /* + * Note that AArch64 Linux keeps the TLS pointer in TPIDR; this is * different from AArch32 Linux, which uses TPIDRRO. */ env->cp15.tpidr_el[0] = newtls; + /* TPIDR2_EL0 is cleared with CLONE_SETTLS. */ + env->cp15.tpidr2_el0 = 0; } static inline abi_ulong get_sp_from_cpustate(CPUARMState *state) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-35-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/cpu_loop.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/linux-user/aarch64/cpu_loop.c b/linux-user/aarch64/cpu_loop.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/cpu_loop.c +++ b/linux-user/aarch64/cpu_loop.c @@ -XXX,XX +XXX,XX @@ void cpu_loop(CPUARMState *env) switch (trapnr) { case EXCP_SWI: + /* + * On syscall, PSTATE.ZA is preserved, along with the ZA matrix. + * PSTATE.SM is cleared, per SMSTOP, which does ResetSVEState. + */ + if (FIELD_EX64(env->svcr, SVCR, SM)) { + env->svcr = FIELD_DP64(env->svcr, SVCR, SM, 0); + arm_rebuild_hflags(env); + arm_reset_sve_state(env); + } ret = do_syscall(env, env->xregs[8], env->xregs[0], -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Make sure to zero the currently reserved fields. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-36-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ struct target_extra_context { struct target_sve_context { struct target_aarch64_ctx head; uint16_t vl; - uint16_t reserved[3]; + uint16_t flags; + uint16_t reserved[2]; /* The actual SVE data immediately follows. It is laid out * according to TARGET_SVE_SIG_{Z,P}REG_OFFSET, based off of * the original struct pointer. @@ -XXX,XX +XXX,XX @@ struct target_sve_context { #define TARGET_SVE_SIG_CONTEXT_SIZE(VQ) \ (TARGET_SVE_SIG_PREG_OFFSET(VQ, 17)) +#define TARGET_SVE_SIG_FLAG_SM 1 + struct target_rt_sigframe { struct target_siginfo info; struct target_ucontext uc; @@ -XXX,XX +XXX,XX @@ static void target_setup_sve_record(struct target_sve_context *sve, { int i, j; + memset(sve, 0, sizeof(*sve)); __put_user(TARGET_SVE_MAGIC, &sve->head.magic); __put_user(size, &sve->head.size); __put_user(vq * TARGET_SVE_VQ_BYTES, &sve->vl); + if (FIELD_EX64(env->svcr, SVCR, SM)) { + __put_user(TARGET_SVE_SIG_FLAG_SM, &sve->flags); + } /* Note that SVE regs are stored as a byte stream, with each byte element * at a subsequent address. This corresponds to a little-endian store -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Fold the return value setting into the goto, so each point of failure need not do both. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-37-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 26 +++++++++++--------------- 1 file changed, 11 insertions(+), 15 deletions(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, struct target_sve_context *sve = NULL; uint64_t extra_datap = 0; bool used_extra = false; - bool err = false; int vq = 0, sve_size = 0; target_restore_general_frame(env, sf); @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, switch (magic) { case 0: if (size != 0) { - err = true; - goto exit; + goto err; } if (used_extra) { ctx = NULL; @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, case TARGET_FPSIMD_MAGIC: if (fpsimd || size != sizeof(struct target_fpsimd_context)) { - err = true; - goto exit; + goto err; } fpsimd = (struct target_fpsimd_context *)ctx; break; @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, break; } } - err = true; - goto exit; + goto err; case TARGET_EXTRA_MAGIC: if (extra || size != sizeof(struct target_extra_context)) { - err = true; - goto exit; + goto err; } __get_user(extra_datap, &((struct target_extra_context *)ctx)->datap); @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, /* Unknown record -- we certainly didn't generate it. * Did we in fact get out of sync? */ - err = true; - goto exit; + goto err; } ctx = (void *)ctx + size; } @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, if (fpsimd) { target_restore_fpsimd_record(env, fpsimd); } else { - err = true; + goto err; } /* SVE data, if present, overwrites FPSIMD data. */ if (sve) { target_restore_sve_record(env, sve, vq); } - - exit: unlock_user(extra, extra_datap, 0); - return err; + return 0; + + err: + unlock_user(extra, extra_datap, 0); + return 1; } static abi_ulong get_sigframe(struct target_sigaction *ka, -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> In parse_user_sigframe, the kernel rejects duplicate sve records, or records that are smaller than the header. We were silently allowing these cases to pass, dropping the record. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-38-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, break; case TARGET_SVE_MAGIC: + if (sve || size < sizeof(struct target_sve_context)) { + goto err; + } if (cpu_isar_feature(aa64_sve, env_archcpu(env))) { vq = sve_vq(env); sve_size = QEMU_ALIGN_UP(TARGET_SVE_SIG_CONTEXT_SIZE(vq), 16); - if (!sve && size == sve_size) { + if (size == sve_size) { sve = (struct target_sve_context *)ctx; break; } -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-39-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, __get_user(extra_size, &((struct target_extra_context *)ctx)->size); extra = lock_user(VERIFY_READ, extra_datap, extra_size, 0); + if (!extra) { + return 1; + } break; default: -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Move the checks out of the parsing loop and into the restore function. This more closely mirrors the code structure in the kernel, and is slightly clearer. Reject rather than silently skip incorrect VL and SVE record sizes, bringing our checks in to line with those the kernel does. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-40-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 51 +++++++++++++++++++++++++------------ 1 file changed, 35 insertions(+), 16 deletions(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ static void target_restore_fpsimd_record(CPUARMState *env, } } -static void target_restore_sve_record(CPUARMState *env, - struct target_sve_context *sve, int vq) +static bool target_restore_sve_record(CPUARMState *env, + struct target_sve_context *sve, + int size) { - int i, j; + int i, j, vl, vq; - /* Note that SVE regs are stored as a byte stream, with each byte element + if (!cpu_isar_feature(aa64_sve, env_archcpu(env))) { + return false; + } + + __get_user(vl, &sve->vl); + vq = sve_vq(env); + + /* Reject mismatched VL. */ + if (vl != vq * TARGET_SVE_VQ_BYTES) { + return false; + } + + /* Accept empty record -- used to clear PSTATE.SM. */ + if (size <= sizeof(*sve)) { + return true; + } + + /* Reject non-empty but incomplete record. */ + if (size < TARGET_SVE_SIG_CONTEXT_SIZE(vq)) { + return false; + } + + /* + * Note that SVE regs are stored as a byte stream, with each byte element * at a subsequent address. This corresponds to a little-endian load * of our 64-bit hunks. */ @@ -XXX,XX +XXX,XX @@ static void target_restore_sve_record(CPUARMState *env, } } } + return true; } static int target_restore_sigframe(CPUARMState *env, @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, struct target_sve_context *sve = NULL; uint64_t extra_datap = 0; bool used_extra = false; - int vq = 0, sve_size = 0; + int sve_size = 0; target_restore_general_frame(env, sf); @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, if (sve || size < sizeof(struct target_sve_context)) { goto err; } - if (cpu_isar_feature(aa64_sve, env_archcpu(env))) { - vq = sve_vq(env); - sve_size = QEMU_ALIGN_UP(TARGET_SVE_SIG_CONTEXT_SIZE(vq), 16); - if (size == sve_size) { - sve = (struct target_sve_context *)ctx; - break; - } - } - goto err; + sve = (struct target_sve_context *)ctx; + sve_size = size; + break; case TARGET_EXTRA_MAGIC: if (extra || size != sizeof(struct target_extra_context)) { @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, } /* SVE data, if present, overwrites FPSIMD data. */ - if (sve) { - target_restore_sve_record(env, sve, vq); + if (sve && !target_restore_sve_record(env, sve, sve_size)) { + goto err; } unlock_user(extra, extra_datap, 0); return 0; -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Set the SM bit in the SVE record on signal delivery, create the ZA record. Restore SM and ZA state according to the records present on return. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-41-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 167 +++++++++++++++++++++++++++++++++--- 1 file changed, 154 insertions(+), 13 deletions(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ struct target_sve_context { #define TARGET_SVE_SIG_FLAG_SM 1 +#define TARGET_ZA_MAGIC 0x54366345 + +struct target_za_context { + struct target_aarch64_ctx head; + uint16_t vl; + uint16_t reserved[3]; + /* The actual ZA data immediately follows. */ +}; + +#define TARGET_ZA_SIG_REGS_OFFSET \ + QEMU_ALIGN_UP(sizeof(struct target_za_context), TARGET_SVE_VQ_BYTES) +#define TARGET_ZA_SIG_ZAV_OFFSET(VQ, N) \ + (TARGET_ZA_SIG_REGS_OFFSET + (VQ) * TARGET_SVE_VQ_BYTES * (N)) +#define TARGET_ZA_SIG_CONTEXT_SIZE(VQ) \ + TARGET_ZA_SIG_ZAV_OFFSET(VQ, VQ * TARGET_SVE_VQ_BYTES) + struct target_rt_sigframe { struct target_siginfo info; struct target_ucontext uc; @@ -XXX,XX +XXX,XX @@ static void target_setup_end_record(struct target_aarch64_ctx *end) } static void target_setup_sve_record(struct target_sve_context *sve, - CPUARMState *env, int vq, int size) + CPUARMState *env, int size) { - int i, j; + int i, j, vq = sve_vq(env); memset(sve, 0, sizeof(*sve)); __put_user(TARGET_SVE_MAGIC, &sve->head.magic); @@ -XXX,XX +XXX,XX @@ static void target_setup_sve_record(struct target_sve_context *sve, } } +static void target_setup_za_record(struct target_za_context *za, + CPUARMState *env, int size) +{ + int vq = sme_vq(env); + int vl = vq * TARGET_SVE_VQ_BYTES; + int i, j; + + memset(za, 0, sizeof(*za)); + __put_user(TARGET_ZA_MAGIC, &za->head.magic); + __put_user(size, &za->head.size); + __put_user(vl, &za->vl); + + if (size == TARGET_ZA_SIG_CONTEXT_SIZE(0)) { + return; + } + assert(size == TARGET_ZA_SIG_CONTEXT_SIZE(vq)); + + /* + * Note that ZA vectors are stored as a byte stream, + * with each byte element at a subsequent address. + */ + for (i = 0; i < vl; ++i) { + uint64_t *z = (void *)za + TARGET_ZA_SIG_ZAV_OFFSET(vq, i); + for (j = 0; j < vq * 2; ++j) { + __put_user_e(env->zarray[i].d[j], z + j, le); + } + } +} + static void target_restore_general_frame(CPUARMState *env, struct target_rt_sigframe *sf) { @@ -XXX,XX +XXX,XX @@ static void target_restore_fpsimd_record(CPUARMState *env, static bool target_restore_sve_record(CPUARMState *env, struct target_sve_context *sve, - int size) + int size, int *svcr) { - int i, j, vl, vq; + int i, j, vl, vq, flags; + bool sm; - if (!cpu_isar_feature(aa64_sve, env_archcpu(env))) { + __get_user(vl, &sve->vl); + __get_user(flags, &sve->flags); + + sm = flags & TARGET_SVE_SIG_FLAG_SM; + + /* The cpu must support Streaming or Non-streaming SVE. */ + if (sm + ? !cpu_isar_feature(aa64_sme, env_archcpu(env)) + : !cpu_isar_feature(aa64_sve, env_archcpu(env))) { return false; } - __get_user(vl, &sve->vl); - vq = sve_vq(env); + /* + * Note that we cannot use sve_vq() because that depends on the + * current setting of PSTATE.SM, not the state to be restored. + */ + vq = sve_vqm1_for_el_sm(env, 0, sm) + 1; /* Reject mismatched VL. */ if (vl != vq * TARGET_SVE_VQ_BYTES) { @@ -XXX,XX +XXX,XX @@ static bool target_restore_sve_record(CPUARMState *env, return false; } + *svcr = FIELD_DP64(*svcr, SVCR, SM, sm); + /* * Note that SVE regs are stored as a byte stream, with each byte element * at a subsequent address. This corresponds to a little-endian load @@ -XXX,XX +XXX,XX @@ static bool target_restore_sve_record(CPUARMState *env, return true; } +static bool target_restore_za_record(CPUARMState *env, + struct target_za_context *za, + int size, int *svcr) +{ + int i, j, vl, vq; + + if (!cpu_isar_feature(aa64_sme, env_archcpu(env))) { + return false; + } + + __get_user(vl, &za->vl); + vq = sme_vq(env); + + /* Reject mismatched VL. */ + if (vl != vq * TARGET_SVE_VQ_BYTES) { + return false; + } + + /* Accept empty record -- used to clear PSTATE.ZA. */ + if (size <= TARGET_ZA_SIG_CONTEXT_SIZE(0)) { + return true; + } + + /* Reject non-empty but incomplete record. */ + if (size < TARGET_ZA_SIG_CONTEXT_SIZE(vq)) { + return false; + } + + *svcr = FIELD_DP64(*svcr, SVCR, ZA, 1); + + for (i = 0; i < vl; ++i) { + uint64_t *z = (void *)za + TARGET_ZA_SIG_ZAV_OFFSET(vq, i); + for (j = 0; j < vq * 2; ++j) { + __get_user_e(env->zarray[i].d[j], z + j, le); + } + } + return true; +} + static int target_restore_sigframe(CPUARMState *env, struct target_rt_sigframe *sf) { struct target_aarch64_ctx *ctx, *extra = NULL; struct target_fpsimd_context *fpsimd = NULL; struct target_sve_context *sve = NULL; + struct target_za_context *za = NULL; uint64_t extra_datap = 0; bool used_extra = false; int sve_size = 0; + int za_size = 0; + int svcr = 0; target_restore_general_frame(env, sf); @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, sve_size = size; break; + case TARGET_ZA_MAGIC: + if (za || size < sizeof(struct target_za_context)) { + goto err; + } + za = (struct target_za_context *)ctx; + za_size = size; + break; + case TARGET_EXTRA_MAGIC: if (extra || size != sizeof(struct target_extra_context)) { goto err; @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, } /* SVE data, if present, overwrites FPSIMD data. */ - if (sve && !target_restore_sve_record(env, sve, sve_size)) { + if (sve && !target_restore_sve_record(env, sve, sve_size, &svcr)) { goto err; } + if (za && !target_restore_za_record(env, za, za_size, &svcr)) { + goto err; + } + if (env->svcr != svcr) { + env->svcr = svcr; + arm_rebuild_hflags(env); + } unlock_user(extra, extra_datap, 0); return 0; @@ -XXX,XX +XXX,XX @@ static void target_setup_frame(int usig, struct target_sigaction *ka, .total_size = offsetof(struct target_rt_sigframe, uc.tuc_mcontext.__reserved), }; - int fpsimd_ofs, fr_ofs, sve_ofs = 0, vq = 0, sve_size = 0; + int fpsimd_ofs, fr_ofs, sve_ofs = 0, za_ofs = 0; + int sve_size = 0, za_size = 0; struct target_rt_sigframe *frame; struct target_rt_frame_record *fr; abi_ulong frame_addr, return_addr; @@ -XXX,XX +XXX,XX @@ static void target_setup_frame(int usig, struct target_sigaction *ka, &layout); /* SVE state needs saving only if it exists. */ - if (cpu_isar_feature(aa64_sve, env_archcpu(env))) { - vq = sve_vq(env); - sve_size = QEMU_ALIGN_UP(TARGET_SVE_SIG_CONTEXT_SIZE(vq), 16); + if (cpu_isar_feature(aa64_sve, env_archcpu(env)) || + cpu_isar_feature(aa64_sme, env_archcpu(env))) { + sve_size = QEMU_ALIGN_UP(TARGET_SVE_SIG_CONTEXT_SIZE(sve_vq(env)), 16); sve_ofs = alloc_sigframe_space(sve_size, &layout); } + if (cpu_isar_feature(aa64_sme, env_archcpu(env))) { + /* ZA state needs saving only if it is enabled. */ + if (FIELD_EX64(env->svcr, SVCR, ZA)) { + za_size = TARGET_ZA_SIG_CONTEXT_SIZE(sme_vq(env)); + } else { + za_size = TARGET_ZA_SIG_CONTEXT_SIZE(0); + } + za_ofs = alloc_sigframe_space(za_size, &layout); + } if (layout.extra_ofs) { /* Reserve space for the extra end marker. The standard end marker @@ -XXX,XX +XXX,XX @@ static void target_setup_frame(int usig, struct target_sigaction *ka, target_setup_end_record((void *)frame + layout.extra_end_ofs); } if (sve_ofs) { - target_setup_sve_record((void *)frame + sve_ofs, env, vq, sve_size); + target_setup_sve_record((void *)frame + sve_ofs, env, sve_size); + } + if (za_ofs) { + target_setup_za_record((void *)frame + za_ofs, env, za_size); } /* Set up the stack frame for unwinding. */ @@ -XXX,XX +XXX,XX @@ static void target_setup_frame(int usig, struct target_sigaction *ka, env->btype = 2; } + /* + * Invoke the signal handler with both SM and ZA disabled. + * When clearing SM, ResetSVEState, per SMSTOP. + */ + if (FIELD_EX64(env->svcr, SVCR, SM)) { + arm_reset_sve_state(env); + } + if (env->svcr) { + env->svcr = 0; + arm_rebuild_hflags(env); + } + if (info) { tswap_siginfo(&frame->info, info); env->xregs[1] = frame_addr + offsetof(struct target_rt_sigframe, info); -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Add "sve" to the sve prctl functions, to distinguish them from the coming "sme" prctls with similar names. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-42-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/target_prctl.h | 8 ++++---- linux-user/syscall.c | 12 ++++++------ 2 files changed, 10 insertions(+), 10 deletions(-) diff --git a/linux-user/aarch64/target_prctl.h b/linux-user/aarch64/target_prctl.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/target_prctl.h +++ b/linux-user/aarch64/target_prctl.h @@ -XXX,XX +XXX,XX @@ #ifndef AARCH64_TARGET_PRCTL_H #define AARCH64_TARGET_PRCTL_H -static abi_long do_prctl_get_vl(CPUArchState *env) +static abi_long do_prctl_sve_get_vl(CPUArchState *env) { ARMCPU *cpu = env_archcpu(env); if (cpu_isar_feature(aa64_sve, cpu)) { @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_get_vl(CPUArchState *env) } return -TARGET_EINVAL; } -#define do_prctl_get_vl do_prctl_get_vl +#define do_prctl_sve_get_vl do_prctl_sve_get_vl -static abi_long do_prctl_set_vl(CPUArchState *env, abi_long arg2) +static abi_long do_prctl_sve_set_vl(CPUArchState *env, abi_long arg2) { /* * We cannot support either PR_SVE_SET_VL_ONEXEC or PR_SVE_VL_INHERIT. @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_set_vl(CPUArchState *env, abi_long arg2) } return -TARGET_EINVAL; } -#define do_prctl_set_vl do_prctl_set_vl +#define do_prctl_sve_set_vl do_prctl_sve_set_vl static abi_long do_prctl_reset_keys(CPUArchState *env, abi_long arg2) { diff --git a/linux-user/syscall.c b/linux-user/syscall.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/syscall.c +++ b/linux-user/syscall.c @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_inval1(CPUArchState *env, abi_long arg2) #ifndef do_prctl_set_fp_mode #define do_prctl_set_fp_mode do_prctl_inval1 #endif -#ifndef do_prctl_get_vl -#define do_prctl_get_vl do_prctl_inval0 +#ifndef do_prctl_sve_get_vl +#define do_prctl_sve_get_vl do_prctl_inval0 #endif -#ifndef do_prctl_set_vl -#define do_prctl_set_vl do_prctl_inval1 +#ifndef do_prctl_sve_set_vl +#define do_prctl_sve_set_vl do_prctl_inval1 #endif #ifndef do_prctl_reset_keys #define do_prctl_reset_keys do_prctl_inval1 @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl(CPUArchState *env, abi_long option, abi_long arg2, case PR_SET_FP_MODE: return do_prctl_set_fp_mode(env, arg2); case PR_SVE_GET_VL: - return do_prctl_get_vl(env); + return do_prctl_sve_get_vl(env); case PR_SVE_SET_VL: - return do_prctl_set_vl(env, arg2); + return do_prctl_sve_set_vl(env, arg2); case PR_PAC_RESET_KEYS: if (arg3 || arg4 || arg5) { return -TARGET_EINVAL; -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> These prctl set the Streaming SVE vector length, which may be completely different from the Normal SVE vector length. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-43-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/target_prctl.h | 54 +++++++++++++++++++++++++++++++ linux-user/syscall.c | 16 +++++++++ 2 files changed, 70 insertions(+) diff --git a/linux-user/aarch64/target_prctl.h b/linux-user/aarch64/target_prctl.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/target_prctl.h +++ b/linux-user/aarch64/target_prctl.h @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_sve_get_vl(CPUArchState *env) { ARMCPU *cpu = env_archcpu(env); if (cpu_isar_feature(aa64_sve, cpu)) { + /* PSTATE.SM is always unset on syscall entry. */ return sve_vq(env) * 16; } return -TARGET_EINVAL; @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_sve_set_vl(CPUArchState *env, abi_long arg2) && arg2 >= 0 && arg2 <= 512 * 16 && !(arg2 & 15)) { uint32_t vq, old_vq; + /* PSTATE.SM is always unset on syscall entry. */ old_vq = sve_vq(env); /* @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_sve_set_vl(CPUArchState *env, abi_long arg2) } #define do_prctl_sve_set_vl do_prctl_sve_set_vl +static abi_long do_prctl_sme_get_vl(CPUArchState *env) +{ + ARMCPU *cpu = env_archcpu(env); + if (cpu_isar_feature(aa64_sme, cpu)) { + return sme_vq(env) * 16; + } + return -TARGET_EINVAL; +} +#define do_prctl_sme_get_vl do_prctl_sme_get_vl + +static abi_long do_prctl_sme_set_vl(CPUArchState *env, abi_long arg2) +{ + /* + * We cannot support either PR_SME_SET_VL_ONEXEC or PR_SME_VL_INHERIT. + * Note the kernel definition of sve_vl_valid allows for VQ=512, + * i.e. VL=8192, even though the architectural maximum is VQ=16. + */ + if (cpu_isar_feature(aa64_sme, env_archcpu(env)) + && arg2 >= 0 && arg2 <= 512 * 16 && !(arg2 & 15)) { + int vq, old_vq; + + old_vq = sme_vq(env); + + /* + * Bound the value of vq, so that we know that it fits into + * the 4-bit field in SMCR_EL1. Because PSTATE.SM is cleared + * on syscall entry, we are not modifying the current SVE + * vector length. + */ + vq = MAX(arg2 / 16, 1); + vq = MIN(vq, 16); + env->vfp.smcr_el[1] = + FIELD_DP64(env->vfp.smcr_el[1], SMCR, LEN, vq - 1); + + /* Delay rebuilding hflags until we know if ZA must change. */ + vq = sve_vqm1_for_el_sm(env, 0, true) + 1; + + if (vq != old_vq) { + /* + * PSTATE.ZA state is cleared on any change to SVL. + * We need not call arm_rebuild_hflags because PSTATE.SM was + * cleared on syscall entry, so this hasn't changed VL. + */ + env->svcr = FIELD_DP64(env->svcr, SVCR, ZA, 0); + arm_rebuild_hflags(env); + } + return vq * 16; + } + return -TARGET_EINVAL; +} +#define do_prctl_sme_set_vl do_prctl_sme_set_vl + static abi_long do_prctl_reset_keys(CPUArchState *env, abi_long arg2) { ARMCPU *cpu = env_archcpu(env); diff --git a/linux-user/syscall.c b/linux-user/syscall.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/syscall.c +++ b/linux-user/syscall.c @@ -XXX,XX +XXX,XX @@ abi_long do_arch_prctl(CPUX86State *env, int code, abi_ulong addr) #ifndef PR_SET_SYSCALL_USER_DISPATCH # define PR_SET_SYSCALL_USER_DISPATCH 59 #endif +#ifndef PR_SME_SET_VL +# define PR_SME_SET_VL 63 +# define PR_SME_GET_VL 64 +# define PR_SME_VL_LEN_MASK 0xffff +# define PR_SME_VL_INHERIT (1 << 17) +#endif #include "target_prctl.h" @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_inval1(CPUArchState *env, abi_long arg2) #ifndef do_prctl_set_unalign #define do_prctl_set_unalign do_prctl_inval1 #endif +#ifndef do_prctl_sme_get_vl +#define do_prctl_sme_get_vl do_prctl_inval0 +#endif +#ifndef do_prctl_sme_set_vl +#define do_prctl_sme_set_vl do_prctl_inval1 +#endif static abi_long do_prctl(CPUArchState *env, abi_long option, abi_long arg2, abi_long arg3, abi_long arg4, abi_long arg5) @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl(CPUArchState *env, abi_long option, abi_long arg2, return do_prctl_sve_get_vl(env); case PR_SVE_SET_VL: return do_prctl_sve_set_vl(env, arg2); + case PR_SME_GET_VL: + return do_prctl_sme_get_vl(env); + case PR_SME_SET_VL: + return do_prctl_sme_set_vl(env, arg2); case PR_PAC_RESET_KEYS: if (arg3 || arg4 || arg5) { return -TARGET_EINVAL; -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> There's no reason to set CPACR_EL1.ZEN if SVE disabled. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-44-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void arm_cpu_reset(DeviceState *dev) /* and to the FP/Neon instructions */ env->cp15.cpacr_el1 = FIELD_DP64(env->cp15.cpacr_el1, CPACR_EL1, FPEN, 3); - /* and to the SVE instructions */ - env->cp15.cpacr_el1 = FIELD_DP64(env->cp15.cpacr_el1, - CPACR_EL1, ZEN, 3); - /* with reasonable vector length */ + /* and to the SVE instructions, with default vector length */ if (cpu_isar_feature(aa64_sve, cpu)) { + env->cp15.cpacr_el1 = FIELD_DP64(env->cp15.cpacr_el1, + CPACR_EL1, ZEN, 3); env->vfp.zcr_el[1] = cpu->sve_default_vq - 1; } /* -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Enable SME, TPIDR2_EL0, and FA64 if supported by the cpu. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-45-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void arm_cpu_reset(DeviceState *dev) CPACR_EL1, ZEN, 3); env->vfp.zcr_el[1] = cpu->sve_default_vq - 1; } + /* and for SME instructions, with default vector length, and TPIDR2 */ + if (cpu_isar_feature(aa64_sme, cpu)) { + env->cp15.sctlr_el[1] |= SCTLR_EnTP2; + env->cp15.cpacr_el1 = FIELD_DP64(env->cp15.cpacr_el1, + CPACR_EL1, SMEN, 3); + env->vfp.smcr_el[1] = cpu->sme_default_vq - 1; + if (cpu_isar_feature(aa64_sme_fa64, cpu)) { + env->vfp.smcr_el[1] = FIELD_DP64(env->vfp.smcr_el[1], + SMCR, FA64, 1); + } + } /* * Enable 48-bit address space (TODO: take reserved_va into account). * Enable TBI0 but not TBI1. -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-46-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/elfload.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/linux-user/elfload.c b/linux-user/elfload.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/elfload.c +++ b/linux-user/elfload.c @@ -XXX,XX +XXX,XX @@ enum { ARM_HWCAP2_A64_RNG = 1 << 16, ARM_HWCAP2_A64_BTI = 1 << 17, ARM_HWCAP2_A64_MTE = 1 << 18, + ARM_HWCAP2_A64_ECV = 1 << 19, + ARM_HWCAP2_A64_AFP = 1 << 20, + ARM_HWCAP2_A64_RPRES = 1 << 21, + ARM_HWCAP2_A64_MTE3 = 1 << 22, + ARM_HWCAP2_A64_SME = 1 << 23, + ARM_HWCAP2_A64_SME_I16I64 = 1 << 24, + ARM_HWCAP2_A64_SME_F64F64 = 1 << 25, + ARM_HWCAP2_A64_SME_I8I32 = 1 << 26, + ARM_HWCAP2_A64_SME_F16F32 = 1 << 27, + ARM_HWCAP2_A64_SME_B16F32 = 1 << 28, + ARM_HWCAP2_A64_SME_F32F32 = 1 << 29, + ARM_HWCAP2_A64_SME_FA64 = 1 << 30, }; #define ELF_HWCAP get_elf_hwcap() @@ -XXX,XX +XXX,XX @@ static uint32_t get_elf_hwcap2(void) GET_FEATURE_ID(aa64_rndr, ARM_HWCAP2_A64_RNG); GET_FEATURE_ID(aa64_bti, ARM_HWCAP2_A64_BTI); GET_FEATURE_ID(aa64_mte, ARM_HWCAP2_A64_MTE); + GET_FEATURE_ID(aa64_sme, (ARM_HWCAP2_A64_SME | + ARM_HWCAP2_A64_SME_F32F32 | + ARM_HWCAP2_A64_SME_B16F32 | + ARM_HWCAP2_A64_SME_F16F32 | + ARM_HWCAP2_A64_SME_I8I32)); + GET_FEATURE_ID(aa64_sme_f64f64, ARM_HWCAP2_A64_SME_F64F64); + GET_FEATURE_ID(aa64_sme_i16i64, ARM_HWCAP2_A64_SME_I16I64); + GET_FEATURE_ID(aa64_sme_fa64, ARM_HWCAP2_A64_SME_FA64); return hwcaps; } -- 2.25.1