From: Alistair Francis <alistair.francis@wdc.com> The following changes since commit c77283dd5d79149f4e7e9edd00f65416c648ee59: Merge tag 'pull-request-2025-07-02' of https://gitlab.com/thuth/qemu into staging (2025-07-03 06:01:41 -0400) are available in the Git repository at: https://github.com/alistair23/qemu.git tags/pull-riscv-to-apply-20250704 for you to fetch changes up to dc8bffc4eb0a93d3266cea1b17f8848dea5b915c: target: riscv: Add Svrsw60t59b extension support (2025-07-04 21:09:49 +1000) ---------------------------------------------------------------- Second RISC-V PR for 10.1 * sstc extension fixes * Fix zama16b order in isa_edata_arr * Profile handling fixes * Extend PMP region up to 64 * Remove capital 'Z' CPU properties * Add missing named features * Support atomic instruction fetch (Ziccif) * Add max_satp_mode from host cpu * Extend and configure PMP region count * Fix PPN field of Translation-reponse register * Use qemu_chr_fe_write_all() in DBCN_CONSOLE_WRITE_BYTE * Fix fcvt.s.bf16 NaN box checking * Avoid infinite delay of async xmit function * Device tree reg cleanups * Add Kunminghu CPU and platform * Fix missing exit TB flow for ldff_trans * Fix migration failure when aia is configured as aplic-imsic * Fix MEPC/SEPC bit masking for IALIGN * Add a property to set vill bit on reserved usage of vsetvli instruction * Add Svrsw60t59b extension support ---------------------------------------------------------------- Alexandre Ghiti (1): target: riscv: Add Svrsw60t59b extension support Anton Blanchard (1): target/riscv: Fix fcvt.s.bf16 NaN box checking Charalampos Mitrodimas (2): target/riscv: Fix MEPC/SEPC bit masking for IALIGN tests/tcg/riscv64: Add test for MEPC bit masking Daniel Henrique Barboza (9): target/riscv/cpu.c: fix zama16b order in isa_edata_arr[] target/riscv/tcg: restrict satp_mode changes in cpu_set_profile target/riscv/tcg: decouple profile enablement from user prop target/riscv: add profile->present flag target/riscv: remove capital 'Z' CPU properties target/riscv/cpu.c: add 'sdtrig' in riscv,isa target/riscv/cpu.c: add 'ssstrict' to riscv, isa target/riscv/cpu.c: do better with 'named features' doc target/riscv: use qemu_chr_fe_write_all() in DBCN_CONSOLE_WRITE_BYTE Florian Lugou (1): hw/char: sifive_uart: Avoid infinite delay of async xmit function Huang Borong (2): target/riscv: Add BOSC's Xiangshan Kunminghu CPU hw/riscv: Initial support for BOSC's Xiangshan Kunminghu FPGA prototype Jay Chang (2): target/riscv: Extend PMP region up to 64 target/riscv: Make PMP region count configurable Jim Shu (5): target/riscv: Add the checking into stimecmp write function. hw/intc: riscv_aclint: Fix mtime write for sstc extension target/riscv: Fix VSTIP bit in sstc extension. target/riscv: Enable/Disable S/VS-mode Timer when STCE bit is changed target/riscv: support atomic instruction fetch (Ziccif) Joel Stanley (12): hw/riscv/virt: Fix clint base address type hw/riscv/virt: Use setprop_sized_cells for clint hw/riscv/virt: Use setprop_sized_cells for memory hw/riscv/virt: Use setprop_sized_cells for aplic hw/riscv/virt: Use setprop_sized_cells for aclint hw/riscv/virt: Use setprop_sized_cells for plic hw/riscv/virt: Use setprop_sized_cells for virtio hw/riscv/virt: Use setprop_sized_cells for reset hw/riscv/virt: Use setprop_sized_cells for uart hw/riscv/virt: Use setprop_sized_cells for rtc hw/riscv/virt: Use setprop_sized_cells for iommu hw/riscv/virt: Use setprop_sized_cells for pcie Max Chou (1): target/riscv: rvv: Fix missing exit TB flow for ldff_trans Meng Zhuo (1): target/riscv/kvm: add max_satp_mode from host cpu Nutty Liu (1): hw/riscv/riscv-iommu: Fix PPN field of Translation-reponse register Vasilis Liaskovitis (1): target/riscv: Add a property to set vill bit on reserved usage of vsetvli instruction liu.xuemei1@zte.com.cn (1): migration: Fix migration failure when aia is configured as aplic-imsic MAINTAINERS | 7 + docs/system/riscv/xiangshan-kunminghu.rst | 39 +++++ docs/system/target-riscv.rst | 1 + configs/devices/riscv64-softmmu/default.mak | 1 + hw/riscv/riscv-iommu-bits.h | 1 + include/hw/riscv/xiangshan_kmh.h | 68 +++++++++ target/riscv/cpu-qom.h | 1 + target/riscv/cpu.h | 19 ++- target/riscv/cpu_bits.h | 63 +++++++- target/riscv/helper.h | 2 +- target/riscv/internals.h | 27 ++++ target/riscv/time_helper.h | 1 + target/riscv/cpu_cfg_fields.h.inc | 3 + hw/char/sifive_uart.c | 6 +- hw/intc/riscv_aclint.c | 5 + hw/intc/riscv_aplic.c | 12 +- hw/intc/riscv_imsic.c | 10 +- hw/riscv/riscv-iommu.c | 9 +- hw/riscv/virt.c | 66 ++++----- hw/riscv/xiangshan_kmh.c | 220 ++++++++++++++++++++++++++++ target/riscv/cpu.c | 144 +++++++++++++++--- target/riscv/cpu_helper.c | 3 +- target/riscv/csr.c | 192 +++++++++++++++++++++++- target/riscv/fpu_helper.c | 2 +- target/riscv/kvm/kvm-cpu.c | 18 ++- target/riscv/machine.c | 3 +- target/riscv/op_helper.c | 4 +- target/riscv/pmp.c | 28 ++-- target/riscv/riscv-qmp-cmds.c | 2 +- target/riscv/tcg/tcg-cpu.c | 186 +++++++++++------------ target/riscv/time_helper.c | 65 +++++++- target/riscv/translate.c | 46 ++++-- target/riscv/vector_helper.c | 12 +- target/riscv/insn_trans/trans_rvv.c.inc | 10 +- hw/riscv/Kconfig | 9 ++ hw/riscv/meson.build | 1 + tests/data/acpi/riscv64/virt/RHCT | Bin 400 -> 416 bytes tests/tcg/riscv64/Makefile.softmmu-target | 4 + tests/tcg/riscv64/test-mepc-masking.S | 73 +++++++++ 39 files changed, 1151 insertions(+), 212 deletions(-) create mode 100644 docs/system/riscv/xiangshan-kunminghu.rst create mode 100644 include/hw/riscv/xiangshan_kmh.h create mode 100644 hw/riscv/xiangshan_kmh.c create mode 100644 tests/tcg/riscv64/test-mepc-masking.S From: Jim Shu <jim.shu@sifive.com> Preparation commit to let aclint timer to use stimecmp write function. Aclint timer doesn't call sstc() predicate so we need to check inside the stimecmp write function. Signed-off-by: Jim Shu <jim.shu@sifive.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250519143518.11086-2-jim.shu@sifive.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/time_helper.c | 19 +++++++++++++++++-- 1 file changed, 17 insertions(+), 2 deletions(-) diff --git a/target/riscv/time_helper.c b/target/riscv/time_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/time_helper.c +++ b/target/riscv/time_helper.c @@ -XXX,XX +XXX,XX @@ void riscv_timer_write_timecmp(CPURISCVState *env, QEMUTimer *timer, { uint64_t diff, ns_diff, next; RISCVAclintMTimerState *mtimer = env->rdtime_fn_arg; - uint32_t timebase_freq = mtimer->timebase_freq; - uint64_t rtc_r = env->rdtime_fn(env->rdtime_fn_arg) + delta; + uint32_t timebase_freq; + uint64_t rtc_r; + + if (!riscv_cpu_cfg(env)->ext_sstc || !env->rdtime_fn || + !env->rdtime_fn_arg || !get_field(env->menvcfg, MENVCFG_STCE)) { + /* S/VS Timer IRQ depends on sstc extension, rdtime_fn(), and STCE. */ + return; + } + + if (timer_irq == MIP_VSTIP && + (!riscv_has_ext(env, RVH) || !get_field(env->henvcfg, HENVCFG_STCE))) { + /* VS Timer IRQ also depends on RVH and henvcfg.STCE. */ + return; + } + + timebase_freq = mtimer->timebase_freq; + rtc_r = env->rdtime_fn(env->rdtime_fn_arg) + delta; if (timecmp <= rtc_r) { /* -- 2.50.0 From: Jim Shu <jim.shu@sifive.com> When changing the mtime value, the period of [s|vs]timecmp timers should also be updated, similar to the period of mtimecmp timer. The period of the stimecmp timer is the time until the next S-mode timer IRQ. The value is calculated as "stimecmp - time". [1] It is equal to "stimecmp - mtime" since the time CSR is a read-only shadow of the memory-mapped mtime register. Thus, changing mtime value will update the period of stimecmp timer. Similarly, the period of vstimecmp timer is calculated as "vstimecmp - (mtime + htimedelta)" [2], so changing mtime value will update the period of vstimecmp timer. [1] RISC-V Priv spec ch 9.1.1. Supervisor Timer (stimecmp) Register A supervisor timer interrupt becomes pending, as reflected in the STIP bit in the mip and sip registers whenever time contains a value greater than or equal to stimecmp. [2] RISC-V Priv spec ch19.2.1. Virtual Supervisor Timer (vstimecmp) Register A virtual supervisor timer interrupt becomes pending, as reflected in the VSTIP bit in the hip register, whenever (time + htimedelta), truncated to 64 bits, contains a value greater than or equal to vstimecmp Signed-off-by: Jim Shu <jim.shu@sifive.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250519143518.11086-3-jim.shu@sifive.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/intc/riscv_aclint.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/hw/intc/riscv_aclint.c b/hw/intc/riscv_aclint.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/riscv_aclint.c +++ b/hw/intc/riscv_aclint.c @@ -XXX,XX +XXX,XX @@ #include "qemu/module.h" #include "hw/sysbus.h" #include "target/riscv/cpu.h" +#include "target/riscv/time_helper.h" #include "hw/qdev-properties.h" #include "hw/intc/riscv_aclint.h" #include "qemu/timer.h" @@ -XXX,XX +XXX,XX @@ static void riscv_aclint_mtimer_write(void *opaque, hwaddr addr, riscv_aclint_mtimer_write_timecmp(mtimer, RISCV_CPU(cpu), mtimer->hartid_base + i, mtimer->timecmp[i]); + riscv_timer_write_timecmp(env, env->stimer, env->stimecmp, 0, MIP_STIP); + riscv_timer_write_timecmp(env, env->vstimer, env->vstimecmp, + env->htimedelta, MIP_VSTIP); + } return; } -- 2.50.0 From: Jim Shu <jim.shu@sifive.com> VSTIP is only writable when both [mh]envcfg.STCE is enabled, or it will revert it's defined behavior as if sstc extension is not implemented. Signed-off-by: Jim Shu <jim.shu@sifive.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250519143518.11086-4-jim.shu@sifive.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/csr.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_mip64(CPURISCVState *env, int csrno, if (riscv_cpu_cfg(env)->ext_sstc && (env->priv == PRV_M) && get_field(env->menvcfg, MENVCFG_STCE)) { /* sstc extension forbids STIP & VSTIP to be writeable in mip */ - mask = mask & ~(MIP_STIP | MIP_VSTIP); + + /* STIP is not writable when menvcfg.STCE is enabled. */ + mask = mask & ~MIP_STIP; + + /* VSTIP is not writable when both [mh]envcfg.STCE are enabled. */ + if (get_field(env->henvcfg, HENVCFG_STCE)) { + mask = mask & ~MIP_VSTIP; + } } if (mask) { -- 2.50.0 From: Jim Shu <jim.shu@sifive.com> Updating STCE will enable/disable SSTC in S-mode or/and VS-mode, so we also need to update S/VS-mode Timer and S/VSTIP bits in $mip CSR. Signed-off-by: Jim Shu <jim.shu@sifive.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250519143518.11086-5-jim.shu@sifive.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/time_helper.h | 1 + target/riscv/csr.c | 46 ++++++++++++++++++++++++++++++++++++++ target/riscv/time_helper.c | 46 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 93 insertions(+) diff --git a/target/riscv/time_helper.h b/target/riscv/time_helper.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/time_helper.h +++ b/target/riscv/time_helper.h @@ -XXX,XX +XXX,XX @@ void riscv_timer_write_timecmp(CPURISCVState *env, QEMUTimer *timer, uint64_t timecmp, uint64_t delta, uint32_t timer_irq); +void riscv_timer_stce_changed(CPURISCVState *env, bool is_m_mode, bool enable); void riscv_timer_init(RISCVCPU *cpu); #endif diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static RISCVException write_menvcfg(CPURISCVState *env, int csrno, const RISCVCPUConfig *cfg = riscv_cpu_cfg(env); uint64_t mask = MENVCFG_FIOM | MENVCFG_CBIE | MENVCFG_CBCFE | MENVCFG_CBZE | MENVCFG_CDE; + bool stce_changed = false; if (riscv_cpu_mxl(env) == MXL_RV64) { mask |= (cfg->ext_svpbmt ? MENVCFG_PBMTE : 0) | @@ -XXX,XX +XXX,XX @@ static RISCVException write_menvcfg(CPURISCVState *env, int csrno, if ((val & MENVCFG_DTE) == 0) { env->mstatus &= ~MSTATUS_SDT; } + + if (cfg->ext_sstc && + ((env->menvcfg & MENVCFG_STCE) != (val & MENVCFG_STCE))) { + stce_changed = true; + } } env->menvcfg = (env->menvcfg & ~mask) | (val & mask); + + if (stce_changed) { + riscv_timer_stce_changed(env, true, !!(val & MENVCFG_STCE)); + } + return write_henvcfg(env, CSR_HENVCFG, env->henvcfg, ra); } @@ -XXX,XX +XXX,XX @@ static RISCVException write_menvcfgh(CPURISCVState *env, int csrno, (cfg->ext_smcdeleg ? MENVCFG_CDE : 0) | (cfg->ext_ssdbltrp ? MENVCFG_DTE : 0); uint64_t valh = (uint64_t)val << 32; + bool stce_changed = false; + + if (cfg->ext_sstc && + ((env->menvcfg & MENVCFG_STCE) != (valh & MENVCFG_STCE))) { + stce_changed = true; + } if ((valh & MENVCFG_DTE) == 0) { env->mstatus &= ~MSTATUS_SDT; } env->menvcfg = (env->menvcfg & ~mask) | (valh & mask); + + if (stce_changed) { + riscv_timer_stce_changed(env, true, !!(valh & MENVCFG_STCE)); + } + return write_henvcfgh(env, CSR_HENVCFGH, env->henvcfg >> 32, ra); } @@ -XXX,XX +XXX,XX @@ static RISCVException read_henvcfg(CPURISCVState *env, int csrno, static RISCVException write_henvcfg(CPURISCVState *env, int csrno, target_ulong val, uintptr_t ra) { + const RISCVCPUConfig *cfg = riscv_cpu_cfg(env); uint64_t mask = HENVCFG_FIOM | HENVCFG_CBIE | HENVCFG_CBCFE | HENVCFG_CBZE; RISCVException ret; + bool stce_changed = false; ret = smstateen_acc_ok(env, 0, SMSTATEEN0_HSENVCFG); if (ret != RISCV_EXCP_NONE) { @@ -XXX,XX +XXX,XX @@ static RISCVException write_henvcfg(CPURISCVState *env, int csrno, get_field(val, HENVCFG_PMM) != PMM_FIELD_RESERVED) { mask |= HENVCFG_PMM; } + + if (cfg->ext_sstc && + ((env->henvcfg & HENVCFG_STCE) != (val & HENVCFG_STCE))) { + stce_changed = true; + } } env->henvcfg = val & mask; @@ -XXX,XX +XXX,XX @@ static RISCVException write_henvcfg(CPURISCVState *env, int csrno, env->vsstatus &= ~MSTATUS_SDT; } + if (stce_changed) { + riscv_timer_stce_changed(env, false, !!(val & HENVCFG_STCE)); + } + return RISCV_EXCP_NONE; } @@ -XXX,XX +XXX,XX @@ static RISCVException read_henvcfgh(CPURISCVState *env, int csrno, static RISCVException write_henvcfgh(CPURISCVState *env, int csrno, target_ulong val, uintptr_t ra) { + const RISCVCPUConfig *cfg = riscv_cpu_cfg(env); uint64_t mask = env->menvcfg & (HENVCFG_PBMTE | HENVCFG_STCE | HENVCFG_ADUE | HENVCFG_DTE); uint64_t valh = (uint64_t)val << 32; RISCVException ret; + bool stce_changed = false; ret = smstateen_acc_ok(env, 0, SMSTATEEN0_HSENVCFG); if (ret != RISCV_EXCP_NONE) { return ret; } + + if (cfg->ext_sstc && + ((env->henvcfg & HENVCFG_STCE) != (valh & HENVCFG_STCE))) { + stce_changed = true; + } + env->henvcfg = (env->henvcfg & 0xFFFFFFFF) | (valh & mask); if ((env->henvcfg & HENVCFG_DTE) == 0) { env->vsstatus &= ~MSTATUS_SDT; } + + if (stce_changed) { + riscv_timer_stce_changed(env, false, !!(val & HENVCFG_STCE)); + } + return RISCV_EXCP_NONE; } diff --git a/target/riscv/time_helper.c b/target/riscv/time_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/time_helper.c +++ b/target/riscv/time_helper.c @@ -XXX,XX +XXX,XX @@ void riscv_timer_write_timecmp(CPURISCVState *env, QEMUTimer *timer, timer_mod(timer, next); } +/* + * When disabling xenvcfg.STCE, the S/VS Timer may be disabled at the same time. + * It is safe to call this function regardless of whether the timer has been + * deleted or not. timer_del() will do nothing if the timer has already + * been deleted. + */ +static void riscv_timer_disable_timecmp(CPURISCVState *env, QEMUTimer *timer, + uint32_t timer_irq) +{ + /* Disable S-mode Timer IRQ and HW-based STIP */ + if ((timer_irq == MIP_STIP) && !get_field(env->menvcfg, MENVCFG_STCE)) { + riscv_cpu_update_mip(env, timer_irq, BOOL_TO_MASK(0)); + timer_del(timer); + return; + } + + /* Disable VS-mode Timer IRQ and HW-based VSTIP */ + if ((timer_irq == MIP_VSTIP) && + (!get_field(env->menvcfg, MENVCFG_STCE) || + !get_field(env->henvcfg, HENVCFG_STCE))) { + env->vstime_irq = 0; + riscv_cpu_update_mip(env, 0, BOOL_TO_MASK(0)); + timer_del(timer); + return; + } +} + +/* Enable or disable S/VS-mode Timer when xenvcfg.STCE is changed */ +void riscv_timer_stce_changed(CPURISCVState *env, bool is_m_mode, bool enable) +{ + if (enable) { + riscv_timer_write_timecmp(env, env->vstimer, env->vstimecmp, + env->htimedelta, MIP_VSTIP); + } else { + riscv_timer_disable_timecmp(env, env->vstimer, MIP_VSTIP); + } + + if (is_m_mode) { + if (enable) { + riscv_timer_write_timecmp(env, env->stimer, env->stimecmp, 0, MIP_STIP); + } else { + riscv_timer_disable_timecmp(env, env->stimer, MIP_STIP); + } + } +} + void riscv_timer_init(RISCVCPU *cpu) { CPURISCVState *env; -- 2.50.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Put it after zalrsc and before zawrs. Cc: qemu-trivial@nongnu.org Fixes: a60ce58fd9 ("target/riscv: Support Zama16b extension") Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250522113344.823294-1-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const RISCVIsaExtData isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(zaamo, PRIV_VERSION_1_12_0, ext_zaamo), ISA_EXT_DATA_ENTRY(zabha, PRIV_VERSION_1_13_0, ext_zabha), ISA_EXT_DATA_ENTRY(zacas, PRIV_VERSION_1_12_0, ext_zacas), - ISA_EXT_DATA_ENTRY(zama16b, PRIV_VERSION_1_13_0, ext_zama16b), ISA_EXT_DATA_ENTRY(zalrsc, PRIV_VERSION_1_12_0, ext_zalrsc), + ISA_EXT_DATA_ENTRY(zama16b, PRIV_VERSION_1_13_0, ext_zama16b), ISA_EXT_DATA_ENTRY(zawrs, PRIV_VERSION_1_12_0, ext_zawrs), ISA_EXT_DATA_ENTRY(zfa, PRIV_VERSION_1_12_0, ext_zfa), ISA_EXT_DATA_ENTRY(zfbfmin, PRIV_VERSION_1_12_0, ext_zfbfmin), -- 2.50.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> We're changing 'mmu' to true regardless of whether the profile is being enabled or not, and at the same time we're changing satp_mode to profile->enabled. This will promote a situation where we'll set mmu=on without a virtual memory mode, which is a mistake. Only touch 'mmu' and satp_mode if the profile is being enabled. Suggested-by: Andrew Jones <ajones@ventanamicro.com> Fixes: 55398025e7 ("target/riscv: add satp_mode profile support") Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Björn Töpel <bjorn@rivosinc.com> Tested-by: Björn Töpel <bjorn@rivosinc.com> Message-ID: <20250528184407.1451983-2-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/tcg/tcg-cpu.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/target/riscv/tcg/tcg-cpu.c b/target/riscv/tcg/tcg-cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/tcg/tcg-cpu.c +++ b/target/riscv/tcg/tcg-cpu.c @@ -XXX,XX +XXX,XX @@ static void cpu_set_profile(Object *obj, Visitor *v, const char *name, if (profile->enabled) { cpu->env.priv_ver = profile->priv_spec; - } #ifndef CONFIG_USER_ONLY - if (profile->satp_mode != RISCV_PROFILE_ATTR_UNUSED) { - object_property_set_bool(obj, "mmu", true, NULL); - const char *satp_prop = satp_mode_str(profile->satp_mode, - riscv_cpu_is_32bit(cpu)); - object_property_set_bool(obj, satp_prop, profile->enabled, NULL); - } + if (profile->satp_mode != RISCV_PROFILE_ATTR_UNUSED) { + object_property_set_bool(obj, "mmu", true, NULL); + const char *satp_prop = satp_mode_str(profile->satp_mode, + riscv_cpu_is_32bit(cpu)); + object_property_set_bool(obj, satp_prop, true, NULL); + } #endif + } for (i = 0; misa_bits[i] != 0; i++) { uint32_t bit = misa_bits[i]; -- 2.50.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> We have code in riscv_cpu_add_profiles() to enable a profile right away in case a CPU chose the profile during its cpu_init(). But we're using the user callback option to do so, setting profile->user_set. Create a new helper that does all the grunt work to enable/disable a given profile. Use this new helper in the cases where we want a CPU to be compatible to a certain profile, leaving the user callback to be used exclusively by users. Fixes: fba92a92e3 ("target/riscv: add 'rva22u64' CPU") Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Björn Töpel <bjorn@rivosinc.com> Tested-by: Björn Töpel <bjorn@rivosinc.com> Message-ID: <20250528184407.1451983-3-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/tcg/tcg-cpu.c | 127 +++++++++++++++++++------------------ 1 file changed, 67 insertions(+), 60 deletions(-) diff --git a/target/riscv/tcg/tcg-cpu.c b/target/riscv/tcg/tcg-cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/tcg/tcg-cpu.c +++ b/target/riscv/tcg/tcg-cpu.c @@ -XXX,XX +XXX,XX @@ static bool riscv_cpu_is_generic(Object *cpu_obj) return object_dynamic_cast(cpu_obj, TYPE_RISCV_DYNAMIC_CPU) != NULL; } +static void riscv_cpu_set_profile(RISCVCPU *cpu, + RISCVCPUProfile *profile, + bool enabled) +{ + int i, ext_offset; + + if (profile->u_parent != NULL) { + riscv_cpu_set_profile(cpu, profile->u_parent, enabled); + } + + if (profile->s_parent != NULL) { + riscv_cpu_set_profile(cpu, profile->s_parent, enabled); + } + + profile->enabled = enabled; + + if (profile->enabled) { + cpu->env.priv_ver = profile->priv_spec; + +#ifndef CONFIG_USER_ONLY + if (profile->satp_mode != RISCV_PROFILE_ATTR_UNUSED) { + object_property_set_bool(OBJECT(cpu), "mmu", true, NULL); + const char *satp_prop = satp_mode_str(profile->satp_mode, + riscv_cpu_is_32bit(cpu)); + object_property_set_bool(OBJECT(cpu), satp_prop, true, NULL); + } +#endif + } + + for (i = 0; misa_bits[i] != 0; i++) { + uint32_t bit = misa_bits[i]; + + if (!(profile->misa_ext & bit)) { + continue; + } + + if (bit == RVI && !profile->enabled) { + /* + * Disabling profiles will not disable the base + * ISA RV64I. + */ + continue; + } + + cpu_misa_ext_add_user_opt(bit, profile->enabled); + riscv_cpu_write_misa_bit(cpu, bit, profile->enabled); + } + + for (i = 0; profile->ext_offsets[i] != RISCV_PROFILE_EXT_LIST_END; i++) { + ext_offset = profile->ext_offsets[i]; + + if (profile->enabled) { + if (cpu_cfg_offset_is_named_feat(ext_offset)) { + riscv_cpu_enable_named_feat(cpu, ext_offset); + } + + cpu_bump_multi_ext_priv_ver(&cpu->env, ext_offset); + } + + cpu_cfg_ext_add_user_opt(ext_offset, profile->enabled); + isa_ext_update_enabled(cpu, ext_offset, profile->enabled); + } +} + /* * We'll get here via the following path: * @@ -XXX,XX +XXX,XX @@ static void cpu_set_profile(Object *obj, Visitor *v, const char *name, RISCVCPUProfile *profile = opaque; RISCVCPU *cpu = RISCV_CPU(obj); bool value; - int i, ext_offset; if (riscv_cpu_is_vendor(obj)) { error_setg(errp, "Profile %s is not available for vendor CPUs", @@ -XXX,XX +XXX,XX @@ static void cpu_set_profile(Object *obj, Visitor *v, const char *name, } profile->user_set = true; - profile->enabled = value; - - if (profile->u_parent != NULL) { - object_property_set_bool(obj, profile->u_parent->name, - profile->enabled, NULL); - } - - if (profile->s_parent != NULL) { - object_property_set_bool(obj, profile->s_parent->name, - profile->enabled, NULL); - } - - if (profile->enabled) { - cpu->env.priv_ver = profile->priv_spec; - -#ifndef CONFIG_USER_ONLY - if (profile->satp_mode != RISCV_PROFILE_ATTR_UNUSED) { - object_property_set_bool(obj, "mmu", true, NULL); - const char *satp_prop = satp_mode_str(profile->satp_mode, - riscv_cpu_is_32bit(cpu)); - object_property_set_bool(obj, satp_prop, true, NULL); - } -#endif - } - - for (i = 0; misa_bits[i] != 0; i++) { - uint32_t bit = misa_bits[i]; - - if (!(profile->misa_ext & bit)) { - continue; - } - if (bit == RVI && !profile->enabled) { - /* - * Disabling profiles will not disable the base - * ISA RV64I. - */ - continue; - } - - cpu_misa_ext_add_user_opt(bit, profile->enabled); - riscv_cpu_write_misa_bit(cpu, bit, profile->enabled); - } - - for (i = 0; profile->ext_offsets[i] != RISCV_PROFILE_EXT_LIST_END; i++) { - ext_offset = profile->ext_offsets[i]; - - if (profile->enabled) { - if (cpu_cfg_offset_is_named_feat(ext_offset)) { - riscv_cpu_enable_named_feat(cpu, ext_offset); - } - - cpu_bump_multi_ext_priv_ver(&cpu->env, ext_offset); - } - - cpu_cfg_ext_add_user_opt(ext_offset, profile->enabled); - isa_ext_update_enabled(cpu, ext_offset, profile->enabled); - } + riscv_cpu_set_profile(cpu, profile, value); } static void cpu_get_profile(Object *obj, Visitor *v, const char *name, @@ -XXX,XX +XXX,XX @@ static void cpu_get_profile(Object *obj, Visitor *v, const char *name, static void riscv_cpu_add_profiles(Object *cpu_obj) { for (int i = 0; riscv_profiles[i] != NULL; i++) { - const RISCVCPUProfile *profile = riscv_profiles[i]; + RISCVCPUProfile *profile = riscv_profiles[i]; object_property_add(cpu_obj, profile->name, "bool", cpu_get_profile, cpu_set_profile, @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_profiles(Object *cpu_obj) * case. */ if (profile->enabled) { - object_property_set_bool(cpu_obj, profile->name, true, NULL); + riscv_cpu_set_profile(RISCV_CPU(cpu_obj), profile, true); } } } -- 2.50.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Björn reported in [1] a case where a rv64 CPU is going through the profile code path to enable satp mode. In this case,the amount of extensions on top of the rv64 CPU made it compliant with the RVA22S64 profile during the validation of CPU 0. When the subsequent CPUs were initialized the static profile object has the 'enable' flag set, enabling the profile code path for those CPUs. This happens because we are initializing and realizing each CPU before going to the next, i.e. init and realize CPU0, then init and realize CPU1 and so on. If we change any persistent state during the validation of CPU N it will interfere with the init/realization of CPU N+1. We're using the 'enabled' profile flag to do two distinct things: inform cpu_init() that we want profile extensions to be enabled, and telling QMP that a profile is currently enabled in the CPU. We want to be flexible enough to recognize profile support for all CPUs that has the extension prerequisites, but we do not want to force the profile code path if a profile wasn't set too. Add a new 'present' flag for profiles that will coexist with the 'enabled' flag. Enabling a profile means "we want to switch on all its mandatory extensions". A profile is 'present' if we asserted during validation that the CPU has the needed prerequisites. This means that the case reported by Björn now results in RVA22S64.enabled=false and RVA22S64.present=true. QMP will recognize it as a RVA22 compliant CPU and we won't force the CPU into the profile path. [1] https://lore.kernel.org/qemu-riscv/87y0usiz22.fsf@all.your.base.are.belong.to.us/ Reported-by: Björn Töpel <bjorn@kernel.org> Fixes: 2af005d610 ("target/riscv/tcg: validate profiles during finalize") Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Björn Töpel <bjorn@rivosinc.com> Tested-by: Björn Töpel <bjorn@rivosinc.com> Message-ID: <20250528184407.1451983-4-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 15 +++++++++++++++ target/riscv/riscv-qmp-cmds.c | 2 +- target/riscv/tcg/tcg-cpu.c | 11 +++-------- 3 files changed, 19 insertions(+), 9 deletions(-) diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.h +++ b/target/riscv/cpu.h @@ -XXX,XX +XXX,XX @@ typedef struct riscv_cpu_profile { struct riscv_cpu_profile *s_parent; const char *name; uint32_t misa_ext; + /* + * The profile is enabled/disabled via command line or + * via cpu_init(). Enabling a profile will add all its + * mandatory extensions in the CPU during init(). + */ bool enabled; + /* + * The profile is present in the CPU, i.e. the current set of + * CPU extensions complies with it. A profile can be enabled + * and not present (e.g. the user disabled a mandatory extension) + * and the other way around (e.g. all mandatory extensions are + * present in a non-profile CPU). + * + * QMP uses this flag. + */ + bool present; bool user_set; int priv_spec; int satp_mode; diff --git a/target/riscv/riscv-qmp-cmds.c b/target/riscv/riscv-qmp-cmds.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/riscv-qmp-cmds.c +++ b/target/riscv/riscv-qmp-cmds.c @@ -XXX,XX +XXX,XX @@ static void riscv_obj_add_profiles_qdict(Object *obj, QDict *qdict_out) for (int i = 0; riscv_profiles[i] != NULL; i++) { profile = riscv_profiles[i]; - value = QOBJECT(qbool_from_bool(profile->enabled)); + value = QOBJECT(qbool_from_bool(profile->present)); qdict_put_obj(qdict_out, profile->name, value); } diff --git a/target/riscv/tcg/tcg-cpu.c b/target/riscv/tcg/tcg-cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/tcg/tcg-cpu.c +++ b/target/riscv/tcg/tcg-cpu.c @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_check_parent_profile(RISCVCPU *cpu, RISCVCPUProfile *profile, RISCVCPUProfile *parent) { - const char *parent_name; - bool parent_enabled; - - if (!profile->enabled || !parent) { + if (!profile->present || !parent) { return; } - parent_name = parent->name; - parent_enabled = object_property_get_bool(OBJECT(cpu), parent_name, NULL); - profile->enabled = parent_enabled; + profile->present = parent->present; } static void riscv_cpu_validate_profile(RISCVCPU *cpu, @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_validate_profile(RISCVCPU *cpu, } } - profile->enabled = profile_impl; + profile->present = profile_impl; riscv_cpu_check_parent_profile(cpu, profile, profile->u_parent); riscv_cpu_check_parent_profile(cpu, profile, profile->s_parent); -- 2.50.0 From: Jay Chang <jay.chang@sifive.com> According to the RISC-V Privileged Specification (version >1.12), RV32 supports 16 CSRs (pmpcfg0–pmpcfg15) to configure 64 PMP regions (pmpaddr0–pmpaddr63). Signed-off-by: Jay Chang <jay.chang@sifive.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Message-ID: <20250522081236.4050-2-jay.chang@sifive.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_bits.h | 60 +++++++++++++++++++ target/riscv/csr.c | 124 +++++++++++++++++++++++++++++++++++++++- 2 files changed, 182 insertions(+), 2 deletions(-) diff --git a/target/riscv/cpu_bits.h b/target/riscv/cpu_bits.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_bits.h +++ b/target/riscv/cpu_bits.h @@ -XXX,XX +XXX,XX @@ #define CSR_PMPCFG1 0x3a1 #define CSR_PMPCFG2 0x3a2 #define CSR_PMPCFG3 0x3a3 +#define CSR_PMPCFG4 0x3a4 +#define CSR_PMPCFG5 0x3a5 +#define CSR_PMPCFG6 0x3a6 +#define CSR_PMPCFG7 0x3a7 +#define CSR_PMPCFG8 0x3a8 +#define CSR_PMPCFG9 0x3a9 +#define CSR_PMPCFG10 0x3aa +#define CSR_PMPCFG11 0x3ab +#define CSR_PMPCFG12 0x3ac +#define CSR_PMPCFG13 0x3ad +#define CSR_PMPCFG14 0x3ae +#define CSR_PMPCFG15 0x3af #define CSR_PMPADDR0 0x3b0 #define CSR_PMPADDR1 0x3b1 #define CSR_PMPADDR2 0x3b2 @@ -XXX,XX +XXX,XX @@ #define CSR_PMPADDR13 0x3bd #define CSR_PMPADDR14 0x3be #define CSR_PMPADDR15 0x3bf +#define CSR_PMPADDR16 0x3c0 +#define CSR_PMPADDR17 0x3c1 +#define CSR_PMPADDR18 0x3c2 +#define CSR_PMPADDR19 0x3c3 +#define CSR_PMPADDR20 0x3c4 +#define CSR_PMPADDR21 0x3c5 +#define CSR_PMPADDR22 0x3c6 +#define CSR_PMPADDR23 0x3c7 +#define CSR_PMPADDR24 0x3c8 +#define CSR_PMPADDR25 0x3c9 +#define CSR_PMPADDR26 0x3ca +#define CSR_PMPADDR27 0x3cb +#define CSR_PMPADDR28 0x3cc +#define CSR_PMPADDR29 0x3cd +#define CSR_PMPADDR30 0x3ce +#define CSR_PMPADDR31 0x3cf +#define CSR_PMPADDR32 0x3d0 +#define CSR_PMPADDR33 0x3d1 +#define CSR_PMPADDR34 0x3d2 +#define CSR_PMPADDR35 0x3d3 +#define CSR_PMPADDR36 0x3d4 +#define CSR_PMPADDR37 0x3d5 +#define CSR_PMPADDR38 0x3d6 +#define CSR_PMPADDR39 0x3d7 +#define CSR_PMPADDR40 0x3d8 +#define CSR_PMPADDR41 0x3d9 +#define CSR_PMPADDR42 0x3da +#define CSR_PMPADDR43 0x3db +#define CSR_PMPADDR44 0x3dc +#define CSR_PMPADDR45 0x3dd +#define CSR_PMPADDR46 0x3de +#define CSR_PMPADDR47 0x3df +#define CSR_PMPADDR48 0x3e0 +#define CSR_PMPADDR49 0x3e1 +#define CSR_PMPADDR50 0x3e2 +#define CSR_PMPADDR51 0x3e3 +#define CSR_PMPADDR52 0x3e4 +#define CSR_PMPADDR53 0x3e5 +#define CSR_PMPADDR54 0x3e6 +#define CSR_PMPADDR55 0x3e7 +#define CSR_PMPADDR56 0x3e8 +#define CSR_PMPADDR57 0x3e9 +#define CSR_PMPADDR58 0x3ea +#define CSR_PMPADDR59 0x3eb +#define CSR_PMPADDR60 0x3ec +#define CSR_PMPADDR61 0x3ed +#define CSR_PMPADDR62 0x3ee +#define CSR_PMPADDR63 0x3ef /* RNMI */ #define CSR_MNSCRATCH 0x740 diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_PMPCFG1] = { "pmpcfg1", pmp, read_pmpcfg, write_pmpcfg }, [CSR_PMPCFG2] = { "pmpcfg2", pmp, read_pmpcfg, write_pmpcfg }, [CSR_PMPCFG3] = { "pmpcfg3", pmp, read_pmpcfg, write_pmpcfg }, + [CSR_PMPCFG4] = { "pmpcfg4", pmp, read_pmpcfg, write_pmpcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPCFG5] = { "pmpcfg5", pmp, read_pmpcfg, write_pmpcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPCFG6] = { "pmpcfg6", pmp, read_pmpcfg, write_pmpcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPCFG7] = { "pmpcfg7", pmp, read_pmpcfg, write_pmpcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPCFG8] = { "pmpcfg8", pmp, read_pmpcfg, write_pmpcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPCFG9] = { "pmpcfg9", pmp, read_pmpcfg, write_pmpcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPCFG10] = { "pmpcfg10", pmp, read_pmpcfg, write_pmpcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPCFG11] = { "pmpcfg11", pmp, read_pmpcfg, write_pmpcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPCFG12] = { "pmpcfg12", pmp, read_pmpcfg, write_pmpcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPCFG13] = { "pmpcfg13", pmp, read_pmpcfg, write_pmpcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPCFG14] = { "pmpcfg14", pmp, read_pmpcfg, write_pmpcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPCFG15] = { "pmpcfg15", pmp, read_pmpcfg, write_pmpcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, [CSR_PMPADDR0] = { "pmpaddr0", pmp, read_pmpaddr, write_pmpaddr }, [CSR_PMPADDR1] = { "pmpaddr1", pmp, read_pmpaddr, write_pmpaddr }, [CSR_PMPADDR2] = { "pmpaddr2", pmp, read_pmpaddr, write_pmpaddr }, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_PMPADDR11] = { "pmpaddr11", pmp, read_pmpaddr, write_pmpaddr }, [CSR_PMPADDR12] = { "pmpaddr12", pmp, read_pmpaddr, write_pmpaddr }, [CSR_PMPADDR13] = { "pmpaddr13", pmp, read_pmpaddr, write_pmpaddr }, - [CSR_PMPADDR14] = { "pmpaddr14", pmp, read_pmpaddr, write_pmpaddr }, - [CSR_PMPADDR15] = { "pmpaddr15", pmp, read_pmpaddr, write_pmpaddr }, + [CSR_PMPADDR14] = { "pmpaddr14", pmp, read_pmpaddr, write_pmpaddr }, + [CSR_PMPADDR15] = { "pmpaddr15", pmp, read_pmpaddr, write_pmpaddr }, + [CSR_PMPADDR16] = { "pmpaddr16", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR17] = { "pmpaddr17", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR18] = { "pmpaddr18", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR19] = { "pmpaddr19", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR20] = { "pmpaddr20", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR21] = { "pmpaddr21", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR22] = { "pmpaddr22", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR23] = { "pmpaddr23", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR24] = { "pmpaddr24", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR25] = { "pmpaddr25", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR26] = { "pmpaddr26", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR27] = { "pmpaddr27", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR28] = { "pmpaddr28", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR29] = { "pmpaddr29", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR30] = { "pmpaddr30", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR31] = { "pmpaddr31", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR32] = { "pmpaddr32", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR33] = { "pmpaddr33", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR34] = { "pmpaddr34", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR35] = { "pmpaddr35", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR36] = { "pmpaddr36", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR37] = { "pmpaddr37", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR38] = { "pmpaddr38", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR39] = { "pmpaddr39", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR40] = { "pmpaddr40", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR41] = { "pmpaddr41", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR42] = { "pmpaddr42", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR43] = { "pmpaddr43", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR44] = { "pmpaddr44", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR45] = { "pmpaddr45", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR46] = { "pmpaddr46", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR47] = { "pmpaddr47", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR48] = { "pmpaddr48", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR49] = { "pmpaddr49", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR50] = { "pmpaddr50", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR51] = { "pmpaddr51", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR52] = { "pmpaddr52", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR53] = { "pmpaddr53", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR54] = { "pmpaddr54", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR55] = { "pmpaddr55", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR56] = { "pmpaddr56", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR57] = { "pmpaddr57", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR58] = { "pmpaddr58", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR59] = { "pmpaddr59", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR60] = { "pmpaddr60", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR61] = { "pmpaddr61", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR62] = { "pmpaddr62", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_PMPADDR63] = { "pmpaddr63", pmp, read_pmpaddr, write_pmpaddr, + .min_priv_ver = PRIV_VERSION_1_12_0 }, /* Debug CSRs */ [CSR_TSELECT] = { "tselect", debug, read_tselect, write_tselect }, -- 2.50.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> These properties were deprecated in QEMU 8.2, commit 8043effd9b. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250530134608.1806922-1-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 1 - target/riscv/cpu.c | 17 ----------------- target/riscv/tcg/tcg-cpu.c | 31 +------------------------------ 3 files changed, 1 insertion(+), 48 deletions(-) diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.h +++ b/target/riscv/cpu.h @@ -XXX,XX +XXX,XX @@ extern const RISCVCPUMultiExtConfig riscv_cpu_extensions[]; extern const RISCVCPUMultiExtConfig riscv_cpu_vendor_exts[]; extern const RISCVCPUMultiExtConfig riscv_cpu_experimental_exts[]; extern const RISCVCPUMultiExtConfig riscv_cpu_named_features[]; -extern const RISCVCPUMultiExtConfig riscv_cpu_deprecated_exts[]; typedef struct isa_ext_data { const char *name; diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const RISCVCPUMultiExtConfig riscv_cpu_named_features[] = { { }, }; -/* Deprecated entries marked for future removal */ -const RISCVCPUMultiExtConfig riscv_cpu_deprecated_exts[] = { - MULTI_EXT_CFG_BOOL("Zifencei", ext_zifencei, true), - MULTI_EXT_CFG_BOOL("Zicsr", ext_zicsr, true), - MULTI_EXT_CFG_BOOL("Zihintntl", ext_zihintntl, true), - MULTI_EXT_CFG_BOOL("Zihintpause", ext_zihintpause, true), - MULTI_EXT_CFG_BOOL("Zawrs", ext_zawrs, true), - MULTI_EXT_CFG_BOOL("Zfa", ext_zfa, true), - MULTI_EXT_CFG_BOOL("Zfh", ext_zfh, false), - MULTI_EXT_CFG_BOOL("Zfhmin", ext_zfhmin, false), - MULTI_EXT_CFG_BOOL("Zve32f", ext_zve32f, false), - MULTI_EXT_CFG_BOOL("Zve64f", ext_zve64f, false), - MULTI_EXT_CFG_BOOL("Zve64d", ext_zve64d, false), - - { }, -}; - static void cpu_set_prop_err(RISCVCPU *cpu, const char *propname, Error **errp) { diff --git a/target/riscv/tcg/tcg-cpu.c b/target/riscv/tcg/tcg-cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/tcg/tcg-cpu.c +++ b/target/riscv/tcg/tcg-cpu.c @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_profiles(Object *cpu_obj) } } -static bool cpu_ext_is_deprecated(const char *ext_name) -{ - return isupper(ext_name[0]); -} - -/* - * String will be allocated in the heap. Caller is responsible - * for freeing it. - */ -static char *cpu_ext_to_lower(const char *ext_name) -{ - char *ret = g_malloc0(strlen(ext_name) + 1); - - strcpy(ret, ext_name); - ret[0] = tolower(ret[0]); - - return ret; -} - static void cpu_set_multi_ext_cfg(Object *obj, Visitor *v, const char *name, void *opaque, Error **errp) { @@ -XXX,XX +XXX,XX @@ static void cpu_set_multi_ext_cfg(Object *obj, Visitor *v, const char *name, return; } - if (cpu_ext_is_deprecated(multi_ext_cfg->name)) { - g_autofree char *lower = cpu_ext_to_lower(multi_ext_cfg->name); - - warn_report("CPU property '%s' is deprecated. Please use '%s' instead", - multi_ext_cfg->name, lower); - } - cpu_cfg_ext_add_user_opt(multi_ext_cfg->offset, value); prev_val = isa_ext_is_enabled(cpu, multi_ext_cfg->offset); @@ -XXX,XX +XXX,XX @@ static void cpu_add_multi_ext_prop(Object *cpu_obj, const RISCVCPUMultiExtConfig *multi_cfg) { bool generic_cpu = riscv_cpu_is_generic(cpu_obj); - bool deprecated_ext = cpu_ext_is_deprecated(multi_cfg->name); object_property_add(cpu_obj, multi_cfg->name, "bool", cpu_get_multi_ext_cfg, cpu_set_multi_ext_cfg, NULL, (void *)multi_cfg); - if (!generic_cpu || deprecated_ext) { + if (!generic_cpu) { return; } @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_user_properties(Object *obj) riscv_cpu_add_multiext_prop_array(obj, riscv_cpu_vendor_exts); riscv_cpu_add_multiext_prop_array(obj, riscv_cpu_experimental_exts); - riscv_cpu_add_multiext_prop_array(obj, riscv_cpu_deprecated_exts); - riscv_cpu_add_profiles(obj); } -- 2.50.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> We have support for sdtrig for awhile but we are not advertising it. It is enabled by default via the 'debug' flag. Use the same flag to also advertise sdtrig. Add an exception in disable_priv_spec_isa_exts() to avoid spamming warnings for 'sdtrig' for vendor CPUs like sifive_u. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250604174329.1147549-2-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 1 + target/riscv/tcg/tcg-cpu.c | 9 +++++++++ tests/data/acpi/riscv64/virt/RHCT | Bin 400 -> 406 bytes 3 files changed, 10 insertions(+) diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const RISCVIsaExtData isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(zvkt, PRIV_VERSION_1_12_0, ext_zvkt), ISA_EXT_DATA_ENTRY(zhinx, PRIV_VERSION_1_12_0, ext_zhinx), ISA_EXT_DATA_ENTRY(zhinxmin, PRIV_VERSION_1_12_0, ext_zhinxmin), + ISA_EXT_DATA_ENTRY(sdtrig, PRIV_VERSION_1_12_0, debug), ISA_EXT_DATA_ENTRY(shcounterenw, PRIV_VERSION_1_12_0, has_priv_1_12), ISA_EXT_DATA_ENTRY(sha, PRIV_VERSION_1_12_0, ext_sha), ISA_EXT_DATA_ENTRY(shgatpa, PRIV_VERSION_1_12_0, has_priv_1_12), diff --git a/target/riscv/tcg/tcg-cpu.c b/target/riscv/tcg/tcg-cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/tcg/tcg-cpu.c +++ b/target/riscv/tcg/tcg-cpu.c @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu) continue; } + /* + * cpu.debug = true is marked as 'sdtrig', priv spec 1.12. + * Skip this warning since existing CPUs with older priv + * spec and debug = true will be impacted. + */ + if (!strcmp(edata->name, "sdtrig")) { + continue; + } + isa_ext_update_enabled(cpu, edata->ext_enable_offset, false); /* diff --git a/tests/data/acpi/riscv64/virt/RHCT b/tests/data/acpi/riscv64/virt/RHCT index XXXXXXX..XXXXXXX 100644 Binary files a/tests/data/acpi/riscv64/virt/RHCT and b/tests/data/acpi/riscv64/virt/RHCT differ -- 2.50.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> 'ssstrict' is a RVA23 profile-defined extension defined as follows: "No non-conforming extensions are present. Attempts to execute unimplemented opcodes or access unimplemented CSRs in the standard or reserved encoding spaces raises an illegal instruction exception that results in a contained trap to the supervisor-mode trap handler." In short, we need to throw an exception when accessing unimplemented CSRs or opcodes. We do that, so let's advertise it. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20250529202315.1684198-3-dbarboza@ventanamicro.com> Message-ID: <20250604174329.1147549-3-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 1 + tests/data/acpi/riscv64/virt/RHCT | Bin 406 -> 416 bytes 2 files changed, 1 insertion(+) diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const RISCVIsaExtData isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(ssnpm, PRIV_VERSION_1_13_0, ext_ssnpm), ISA_EXT_DATA_ENTRY(sspm, PRIV_VERSION_1_13_0, ext_sspm), ISA_EXT_DATA_ENTRY(ssstateen, PRIV_VERSION_1_12_0, ext_ssstateen), + ISA_EXT_DATA_ENTRY(ssstrict, PRIV_VERSION_1_12_0, has_priv_1_12), ISA_EXT_DATA_ENTRY(sstc, PRIV_VERSION_1_12_0, ext_sstc), ISA_EXT_DATA_ENTRY(sstvala, PRIV_VERSION_1_12_0, has_priv_1_12), ISA_EXT_DATA_ENTRY(sstvecd, PRIV_VERSION_1_12_0, has_priv_1_12), diff --git a/tests/data/acpi/riscv64/virt/RHCT b/tests/data/acpi/riscv64/virt/RHCT index XXXXXXX..XXXXXXX 100644 Binary files a/tests/data/acpi/riscv64/virt/RHCT and b/tests/data/acpi/riscv64/virt/RHCT differ -- 2.50.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Most of the named features are added directly in isa_edata_arr[], some of them are also added in riscv_cpu_named_features(). There is a reason for that, and the existing docs can do better explaining it. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Message-ID: <20250529202315.1684198-4-dbarboza@ventanamicro.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250604174329.1147549-4-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const RISCVCPUMultiExtConfig riscv_cpu_experimental_exts[] = { * 'Named features' is the name we give to extensions that we * don't want to expose to users. They are either immutable * (always enabled/disable) or they'll vary depending on - * the resulting CPU state. They have riscv,isa strings - * and priv_ver like regular extensions. + * the resulting CPU state. + * + * Some of them are always enabled depending on priv version + * of the CPU and are declared directly in isa_edata_arr[]. + * The ones listed here have special checks during finalize() + * time and require their own flags like regular extensions. + * See riscv_cpu_update_named_features() for more info. */ const RISCVCPUMultiExtConfig riscv_cpu_named_features[] = { MULTI_EXT_CFG_BOOL("zic64b", ext_zic64b, true), MULTI_EXT_CFG_BOOL("ssstateen", ext_ssstateen, true), MULTI_EXT_CFG_BOOL("sha", ext_sha, true), + + /* + * 'ziccrse' has its own flag because the KVM driver + * wants to enable/disable it on its own accord. + */ MULTI_EXT_CFG_BOOL("ziccrse", ext_ziccrse, true), { }, -- 2.50.0 From: Jim Shu <jim.shu@sifive.com> Support 4-byte atomic instruction fetch when instruction is natural aligned. Current implementation is not atomic because it loads instruction twice for first and last 2 bytes. We load 4 bytes at once to keep the atomicity. This instruction preload method only applys when instruction is 4-byte aligned. If instruction is unaligned, it could be across pages so that preload will trigger additional page fault. We encounter this issue when doing pressure test of enabling & disabling Linux kernel ftrace. Ftrace with kernel preemption requires concurrent modification and execution of instruction, so non-atomic instruction fetch will cause the race condition. We may fetch the wrong instruction which is the mixing of 2 instructions. Also, RISC-V Profile wants to provide this feature by HW. RVA20U64 Ziccif protects the atomicity of instruction fetch when it is natural aligned. This commit depends on the atomic read support of translator_ld in the commit 6a9dfe1984b0c593fb0ddb52d4e70832e6201dd6. Signed-off-by: Jim Shu <jim.shu@sifive.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250508094838.19394-1-jim.shu@sifive.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/translate.c | 46 +++++++++++++++++++++++++++++----------- 1 file changed, 34 insertions(+), 12 deletions(-) diff --git a/target/riscv/translate.c b/target/riscv/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/translate.c +++ b/target/riscv/translate.c @@ -XXX,XX +XXX,XX @@ const RISCVDecoder decoder_table[] = { const size_t decoder_table_size = ARRAY_SIZE(decoder_table); -static void decode_opc(CPURISCVState *env, DisasContext *ctx, uint16_t opcode) +static void decode_opc(CPURISCVState *env, DisasContext *ctx) { + uint32_t opcode; + bool pc_is_4byte_align = ((ctx->base.pc_next % 4) == 0); + ctx->virt_inst_excp = false; - ctx->cur_insn_len = insn_len(opcode); + if (pc_is_4byte_align) { + /* + * Load 4 bytes at once to make instruction fetch atomically. + * + * Note: When pc is 4-byte aligned, 4-byte instruction wouldn't be + * across pages. We could preload 4 bytes instruction no matter + * real one is 2 or 4 bytes. Instruction preload wouldn't trigger + * additional page fault. + */ + opcode = translator_ldl(env, &ctx->base, ctx->base.pc_next); + } else { + /* + * For unaligned pc, instruction preload may trigger additional + * page fault so we only load 2 bytes here. + */ + opcode = (uint32_t) translator_lduw(env, &ctx->base, ctx->base.pc_next); + } + ctx->ol = ctx->xl; + + ctx->cur_insn_len = insn_len((uint16_t)opcode); /* Check for compressed insn */ if (ctx->cur_insn_len == 2) { - ctx->opcode = opcode; + ctx->opcode = (uint16_t)opcode; /* * The Zca extension is added as way to refer to instructions in the C * extension that do not include the floating-point loads and stores @@ -XXX,XX +XXX,XX @@ static void decode_opc(CPURISCVState *env, DisasContext *ctx, uint16_t opcode) return; } } else { - uint32_t opcode32 = opcode; - opcode32 = deposit32(opcode32, 16, 16, - translator_lduw(env, &ctx->base, - ctx->base.pc_next + 2)); - ctx->opcode = opcode32; + if (!pc_is_4byte_align) { + /* Load last 2 bytes of instruction here */ + opcode = deposit32(opcode, 16, 16, + translator_lduw(env, &ctx->base, + ctx->base.pc_next + 2)); + } + ctx->opcode = opcode; for (guint i = 0; i < ctx->decoders->len; ++i) { riscv_cpu_decode_fn func = g_ptr_array_index(ctx->decoders, i); - if (func(ctx, opcode32)) { + if (func(ctx, opcode)) { return; } } @@ -XXX,XX +XXX,XX @@ static void riscv_tr_translate_insn(DisasContextBase *dcbase, CPUState *cpu) { DisasContext *ctx = container_of(dcbase, DisasContext, base); CPURISCVState *env = cpu_env(cpu); - uint16_t opcode16 = translator_lduw(env, &ctx->base, ctx->base.pc_next); - ctx->ol = ctx->xl; - decode_opc(env, ctx, opcode16); + decode_opc(env, ctx); ctx->base.pc_next += ctx->cur_insn_len; /* -- 2.50.0 From: Meng Zhuo <mengzhuo@iscas.ac.cn> This patch adds max_satp_mode from host kvm cpu setting. Tested on: Milkv Megrez (Eswin 7700x) Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2931 Signed-off-by: Meng Zhuo <mengzhuo@iscas.ac.cn> Message-ID: <20250606034250.181707-1-mengzhuo@iscas.ac.cn> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/kvm/kvm-cpu.c | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-) diff --git a/target/riscv/kvm/kvm-cpu.c b/target/riscv/kvm/kvm-cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm/kvm-cpu.c +++ b/target/riscv/kvm/kvm-cpu.c @@ -XXX,XX +XXX,XX @@ static void kvm_riscv_destroy_scratch_vcpu(KVMScratchCPU *scratch) close(scratch->kvmfd); } +static void kvm_riscv_init_max_satp_mode(RISCVCPU *cpu, KVMScratchCPU *kvmcpu) +{ + struct kvm_one_reg reg; + int ret; + + reg.id = RISCV_CONFIG_REG(satp_mode); + reg.addr = (uint64_t)&cpu->cfg.max_satp_mode; + ret = ioctl(kvmcpu->cpufd, KVM_GET_ONE_REG, &reg); + if (ret != 0) { + error_report("Unable to retrieve satp mode from host, error %d", ret); + } +} + static void kvm_riscv_init_machine_ids(RISCVCPU *cpu, KVMScratchCPU *kvmcpu) { struct kvm_one_reg reg; @@ -XXX,XX +XXX,XX @@ static void riscv_init_kvm_registers(Object *cpu_obj) kvm_riscv_init_machine_ids(cpu, &kvmcpu); kvm_riscv_init_misa_ext_mask(cpu, &kvmcpu); kvm_riscv_init_cfg(cpu, &kvmcpu); + kvm_riscv_init_max_satp_mode(cpu, &kvmcpu); kvm_riscv_destroy_scratch_vcpu(&kvmcpu); } @@ -XXX,XX +XXX,XX @@ static bool kvm_cpu_realize(CPUState *cs, Error **errp) } } - return true; + return true; } void riscv_kvm_cpu_finalize_features(RISCVCPU *cpu, Error **errp) -- 2.50.0 From: Jay Chang <jay.chang@sifive.com> Previously, the number of PMP regions was hardcoded to 16 in QEMU. This patch replaces the fixed value with a new `pmp_regions` field, allowing platforms to configure the number of PMP regions. If no specific value is provided, the default number of PMP regions remains 16 to preserve the existing behavior. A new CPU parameter num-pmp-regions has been introduced to the QEMU command line. For example: -cpu rv64, g=true, c=true, pmp=true, num-pmp-regions=8 Signed-off-by: Jay Chang <jay.chang@sifive.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250606072525.17313-3-jay.chang@sifive.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 3 +- target/riscv/cpu_cfg_fields.h.inc | 1 + target/riscv/cpu.c | 48 +++++++++++++++++++++++++++++-- target/riscv/csr.c | 5 +++- target/riscv/machine.c | 3 +- target/riscv/pmp.c | 28 ++++++++++++------ 6 files changed, 74 insertions(+), 14 deletions(-) diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.h +++ b/target/riscv/cpu.h @@ -XXX,XX +XXX,XX @@ extern RISCVCPUImpliedExtsRule *riscv_multi_ext_implied_rules[]; #define MMU_USER_IDX 3 -#define MAX_RISCV_PMPS (16) +#define MAX_RISCV_PMPS (64) +#define OLD_MAX_RISCV_PMPS (16) #if !defined(CONFIG_USER_ONLY) #include "pmp.h" diff --git a/target/riscv/cpu_cfg_fields.h.inc b/target/riscv/cpu_cfg_fields.h.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_cfg_fields.h.inc +++ b/target/riscv/cpu_cfg_fields.h.inc @@ -XXX,XX +XXX,XX @@ TYPED_FIELD(uint16_t, elen, 0) TYPED_FIELD(uint16_t, cbom_blocksize, 0) TYPED_FIELD(uint16_t, cbop_blocksize, 0) TYPED_FIELD(uint16_t, cboz_blocksize, 0) +TYPED_FIELD(uint8_t, pmp_regions, 0) TYPED_FIELD(int8_t, max_satp_mode, -1) diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_init(Object *obj) cpu->cfg.cbom_blocksize = 64; cpu->cfg.cbop_blocksize = 64; cpu->cfg.cboz_blocksize = 64; + cpu->cfg.pmp_regions = 16; cpu->env.vext_ver = VEXT_VERSION_1_00_0; cpu->cfg.max_satp_mode = -1; @@ -XXX,XX +XXX,XX @@ static const PropertyInfo prop_pmp = { .set = prop_pmp_set, }; +static void prop_num_pmp_regions_set(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + RISCVCPU *cpu = RISCV_CPU(obj); + uint8_t value; + + visit_type_uint8(v, name, &value, errp); + + if (cpu->cfg.pmp_regions != value && riscv_cpu_is_vendor(obj)) { + cpu_set_prop_err(cpu, name, errp); + return; + } + + if (cpu->env.priv_ver < PRIV_VERSION_1_12_0 && value > OLD_MAX_RISCV_PMPS) { + error_setg(errp, "Number of PMP regions exceeds maximum available"); + return; + } else if (value > MAX_RISCV_PMPS) { + error_setg(errp, "Number of PMP regions exceeds maximum available"); + return; + } + + cpu_option_add_user_setting(name, value); + cpu->cfg.pmp_regions = value; +} + +static void prop_num_pmp_regions_get(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + uint8_t value = RISCV_CPU(obj)->cfg.pmp_regions; + + visit_type_uint8(v, name, &value, errp); +} + +static const PropertyInfo prop_num_pmp_regions = { + .type = "uint8", + .description = "num-pmp-regions", + .get = prop_num_pmp_regions_get, + .set = prop_num_pmp_regions_set, +}; + static int priv_spec_from_str(const char *priv_spec_str) { int priv_version = -1; @@ -XXX,XX +XXX,XX @@ static const Property riscv_cpu_properties[] = { {.name = "mmu", .info = &prop_mmu}, {.name = "pmp", .info = &prop_pmp}, + {.name = "num-pmp-regions", .info = &prop_num_pmp_regions}, {.name = "priv_spec", .info = &prop_priv_spec}, {.name = "vext_spec", .info = &prop_vext_spec}, @@ -XXX,XX +XXX,XX @@ static const TypeInfo riscv_cpu_type_infos[] = { .cfg.max_satp_mode = VM_1_10_MBARE, .cfg.ext_zifencei = true, .cfg.ext_zicsr = true, - .cfg.pmp = true + .cfg.pmp = true, + .cfg.pmp_regions = 8 ), DEFINE_ABSTRACT_RISCV_CPU(TYPE_RISCV_CPU_SIFIVE_U, TYPE_RISCV_VENDOR_CPU, @@ -XXX,XX +XXX,XX @@ static const TypeInfo riscv_cpu_type_infos[] = { .cfg.ext_zifencei = true, .cfg.ext_zicsr = true, .cfg.mmu = true, - .cfg.pmp = true + .cfg.pmp = true, + .cfg.pmp_regions = 8 ), #if defined(TARGET_RISCV32) || \ diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static RISCVException dbltrp_hmode(CPURISCVState *env, int csrno) static RISCVException pmp(CPURISCVState *env, int csrno) { if (riscv_cpu_cfg(env)->pmp) { - if (csrno <= CSR_PMPCFG3) { + int max_pmpcfg = (env->priv_ver >= PRIV_VERSION_1_12_0) ? ++ CSR_PMPCFG15 : CSR_PMPCFG3; + + if (csrno <= max_pmpcfg) { uint32_t reg_index = csrno - CSR_PMPCFG0; /* TODO: RV128 restriction check */ diff --git a/target/riscv/machine.c b/target/riscv/machine.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/machine.c +++ b/target/riscv/machine.c @@ -XXX,XX +XXX,XX @@ static int pmp_post_load(void *opaque, int version_id) RISCVCPU *cpu = opaque; CPURISCVState *env = &cpu->env; int i; + uint8_t pmp_regions = riscv_cpu_cfg(env)->pmp_regions; - for (i = 0; i < MAX_RISCV_PMPS; i++) { + for (i = 0; i < pmp_regions; i++) { pmp_update_rule_addr(env, i); } pmp_update_rule_nums(env); diff --git a/target/riscv/pmp.c b/target/riscv/pmp.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/pmp.c +++ b/target/riscv/pmp.c @@ -XXX,XX +XXX,XX @@ uint32_t pmp_get_num_rules(CPURISCVState *env) */ static inline uint8_t pmp_read_cfg(CPURISCVState *env, uint32_t pmp_index) { - if (pmp_index < MAX_RISCV_PMPS) { + uint8_t pmp_regions = riscv_cpu_cfg(env)->pmp_regions; + + if (pmp_index < pmp_regions) { return env->pmp_state.pmp[pmp_index].cfg_reg; } @@ -XXX,XX +XXX,XX @@ static inline uint8_t pmp_read_cfg(CPURISCVState *env, uint32_t pmp_index) */ static bool pmp_write_cfg(CPURISCVState *env, uint32_t pmp_index, uint8_t val) { - if (pmp_index < MAX_RISCV_PMPS) { + uint8_t pmp_regions = riscv_cpu_cfg(env)->pmp_regions; + + if (pmp_index < pmp_regions) { if (env->pmp_state.pmp[pmp_index].cfg_reg == val) { /* no change */ return false; @@ -XXX,XX +XXX,XX @@ void pmp_update_rule_addr(CPURISCVState *env, uint32_t pmp_index) void pmp_update_rule_nums(CPURISCVState *env) { int i; + uint8_t pmp_regions = riscv_cpu_cfg(env)->pmp_regions; env->pmp_state.num_rules = 0; - for (i = 0; i < MAX_RISCV_PMPS; i++) { + for (i = 0; i < pmp_regions; i++) { const uint8_t a_field = pmp_get_a_field(env->pmp_state.pmp[i].cfg_reg); if (PMP_AMATCH_OFF != a_field) { @@ -XXX,XX +XXX,XX @@ bool pmp_hart_has_privs(CPURISCVState *env, hwaddr addr, int pmp_size = 0; hwaddr s = 0; hwaddr e = 0; + uint8_t pmp_regions = riscv_cpu_cfg(env)->pmp_regions; /* Short cut if no rules */ if (0 == pmp_get_num_rules(env)) { @@ -XXX,XX +XXX,XX @@ bool pmp_hart_has_privs(CPURISCVState *env, hwaddr addr, * 1.10 draft priv spec states there is an implicit order * from low to high */ - for (i = 0; i < MAX_RISCV_PMPS; i++) { + for (i = 0; i < pmp_regions; i++) { s = pmp_is_in_range(env, i, addr); e = pmp_is_in_range(env, i, addr + pmp_size - 1); @@ -XXX,XX +XXX,XX @@ void pmpaddr_csr_write(CPURISCVState *env, uint32_t addr_index, { trace_pmpaddr_csr_write(env->mhartid, addr_index, val); bool is_next_cfg_tor = false; + uint8_t pmp_regions = riscv_cpu_cfg(env)->pmp_regions; - if (addr_index < MAX_RISCV_PMPS) { + if (addr_index < pmp_regions) { if (env->pmp_state.pmp[addr_index].addr_reg == val) { /* no change */ return; @@ -XXX,XX +XXX,XX @@ void pmpaddr_csr_write(CPURISCVState *env, uint32_t addr_index, * In TOR mode, need to check the lock bit of the next pmp * (if there is a next). */ - if (addr_index + 1 < MAX_RISCV_PMPS) { + if (addr_index + 1 < pmp_regions) { uint8_t pmp_cfg = env->pmp_state.pmp[addr_index + 1].cfg_reg; is_next_cfg_tor = PMP_AMATCH_TOR == pmp_get_a_field(pmp_cfg); @@ -XXX,XX +XXX,XX @@ void pmpaddr_csr_write(CPURISCVState *env, uint32_t addr_index, target_ulong pmpaddr_csr_read(CPURISCVState *env, uint32_t addr_index) { target_ulong val = 0; + uint8_t pmp_regions = riscv_cpu_cfg(env)->pmp_regions; - if (addr_index < MAX_RISCV_PMPS) { + if (addr_index < pmp_regions) { val = env->pmp_state.pmp[addr_index].addr_reg; trace_pmpaddr_csr_read(env->mhartid, addr_index, val); } else { @@ -XXX,XX +XXX,XX @@ void mseccfg_csr_write(CPURISCVState *env, target_ulong val) { int i; uint64_t mask = MSECCFG_MMWP | MSECCFG_MML; + uint8_t pmp_regions = riscv_cpu_cfg(env)->pmp_regions; /* Update PMM field only if the value is valid according to Zjpm v1.0 */ if (riscv_cpu_cfg(env)->ext_smmpm && riscv_cpu_mxl(env) == MXL_RV64 && @@ -XXX,XX +XXX,XX @@ void mseccfg_csr_write(CPURISCVState *env, target_ulong val) /* RLB cannot be enabled if it's already 0 and if any regions are locked */ if (!MSECCFG_RLB_ISSET(env)) { - for (i = 0; i < MAX_RISCV_PMPS; i++) { + for (i = 0; i < pmp_regions; i++) { if (pmp_is_locked(env, i)) { val &= ~MSECCFG_RLB; break; @@ -XXX,XX +XXX,XX @@ target_ulong pmp_get_tlb_size(CPURISCVState *env, hwaddr addr) hwaddr tlb_sa = addr & ~(TARGET_PAGE_SIZE - 1); hwaddr tlb_ea = tlb_sa + TARGET_PAGE_SIZE - 1; int i; + uint8_t pmp_regions = riscv_cpu_cfg(env)->pmp_regions; /* * If PMP is not supported or there are no PMP rules, the TLB page will not @@ -XXX,XX +XXX,XX @@ target_ulong pmp_get_tlb_size(CPURISCVState *env, hwaddr addr) return TARGET_PAGE_SIZE; } - for (i = 0; i < MAX_RISCV_PMPS; i++) { + for (i = 0; i < pmp_regions; i++) { if (pmp_get_a_field(env->pmp_state.pmp[i].cfg_reg) == PMP_AMATCH_OFF) { continue; } -- 2.50.0 From: Nutty Liu <liujingqi@lanxincomputing.com> The original implementation incorrectly performed a bitwise AND operation between the PPN of iova and PPN Mask, leading to an incorrect PPN field in Translation-reponse register. The PPN of iova should be set entirely in the PPN field of Translation-reponse register. Also remove the code that was used to clear S field since this field is already zero. Signed-off-by: Nutty Liu <liujingqi@lanxincomputing.com> Reviewed-by: Tomasz Jeznach <tjeznach@rivosinc.com> Message-ID: <20250605124848.1248-1-liujingqi@lanxincomputing.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/riscv-iommu.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/hw/riscv/riscv-iommu.c b/hw/riscv/riscv-iommu.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/riscv-iommu.c +++ b/hw/riscv/riscv-iommu.c @@ -XXX,XX +XXX,XX @@ static void riscv_iommu_process_dbg(RISCVIOMMUState *s) iova = RISCV_IOMMU_TR_RESPONSE_FAULT | (((uint64_t) fault) << 10); } else { iova = iotlb.translated_addr & ~iotlb.addr_mask; - iova >>= TARGET_PAGE_BITS; - iova &= RISCV_IOMMU_TR_RESPONSE_PPN; - - /* We do not support superpages (> 4kbs) for now */ - iova &= ~RISCV_IOMMU_TR_RESPONSE_S; + iova = set_field(0, RISCV_IOMMU_TR_RESPONSE_PPN, PPN_DOWN(iova)); } riscv_iommu_reg_set64(s, RISCV_IOMMU_REG_TR_RESPONSE, iova); } -- 2.50.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> The SBI spec states, for console write byte: "This is a blocking SBI call and it will only return after writing the specified byte to the debug console. It will also return, with SBI_ERR_FAILED, if there are I/O errors." Being a blocker call will either succeed writing the byte or error out, it's feasible to use the blocking qemu_chr_fe_write_all() instead of qemu_chr_fe_write(). Last but not the least, we will duck possible changes in qemu_chr_fe_write() where ret = 0 will have a 'zero byte written' semantic [1] - something that we're not ready to deal in this current state. [1] https://lore.kernel.org/qemu-devel/CAFEAcA_kEndvNtw4EHySXWwQPoGs029yAzZGGBcV=zGHaj7KUQ@mail.gmail.com/ Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-ID: <20250605094456.1385105-2-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/kvm/kvm-cpu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/target/riscv/kvm/kvm-cpu.c b/target/riscv/kvm/kvm-cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm/kvm-cpu.c +++ b/target/riscv/kvm/kvm-cpu.c @@ -XXX,XX +XXX,XX @@ static void kvm_riscv_handle_sbi_dbcn(CPUState *cs, struct kvm_run *run) break; case SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: ch = run->riscv_sbi.args[0]; - ret = qemu_chr_fe_write(serial_hd(0)->be, &ch, sizeof(ch)); + ret = qemu_chr_fe_write_all(serial_hd(0)->be, &ch, sizeof(ch)); if (ret < 0) { error_report("SBI_EXT_DBCN_CONSOLE_WRITE_BYTE: error when " -- 2.50.0 From: Anton Blanchard <antonb@tenstorrent.com> fcvt.s.bf16 uses the FP16 check_nanbox_h() which returns an FP16 quiet NaN. Add check_nanbox_bf16() which returns a BF16 quiet NaN. Signed-off-by: Anton Blanchard <antonb@tenstorrent.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250501114253.594887-1-antonb@tenstorrent.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/internals.h | 16 ++++++++++++++++ target/riscv/fpu_helper.c | 2 +- 2 files changed, 17 insertions(+), 1 deletion(-) diff --git a/target/riscv/internals.h b/target/riscv/internals.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/internals.h +++ b/target/riscv/internals.h @@ -XXX,XX +XXX,XX @@ static inline float16 check_nanbox_h(CPURISCVState *env, uint64_t f) } } +static inline float16 check_nanbox_bf16(CPURISCVState *env, uint64_t f) +{ + /* Disable nanbox check when enable zfinx */ + if (env_archcpu(env)->cfg.ext_zfinx) { + return (uint16_t)f; + } + + uint64_t mask = MAKE_64BIT_MASK(16, 48); + + if (likely((f & mask) == mask)) { + return (uint16_t)f; + } else { + return 0x7FC0u; /* default qnan */ + } +} + #ifndef CONFIG_USER_ONLY /* Our implementation of SysemuCPUOps::has_work */ bool riscv_cpu_has_work(CPUState *cs); diff --git a/target/riscv/fpu_helper.c b/target/riscv/fpu_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/fpu_helper.c +++ b/target/riscv/fpu_helper.c @@ -XXX,XX +XXX,XX @@ uint64_t helper_fcvt_bf16_s(CPURISCVState *env, uint64_t rs1) uint64_t helper_fcvt_s_bf16(CPURISCVState *env, uint64_t rs1) { - float16 frs1 = check_nanbox_h(env, rs1); + float16 frs1 = check_nanbox_bf16(env, rs1); return nanbox_s(env, bfloat16_to_float32(frs1, &env->fp_status)); } -- 2.50.0 From: Florian Lugou <florian.lugou@provenrun.com> The current handler for TXFIFO writes schedules an async callback to pop characters from the queue. When software writes to TXFIFO faster than the async callback delay (100ns), the timer may be pushed back while the previous character has not be dequeued yet. This happens in particular when using -icount with small shift values. This is especially worrysome when software repetitively issues amoor.w instructions (as suggested by SiFive specification) and the FIFO is full, leading to the callback being infinitly pushed back. This commit fixes the issue by never pushing back the timer, only updating it if it is not already active. Signed-off-by: Florian Lugou <florian.lugou@provenrun.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250605101255.797162-1-florian.lugou@provenrun.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/char/sifive_uart.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/hw/char/sifive_uart.c b/hw/char/sifive_uart.c index XXXXXXX..XXXXXXX 100644 --- a/hw/char/sifive_uart.c +++ b/hw/char/sifive_uart.c @@ -XXX,XX +XXX,XX @@ static void sifive_uart_write_tx_fifo(SiFiveUARTState *s, const uint8_t *buf, s->txfifo |= SIFIVE_UART_TXFIFO_FULL; } - timer_mod(s->fifo_trigger_handle, current_time + - TX_INTERRUPT_TRIGGER_DELAY_NS); + if (!timer_pending(s->fifo_trigger_handle)) { + timer_mod(s->fifo_trigger_handle, current_time + + TX_INTERRUPT_TRIGGER_DELAY_NS); + } } static uint64_t -- 2.50.0 From: Joel Stanley <joel@jms.id.au> The address is a hardware address, so use hwaddr for consistency with the rest of the machine. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Joel Stanley <joel@jms.id.au> Message-ID: <20250604025450.85327-2-joel@jms.id.au> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_socket_clint(RISCVVirtState *s, int cpu; g_autofree char *clint_name = NULL; g_autofree uint32_t *clint_cells = NULL; - unsigned long clint_addr; + hwaddr clint_addr; MachineState *ms = MACHINE(s); static const char * const clint_compat[2] = { "sifive,clint0", "riscv,clint0" @@ -XXX,XX +XXX,XX @@ static void create_fdt_socket_clint(RISCVVirtState *s, } clint_addr = s->memmap[VIRT_CLINT].base + - (s->memmap[VIRT_CLINT].size * socket); - clint_name = g_strdup_printf("/soc/clint@%lx", clint_addr); + s->memmap[VIRT_CLINT].size * socket; + clint_name = g_strdup_printf("/soc/clint@%"HWADDR_PRIx, clint_addr); qemu_fdt_add_subnode(ms->fdt, clint_name); qemu_fdt_setprop_string_array(ms->fdt, clint_name, "compatible", (char **)&clint_compat, -- 2.50.0 From: Joel Stanley <joel@jms.id.au> The current device tree property uses two cells for the address (and for the size), but assumes the they are less than 32 bits by hard coding the high cell to zero. Use qemu_fdt_setprop_sized_cells to do the job of splitting the upper and lower 32 bits across cells. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Joel Stanley <joel@jms.id.au> Message-ID: <20250604025450.85327-3-joel@jms.id.au> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_socket_clint(RISCVVirtState *s, qemu_fdt_setprop_string_array(ms->fdt, clint_name, "compatible", (char **)&clint_compat, ARRAY_SIZE(clint_compat)); - qemu_fdt_setprop_cells(ms->fdt, clint_name, "reg", - 0x0, clint_addr, 0x0, s->memmap[VIRT_CLINT].size); + qemu_fdt_setprop_sized_cells(ms->fdt, clint_name, "reg", + 2, clint_addr, 2, s->memmap[VIRT_CLINT].size); qemu_fdt_setprop(ms->fdt, clint_name, "interrupts-extended", clint_cells, s->soc[socket].num_harts * sizeof(uint32_t) * 4); riscv_socket_fdt_write_id(ms, clint_name, socket); -- 2.50.0 From: Joel Stanley <joel@jms.id.au> Use qemu_fdt_setprop_sized_cells to do the job of splitting the upper and lower 32 bits across cells. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Joel Stanley <joel@jms.id.au> Message-ID: <20250604025450.85327-4-joel@jms.id.au> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_socket_memory(RISCVVirtState *s, int socket) size = riscv_socket_mem_size(ms, socket); mem_name = g_strdup_printf("/memory@%"HWADDR_PRIx, addr); qemu_fdt_add_subnode(ms->fdt, mem_name); - qemu_fdt_setprop_cells(ms->fdt, mem_name, "reg", - addr >> 32, addr, size >> 32, size); + qemu_fdt_setprop_sized_cells(ms->fdt, mem_name, "reg", 2, addr, 2, size); qemu_fdt_setprop_string(ms->fdt, mem_name, "device_type", "memory"); riscv_socket_fdt_write_id(ms, mem_name, socket); } -- 2.50.0 From: Joel Stanley <joel@jms.id.au> The current device tree property uses two cells for the address (and for the size), but assumes the they are less than 32 bits by hard coding the high cell to zero. Use qemu_fdt_setprop_sized_cells to do the job of splitting the upper and lower 32 bits across cells. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Joel Stanley <joel@jms.id.au> Message-ID: <20250604025450.85327-5-joel@jms.id.au> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_one_aplic(RISCVVirtState *s, int socket, qemu_fdt_setprop_cell(ms->fdt, aplic_name, "msi-parent", msi_phandle); } - qemu_fdt_setprop_cells(ms->fdt, aplic_name, "reg", - 0x0, aplic_addr, 0x0, aplic_size); + qemu_fdt_setprop_sized_cells(ms->fdt, aplic_name, "reg", + 2, aplic_addr, 2, aplic_size); qemu_fdt_setprop_cell(ms->fdt, aplic_name, "riscv,num-sources", VIRT_IRQCHIP_NUM_SOURCES); -- 2.50.0 From: Joel Stanley <joel@jms.id.au> The current device tree property uses two cells for the address (and for the size), but assumes the they are less than 32 bits by hard coding the high cell to zero. Use qemu_fdt_setprop_sized_cells to do the job of splitting the upper and lower 32 bits across cells. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Joel Stanley <joel@jms.id.au> Message-ID: <20250604025450.85327-6-joel@jms.id.au> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_socket_aclint(RISCVVirtState *s, qemu_fdt_add_subnode(ms->fdt, name); qemu_fdt_setprop_string(ms->fdt, name, "compatible", "riscv,aclint-mswi"); - qemu_fdt_setprop_cells(ms->fdt, name, "reg", - 0x0, addr, 0x0, RISCV_ACLINT_SWI_SIZE); + qemu_fdt_setprop_sized_cells(ms->fdt, name, "reg", + 2, addr, 2, RISCV_ACLINT_SWI_SIZE); qemu_fdt_setprop(ms->fdt, name, "interrupts-extended", aclint_mswi_cells, aclint_cells_size); qemu_fdt_setprop(ms->fdt, name, "interrupt-controller", NULL, 0); @@ -XXX,XX +XXX,XX @@ static void create_fdt_socket_aclint(RISCVVirtState *s, qemu_fdt_add_subnode(ms->fdt, name); qemu_fdt_setprop_string(ms->fdt, name, "compatible", "riscv,aclint-mtimer"); - qemu_fdt_setprop_cells(ms->fdt, name, "reg", - 0x0, addr + RISCV_ACLINT_DEFAULT_MTIME, - 0x0, size - RISCV_ACLINT_DEFAULT_MTIME, - 0x0, addr + RISCV_ACLINT_DEFAULT_MTIMECMP, - 0x0, RISCV_ACLINT_DEFAULT_MTIME); + qemu_fdt_setprop_sized_cells(ms->fdt, name, "reg", + 2, addr + RISCV_ACLINT_DEFAULT_MTIME, + 2, size - RISCV_ACLINT_DEFAULT_MTIME, + 2, addr + RISCV_ACLINT_DEFAULT_MTIMECMP, + 2, RISCV_ACLINT_DEFAULT_MTIME); qemu_fdt_setprop(ms->fdt, name, "interrupts-extended", aclint_mtimer_cells, aclint_cells_size); riscv_socket_fdt_write_id(ms, name, socket); @@ -XXX,XX +XXX,XX @@ static void create_fdt_socket_aclint(RISCVVirtState *s, qemu_fdt_add_subnode(ms->fdt, name); qemu_fdt_setprop_string(ms->fdt, name, "compatible", "riscv,aclint-sswi"); - qemu_fdt_setprop_cells(ms->fdt, name, "reg", - 0x0, addr, 0x0, s->memmap[VIRT_ACLINT_SSWI].size); + qemu_fdt_setprop_sized_cells(ms->fdt, name, "reg", + 2, addr, 2, s->memmap[VIRT_ACLINT_SSWI].size); qemu_fdt_setprop(ms->fdt, name, "interrupts-extended", aclint_sswi_cells, aclint_cells_size); qemu_fdt_setprop(ms->fdt, name, "interrupt-controller", NULL, 0); -- 2.50.0 From: Joel Stanley <joel@jms.id.au> The current device tree property uses two cells for the address (and for the size), but assumes the they are less than 32 bits by hard coding the high cell to zero. Use qemu_fdt_setprop_sized_cells to do the job of splitting the upper and lower 32 bits across cells. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Joel Stanley <joel@jms.id.au> Message-ID: <20250604025450.85327-7-joel@jms.id.au> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_socket_plic(RISCVVirtState *s, s->soc[socket].num_harts * sizeof(uint32_t) * 4); } - qemu_fdt_setprop_cells(ms->fdt, plic_name, "reg", - 0x0, plic_addr, 0x0, s->memmap[VIRT_PLIC].size); + qemu_fdt_setprop_sized_cells(ms->fdt, plic_name, "reg", + 2, plic_addr, 2, s->memmap[VIRT_PLIC].size); qemu_fdt_setprop_cell(ms->fdt, plic_name, "riscv,ndev", VIRT_IRQCHIP_NUM_SOURCES - 1); riscv_socket_fdt_write_id(ms, plic_name, socket); -- 2.50.0 From: Joel Stanley <joel@jms.id.au> The current device tree property uses two cells for the address (and for the size), but assumes the they are less than 32 bits by hard coding the high cell to zero. Use qemu_fdt_setprop_sized_cells to do the job of splitting the upper and lower 32 bits across cells. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Joel Stanley <joel@jms.id.au> Message-ID: <20250604025450.85327-8-joel@jms.id.au> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_virtio(RISCVVirtState *s, uint32_t irq_virtio_phandle) qemu_fdt_add_subnode(ms->fdt, name); qemu_fdt_setprop_string(ms->fdt, name, "compatible", "virtio,mmio"); - qemu_fdt_setprop_cells(ms->fdt, name, "reg", - 0x0, addr, - 0x0, size); + qemu_fdt_setprop_sized_cells(ms->fdt, name, "reg", 2, addr, 2, size); qemu_fdt_setprop_cell(ms->fdt, name, "interrupt-parent", irq_virtio_phandle); if (s->aia_type == VIRT_AIA_TYPE_NONE) { -- 2.50.0 From: Joel Stanley <joel@jms.id.au> The current device tree property uses two cells for the address (and for the size), but assumes the they are less than 32 bits by hard coding the high cell to zero. Use qemu_fdt_setprop_sized_cells to do the job of splitting the upper and lower 32 bits across cells. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Joel Stanley <joel@jms.id.au> Message-ID: <20250604025450.85327-9-joel@jms.id.au> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_reset(RISCVVirtState *s, uint32_t *phandle) qemu_fdt_setprop_string_array(ms->fdt, name, "compatible", (char **)&compat, ARRAY_SIZE(compat)); } - qemu_fdt_setprop_cells(ms->fdt, name, "reg", - 0x0, s->memmap[VIRT_TEST].base, 0x0, s->memmap[VIRT_TEST].size); + qemu_fdt_setprop_sized_cells(ms->fdt, name, "reg", + 2, s->memmap[VIRT_TEST].base, + 2, s->memmap[VIRT_TEST].size); qemu_fdt_setprop_cell(ms->fdt, name, "phandle", test_phandle); test_phandle = qemu_fdt_get_phandle(ms->fdt, name); g_free(name); -- 2.50.0 From: Joel Stanley <joel@jms.id.au> The current device tree property uses two cells for the address (and for the size), but assumes the they are less than 32 bits by hard coding the high cell to zero. Use qemu_fdt_setprop_sized_cells to do the job of splitting the upper and lower 32 bits across cells. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Joel Stanley <joel@jms.id.au> Message-ID: <20250604025450.85327-10-joel@jms.id.au> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_uart(RISCVVirtState *s, s->memmap[VIRT_UART0].base); qemu_fdt_add_subnode(ms->fdt, name); qemu_fdt_setprop_string(ms->fdt, name, "compatible", "ns16550a"); - qemu_fdt_setprop_cells(ms->fdt, name, "reg", - 0x0, s->memmap[VIRT_UART0].base, - 0x0, s->memmap[VIRT_UART0].size); + qemu_fdt_setprop_sized_cells(ms->fdt, name, "reg", + 2, s->memmap[VIRT_UART0].base, + 2, s->memmap[VIRT_UART0].size); qemu_fdt_setprop_cell(ms->fdt, name, "clock-frequency", 3686400); qemu_fdt_setprop_cell(ms->fdt, name, "interrupt-parent", irq_mmio_phandle); if (s->aia_type == VIRT_AIA_TYPE_NONE) { -- 2.50.0 From: Joel Stanley <joel@jms.id.au> The current device tree property uses two cells for the address (and for the size), but assumes the they are less than 32 bits by hard coding the high cell to zero. Use qemu_fdt_setprop_sized_cells to do the job of splitting the upper and lower 32 bits across cells. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Joel Stanley <joel@jms.id.au> Message-ID: <20250604025450.85327-11-joel@jms.id.au> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_rtc(RISCVVirtState *s, qemu_fdt_add_subnode(ms->fdt, name); qemu_fdt_setprop_string(ms->fdt, name, "compatible", "google,goldfish-rtc"); - qemu_fdt_setprop_cells(ms->fdt, name, "reg", - 0x0, s->memmap[VIRT_RTC].base, 0x0, s->memmap[VIRT_RTC].size); + qemu_fdt_setprop_sized_cells(ms->fdt, name, "reg", + 2, s->memmap[VIRT_RTC].base, + 2, s->memmap[VIRT_RTC].size); qemu_fdt_setprop_cell(ms->fdt, name, "interrupt-parent", irq_mmio_phandle); if (s->aia_type == VIRT_AIA_TYPE_NONE) { -- 2.50.0 From: Joel Stanley <joel@jms.id.au> The current device tree property uses two cells for the address (and for the size), but assumes the they are less than 32 bits by hard coding the high cell to zero. Use qemu_fdt_setprop_sized_cells to do the job of splitting the upper and lower 32 bits across cells. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Joel Stanley <joel@jms.id.au> Message-ID: <20250604025450.85327-12-joel@jms.id.au> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_iommu_sys(RISCVVirtState *s, uint32_t irq_chip, qemu_fdt_setprop_cell(fdt, iommu_node, "#iommu-cells", 1); qemu_fdt_setprop_cell(fdt, iommu_node, "phandle", iommu_phandle); - qemu_fdt_setprop_cells(fdt, iommu_node, "reg", - addr >> 32, addr, size >> 32, size); + qemu_fdt_setprop_sized_cells(fdt, iommu_node, "reg", 2, addr, 2, size); qemu_fdt_setprop_cell(fdt, iommu_node, "interrupt-parent", irq_chip); qemu_fdt_setprop_cells(fdt, iommu_node, "interrupts", -- 2.50.0 From: Joel Stanley <joel@jms.id.au> The current device tree property uses two cells for the address (and for the size), but assumes the they are less than 32 bits by hard coding the high cell to zero. Use qemu_fdt_setprop_sized_cells to do the job of splitting the upper and lower 32 bits across cells. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Joel Stanley <joel@jms.id.au> Message-ID: <20250604025450.85327-13-joel@jms.id.au> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_pcie(RISCVVirtState *s, if (s->aia_type == VIRT_AIA_TYPE_APLIC_IMSIC) { qemu_fdt_setprop_cell(ms->fdt, name, "msi-parent", msi_pcie_phandle); } - qemu_fdt_setprop_cells(ms->fdt, name, "reg", 0, - s->memmap[VIRT_PCIE_ECAM].base, 0, s->memmap[VIRT_PCIE_ECAM].size); + qemu_fdt_setprop_sized_cells(ms->fdt, name, "reg", 2, + s->memmap[VIRT_PCIE_ECAM].base, 2, s->memmap[VIRT_PCIE_ECAM].size); qemu_fdt_setprop_sized_cells(ms->fdt, name, "ranges", 1, FDT_PCI_RANGE_IOPORT, 2, 0, 2, s->memmap[VIRT_PCIE_PIO].base, 2, s->memmap[VIRT_PCIE_PIO].size, -- 2.50.0 From: Huang Borong <3543977024@qq.com> Add a CPU entry for the Xiangshan Kunminghu CPU, an open-source, high-performance RISC-V processor. More details can be found at: https://github.com/OpenXiangShan/XiangShan Note: The ISA extensions supported by the Xiangshan Kunminghu CPU are categorized based on four RISC-V specifications: Volume I: Unprivileged Architecture, Volume II: Privileged Architecture, AIA, and RVA23. The extensions within each category are organized according to the chapter order in the specifications. Signed-off-by: Yu Hu <huyu@bosc.ac.cn> Signed-off-by: Ran Wang <wangran@bosc.ac.cn> Signed-off-by: Borong Huang <3543977024@qq.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250425122212.364-1-wangran@bosc.ac.cn> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu-qom.h | 1 + target/riscv/cpu.c | 58 ++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 59 insertions(+) diff --git a/target/riscv/cpu-qom.h b/target/riscv/cpu-qom.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu-qom.h +++ b/target/riscv/cpu-qom.h @@ -XXX,XX +XXX,XX @@ #define TYPE_RISCV_CPU_VEYRON_V1 RISCV_CPU_TYPE_NAME("veyron-v1") #define TYPE_RISCV_CPU_TT_ASCALON RISCV_CPU_TYPE_NAME("tt-ascalon") #define TYPE_RISCV_CPU_XIANGSHAN_NANHU RISCV_CPU_TYPE_NAME("xiangshan-nanhu") +#define TYPE_RISCV_CPU_XIANGSHAN_KMH RISCV_CPU_TYPE_NAME("xiangshan-kunminghu") #define TYPE_RISCV_CPU_HOST RISCV_CPU_TYPE_NAME("host") OBJECT_DECLARE_CPU_TYPE(RISCVCPU, RISCVCPUClass, RISCV_CPU) diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ static const TypeInfo riscv_cpu_type_infos[] = { .cfg.max_satp_mode = VM_1_10_SV39, ), + DEFINE_RISCV_CPU(TYPE_RISCV_CPU_XIANGSHAN_KMH, TYPE_RISCV_VENDOR_CPU, + .misa_mxl_max = MXL_RV64, + .misa_ext = RVG | RVC | RVB | RVS | RVU | RVH | RVV, + .priv_spec = PRIV_VERSION_1_13_0, + /* + * The RISC-V Instruction Set Manual: Volume I + * Unprivileged Architecture + */ + .cfg.ext_zicntr = true, + .cfg.ext_zihpm = true, + .cfg.ext_zihintntl = true, + .cfg.ext_zihintpause = true, + .cfg.ext_zimop = true, + .cfg.ext_zcmop = true, + .cfg.ext_zicond = true, + .cfg.ext_zawrs = true, + .cfg.ext_zacas = true, + .cfg.ext_zfh = true, + .cfg.ext_zfa = true, + .cfg.ext_zcb = true, + .cfg.ext_zbc = true, + .cfg.ext_zvfh = true, + .cfg.ext_zkn = true, + .cfg.ext_zks = true, + .cfg.ext_zkt = true, + .cfg.ext_zvbb = true, + .cfg.ext_zvkt = true, + /* + * The RISC-V Instruction Set Manual: Volume II + * Privileged Architecture + */ + .cfg.ext_smstateen = true, + .cfg.ext_smcsrind = true, + .cfg.ext_sscsrind = true, + .cfg.ext_svnapot = true, + .cfg.ext_svpbmt = true, + .cfg.ext_svinval = true, + .cfg.ext_sstc = true, + .cfg.ext_sscofpmf = true, + .cfg.ext_ssdbltrp = true, + .cfg.ext_ssnpm = true, + .cfg.ext_smnpm = true, + .cfg.ext_smmpm = true, + .cfg.ext_sspm = true, + .cfg.ext_supm = true, + /* The RISC-V Advanced Interrupt Architecture */ + .cfg.ext_smaia = true, + .cfg.ext_ssaia = true, + /* RVA23 Profiles */ + .cfg.ext_zicbom = true, + .cfg.ext_zicbop = true, + .cfg.ext_zicboz = true, + .cfg.ext_svade = true, + .cfg.mmu = true, + .cfg.pmp = true, + .cfg.max_satp_mode = VM_1_10_SV48, + ), + #if defined(CONFIG_TCG) && !defined(CONFIG_USER_ONLY) DEFINE_RISCV_CPU(TYPE_RISCV_CPU_BASE128, TYPE_RISCV_DYNAMIC_CPU, .cfg.max_satp_mode = VM_1_10_SV57, -- 2.50.0 From: Huang Borong <3543977024@qq.com> This implementation provides emulation for the Xiangshan Kunminghu FPGA prototype platform, including support for UART, CLINT, IMSIC, and APLIC devices. More details can be found at https://github.com/OpenXiangShan/XiangShan Signed-off-by: qinshaoqing <qinshaoqing@bosc.ac.cn> Signed-off-by: Yang Wang <wangyang@bosc.ac.cn> Signed-off-by: Yu Hu <819258943@qq.com> Signed-off-by: Ran Wang <wangran@bosc.ac.cn> Signed-off-by: Borong Huang <3543977024@qq.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Message-ID: <20250617074222.17618-1-wangran@bosc.ac.cn> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- MAINTAINERS | 7 + docs/system/riscv/xiangshan-kunminghu.rst | 39 ++++ docs/system/target-riscv.rst | 1 + configs/devices/riscv64-softmmu/default.mak | 1 + include/hw/riscv/xiangshan_kmh.h | 68 ++++++ hw/riscv/xiangshan_kmh.c | 220 ++++++++++++++++++++ hw/riscv/Kconfig | 9 + hw/riscv/meson.build | 1 + 8 files changed, 346 insertions(+) create mode 100644 docs/system/riscv/xiangshan-kunminghu.rst create mode 100644 include/hw/riscv/xiangshan_kmh.h create mode 100644 hw/riscv/xiangshan_kmh.c diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ S: Maintained F: hw/riscv/microblaze-v-generic.c F: docs/system/riscv/microblaze-v-generic.rst +Xiangshan Kunminghu +M: Ran Wang <wangran@bosc.ac.cn> +S: Maintained +F: docs/system/riscv/xiangshan-kunminghu.rst +F: hw/riscv/xiangshan_kmh.c +F: include/hw/riscv/xiangshan_kmh.h + RX Machines ----------- rx-gdbsim diff --git a/docs/system/riscv/xiangshan-kunminghu.rst b/docs/system/riscv/xiangshan-kunminghu.rst new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/docs/system/riscv/xiangshan-kunminghu.rst @@ -XXX,XX +XXX,XX @@ +BOSC Xiangshan Kunminghu FPGA prototype platform (``xiangshan-kunminghu``) +========================================================================== +The ``xiangshan-kunminghu`` machine is compatible with our FPGA prototype +platform. + +XiangShan is an open-source high-performance RISC-V processor project. +The third generation processor is called Kunminghu. Kunminghu is a 64-bit +RV64GCBSUHV processor core. More information can be found in our Github +repository: +https://github.com/OpenXiangShan/XiangShan + +Supported devices +----------------- +The ``xiangshan-kunminghu`` machine supports the following devices: + +* Up to 16 xiangshan-kunminghu cores +* Core Local Interruptor (CLINT) +* Incoming MSI Controller (IMSIC) +* Advanced Platform-Level Interrupt Controller (APLIC) +* 1 UART + +Boot options +------------ +The ``xiangshan-kunminghu`` machine can start using the standard ``-bios`` +functionality for loading the boot image. You need to compile and link +the firmware, kernel, and Device Tree (FDT) into a single binary file, +such as ``fw_payload.bin``. + +Running +------- +Below is an example command line for running the ``xiangshan-kunminghu`` +machine: + +.. code-block:: bash + + $ qemu-system-riscv64 -machine xiangshan-kunminghu \ + -smp 16 -m 16G \ + -bios path/to/opensbi/platform/generic/firmware/fw_payload.bin \ + -nographic diff --git a/docs/system/target-riscv.rst b/docs/system/target-riscv.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/target-riscv.rst +++ b/docs/system/target-riscv.rst @@ -XXX,XX +XXX,XX @@ undocumented; you can get a complete list by running riscv/shakti-c riscv/sifive_u riscv/virt + riscv/xiangshan-kunminghu RISC-V CPU firmware ------------------- diff --git a/configs/devices/riscv64-softmmu/default.mak b/configs/devices/riscv64-softmmu/default.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/devices/riscv64-softmmu/default.mak +++ b/configs/devices/riscv64-softmmu/default.mak @@ -XXX,XX +XXX,XX @@ # CONFIG_RISCV_VIRT=n # CONFIG_MICROCHIP_PFSOC=n # CONFIG_SHAKTI_C=n +# CONFIG_XIANGSHAN_KUNMINGHU=n diff --git a/include/hw/riscv/xiangshan_kmh.h b/include/hw/riscv/xiangshan_kmh.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/riscv/xiangshan_kmh.h @@ -XXX,XX +XXX,XX @@ +/* SPDX-License-Identifier: BSD-2-Clause */ +/* + * QEMU RISC-V Board Compatible with the Xiangshan Kunminghu + * FPGA prototype platform + * + * Copyright (c) 2025 Beijing Institute of Open Source Chip (BOSC) + * + */ + +#ifndef HW_XIANGSHAN_KMH_H +#define HW_XIANGSHAN_KMH_H + +#include "hw/boards.h" +#include "hw/riscv/riscv_hart.h" + +#define XIANGSHAN_KMH_MAX_CPUS 16 + +typedef struct XiangshanKmhSoCState { + /*< private >*/ + DeviceState parent_obj; + + /*< public >*/ + RISCVHartArrayState cpus; + DeviceState *irqchip; + MemoryRegion rom; +} XiangshanKmhSoCState; + +#define TYPE_XIANGSHAN_KMH_SOC "xiangshan.kunminghu.soc" +DECLARE_INSTANCE_CHECKER(XiangshanKmhSoCState, XIANGSHAN_KMH_SOC, + TYPE_XIANGSHAN_KMH_SOC) + +typedef struct XiangshanKmhState { + /*< private >*/ + MachineState parent_obj; + + /*< public >*/ + XiangshanKmhSoCState soc; +} XiangshanKmhState; + +#define TYPE_XIANGSHAN_KMH_MACHINE MACHINE_TYPE_NAME("xiangshan-kunminghu") +DECLARE_INSTANCE_CHECKER(XiangshanKmhState, XIANGSHAN_KMH_MACHINE, + TYPE_XIANGSHAN_KMH_MACHINE) + +enum { + XIANGSHAN_KMH_ROM, + XIANGSHAN_KMH_UART0, + XIANGSHAN_KMH_CLINT, + XIANGSHAN_KMH_APLIC_M, + XIANGSHAN_KMH_APLIC_S, + XIANGSHAN_KMH_IMSIC_M, + XIANGSHAN_KMH_IMSIC_S, + XIANGSHAN_KMH_DRAM, +}; + +enum { + XIANGSHAN_KMH_UART0_IRQ = 10, +}; + +/* Indicating Timebase-freq (1MHZ) */ +#define XIANGSHAN_KMH_CLINT_TIMEBASE_FREQ 1000000 + +#define XIANGSHAN_KMH_IMSIC_NUM_IDS 255 +#define XIANGSHAN_KMH_IMSIC_NUM_GUESTS 7 +#define XIANGSHAN_KMH_IMSIC_GUEST_BITS 3 + +#define XIANGSHAN_KMH_APLIC_NUM_SOURCES 96 + +#endif diff --git a/hw/riscv/xiangshan_kmh.c b/hw/riscv/xiangshan_kmh.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/riscv/xiangshan_kmh.c @@ -XXX,XX +XXX,XX @@ +/* + * QEMU RISC-V Board Compatible with the Xiangshan Kunminghu + * FPGA prototype platform + * + * Copyright (c) 2025 Beijing Institute of Open Source Chip (BOSC) + * SPDX-License-Identifier: GPL-2.0-or-later + * + * Provides a board compatible with the Xiangshan Kunminghu + * FPGA prototype platform: + * + * 0) UART (16550A) + * 1) CLINT (Core-Local Interruptor) + * 2) IMSIC (Incoming MSI Controller) + * 3) APLIC (Advanced Platform-Level Interrupt Controller) + * + * More information can be found in our Github repository: + * https://github.com/OpenXiangShan/XiangShan + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2 or later, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "qapi/error.h" +#include "system/address-spaces.h" +#include "hw/boards.h" +#include "hw/char/serial-mm.h" +#include "hw/intc/riscv_aclint.h" +#include "hw/intc/riscv_aplic.h" +#include "hw/intc/riscv_imsic.h" +#include "hw/qdev-properties.h" +#include "hw/riscv/boot.h" +#include "hw/riscv/xiangshan_kmh.h" +#include "hw/riscv/riscv_hart.h" +#include "system/system.h" + +static const MemMapEntry xiangshan_kmh_memmap[] = { + [XIANGSHAN_KMH_ROM] = { 0x1000, 0xF000 }, + [XIANGSHAN_KMH_UART0] = { 0x310B0000, 0x10000 }, + [XIANGSHAN_KMH_CLINT] = { 0x38000000, 0x10000 }, + [XIANGSHAN_KMH_APLIC_M] = { 0x31100000, 0x4000 }, + [XIANGSHAN_KMH_APLIC_S] = { 0x31120000, 0x4000 }, + [XIANGSHAN_KMH_IMSIC_M] = { 0x3A800000, 0x10000 }, + [XIANGSHAN_KMH_IMSIC_S] = { 0x3B000000, 0x80000 }, + [XIANGSHAN_KMH_DRAM] = { 0x80000000, 0x0 }, +}; + +static DeviceState *xiangshan_kmh_create_aia(uint32_t num_harts) +{ + int i; + const MemMapEntry *memmap = xiangshan_kmh_memmap; + hwaddr addr = 0; + DeviceState *aplic_m = NULL; + + /* M-level IMSICs */ + addr = memmap[XIANGSHAN_KMH_IMSIC_M].base; + for (i = 0; i < num_harts; i++) { + riscv_imsic_create(addr + i * IMSIC_HART_SIZE(0), i, true, + 1, XIANGSHAN_KMH_IMSIC_NUM_IDS); + } + + /* S-level IMSICs */ + addr = memmap[XIANGSHAN_KMH_IMSIC_S].base; + for (i = 0; i < num_harts; i++) { + riscv_imsic_create(addr + + i * IMSIC_HART_SIZE(XIANGSHAN_KMH_IMSIC_GUEST_BITS), + i, false, 1 + XIANGSHAN_KMH_IMSIC_GUEST_BITS, + XIANGSHAN_KMH_IMSIC_NUM_IDS); + } + + /* M-level APLIC */ + aplic_m = riscv_aplic_create(memmap[XIANGSHAN_KMH_APLIC_M].base, + memmap[XIANGSHAN_KMH_APLIC_M].size, + 0, 0, XIANGSHAN_KMH_APLIC_NUM_SOURCES, + 1, true, true, NULL); + + /* S-level APLIC */ + riscv_aplic_create(memmap[XIANGSHAN_KMH_APLIC_S].base, + memmap[XIANGSHAN_KMH_APLIC_S].size, + 0, 0, XIANGSHAN_KMH_APLIC_NUM_SOURCES, + 1, true, false, aplic_m); + + return aplic_m; +} + +static void xiangshan_kmh_soc_realize(DeviceState *dev, Error **errp) +{ + MachineState *ms = MACHINE(qdev_get_machine()); + XiangshanKmhSoCState *s = XIANGSHAN_KMH_SOC(dev); + const MemMapEntry *memmap = xiangshan_kmh_memmap; + MemoryRegion *system_memory = get_system_memory(); + uint32_t num_harts = ms->smp.cpus; + + qdev_prop_set_uint32(DEVICE(&s->cpus), "num-harts", num_harts); + qdev_prop_set_uint32(DEVICE(&s->cpus), "hartid-base", 0); + qdev_prop_set_string(DEVICE(&s->cpus), "cpu-type", + TYPE_RISCV_CPU_XIANGSHAN_KMH); + sysbus_realize(SYS_BUS_DEVICE(&s->cpus), &error_fatal); + + /* AIA */ + s->irqchip = xiangshan_kmh_create_aia(num_harts); + + /* UART */ + serial_mm_init(system_memory, memmap[XIANGSHAN_KMH_UART0].base, 2, + qdev_get_gpio_in(s->irqchip, XIANGSHAN_KMH_UART0_IRQ), + 115200, serial_hd(0), DEVICE_LITTLE_ENDIAN); + + /* CLINT */ + riscv_aclint_swi_create(memmap[XIANGSHAN_KMH_CLINT].base, + 0, num_harts, false); + riscv_aclint_mtimer_create(memmap[XIANGSHAN_KMH_CLINT].base + + RISCV_ACLINT_SWI_SIZE, + RISCV_ACLINT_DEFAULT_MTIMER_SIZE, + 0, num_harts, RISCV_ACLINT_DEFAULT_MTIMECMP, + RISCV_ACLINT_DEFAULT_MTIME, + XIANGSHAN_KMH_CLINT_TIMEBASE_FREQ, true); + + /* ROM */ + memory_region_init_rom(&s->rom, OBJECT(dev), "xiangshan.kunminghu.rom", + memmap[XIANGSHAN_KMH_ROM].size, &error_fatal); + memory_region_add_subregion(system_memory, + memmap[XIANGSHAN_KMH_ROM].base, &s->rom); +} + +static void xiangshan_kmh_soc_class_init(ObjectClass *klass, const void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->realize = xiangshan_kmh_soc_realize; + dc->user_creatable = false; +} + +static void xiangshan_kmh_soc_instance_init(Object *obj) +{ + XiangshanKmhSoCState *s = XIANGSHAN_KMH_SOC(obj); + + object_initialize_child(obj, "cpus", &s->cpus, TYPE_RISCV_HART_ARRAY); +} + +static const TypeInfo xiangshan_kmh_soc_info = { + .name = TYPE_XIANGSHAN_KMH_SOC, + .parent = TYPE_DEVICE, + .instance_size = sizeof(XiangshanKmhSoCState), + .instance_init = xiangshan_kmh_soc_instance_init, + .class_init = xiangshan_kmh_soc_class_init, +}; + +static void xiangshan_kmh_soc_register_types(void) +{ + type_register_static(&xiangshan_kmh_soc_info); +} +type_init(xiangshan_kmh_soc_register_types) + +static void xiangshan_kmh_machine_init(MachineState *machine) +{ + XiangshanKmhState *s = XIANGSHAN_KMH_MACHINE(machine); + const MemMapEntry *memmap = xiangshan_kmh_memmap; + MemoryRegion *system_memory = get_system_memory(); + hwaddr start_addr = memmap[XIANGSHAN_KMH_DRAM].base; + + /* Initialize SoC */ + object_initialize_child(OBJECT(machine), "soc", &s->soc, + TYPE_XIANGSHAN_KMH_SOC); + qdev_realize(DEVICE(&s->soc), NULL, &error_fatal); + + /* Register RAM */ + memory_region_add_subregion(system_memory, + memmap[XIANGSHAN_KMH_DRAM].base, + machine->ram); + + /* ROM reset vector */ + riscv_setup_rom_reset_vec(machine, &s->soc.cpus, + start_addr, + memmap[XIANGSHAN_KMH_ROM].base, + memmap[XIANGSHAN_KMH_ROM].size, 0, 0); + if (machine->firmware) { + riscv_load_firmware(machine->firmware, &start_addr, NULL); + } + + /* Note: dtb has been integrated into firmware(OpenSBI) when compiling */ +} + +static void xiangshan_kmh_machine_class_init(ObjectClass *klass, const void *data) +{ + MachineClass *mc = MACHINE_CLASS(klass); + static const char *const valid_cpu_types[] = { + TYPE_RISCV_CPU_XIANGSHAN_KMH, + NULL + }; + + mc->desc = "RISC-V Board compatible with the Xiangshan " \ + "Kunminghu FPGA prototype platform"; + mc->init = xiangshan_kmh_machine_init; + mc->max_cpus = XIANGSHAN_KMH_MAX_CPUS; + mc->default_cpu_type = TYPE_RISCV_CPU_XIANGSHAN_KMH; + mc->valid_cpu_types = valid_cpu_types; + mc->default_ram_id = "xiangshan.kunminghu.ram"; +} + +static const TypeInfo xiangshan_kmh_machine_info = { + .name = TYPE_XIANGSHAN_KMH_MACHINE, + .parent = TYPE_MACHINE, + .instance_size = sizeof(XiangshanKmhState), + .class_init = xiangshan_kmh_machine_class_init, +}; + +static void xiangshan_kmh_machine_register_types(void) +{ + type_register_static(&xiangshan_kmh_machine_info); +} +type_init(xiangshan_kmh_machine_register_types) diff --git a/hw/riscv/Kconfig b/hw/riscv/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/Kconfig +++ b/hw/riscv/Kconfig @@ -XXX,XX +XXX,XX @@ config SPIKE select HTIF select RISCV_ACLINT select SIFIVE_PLIC + +config XIANGSHAN_KUNMINGHU + bool + default y + depends on RISCV64 + select RISCV_ACLINT + select RISCV_APLIC + select RISCV_IMSIC + select SERIAL_MM diff --git a/hw/riscv/meson.build b/hw/riscv/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/meson.build +++ b/hw/riscv/meson.build @@ -XXX,XX +XXX,XX @@ riscv_ss.add(when: 'CONFIG_ACPI', if_true: files('virt-acpi-build.c')) riscv_ss.add(when: 'CONFIG_RISCV_IOMMU', if_true: files( 'riscv-iommu.c', 'riscv-iommu-pci.c', 'riscv-iommu-sys.c', 'riscv-iommu-hpm.c')) riscv_ss.add(when: 'CONFIG_MICROBLAZE_V', if_true: files('microblaze-v-generic.c')) +riscv_ss.add(when: 'CONFIG_XIANGSHAN_KUNMINGHU', if_true: files('xiangshan_kmh.c')) hw_arch += {'riscv': riscv_ss} -- 2.50.0 From: Max Chou <max.chou@sifive.com> According to the V spec, the vector fault-only-first load instructions may change the VL CSR. So the ldff_trans TCG translation function should generate the lookup_and_goto_ptr flow as the vsetvl/vsetvli translation function to make sure the vl_eq_vlmax TB flag is correct. Signed-off-by: Max Chou <max.chou@sifive.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-ID: <20250627133013.443997-1-max.chou@sifive.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/insn_trans/trans_rvv.c.inc | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn_trans/trans_rvv.c.inc +++ b/target/riscv/insn_trans/trans_rvv.c.inc @@ -XXX,XX +XXX,XX @@ static bool ldff_trans(uint32_t vd, uint32_t rs1, uint32_t data, fn(dest, mask, base, tcg_env, desc); finalize_rvv_inst(s); + + /* vector unit-stride fault-only-first load may modify vl CSR */ + gen_update_pc(s, s->cur_insn_len); + lookup_and_goto_ptr(s); + s->base.is_jmp = DISAS_NORETURN; + return true; } -- 2.50.0 From: "liu.xuemei1@zte.com.cn" <liu.xuemei1@zte.com.cn> Address an error in migration when aia is configured as 'aplic-imsic' in riscv kvm vm by adding riscv_aplic_state_needed() and riscv_imsic_state_needed() to determine whether the corresponding sates are needed. Previously, the fields in the vmsds of 'riscv_aplic' and 'riscv_imsic' can only be initialized under certain special conditions in commit 95a97b3fd2. However, the corresponding ses of these vmsds are inserted into the savevm_state.handlers unconditionally. This led to migration failure characterized by uninitialized fields when save vm state: qemu-system-riscv64: ../migration/vmstate.c:433: vmstate_save_state_v: Assertion 'first_elem || !n_elems || !size' failed. Fixes: 95a97b3fd2 ("target/riscv: update APLIC and IMSIC to support KVM AIA") Signed-off-by: Xuemei Liu <liu.xuemei1@zte.com.cn> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20250616150034827wuHs_ffe3Qm8cqFXT7HeW@zte.com.cn> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/intc/riscv_aplic.c | 12 ++++++++++-- hw/intc/riscv_imsic.c | 10 ++++++++-- 2 files changed, 18 insertions(+), 4 deletions(-) diff --git a/hw/intc/riscv_aplic.c b/hw/intc/riscv_aplic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/riscv_aplic.c +++ b/hw/intc/riscv_aplic.c @@ -XXX,XX +XXX,XX @@ static const Property riscv_aplic_properties[] = { DEFINE_PROP_BOOL("mmode", RISCVAPLICState, mmode, 0), }; +static bool riscv_aplic_state_needed(void *opaque) +{ + RISCVAPLICState *aplic = opaque; + + return riscv_use_emulated_aplic(aplic->msimode); +} + static const VMStateDescription vmstate_riscv_aplic = { .name = "riscv_aplic", - .version_id = 2, - .minimum_version_id = 2, + .version_id = 3, + .minimum_version_id = 3, + .needed = riscv_aplic_state_needed, .fields = (const VMStateField[]) { VMSTATE_UINT32(domaincfg, RISCVAPLICState), VMSTATE_UINT32(mmsicfgaddr, RISCVAPLICState), diff --git a/hw/intc/riscv_imsic.c b/hw/intc/riscv_imsic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/riscv_imsic.c +++ b/hw/intc/riscv_imsic.c @@ -XXX,XX +XXX,XX @@ static const Property riscv_imsic_properties[] = { DEFINE_PROP_UINT32("num-irqs", RISCVIMSICState, num_irqs, 0), }; +static bool riscv_imsic_state_needed(void *opaque) +{ + return !kvm_irqchip_in_kernel(); +} + static const VMStateDescription vmstate_riscv_imsic = { .name = "riscv_imsic", - .version_id = 1, - .minimum_version_id = 1, + .version_id = 2, + .minimum_version_id = 2, + .needed = riscv_imsic_state_needed, .fields = (const VMStateField[]) { VMSTATE_VARRAY_UINT32(eidelivery, RISCVIMSICState, num_pages, 0, -- 2.50.0 From: Charalampos Mitrodimas <charmitro@posteo.net> According to the RISC-V Privileged Architecture specification, the low bit of MEPC/SEPC must always be zero. When IALIGN=32, the two low bits must be zero. This commit fixes the behavior of MEPC/SEPC CSR reads and writes, and the implicit reads by MRET/SRET instructions to properly mask the lowest bit(s) based on whether the C extension is enabled: - When C extension is enabled (IALIGN=16): mask bit 0 - When C extension is disabled (IALIGN=32): mask bits [1:0] Previously, when vectored mode bits from STVEC (which sets bit 0 for vectored mode) were written to MEPC, the bits would not be cleared correctly, causing incorrect behavior on MRET. Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2855 Signed-off-by: Charalampos Mitrodimas <charmitro@posteo.net> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Message-ID: <20250703182157.281320-2-charmitro@posteo.net> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/internals.h | 11 +++++++++++ target/riscv/csr.c | 8 ++++---- target/riscv/op_helper.c | 4 ++-- 3 files changed, 17 insertions(+), 6 deletions(-) diff --git a/target/riscv/internals.h b/target/riscv/internals.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/internals.h +++ b/target/riscv/internals.h @@ -XXX,XX +XXX,XX @@ static inline float16 check_nanbox_bf16(CPURISCVState *env, uint64_t f) } } +static inline target_ulong get_xepc_mask(CPURISCVState *env) +{ + /* When IALIGN=32, both low bits must be zero. + * When IALIGN=16 (has C extension), only bit 0 must be zero. */ + if (riscv_has_ext(env, RVC)) { + return ~(target_ulong)1; + } else { + return ~(target_ulong)3; + } +} + #ifndef CONFIG_USER_ONLY /* Our implementation of SysemuCPUOps::has_work */ bool riscv_cpu_has_work(CPUState *cs); diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static RISCVException write_mscratch(CPURISCVState *env, int csrno, static RISCVException read_mepc(CPURISCVState *env, int csrno, target_ulong *val) { - *val = env->mepc; + *val = env->mepc & get_xepc_mask(env); return RISCV_EXCP_NONE; } static RISCVException write_mepc(CPURISCVState *env, int csrno, target_ulong val, uintptr_t ra) { - env->mepc = val; + env->mepc = val & get_xepc_mask(env); return RISCV_EXCP_NONE; } @@ -XXX,XX +XXX,XX @@ static RISCVException write_sscratch(CPURISCVState *env, int csrno, static RISCVException read_sepc(CPURISCVState *env, int csrno, target_ulong *val) { - *val = env->sepc; + *val = env->sepc & get_xepc_mask(env); return RISCV_EXCP_NONE; } static RISCVException write_sepc(CPURISCVState *env, int csrno, target_ulong val, uintptr_t ra) { - env->sepc = val; + env->sepc = val & get_xepc_mask(env); return RISCV_EXCP_NONE; } diff --git a/target/riscv/op_helper.c b/target/riscv/op_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/op_helper.c +++ b/target/riscv/op_helper.c @@ -XXX,XX +XXX,XX @@ target_ulong helper_sret(CPURISCVState *env) riscv_raise_exception(env, RISCV_EXCP_ILLEGAL_INST, GETPC()); } - target_ulong retpc = env->sepc; + target_ulong retpc = env->sepc & get_xepc_mask(env); if (!riscv_cpu_allow_16bit_insn(&env_archcpu(env)->cfg, env->priv_ver, env->misa_ext) && (retpc & 0x3)) { @@ -XXX,XX +XXX,XX @@ static target_ulong ssdbltrp_mxret(CPURISCVState *env, target_ulong mstatus, target_ulong helper_mret(CPURISCVState *env) { - target_ulong retpc = env->mepc; + target_ulong retpc = env->mepc & get_xepc_mask(env); uint64_t mstatus = env->mstatus; target_ulong prev_priv = get_field(mstatus, MSTATUS_MPP); -- 2.50.0 From: Charalampos Mitrodimas <charmitro@posteo.net> Add a regression test to verify that MEPC properly masks the lower bits when an address with mode bits is written to it, as required by the RISC-V Privileged Architecture specification. The test sets STVEC to an address with bit 0 set (vectored mode), triggers an illegal instruction exception, copies STVEC to MEPC in the trap handler, and verifies that MEPC masks bits [1:0] correctly for IALIGN=32. Without the fix, MEPC retains the mode bits (returns non-zero/FAIL). With the fix, MEPC clears bits [1:0] (returns 0/PASS). Signed-off-by: Charalampos Mitrodimas <charmitro@posteo.net> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Message-ID: <20250703182157.281320-3-charmitro@posteo.net> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- tests/tcg/riscv64/Makefile.softmmu-target | 4 ++ tests/tcg/riscv64/test-mepc-masking.S | 73 +++++++++++++++++++++++ 2 files changed, 77 insertions(+) create mode 100644 tests/tcg/riscv64/test-mepc-masking.S diff --git a/tests/tcg/riscv64/Makefile.softmmu-target b/tests/tcg/riscv64/Makefile.softmmu-target index XXXXXXX..XXXXXXX 100644 --- a/tests/tcg/riscv64/Makefile.softmmu-target +++ b/tests/tcg/riscv64/Makefile.softmmu-target @@ -XXX,XX +XXX,XX @@ EXTRA_RUNS += run-issue1060 run-issue1060: issue1060 $(call run-test, $<, $(QEMU) $(QEMU_OPTS)$<) +EXTRA_RUNS += run-test-mepc-masking +run-test-mepc-masking: test-mepc-masking + $(call run-test, $<, $(QEMU) $(QEMU_OPTS)$<) + # We don't currently support the multiarch system tests undefine MULTIARCH_TESTS diff --git a/tests/tcg/riscv64/test-mepc-masking.S b/tests/tcg/riscv64/test-mepc-masking.S new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/tcg/riscv64/test-mepc-masking.S @@ -XXX,XX +XXX,XX @@ +/* + * Test for MEPC masking bug fix + * + * This test verifies that MEPC properly masks the lower bits according + * to the RISC-V specification when vectored mode bits from STVEC are + * written to MEPC. + */ + + .option norvc + + .text + .global _start +_start: + /* Set up machine trap vector */ + lla t0, machine_trap_handler + csrw mtvec, t0 + + /* Set STVEC with vectored mode (mode bits = 01) */ + li t0, 0x80004001 + csrw stvec, t0 + + /* Clear medeleg to handle exceptions in M-mode */ + csrw medeleg, zero + + /* Trigger illegal instruction exception */ + .word 0xffffffff + +test_completed: + /* Exit with result in a0 */ + /* a0 = 0: success (bits [1:0] were masked) */ + /* a0 != 0: failure (some bits were not masked) */ + j _exit + +machine_trap_handler: + /* Check if illegal instruction (mcause = 2) */ + csrr t0, mcause + li t1, 2 + bne t0, t1, skip_test + + /* Test: Copy STVEC (with mode bits) to MEPC */ + csrr t0, stvec /* t0 = 0x80004001 */ + csrw mepc, t0 /* Write to MEPC */ + csrr t1, mepc /* Read back MEPC */ + + /* Check if bits [1:0] are masked (IALIGN=32 without RVC) */ + andi a0, t1, 3 /* a0 = 0 if both bits masked correctly */ + + /* Set correct return address */ + lla t0, test_completed + csrw mepc, t0 + +skip_test: + mret + +/* Exit with semihosting */ +_exit: + lla a1, semiargs + li t0, 0x20026 /* ADP_Stopped_ApplicationExit */ + sd t0, 0(a1) + sd a0, 8(a1) + li a0, 0x20 /* TARGET_SYS_EXIT_EXTENDED */ + + /* Semihosting call sequence */ + .balign 16 + slli zero, zero, 0x1f + ebreak + srai zero, zero, 0x7 + j . + + .data + .balign 8 +semiargs: + .space 16 -- 2.50.0 From: Vasilis Liaskovitis <vliaskovitis@suse.com> Usage of vsetvli instruction is reserved if VLMAX is changed when vsetvli rs1 and rd arguments are x0. In this case, if the new property is true, only the vill bit will be set. See https://github.com/riscv/riscv-isa-manual/blob/main/src/v-st-ext.adoc#avl-encoding According to the spec, the above use cases are reserved, and "Implementations may set vill in either case." Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2422 Signed-off-by: Vasilis Liaskovitis <vliaskovitis@suse.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Message-ID: <20250618213542.22873-1-vliaskovitis@suse.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/helper.h | 2 +- target/riscv/cpu_cfg_fields.h.inc | 1 + target/riscv/cpu.c | 1 + target/riscv/vector_helper.c | 12 +++++++++++- target/riscv/insn_trans/trans_rvv.c.inc | 4 ++-- 5 files changed, 16 insertions(+), 4 deletions(-) diff --git a/target/riscv/helper.h b/target/riscv/helper.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/helper.h +++ b/target/riscv/helper.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_3(hyp_hsv_d, TCG_CALL_NO_WG, void, env, tl, tl) #endif /* Vector functions */ -DEF_HELPER_3(vsetvl, tl, env, tl, tl) +DEF_HELPER_4(vsetvl, tl, env, tl, tl, tl) DEF_HELPER_5(vle8_v, void, ptr, ptr, tl, env, i32) DEF_HELPER_5(vle16_v, void, ptr, ptr, tl, env, i32) DEF_HELPER_5(vle32_v, void, ptr, ptr, tl, env, i32) diff --git a/target/riscv/cpu_cfg_fields.h.inc b/target/riscv/cpu_cfg_fields.h.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_cfg_fields.h.inc +++ b/target/riscv/cpu_cfg_fields.h.inc @@ -XXX,XX +XXX,XX @@ BOOL_FIELD(ext_supm) BOOL_FIELD(rvv_ta_all_1s) BOOL_FIELD(rvv_ma_all_1s) BOOL_FIELD(rvv_vl_half_avl) +BOOL_FIELD(rvv_vsetvl_x0_vill) /* Named features */ BOOL_FIELD(ext_svade) BOOL_FIELD(ext_zic64b) diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ static const Property riscv_cpu_properties[] = { DEFINE_PROP_BOOL("rvv_ta_all_1s", RISCVCPU, cfg.rvv_ta_all_1s, false), DEFINE_PROP_BOOL("rvv_ma_all_1s", RISCVCPU, cfg.rvv_ma_all_1s, false), DEFINE_PROP_BOOL("rvv_vl_half_avl", RISCVCPU, cfg.rvv_vl_half_avl, false), + DEFINE_PROP_BOOL("rvv_vsetvl_x0_vill", RISCVCPU, cfg.rvv_vsetvl_x0_vill, false), /* * write_misa() is marked as experimental for now so mark diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/vector_helper.c +++ b/target/riscv/vector_helper.c @@ -XXX,XX +XXX,XX @@ #include <math.h> target_ulong HELPER(vsetvl)(CPURISCVState *env, target_ulong s1, - target_ulong s2) + target_ulong s2, target_ulong x0) { int vlmax, vl; RISCVCPU *cpu = env_archcpu(env); @@ -XXX,XX +XXX,XX @@ target_ulong HELPER(vsetvl)(CPURISCVState *env, target_ulong s1, } else { vl = vlmax; } + + if (cpu->cfg.rvv_vsetvl_x0_vill && x0 && (env->vl != vl)) { + /* only set vill bit. */ + env->vill = 1; + env->vtype = 0; + env->vl = 0; + env->vstart = 0; + return 0; + } + env->vl = vl; env->vtype = s2; env->vstart = 0; diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn_trans/trans_rvv.c.inc +++ b/target/riscv/insn_trans/trans_rvv.c.inc @@ -XXX,XX +XXX,XX @@ static bool do_vsetvl(DisasContext *s, int rd, int rs1, TCGv s2) s1 = get_gpr(s, rs1, EXT_ZERO); } - gen_helper_vsetvl(dst, tcg_env, s1, s2); + gen_helper_vsetvl(dst, tcg_env, s1, s2, tcg_constant_tl((int) (rd == 0 && rs1 == 0))); gen_set_gpr(s, rd, dst); finalize_rvv_inst(s); @@ -XXX,XX +XXX,XX @@ static bool do_vsetivli(DisasContext *s, int rd, TCGv s1, TCGv s2) dst = dest_gpr(s, rd); - gen_helper_vsetvl(dst, tcg_env, s1, s2); + gen_helper_vsetvl(dst, tcg_env, s1, s2, tcg_constant_tl(0)); gen_set_gpr(s, rd, dst); finalize_rvv_inst(s); gen_update_pc(s, s->cur_insn_len); -- 2.50.0 From: Alexandre Ghiti <alexghiti@rivosinc.com> The Svrsw60t59b extension allows to free the PTE reserved bits 60 and 59 for software to use. Reviewed-by: Deepak Gupta <debug@rivosinc.com> Signed-off-by: Alexandre Ghiti <alexghiti@rivosinc.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Nutty Liu<liujingqi@lanxincomputing.com> Message-ID: <20250702-dev-alex-svrsw60b59b_v2-v2-1-504ddf0f8530@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/riscv-iommu-bits.h | 1 + target/riscv/cpu_bits.h | 3 ++- target/riscv/cpu_cfg_fields.h.inc | 1 + hw/riscv/riscv-iommu.c | 3 ++- target/riscv/cpu.c | 2 ++ target/riscv/cpu_helper.c | 3 ++- target/riscv/tcg/tcg-cpu.c | 8 ++++++++ 7 files changed, 18 insertions(+), 3 deletions(-) diff --git a/hw/riscv/riscv-iommu-bits.h b/hw/riscv/riscv-iommu-bits.h index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/riscv-iommu-bits.h +++ b/hw/riscv/riscv-iommu-bits.h @@ -XXX,XX +XXX,XX @@ struct riscv_iommu_pq_record { #define RISCV_IOMMU_CAP_SV39 BIT_ULL(9) #define RISCV_IOMMU_CAP_SV48 BIT_ULL(10) #define RISCV_IOMMU_CAP_SV57 BIT_ULL(11) +#define RISCV_IOMMU_CAP_SVRSW60T59B BIT_ULL(14) #define RISCV_IOMMU_CAP_SV32X4 BIT_ULL(16) #define RISCV_IOMMU_CAP_SV39X4 BIT_ULL(17) #define RISCV_IOMMU_CAP_SV48X4 BIT_ULL(18) diff --git a/target/riscv/cpu_bits.h b/target/riscv/cpu_bits.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_bits.h +++ b/target/riscv/cpu_bits.h @@ -XXX,XX +XXX,XX @@ typedef enum { #define PTE_SOFT 0x300 /* Reserved for Software */ #define PTE_PBMT 0x6000000000000000ULL /* Page-based memory types */ #define PTE_N 0x8000000000000000ULL /* NAPOT translation */ -#define PTE_RESERVED 0x1FC0000000000000ULL /* Reserved bits */ +#define PTE_RESERVED(svrsw60t59b) \ + (svrsw60t59b ? 0x07C0000000000000ULL : 0x1FC0000000000000ULL) /* Reserved bits */ #define PTE_ATTR (PTE_N | PTE_PBMT) /* All attributes bits */ /* Page table PPN shift amount */ diff --git a/target/riscv/cpu_cfg_fields.h.inc b/target/riscv/cpu_cfg_fields.h.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_cfg_fields.h.inc +++ b/target/riscv/cpu_cfg_fields.h.inc @@ -XXX,XX +XXX,XX @@ BOOL_FIELD(ext_svadu) BOOL_FIELD(ext_svinval) BOOL_FIELD(ext_svnapot) BOOL_FIELD(ext_svpbmt) +BOOL_FIELD(ext_svrsw60t59b) BOOL_FIELD(ext_svvptc) BOOL_FIELD(ext_svukte) BOOL_FIELD(ext_zdinx) diff --git a/hw/riscv/riscv-iommu.c b/hw/riscv/riscv-iommu.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/riscv-iommu.c +++ b/hw/riscv/riscv-iommu.c @@ -XXX,XX +XXX,XX @@ static void riscv_iommu_realize(DeviceState *dev, Error **errp) } if (s->enable_g_stage) { s->cap |= RISCV_IOMMU_CAP_SV32X4 | RISCV_IOMMU_CAP_SV39X4 | - RISCV_IOMMU_CAP_SV48X4 | RISCV_IOMMU_CAP_SV57X4; + RISCV_IOMMU_CAP_SV48X4 | RISCV_IOMMU_CAP_SV57X4 | + RISCV_IOMMU_CAP_SVRSW60T59B; } if (s->hpm_cntrs > 0) { diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const RISCVIsaExtData isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(svinval, PRIV_VERSION_1_12_0, ext_svinval), ISA_EXT_DATA_ENTRY(svnapot, PRIV_VERSION_1_12_0, ext_svnapot), ISA_EXT_DATA_ENTRY(svpbmt, PRIV_VERSION_1_12_0, ext_svpbmt), + ISA_EXT_DATA_ENTRY(svrsw60t59b, PRIV_VERSION_1_13_0, ext_svrsw60t59b), ISA_EXT_DATA_ENTRY(svukte, PRIV_VERSION_1_13_0, ext_svukte), ISA_EXT_DATA_ENTRY(svvptc, PRIV_VERSION_1_13_0, ext_svvptc), ISA_EXT_DATA_ENTRY(xtheadba, PRIV_VERSION_1_11_0, ext_xtheadba), @@ -XXX,XX +XXX,XX @@ const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { MULTI_EXT_CFG_BOOL("svinval", ext_svinval, false), MULTI_EXT_CFG_BOOL("svnapot", ext_svnapot, false), MULTI_EXT_CFG_BOOL("svpbmt", ext_svpbmt, false), + MULTI_EXT_CFG_BOOL("svrsw60t59b", ext_svrsw60t59b, false), MULTI_EXT_CFG_BOOL("svvptc", ext_svvptc, true), MULTI_EXT_CFG_BOOL("zicntr", ext_zicntr, true), diff --git a/target/riscv/cpu_helper.c b/target/riscv/cpu_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_helper.c +++ b/target/riscv/cpu_helper.c @@ -XXX,XX +XXX,XX @@ static int get_physical_address(CPURISCVState *env, hwaddr *physical, bool svade = riscv_cpu_cfg(env)->ext_svade; bool svadu = riscv_cpu_cfg(env)->ext_svadu; bool adue = svadu ? env->menvcfg & MENVCFG_ADUE : !svade; + bool svrsw60t59b = riscv_cpu_cfg(env)->ext_svrsw60t59b; if (first_stage && two_stage && env->virt_enabled) { pbmte = pbmte && (env->henvcfg & HENVCFG_PBMTE); @@ -XXX,XX +XXX,XX @@ static int get_physical_address(CPURISCVState *env, hwaddr *physical, if (riscv_cpu_sxl(env) == MXL_RV32) { ppn = pte >> PTE_PPN_SHIFT; } else { - if (pte & PTE_RESERVED) { + if (pte & PTE_RESERVED(svrsw60t59b)) { qemu_log_mask(LOG_GUEST_ERROR, "%s: reserved bits set in PTE: " "addr: 0x%" HWADDR_PRIx " pte: 0x" TARGET_FMT_lx "\n", __func__, pte_addr, pte); diff --git a/target/riscv/tcg/tcg-cpu.c b/target/riscv/tcg/tcg-cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/tcg/tcg-cpu.c +++ b/target/riscv/tcg/tcg-cpu.c @@ -XXX,XX +XXX,XX @@ void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) cpu->cfg.ext_ssctr = false; } + if (cpu->cfg.ext_svrsw60t59b && + (!cpu->cfg.mmu || mcc->def->misa_mxl_max == MXL_RV32)) { + error_setg(errp, "svrsw60t59b is not supported on RV32 and MMU-less platforms"); + return; + } + /* * Disable isa extensions based on priv spec after we * validated and set everything we need. @@ -XXX,XX +XXX,XX @@ static void riscv_init_max_cpu_extensions(Object *obj) if (env->misa_mxl != MXL_RV32) { isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zcf), false); + } else { + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_svrsw60t59b), false); } /* -- 2.50.0 The following changes since commit 58ee924b97d1c0898555647a31820c5a20d55a73: Merge tag 'for-upstream' of https://gitlab.com/bonzini/qemu into staging (2024-07-17 15:40:28 +1000) are available in the Git repository at: https://github.com/alistair23/qemu.git tags/pull-riscv-to-apply-20240718-1 for you to fetch changes up to daff9f7f7a457f78ce455e6abf19c2a37dfe7630: roms/opensbi: Update to v1.5 (2024-07-18 12:08:45 +1000) ---------------------------------------------------------------- RISC-V PR for 9.1 * Support the zimop, zcmop, zama16b and zabha extensions * Validate the mode when setting vstvec CSR * Add decode support for Zawrs extension * Update the KVM regs to Linux 6.10-rc5 * Add smcntrpmf extension support * Raise an exception when CSRRS/CSRRC writes a read-only CSR * Re-insert and deprecate 'riscv,delegate' in virt machine device tree * roms/opensbi: Update to v1.5 ---------------------------------------------------------------- Atish Patra (7): target/riscv: Fix the predicate functions for mhpmeventhX CSRs target/riscv: Only set INH fields if priv mode is available target/riscv: Implement privilege mode filtering for cycle/instret target/riscv: Save counter values during countinhibit update target/riscv: Enforce WARL behavior for scounteren/hcounteren target/riscv: Do not setup pmu timer if OF is disabled target/riscv: Expose the Smcntrpmf config Balaji Ravikumar (1): disas/riscv: Add decode for Zawrs extension Daniel Henrique Barboza (3): target/riscv/kvm: update KVM regs to Linux 6.10-rc5 hw/riscv/virt.c: re-insert and deprecate 'riscv,delegate' roms/opensbi: Update to v1.5 Jiayi Li (1): target/riscv: Validate the mode in write_vstvec Kaiwen Xue (3): target/riscv: Add cycle & instret privilege mode filtering properties target/riscv: Add cycle & instret privilege mode filtering definitions target/riscv: Add cycle & instret privilege mode filtering support LIU Zhiwei (11): target/riscv: Add zimop extension disas/riscv: Support zimop disassemble target/riscv: Add zcmop extension disas/riscv: Support zcmop disassemble target/riscv: Support Zama16b extension target/riscv: Move gen_amo before implement Zabha target/riscv: Add AMO instructions for Zabha target/riscv: Move gen_cmpxchg before adding amocas.[b|h] target/riscv: Add amocas.[b|h] for Zabha target/riscv: Expose zabha extension as a cpu property disas/riscv: Support zabha disassemble Rajnesh Kanwal (3): target/riscv: Combine set_mode and set_virt functions. target/riscv: Start counters from both mhpmcounter and mcountinhibit target/riscv: More accurately model priv mode filtering. Yu-Ming Chang (1): target/riscv: raise an exception when CSRRS/CSRRC writes a read-only CSR docs/about/deprecated.rst | 11 + target/riscv/cpu.h | 24 +- target/riscv/cpu_bits.h | 41 ++ target/riscv/cpu_cfg.h | 5 + target/riscv/pmu.h | 4 + target/riscv/insn16.decode | 1 + target/riscv/insn32.decode | 33 ++ disas/riscv.c | 187 +++++++++ hw/riscv/virt.c | 9 + target/riscv/cpu.c | 10 + target/riscv/cpu_helper.c | 66 ++-- target/riscv/csr.c | 501 ++++++++++++++++++++----- target/riscv/kvm/kvm-cpu.c | 2 + target/riscv/machine.c | 5 +- target/riscv/op_helper.c | 23 +- target/riscv/pmu.c | 181 ++++++++- target/riscv/tcg/tcg-cpu.c | 5 + target/riscv/translate.c | 38 ++ target/riscv/insn_trans/trans_rva.c.inc | 51 +-- target/riscv/insn_trans/trans_rvd.c.inc | 14 +- target/riscv/insn_trans/trans_rvf.c.inc | 14 +- target/riscv/insn_trans/trans_rvi.c.inc | 6 + target/riscv/insn_trans/trans_rvzabha.c.inc | 145 +++++++ target/riscv/insn_trans/trans_rvzacas.c.inc | 13 - target/riscv/insn_trans/trans_rvzcmop.c.inc | 29 ++ target/riscv/insn_trans/trans_rvzimop.c.inc | 37 ++ pc-bios/opensbi-riscv32-generic-fw_dynamic.bin | Bin 267416 -> 268312 bytes pc-bios/opensbi-riscv64-generic-fw_dynamic.bin | Bin 270808 -> 272504 bytes roms/opensbi | 2 +- 29 files changed, 1246 insertions(+), 211 deletions(-) create mode 100644 target/riscv/insn_trans/trans_rvzabha.c.inc create mode 100644 target/riscv/insn_trans/trans_rvzcmop.c.inc create mode 100644 target/riscv/insn_trans/trans_rvzimop.c.inc From: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Zimop extension defines an encoding space for 40 MOPs.The Zimop extension defines 32 MOP instructions named MOP.R.n, where n is an integer between 0 and 31, inclusive. The Zimop extension additionally defines 8 MOP instructions named MOP.RR.n, where n is an integer between 0 and 7. These 40 MOPs initially are defined to simply write zero to x[rd], but are designed to be redefined by later extensions to perform some other action. Signed-off-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Deepak Gupta <debug@rivosinc.com> Message-ID: <20240709113652.1239-2-zhiwei_liu@linux.alibaba.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_cfg.h | 1 + target/riscv/insn32.decode | 11 ++++++ target/riscv/cpu.c | 2 ++ target/riscv/translate.c | 1 + target/riscv/insn_trans/trans_rvzimop.c.inc | 37 +++++++++++++++++++++ 5 files changed, 52 insertions(+) create mode 100644 target/riscv/insn_trans/trans_rvzimop.c.inc diff --git a/target/riscv/cpu_cfg.h b/target/riscv/cpu_cfg.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_cfg.h +++ b/target/riscv/cpu_cfg.h @@ -XXX,XX +XXX,XX @@ struct RISCVCPUConfig { bool ext_zihintntl; bool ext_zihintpause; bool ext_zihpm; + bool ext_zimop; bool ext_ztso; bool ext_smstateen; bool ext_sstc; diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn32.decode +++ b/target/riscv/insn32.decode @@ -XXX,XX +XXX,XX @@ %imm_bs 30:2 !function=ex_shift_3 %imm_rnum 20:4 %imm_z6 26:1 15:5 +%imm_mop5 30:1 26:2 20:2 +%imm_mop3 30:1 26:2 # Argument sets: &empty @@ -XXX,XX +XXX,XX @@ &r2nfvm vm rd rs1 nf &rnfvm vm rd rs1 rs2 nf &k_aes shamt rs2 rs1 rd +&mop5 imm rd rs1 +&mop3 imm rd rs1 rs2 # Formats 32: @r ....... ..... ..... ... ..... ....... &r %rs2 %rs1 %rd @@ -XXX,XX +XXX,XX @@ @k_aes .. ..... ..... ..... ... ..... ....... &k_aes shamt=%imm_bs %rs2 %rs1 %rd @i_aes .. ..... ..... ..... ... ..... ....... &i imm=%imm_rnum %rs1 %rd +@mop5 . . .. .. .... .. ..... ... ..... ....... &mop5 imm=%imm_mop5 %rd %rs1 +@mop3 . . .. .. . ..... ..... ... ..... ....... &mop3 imm=%imm_mop3 %rd %rs1 %rs2 + # Formats 64: @sh5 ....... ..... ..... ... ..... ....... &shift shamt=%sh5 %rs1 %rd @@ -XXX,XX +XXX,XX @@ amocas_w 00101 . . ..... ..... 010 ..... 0101111 @atom_st amocas_d 00101 . . ..... ..... 011 ..... 0101111 @atom_st # *** RV64 Zacas Standard Extension *** amocas_q 00101 . . ..... ..... 100 ..... 0101111 @atom_st + +# *** Zimop may-be-operation extension *** +mop_r_n 1 . 00 .. 0111 .. ..... 100 ..... 1110011 @mop5 +mop_rr_n 1 . 00 .. 1 ..... ..... 100 ..... 1110011 @mop3 diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const RISCVIsaExtData isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(zihintntl, PRIV_VERSION_1_10_0, ext_zihintntl), ISA_EXT_DATA_ENTRY(zihintpause, PRIV_VERSION_1_10_0, ext_zihintpause), ISA_EXT_DATA_ENTRY(zihpm, PRIV_VERSION_1_12_0, ext_zihpm), + ISA_EXT_DATA_ENTRY(zimop, PRIV_VERSION_1_13_0, ext_zimop), ISA_EXT_DATA_ENTRY(zmmul, PRIV_VERSION_1_12_0, ext_zmmul), ISA_EXT_DATA_ENTRY(za64rs, PRIV_VERSION_1_12_0, has_priv_1_11), ISA_EXT_DATA_ENTRY(zaamo, PRIV_VERSION_1_12_0, ext_zaamo), @@ -XXX,XX +XXX,XX @@ const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { MULTI_EXT_CFG_BOOL("zicsr", ext_zicsr, true), MULTI_EXT_CFG_BOOL("zihintntl", ext_zihintntl, true), MULTI_EXT_CFG_BOOL("zihintpause", ext_zihintpause, true), + MULTI_EXT_CFG_BOOL("zimop", ext_zimop, false), MULTI_EXT_CFG_BOOL("zacas", ext_zacas, false), MULTI_EXT_CFG_BOOL("zaamo", ext_zaamo, false), MULTI_EXT_CFG_BOOL("zalrsc", ext_zalrsc, false), diff --git a/target/riscv/translate.c b/target/riscv/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/translate.c +++ b/target/riscv/translate.c @@ -XXX,XX +XXX,XX @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc) #include "insn_trans/trans_rvzacas.c.inc" #include "insn_trans/trans_rvzawrs.c.inc" #include "insn_trans/trans_rvzicbo.c.inc" +#include "insn_trans/trans_rvzimop.c.inc" #include "insn_trans/trans_rvzfa.c.inc" #include "insn_trans/trans_rvzfh.c.inc" #include "insn_trans/trans_rvk.c.inc" diff --git a/target/riscv/insn_trans/trans_rvzimop.c.inc b/target/riscv/insn_trans/trans_rvzimop.c.inc new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/riscv/insn_trans/trans_rvzimop.c.inc @@ -XXX,XX +XXX,XX @@ +/* + * RISC-V translation routines for May-Be-Operation(zimop). + * + * Copyright (c) 2024 Alibaba Group. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2 or later, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#define REQUIRE_ZIMOP(ctx) do { \ + if (!ctx->cfg_ptr->ext_zimop) { \ + return false; \ + } \ +} while (0) + +static bool trans_mop_r_n(DisasContext *ctx, arg_mop_r_n *a) +{ + REQUIRE_ZIMOP(ctx); + gen_set_gpr(ctx, a->rd, ctx->zero); + return true; +} + +static bool trans_mop_rr_n(DisasContext *ctx, arg_mop_rr_n *a) +{ + REQUIRE_ZIMOP(ctx); + gen_set_gpr(ctx, a->rd, ctx->zero); + return true; +} -- 2.45.2 From: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Signed-off-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Deepak Gupta <debug@rivosinc.com> Message-ID: <20240709113652.1239-3-zhiwei_liu@linux.alibaba.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- disas/riscv.c | 98 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 98 insertions(+) diff --git a/disas/riscv.c b/disas/riscv.c index XXXXXXX..XXXXXXX 100644 --- a/disas/riscv.c +++ b/disas/riscv.c @@ -XXX,XX +XXX,XX @@ typedef enum { rv_op_amocas_w = 875, rv_op_amocas_d = 876, rv_op_amocas_q = 877, + rv_mop_r_0 = 878, + rv_mop_r_1 = 879, + rv_mop_r_2 = 880, + rv_mop_r_3 = 881, + rv_mop_r_4 = 882, + rv_mop_r_5 = 883, + rv_mop_r_6 = 884, + rv_mop_r_7 = 885, + rv_mop_r_8 = 886, + rv_mop_r_9 = 887, + rv_mop_r_10 = 888, + rv_mop_r_11 = 889, + rv_mop_r_12 = 890, + rv_mop_r_13 = 891, + rv_mop_r_14 = 892, + rv_mop_r_15 = 893, + rv_mop_r_16 = 894, + rv_mop_r_17 = 895, + rv_mop_r_18 = 896, + rv_mop_r_19 = 897, + rv_mop_r_20 = 898, + rv_mop_r_21 = 899, + rv_mop_r_22 = 900, + rv_mop_r_23 = 901, + rv_mop_r_24 = 902, + rv_mop_r_25 = 903, + rv_mop_r_26 = 904, + rv_mop_r_27 = 905, + rv_mop_r_28 = 906, + rv_mop_r_29 = 907, + rv_mop_r_30 = 908, + rv_mop_r_31 = 909, + rv_mop_rr_0 = 910, + rv_mop_rr_1 = 911, + rv_mop_rr_2 = 912, + rv_mop_rr_3 = 913, + rv_mop_rr_4 = 914, + rv_mop_rr_5 = 915, + rv_mop_rr_6 = 916, + rv_mop_rr_7 = 917, } rv_op; /* register names */ @@ -XXX,XX +XXX,XX @@ const rv_opcode_data rvi_opcode_data[] = { { "amocas.w", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, { "amocas.d", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, { "amocas.q", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "mop.r.0", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.1", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.2", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.3", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.4", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.5", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.6", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.7", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.8", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.9", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.10", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.11", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.12", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.13", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.14", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.15", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.16", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.17", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.18", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.19", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.20", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.21", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.22", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.23", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.24", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.25", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.26", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.27", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.28", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.29", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.30", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.r.31", rv_codec_r, rv_fmt_rd_rs1, NULL, 0, 0 }, + { "mop.rr.0", rv_codec_r, rv_fmt_rd_rs1_rs2, NULL, 0, 0, 0 }, + { "mop.rr.1", rv_codec_r, rv_fmt_rd_rs1_rs2, NULL, 0, 0, 0 }, + { "mop.rr.2", rv_codec_r, rv_fmt_rd_rs1_rs2, NULL, 0, 0, 0 }, + { "mop.rr.3", rv_codec_r, rv_fmt_rd_rs1_rs2, NULL, 0, 0, 0 }, + { "mop.rr.4", rv_codec_r, rv_fmt_rd_rs1_rs2, NULL, 0, 0, 0 }, + { "mop.rr.5", rv_codec_r, rv_fmt_rd_rs1_rs2, NULL, 0, 0, 0 }, + { "mop.rr.6", rv_codec_r, rv_fmt_rd_rs1_rs2, NULL, 0, 0, 0 }, + { "mop.rr.7", rv_codec_r, rv_fmt_rd_rs1_rs2, NULL, 0, 0, 0 }, }; /* CSR names */ @@ -XXX,XX +XXX,XX @@ static void decode_inst_opcode(rv_decode *dec, rv_isa isa) case 1: op = rv_op_csrrw; break; case 2: op = rv_op_csrrs; break; case 3: op = rv_op_csrrc; break; + case 4: + if (dec->cfg->ext_zimop) { + int imm_mop5, imm_mop3; + if ((extract32(inst, 22, 10) & 0b1011001111) + == 0b1000000111) { + imm_mop5 = deposit32(deposit32(extract32(inst, 20, 2), + 2, 2, + extract32(inst, 26, 2)), + 4, 1, extract32(inst, 30, 1)); + op = rv_mop_r_0 + imm_mop5; + } else if ((extract32(inst, 25, 7) & 0b1011001) + == 0b1000001) { + imm_mop3 = deposit32(extract32(inst, 26, 2), + 2, 1, extract32(inst, 30, 1)); + op = rv_mop_rr_0 + imm_mop3; + } + } + break; case 5: op = rv_op_csrrwi; break; case 6: op = rv_op_csrrsi; break; case 7: op = rv_op_csrrci; break; -- 2.45.2 From: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Zcmop defines eight 16-bit MOP instructions named C.MOP.n, where n is an odd integer between 1 and 15, inclusive. C.MOP.n is encoded in the reserved encoding space corresponding to C.LUI xn, 0. Unlike the MOPs defined in the Zimop extension, the C.MOP.n instructions are defined to not write any register. In current implementation, C.MOP.n only has an check function, without any other more behavior. Signed-off-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Deepak Gupta <debug@rivosinc.com> Message-ID: <20240709113652.1239-4-zhiwei_liu@linux.alibaba.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_cfg.h | 1 + target/riscv/insn16.decode | 1 + target/riscv/cpu.c | 2 ++ target/riscv/tcg/tcg-cpu.c | 5 ++++ target/riscv/translate.c | 1 + target/riscv/insn_trans/trans_rvzcmop.c.inc | 29 +++++++++++++++++++++ 6 files changed, 39 insertions(+) create mode 100644 target/riscv/insn_trans/trans_rvzcmop.c.inc diff --git a/target/riscv/cpu_cfg.h b/target/riscv/cpu_cfg.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_cfg.h +++ b/target/riscv/cpu_cfg.h @@ -XXX,XX +XXX,XX @@ struct RISCVCPUConfig { bool ext_zihintpause; bool ext_zihpm; bool ext_zimop; + bool ext_zcmop; bool ext_ztso; bool ext_smstateen; bool ext_sstc; diff --git a/target/riscv/insn16.decode b/target/riscv/insn16.decode index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn16.decode +++ b/target/riscv/insn16.decode @@ -XXX,XX +XXX,XX @@ sw 110 ... ... .. ... 00 @cs_w addi 000 . ..... ..... 01 @ci addi 010 . ..... ..... 01 @c_li { + c_mop_n 011 0 0 n:3 1 00000 01 illegal 011 0 ----- 00000 01 # c.addi16sp and c.lui, RES nzimm=0 addi 011 . 00010 ..... 01 @c_addi16sp lui 011 . ..... ..... 01 @c_lui diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const RISCVIsaExtData isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(zcf, PRIV_VERSION_1_12_0, ext_zcf), ISA_EXT_DATA_ENTRY(zcd, PRIV_VERSION_1_12_0, ext_zcd), ISA_EXT_DATA_ENTRY(zce, PRIV_VERSION_1_12_0, ext_zce), + ISA_EXT_DATA_ENTRY(zcmop, PRIV_VERSION_1_13_0, ext_zcmop), ISA_EXT_DATA_ENTRY(zcmp, PRIV_VERSION_1_12_0, ext_zcmp), ISA_EXT_DATA_ENTRY(zcmt, PRIV_VERSION_1_12_0, ext_zcmt), ISA_EXT_DATA_ENTRY(zba, PRIV_VERSION_1_12_0, ext_zba), @@ -XXX,XX +XXX,XX @@ const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { MULTI_EXT_CFG_BOOL("zihintntl", ext_zihintntl, true), MULTI_EXT_CFG_BOOL("zihintpause", ext_zihintpause, true), MULTI_EXT_CFG_BOOL("zimop", ext_zimop, false), + MULTI_EXT_CFG_BOOL("zcmop", ext_zcmop, false), MULTI_EXT_CFG_BOOL("zacas", ext_zacas, false), MULTI_EXT_CFG_BOOL("zaamo", ext_zaamo, false), MULTI_EXT_CFG_BOOL("zalrsc", ext_zalrsc, false), diff --git a/target/riscv/tcg/tcg-cpu.c b/target/riscv/tcg/tcg-cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/tcg/tcg-cpu.c +++ b/target/riscv/tcg/tcg-cpu.c @@ -XXX,XX +XXX,XX @@ void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) } } + if (cpu->cfg.ext_zcmop && !cpu->cfg.ext_zca) { + error_setg(errp, "Zcmop extensions require Zca"); + return; + } + if (mcc->misa_mxl_max != MXL_RV32 && cpu->cfg.ext_zcf) { error_setg(errp, "Zcf extension is only relevant to RV32"); return; diff --git a/target/riscv/translate.c b/target/riscv/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/translate.c +++ b/target/riscv/translate.c @@ -XXX,XX +XXX,XX @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc) /* Include the auto-generated decoder for 16 bit insn */ #include "decode-insn16.c.inc" #include "insn_trans/trans_rvzce.c.inc" +#include "insn_trans/trans_rvzcmop.c.inc" /* Include decoders for factored-out extensions */ #include "decode-XVentanaCondOps.c.inc" diff --git a/target/riscv/insn_trans/trans_rvzcmop.c.inc b/target/riscv/insn_trans/trans_rvzcmop.c.inc new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/riscv/insn_trans/trans_rvzcmop.c.inc @@ -XXX,XX +XXX,XX @@ +/* + * RISC-V translation routines for compressed May-Be-Operation(zcmop). + * + * Copyright (c) 2024 Alibaba Group. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2 or later, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#define REQUIRE_ZCMOP(ctx) do { \ + if (!ctx->cfg_ptr->ext_zcmop) { \ + return false; \ + } \ +} while (0) + +static bool trans_c_mop_n(DisasContext *ctx, arg_c_mop_n *a) +{ + REQUIRE_ZCMOP(ctx); + return true; +} -- 2.45.2 From: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Although in QEMU disassemble, we usually lift compressed instruction to an normal format when display the instruction name. For C.MOP.n, it is more reasonable to directly display its compressed name, because its behavior can be redefined by later extension. Signed-off-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Deepak Gupta <debug@rivosinc.com> Message-ID: <20240709113652.1239-5-zhiwei_liu@linux.alibaba.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- disas/riscv.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/disas/riscv.c b/disas/riscv.c index XXXXXXX..XXXXXXX 100644 --- a/disas/riscv.c +++ b/disas/riscv.c @@ -XXX,XX +XXX,XX @@ typedef enum { rv_mop_rr_5 = 915, rv_mop_rr_6 = 916, rv_mop_rr_7 = 917, + rv_c_mop_1 = 918, + rv_c_mop_3 = 919, + rv_c_mop_5 = 920, + rv_c_mop_7 = 921, + rv_c_mop_9 = 922, + rv_c_mop_11 = 923, + rv_c_mop_13 = 924, + rv_c_mop_15 = 925, } rv_op; /* register names */ @@ -XXX,XX +XXX,XX @@ const rv_opcode_data rvi_opcode_data[] = { { "mop.rr.5", rv_codec_r, rv_fmt_rd_rs1_rs2, NULL, 0, 0, 0 }, { "mop.rr.6", rv_codec_r, rv_fmt_rd_rs1_rs2, NULL, 0, 0, 0 }, { "mop.rr.7", rv_codec_r, rv_fmt_rd_rs1_rs2, NULL, 0, 0, 0 }, + { "c.mop.1", rv_codec_ci_none, rv_fmt_none, NULL, 0, 0, 0 }, + { "c.mop.3", rv_codec_ci_none, rv_fmt_none, NULL, 0, 0, 0 }, + { "c.mop.5", rv_codec_ci_none, rv_fmt_none, NULL, 0, 0, 0 }, + { "c.mop.7", rv_codec_ci_none, rv_fmt_none, NULL, 0, 0, 0 }, + { "c.mop.9", rv_codec_ci_none, rv_fmt_none, NULL, 0, 0, 0 }, + { "c.mop.11", rv_codec_ci_none, rv_fmt_none, NULL, 0, 0, 0 }, + { "c.mop.13", rv_codec_ci_none, rv_fmt_none, NULL, 0, 0, 0 }, + { "c.mop.15", rv_codec_ci_none, rv_fmt_none, NULL, 0, 0, 0 }, }; /* CSR names */ @@ -XXX,XX +XXX,XX @@ static void decode_inst_opcode(rv_decode *dec, rv_isa isa) break; case 2: op = rv_op_c_li; break; case 3: + if (dec->cfg->ext_zcmop) { + if ((((inst >> 2) & 0b111111) == 0b100000) && + (((inst >> 11) & 0b11) == 0b0)) { + op = rv_c_mop_1 + ((inst >> 8) & 0b111); + break; + } + } switch ((inst >> 7) & 0b11111) { case 2: op = rv_op_c_addi16sp; break; default: op = rv_op_c_lui; break; -- 2.45.2 From: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Zama16b is the property that misaligned load/stores/atomics within a naturally aligned 16-byte region are atomic. According to the specification, Zama16b applies only to AMOs, loads and stores defined in the base ISAs, and loads and stores of no more than XLEN bits defined in the F, D, and Q extensions. Thus it should not apply to zacas or RVC instructions. For an instruction in that set, if all accessed bytes lie within 16B granule, the instruction will not raise an exception for reasons of address alignment, and the instruction will give rise to only one memory operation for the purposes of RVWMO—i.e., it will execute atomically. Signed-off-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240709113652.1239-6-zhiwei_liu@linux.alibaba.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_cfg.h | 1 + target/riscv/cpu.c | 2 ++ target/riscv/insn_trans/trans_rva.c.inc | 42 ++++++++++++++----------- target/riscv/insn_trans/trans_rvd.c.inc | 14 +++++++-- target/riscv/insn_trans/trans_rvf.c.inc | 14 +++++++-- target/riscv/insn_trans/trans_rvi.c.inc | 6 ++++ 6 files changed, 57 insertions(+), 22 deletions(-) diff --git a/target/riscv/cpu_cfg.h b/target/riscv/cpu_cfg.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_cfg.h +++ b/target/riscv/cpu_cfg.h @@ -XXX,XX +XXX,XX @@ struct RISCVCPUConfig { bool ext_zdinx; bool ext_zaamo; bool ext_zacas; + bool ext_zama16b; bool ext_zalrsc; bool ext_zawrs; bool ext_zfa; diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const RISCVIsaExtData isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(za64rs, PRIV_VERSION_1_12_0, has_priv_1_11), ISA_EXT_DATA_ENTRY(zaamo, PRIV_VERSION_1_12_0, ext_zaamo), ISA_EXT_DATA_ENTRY(zacas, PRIV_VERSION_1_12_0, ext_zacas), + ISA_EXT_DATA_ENTRY(zama16b, PRIV_VERSION_1_13_0, ext_zama16b), ISA_EXT_DATA_ENTRY(zalrsc, PRIV_VERSION_1_12_0, ext_zalrsc), ISA_EXT_DATA_ENTRY(zawrs, PRIV_VERSION_1_12_0, ext_zawrs), ISA_EXT_DATA_ENTRY(zfa, PRIV_VERSION_1_12_0, ext_zfa), @@ -XXX,XX +XXX,XX @@ const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { MULTI_EXT_CFG_BOOL("zimop", ext_zimop, false), MULTI_EXT_CFG_BOOL("zcmop", ext_zcmop, false), MULTI_EXT_CFG_BOOL("zacas", ext_zacas, false), + MULTI_EXT_CFG_BOOL("zama16b", ext_zama16b, false), MULTI_EXT_CFG_BOOL("zaamo", ext_zaamo, false), MULTI_EXT_CFG_BOOL("zalrsc", ext_zalrsc, false), MULTI_EXT_CFG_BOOL("zawrs", ext_zawrs, true), diff --git a/target/riscv/insn_trans/trans_rva.c.inc b/target/riscv/insn_trans/trans_rva.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn_trans/trans_rva.c.inc +++ b/target/riscv/insn_trans/trans_rva.c.inc @@ -XXX,XX +XXX,XX @@ static bool gen_amo(DisasContext *ctx, arg_atomic *a, TCGv dest = dest_gpr(ctx, a->rd); TCGv src1, src2 = get_gpr(ctx, a->rs2, EXT_NONE); + if (ctx->cfg_ptr->ext_zama16b) { + mop |= MO_ATOM_WITHIN16; + } else { + mop |= MO_ALIGN; + } + decode_save_opc(ctx); src1 = get_address(ctx, a->rs1, 0); func(dest, src1, src2, ctx->mem_idx, mop); @@ -XXX,XX +XXX,XX @@ static bool trans_sc_w(DisasContext *ctx, arg_sc_w *a) static bool trans_amoswap_w(DisasContext *ctx, arg_amoswap_w *a) { REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_xchg_tl, (MO_ALIGN | MO_TESL)); + return gen_amo(ctx, a, &tcg_gen_atomic_xchg_tl, MO_TESL); } static bool trans_amoadd_w(DisasContext *ctx, arg_amoadd_w *a) { REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_add_tl, (MO_ALIGN | MO_TESL)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_add_tl, MO_TESL); } static bool trans_amoxor_w(DisasContext *ctx, arg_amoxor_w *a) { REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_xor_tl, (MO_ALIGN | MO_TESL)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_xor_tl, MO_TESL); } static bool trans_amoand_w(DisasContext *ctx, arg_amoand_w *a) { REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_and_tl, (MO_ALIGN | MO_TESL)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_and_tl, MO_TESL); } static bool trans_amoor_w(DisasContext *ctx, arg_amoor_w *a) { REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_or_tl, (MO_ALIGN | MO_TESL)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_or_tl, MO_TESL); } static bool trans_amomin_w(DisasContext *ctx, arg_amomin_w *a) { REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smin_tl, (MO_ALIGN | MO_TESL)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smin_tl, MO_TESL); } static bool trans_amomax_w(DisasContext *ctx, arg_amomax_w *a) { REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smax_tl, (MO_ALIGN | MO_TESL)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smax_tl, MO_TESL); } static bool trans_amominu_w(DisasContext *ctx, arg_amominu_w *a) { REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umin_tl, (MO_ALIGN | MO_TESL)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umin_tl, MO_TESL); } static bool trans_amomaxu_w(DisasContext *ctx, arg_amomaxu_w *a) { REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umax_tl, (MO_ALIGN | MO_TESL)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umax_tl, MO_TESL); } static bool trans_lr_d(DisasContext *ctx, arg_lr_d *a) @@ -XXX,XX +XXX,XX @@ static bool trans_amoswap_d(DisasContext *ctx, arg_amoswap_d *a) { REQUIRE_64BIT(ctx); REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_xchg_tl, (MO_ALIGN | MO_TEUQ)); + return gen_amo(ctx, a, &tcg_gen_atomic_xchg_tl, MO_TEUQ); } static bool trans_amoadd_d(DisasContext *ctx, arg_amoadd_d *a) { REQUIRE_64BIT(ctx); REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_add_tl, (MO_ALIGN | MO_TEUQ)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_add_tl, MO_TEUQ); } static bool trans_amoxor_d(DisasContext *ctx, arg_amoxor_d *a) { REQUIRE_64BIT(ctx); REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_xor_tl, (MO_ALIGN | MO_TEUQ)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_xor_tl, MO_TEUQ); } static bool trans_amoand_d(DisasContext *ctx, arg_amoand_d *a) { REQUIRE_64BIT(ctx); REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_and_tl, (MO_ALIGN | MO_TEUQ)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_and_tl, MO_TEUQ); } static bool trans_amoor_d(DisasContext *ctx, arg_amoor_d *a) { REQUIRE_64BIT(ctx); REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_or_tl, (MO_ALIGN | MO_TEUQ)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_or_tl, MO_TEUQ); } static bool trans_amomin_d(DisasContext *ctx, arg_amomin_d *a) { REQUIRE_64BIT(ctx); REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smin_tl, (MO_ALIGN | MO_TEUQ)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smin_tl, MO_TEUQ); } static bool trans_amomax_d(DisasContext *ctx, arg_amomax_d *a) { REQUIRE_64BIT(ctx); REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smax_tl, (MO_ALIGN | MO_TEUQ)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smax_tl, MO_TEUQ); } static bool trans_amominu_d(DisasContext *ctx, arg_amominu_d *a) { REQUIRE_64BIT(ctx); REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umin_tl, (MO_ALIGN | MO_TEUQ)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umin_tl, MO_TEUQ); } static bool trans_amomaxu_d(DisasContext *ctx, arg_amomaxu_d *a) { REQUIRE_64BIT(ctx); REQUIRE_A_OR_ZAAMO(ctx); - return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umax_tl, (MO_ALIGN | MO_TEUQ)); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umax_tl, MO_TEUQ); } diff --git a/target/riscv/insn_trans/trans_rvd.c.inc b/target/riscv/insn_trans/trans_rvd.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn_trans/trans_rvd.c.inc +++ b/target/riscv/insn_trans/trans_rvd.c.inc @@ -XXX,XX +XXX,XX @@ static bool trans_fld(DisasContext *ctx, arg_fld *a) { TCGv addr; + MemOp memop = MO_TEUQ; REQUIRE_FPU; REQUIRE_EXT(ctx, RVD); + if (ctx->cfg_ptr->ext_zama16b && (ctx->cur_insn_len != 2)) { + memop |= MO_ATOM_WITHIN16; + } + decode_save_opc(ctx); addr = get_address(ctx, a->rs1, a->imm); - tcg_gen_qemu_ld_i64(cpu_fpr[a->rd], addr, ctx->mem_idx, MO_TEUQ); + tcg_gen_qemu_ld_i64(cpu_fpr[a->rd], addr, ctx->mem_idx, memop); mark_fs_dirty(ctx); return true; @@ -XXX,XX +XXX,XX @@ static bool trans_fld(DisasContext *ctx, arg_fld *a) static bool trans_fsd(DisasContext *ctx, arg_fsd *a) { TCGv addr; + MemOp memop = MO_TEUQ; REQUIRE_FPU; REQUIRE_EXT(ctx, RVD); + if (ctx->cfg_ptr->ext_zama16b && (ctx->cur_insn_len != 2)) { + memop |= MO_ATOM_WITHIN16; + } + decode_save_opc(ctx); addr = get_address(ctx, a->rs1, a->imm); - tcg_gen_qemu_st_i64(cpu_fpr[a->rs2], addr, ctx->mem_idx, MO_TEUQ); + tcg_gen_qemu_st_i64(cpu_fpr[a->rs2], addr, ctx->mem_idx, memop); return true; } diff --git a/target/riscv/insn_trans/trans_rvf.c.inc b/target/riscv/insn_trans/trans_rvf.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn_trans/trans_rvf.c.inc +++ b/target/riscv/insn_trans/trans_rvf.c.inc @@ -XXX,XX +XXX,XX @@ static bool trans_flw(DisasContext *ctx, arg_flw *a) { TCGv_i64 dest; TCGv addr; + MemOp memop = MO_TEUL; REQUIRE_FPU; REQUIRE_EXT(ctx, RVF); + if (ctx->cfg_ptr->ext_zama16b && (ctx->cur_insn_len != 2)) { + memop |= MO_ATOM_WITHIN16; + } + decode_save_opc(ctx); addr = get_address(ctx, a->rs1, a->imm); dest = cpu_fpr[a->rd]; - tcg_gen_qemu_ld_i64(dest, addr, ctx->mem_idx, MO_TEUL); + tcg_gen_qemu_ld_i64(dest, addr, ctx->mem_idx, memop); gen_nanbox_s(dest, dest); mark_fs_dirty(ctx); @@ -XXX,XX +XXX,XX @@ static bool trans_flw(DisasContext *ctx, arg_flw *a) static bool trans_fsw(DisasContext *ctx, arg_fsw *a) { TCGv addr; + MemOp memop = MO_TEUL; REQUIRE_FPU; REQUIRE_EXT(ctx, RVF); + if (ctx->cfg_ptr->ext_zama16b && (ctx->cur_insn_len != 2)) { + memop |= MO_ATOM_WITHIN16; + } + decode_save_opc(ctx); addr = get_address(ctx, a->rs1, a->imm); - tcg_gen_qemu_st_i64(cpu_fpr[a->rs2], addr, ctx->mem_idx, MO_TEUL); + tcg_gen_qemu_st_i64(cpu_fpr[a->rs2], addr, ctx->mem_idx, memop); return true; } diff --git a/target/riscv/insn_trans/trans_rvi.c.inc b/target/riscv/insn_trans/trans_rvi.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn_trans/trans_rvi.c.inc +++ b/target/riscv/insn_trans/trans_rvi.c.inc @@ -XXX,XX +XXX,XX @@ static bool gen_load(DisasContext *ctx, arg_lb *a, MemOp memop) { bool out; + if (ctx->cfg_ptr->ext_zama16b && (ctx->cur_insn_len != 2)) { + memop |= MO_ATOM_WITHIN16; + } decode_save_opc(ctx); if (get_xl(ctx) == MXL_RV128) { out = gen_load_i128(ctx, a, memop); @@ -XXX,XX +XXX,XX @@ static bool gen_store_i128(DisasContext *ctx, arg_sb *a, MemOp memop) static bool gen_store(DisasContext *ctx, arg_sb *a, MemOp memop) { + if (ctx->cfg_ptr->ext_zama16b && (ctx->cur_insn_len != 2)) { + memop |= MO_ATOM_WITHIN16; + } decode_save_opc(ctx); if (get_xl(ctx) == MXL_RV128) { return gen_store_i128(ctx, a, memop); -- 2.45.2 From: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Signed-off-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240709113652.1239-7-zhiwei_liu@linux.alibaba.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/translate.c | 21 +++++++++++++++++++++ target/riscv/insn_trans/trans_rva.c.inc | 21 --------------------- 2 files changed, 21 insertions(+), 21 deletions(-) diff --git a/target/riscv/translate.c b/target/riscv/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/translate.c +++ b/target/riscv/translate.c @@ -XXX,XX +XXX,XX @@ static bool gen_unary_per_ol(DisasContext *ctx, arg_r2 *a, DisasExtend ext, return gen_unary(ctx, a, ext, f_tl); } +static bool gen_amo(DisasContext *ctx, arg_atomic *a, + void(*func)(TCGv, TCGv, TCGv, TCGArg, MemOp), + MemOp mop) +{ + TCGv dest = dest_gpr(ctx, a->rd); + TCGv src1, src2 = get_gpr(ctx, a->rs2, EXT_NONE); + + if (ctx->cfg_ptr->ext_zama16b) { + mop |= MO_ATOM_WITHIN16; + } else { + mop |= MO_ALIGN; + } + + decode_save_opc(ctx); + src1 = get_address(ctx, a->rs1, 0); + func(dest, src1, src2, ctx->mem_idx, mop); + + gen_set_gpr(ctx, a->rd, dest); + return true; +} + static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc) { DisasContext *ctx = container_of(dcbase, DisasContext, base); diff --git a/target/riscv/insn_trans/trans_rva.c.inc b/target/riscv/insn_trans/trans_rva.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn_trans/trans_rva.c.inc +++ b/target/riscv/insn_trans/trans_rva.c.inc @@ -XXX,XX +XXX,XX @@ static bool gen_sc(DisasContext *ctx, arg_atomic *a, MemOp mop) return true; } -static bool gen_amo(DisasContext *ctx, arg_atomic *a, - void(*func)(TCGv, TCGv, TCGv, TCGArg, MemOp), - MemOp mop) -{ - TCGv dest = dest_gpr(ctx, a->rd); - TCGv src1, src2 = get_gpr(ctx, a->rs2, EXT_NONE); - - if (ctx->cfg_ptr->ext_zama16b) { - mop |= MO_ATOM_WITHIN16; - } else { - mop |= MO_ALIGN; - } - - decode_save_opc(ctx); - src1 = get_address(ctx, a->rs1, 0); - func(dest, src1, src2, ctx->mem_idx, mop); - - gen_set_gpr(ctx, a->rd, dest); - return true; -} - static bool trans_lr_w(DisasContext *ctx, arg_lr_w *a) { REQUIRE_A_OR_ZALRSC(ctx); -- 2.45.2 From: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Signed-off-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240709113652.1239-8-zhiwei_liu@linux.alibaba.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_cfg.h | 1 + target/riscv/insn32.decode | 20 +++ target/riscv/translate.c | 4 +- target/riscv/insn_trans/trans_rvzabha.c.inc | 131 ++++++++++++++++++++ 4 files changed, 155 insertions(+), 1 deletion(-) create mode 100644 target/riscv/insn_trans/trans_rvzabha.c.inc diff --git a/target/riscv/cpu_cfg.h b/target/riscv/cpu_cfg.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_cfg.h +++ b/target/riscv/cpu_cfg.h @@ -XXX,XX +XXX,XX @@ struct RISCVCPUConfig { bool ext_zaamo; bool ext_zacas; bool ext_zama16b; + bool ext_zabha; bool ext_zalrsc; bool ext_zawrs; bool ext_zfa; diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn32.decode +++ b/target/riscv/insn32.decode @@ -XXX,XX +XXX,XX @@ amocas_q 00101 . . ..... ..... 100 ..... 0101111 @atom_st # *** Zimop may-be-operation extension *** mop_r_n 1 . 00 .. 0111 .. ..... 100 ..... 1110011 @mop5 mop_rr_n 1 . 00 .. 1 ..... ..... 100 ..... 1110011 @mop3 + +# *** Zabhb Standard Extension *** +amoswap_b 00001 . . ..... ..... 000 ..... 0101111 @atom_st +amoadd_b 00000 . . ..... ..... 000 ..... 0101111 @atom_st +amoxor_b 00100 . . ..... ..... 000 ..... 0101111 @atom_st +amoand_b 01100 . . ..... ..... 000 ..... 0101111 @atom_st +amoor_b 01000 . . ..... ..... 000 ..... 0101111 @atom_st +amomin_b 10000 . . ..... ..... 000 ..... 0101111 @atom_st +amomax_b 10100 . . ..... ..... 000 ..... 0101111 @atom_st +amominu_b 11000 . . ..... ..... 000 ..... 0101111 @atom_st +amomaxu_b 11100 . . ..... ..... 000 ..... 0101111 @atom_st +amoswap_h 00001 . . ..... ..... 001 ..... 0101111 @atom_st +amoadd_h 00000 . . ..... ..... 001 ..... 0101111 @atom_st +amoxor_h 00100 . . ..... ..... 001 ..... 0101111 @atom_st +amoand_h 01100 . . ..... ..... 001 ..... 0101111 @atom_st +amoor_h 01000 . . ..... ..... 001 ..... 0101111 @atom_st +amomin_h 10000 . . ..... ..... 001 ..... 0101111 @atom_st +amomax_h 10100 . . ..... ..... 001 ..... 0101111 @atom_st +amominu_h 11000 . . ..... ..... 001 ..... 0101111 @atom_st +amomaxu_h 11100 . . ..... ..... 001 ..... 0101111 @atom_st diff --git a/target/riscv/translate.c b/target/riscv/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/translate.c +++ b/target/riscv/translate.c @@ -XXX,XX +XXX,XX @@ static bool gen_amo(DisasContext *ctx, arg_atomic *a, { TCGv dest = dest_gpr(ctx, a->rd); TCGv src1, src2 = get_gpr(ctx, a->rs2, EXT_NONE); + MemOp size = mop & MO_SIZE; - if (ctx->cfg_ptr->ext_zama16b) { + if (ctx->cfg_ptr->ext_zama16b && size >= MO_32) { mop |= MO_ATOM_WITHIN16; } else { mop |= MO_ALIGN; @@ -XXX,XX +XXX,XX @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc) #include "insn_trans/trans_rvb.c.inc" #include "insn_trans/trans_rvzicond.c.inc" #include "insn_trans/trans_rvzacas.c.inc" +#include "insn_trans/trans_rvzabha.c.inc" #include "insn_trans/trans_rvzawrs.c.inc" #include "insn_trans/trans_rvzicbo.c.inc" #include "insn_trans/trans_rvzimop.c.inc" diff --git a/target/riscv/insn_trans/trans_rvzabha.c.inc b/target/riscv/insn_trans/trans_rvzabha.c.inc new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/riscv/insn_trans/trans_rvzabha.c.inc @@ -XXX,XX +XXX,XX @@ +/* + * RISC-V translation routines for the Zabha Standard Extension. + * + * Copyright (c) 2024 Alibaba Group + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2 or later, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#define REQUIRE_ZABHA(ctx) do { \ + if (!ctx->cfg_ptr->ext_zabha) { \ + return false; \ + } \ +} while (0) + +static bool trans_amoswap_b(DisasContext *ctx, arg_amoswap_b *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_xchg_tl, MO_SB); +} + +static bool trans_amoadd_b(DisasContext *ctx, arg_amoadd_b *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_add_tl, MO_SB); +} + +static bool trans_amoxor_b(DisasContext *ctx, arg_amoxor_b *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_xor_tl, MO_SB); +} + +static bool trans_amoand_b(DisasContext *ctx, arg_amoand_b *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_and_tl, MO_SB); +} + +static bool trans_amoor_b(DisasContext *ctx, arg_amoor_b *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_or_tl, MO_SB); +} + +static bool trans_amomin_b(DisasContext *ctx, arg_amomin_b *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smin_tl, MO_SB); +} + +static bool trans_amomax_b(DisasContext *ctx, arg_amomax_b *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smax_tl, MO_SB); +} + +static bool trans_amominu_b(DisasContext *ctx, arg_amominu_b *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umin_tl, MO_SB); +} + +static bool trans_amomaxu_b(DisasContext *ctx, arg_amomaxu_b *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umax_tl, MO_SB); +} + +static bool trans_amoswap_h(DisasContext *ctx, arg_amoswap_h *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_xchg_tl, MO_TESW); +} + +static bool trans_amoadd_h(DisasContext *ctx, arg_amoadd_h *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_add_tl, MO_TESW); +} + +static bool trans_amoxor_h(DisasContext *ctx, arg_amoxor_h *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_xor_tl, MO_TESW); +} + +static bool trans_amoand_h(DisasContext *ctx, arg_amoand_h *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_and_tl, MO_TESW); +} + +static bool trans_amoor_h(DisasContext *ctx, arg_amoor_h *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_or_tl, MO_TESW); +} + +static bool trans_amomin_h(DisasContext *ctx, arg_amomin_h *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smin_tl, MO_TESW); +} + +static bool trans_amomax_h(DisasContext *ctx, arg_amomax_h *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smax_tl, MO_TESW); +} + +static bool trans_amominu_h(DisasContext *ctx, arg_amominu_h *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umin_tl, MO_TESW); +} + +static bool trans_amomaxu_h(DisasContext *ctx, arg_amomaxu_h *a) +{ + REQUIRE_ZABHA(ctx); + return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umax_tl, MO_TESW); +} -- 2.45.2 From: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Signed-off-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240709113652.1239-9-zhiwei_liu@linux.alibaba.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/translate.c | 13 +++++++++++++ target/riscv/insn_trans/trans_rvzacas.c.inc | 13 ------------- 2 files changed, 13 insertions(+), 13 deletions(-) diff --git a/target/riscv/translate.c b/target/riscv/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/translate.c +++ b/target/riscv/translate.c @@ -XXX,XX +XXX,XX @@ static bool gen_amo(DisasContext *ctx, arg_atomic *a, return true; } +static bool gen_cmpxchg(DisasContext *ctx, arg_atomic *a, MemOp mop) +{ + TCGv dest = get_gpr(ctx, a->rd, EXT_NONE); + TCGv src1 = get_address(ctx, a->rs1, 0); + TCGv src2 = get_gpr(ctx, a->rs2, EXT_NONE); + + decode_save_opc(ctx); + tcg_gen_atomic_cmpxchg_tl(dest, src1, dest, src2, ctx->mem_idx, mop); + + gen_set_gpr(ctx, a->rd, dest); + return true; +} + static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc) { DisasContext *ctx = container_of(dcbase, DisasContext, base); diff --git a/target/riscv/insn_trans/trans_rvzacas.c.inc b/target/riscv/insn_trans/trans_rvzacas.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn_trans/trans_rvzacas.c.inc +++ b/target/riscv/insn_trans/trans_rvzacas.c.inc @@ -XXX,XX +XXX,XX @@ } \ } while (0) -static bool gen_cmpxchg(DisasContext *ctx, arg_atomic *a, MemOp mop) -{ - TCGv dest = get_gpr(ctx, a->rd, EXT_NONE); - TCGv src1 = get_address(ctx, a->rs1, 0); - TCGv src2 = get_gpr(ctx, a->rs2, EXT_NONE); - - decode_save_opc(ctx); - tcg_gen_atomic_cmpxchg_tl(dest, src1, dest, src2, ctx->mem_idx, mop); - - gen_set_gpr(ctx, a->rd, dest); - return true; -} - static bool trans_amocas_w(DisasContext *ctx, arg_amocas_w *a) { REQUIRE_ZACAS(ctx); -- 2.45.2 From: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Signed-off-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240709113652.1239-10-zhiwei_liu@linux.alibaba.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/insn32.decode | 2 ++ target/riscv/insn_trans/trans_rvzabha.c.inc | 14 ++++++++++++++ 2 files changed, 16 insertions(+) diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn32.decode +++ b/target/riscv/insn32.decode @@ -XXX,XX +XXX,XX @@ amomin_h 10000 . . ..... ..... 001 ..... 0101111 @atom_st amomax_h 10100 . . ..... ..... 001 ..... 0101111 @atom_st amominu_h 11000 . . ..... ..... 001 ..... 0101111 @atom_st amomaxu_h 11100 . . ..... ..... 001 ..... 0101111 @atom_st +amocas_b 00101 . . ..... ..... 000 ..... 0101111 @atom_st +amocas_h 00101 . . ..... ..... 001 ..... 0101111 @atom_st diff --git a/target/riscv/insn_trans/trans_rvzabha.c.inc b/target/riscv/insn_trans/trans_rvzabha.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn_trans/trans_rvzabha.c.inc +++ b/target/riscv/insn_trans/trans_rvzabha.c.inc @@ -XXX,XX +XXX,XX @@ static bool trans_amomaxu_h(DisasContext *ctx, arg_amomaxu_h *a) REQUIRE_ZABHA(ctx); return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umax_tl, MO_TESW); } + +static bool trans_amocas_b(DisasContext *ctx, arg_amocas_b *a) +{ + REQUIRE_ZACAS(ctx); + REQUIRE_ZABHA(ctx); + return gen_cmpxchg(ctx, a, MO_SB); +} + +static bool trans_amocas_h(DisasContext *ctx, arg_amocas_h *a) +{ + REQUIRE_ZACAS(ctx); + REQUIRE_ZABHA(ctx); + return gen_cmpxchg(ctx, a, MO_ALIGN | MO_TESW); +} -- 2.45.2 From: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Signed-off-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240709113652.1239-11-zhiwei_liu@linux.alibaba.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const RISCVIsaExtData isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(zmmul, PRIV_VERSION_1_12_0, ext_zmmul), ISA_EXT_DATA_ENTRY(za64rs, PRIV_VERSION_1_12_0, has_priv_1_11), ISA_EXT_DATA_ENTRY(zaamo, PRIV_VERSION_1_12_0, ext_zaamo), + ISA_EXT_DATA_ENTRY(zabha, PRIV_VERSION_1_13_0, ext_zabha), ISA_EXT_DATA_ENTRY(zacas, PRIV_VERSION_1_12_0, ext_zacas), ISA_EXT_DATA_ENTRY(zama16b, PRIV_VERSION_1_13_0, ext_zama16b), ISA_EXT_DATA_ENTRY(zalrsc, PRIV_VERSION_1_12_0, ext_zalrsc), @@ -XXX,XX +XXX,XX @@ const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { MULTI_EXT_CFG_BOOL("zcmop", ext_zcmop, false), MULTI_EXT_CFG_BOOL("zacas", ext_zacas, false), MULTI_EXT_CFG_BOOL("zama16b", ext_zama16b, false), + MULTI_EXT_CFG_BOOL("zabha", ext_zabha, false), MULTI_EXT_CFG_BOOL("zaamo", ext_zaamo, false), MULTI_EXT_CFG_BOOL("zalrsc", ext_zalrsc, false), MULTI_EXT_CFG_BOOL("zawrs", ext_zawrs, true), -- 2.45.2 From: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Signed-off-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240709113652.1239-12-zhiwei_liu@linux.alibaba.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- disas/riscv.c | 60 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 60 insertions(+) diff --git a/disas/riscv.c b/disas/riscv.c index XXXXXXX..XXXXXXX 100644 --- a/disas/riscv.c +++ b/disas/riscv.c @@ -XXX,XX +XXX,XX @@ typedef enum { rv_c_mop_11 = 923, rv_c_mop_13 = 924, rv_c_mop_15 = 925, + rv_op_amoswap_b = 926, + rv_op_amoadd_b = 927, + rv_op_amoxor_b = 928, + rv_op_amoor_b = 929, + rv_op_amoand_b = 930, + rv_op_amomin_b = 931, + rv_op_amomax_b = 932, + rv_op_amominu_b = 933, + rv_op_amomaxu_b = 934, + rv_op_amoswap_h = 935, + rv_op_amoadd_h = 936, + rv_op_amoxor_h = 937, + rv_op_amoor_h = 938, + rv_op_amoand_h = 939, + rv_op_amomin_h = 940, + rv_op_amomax_h = 941, + rv_op_amominu_h = 942, + rv_op_amomaxu_h = 943, + rv_op_amocas_b = 944, + rv_op_amocas_h = 945, } rv_op; /* register names */ @@ -XXX,XX +XXX,XX @@ const rv_opcode_data rvi_opcode_data[] = { { "c.mop.11", rv_codec_ci_none, rv_fmt_none, NULL, 0, 0, 0 }, { "c.mop.13", rv_codec_ci_none, rv_fmt_none, NULL, 0, 0, 0 }, { "c.mop.15", rv_codec_ci_none, rv_fmt_none, NULL, 0, 0, 0 }, + { "amoswap.b", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amoadd.b", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amoxor.b", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amoor.b", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amoand.b", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amomin.b", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amomax.b", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amominu.b", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amomaxu.b", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amoswap.h", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amoadd.h", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amoxor.h", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amoor.h", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amoand.h", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amomin.h", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amomax.h", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amominu.h", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amomaxu.h", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amocas.b", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "amocas.h", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, }; /* CSR names */ @@ -XXX,XX +XXX,XX @@ static void decode_inst_opcode(rv_decode *dec, rv_isa isa) case 11: switch (((inst >> 24) & 0b11111000) | ((inst >> 12) & 0b00000111)) { + case 0: op = rv_op_amoadd_b; break; + case 1: op = rv_op_amoadd_h; break; case 2: op = rv_op_amoadd_w; break; case 3: op = rv_op_amoadd_d; break; case 4: op = rv_op_amoadd_q; break; + case 8: op = rv_op_amoswap_b; break; + case 9: op = rv_op_amoswap_h; break; case 10: op = rv_op_amoswap_w; break; case 11: op = rv_op_amoswap_d; break; case 12: op = rv_op_amoswap_q; break; @@ -XXX,XX +XXX,XX @@ static void decode_inst_opcode(rv_decode *dec, rv_isa isa) case 26: op = rv_op_sc_w; break; case 27: op = rv_op_sc_d; break; case 28: op = rv_op_sc_q; break; + case 32: op = rv_op_amoxor_b; break; + case 33: op = rv_op_amoxor_h; break; case 34: op = rv_op_amoxor_w; break; case 35: op = rv_op_amoxor_d; break; case 36: op = rv_op_amoxor_q; break; + case 40: op = rv_op_amocas_b; break; + case 41: op = rv_op_amocas_h; break; case 42: op = rv_op_amocas_w; break; case 43: op = rv_op_amocas_d; break; case 44: op = rv_op_amocas_q; break; + case 64: op = rv_op_amoor_b; break; + case 65: op = rv_op_amoor_h; break; case 66: op = rv_op_amoor_w; break; case 67: op = rv_op_amoor_d; break; case 68: op = rv_op_amoor_q; break; + case 96: op = rv_op_amoand_b; break; + case 97: op = rv_op_amoand_h; break; case 98: op = rv_op_amoand_w; break; case 99: op = rv_op_amoand_d; break; case 100: op = rv_op_amoand_q; break; + case 128: op = rv_op_amomin_b; break; + case 129: op = rv_op_amomin_h; break; case 130: op = rv_op_amomin_w; break; case 131: op = rv_op_amomin_d; break; case 132: op = rv_op_amomin_q; break; + case 160: op = rv_op_amomax_b; break; + case 161: op = rv_op_amomax_h; break; case 162: op = rv_op_amomax_w; break; case 163: op = rv_op_amomax_d; break; case 164: op = rv_op_amomax_q; break; + case 192: op = rv_op_amominu_b; break; + case 193: op = rv_op_amominu_h; break; case 194: op = rv_op_amominu_w; break; case 195: op = rv_op_amominu_d; break; case 196: op = rv_op_amominu_q; break; + case 224: op = rv_op_amomaxu_b; break; + case 225: op = rv_op_amomaxu_h; break; case 226: op = rv_op_amomaxu_w; break; case 227: op = rv_op_amomaxu_d; break; case 228: op = rv_op_amomaxu_q; break; -- 2.45.2 From: Jiayi Li <lijiayi@eswincomputing.com> Base on the riscv-privileged spec, vstvec substitutes for the usual stvec. Therefore, the encoding of the MODE should also be restricted to 0 and 1. Signed-off-by: Jiayi Li <lijiayi@eswincomputing.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Message-ID: <20240701022553.1982-1-lijiayi@eswincomputing.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/csr.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static RISCVException read_vstvec(CPURISCVState *env, int csrno, static RISCVException write_vstvec(CPURISCVState *env, int csrno, target_ulong val) { - env->vstvec = val; + /* bits [1:0] encode mode; 0 = direct, 1 = vectored, 2 >= reserved */ + if ((val & 3) < 2) { + env->vstvec = val; + } else { + qemu_log_mask(LOG_UNIMP, "CSR_VSTVEC: reserved mode not supported\n"); + } return RISCV_EXCP_NONE; } -- 2.45.2 From: Balaji Ravikumar <bravikumar@rivosinc.com> Add disassembly support for these instructions from Zawrs: * wrs.sto * wrs.nto Signed-off-by: Balaji Ravikumar <bravikumar@rivosinc.com> Signed-off-by: Rob Bradford <rbradford@rivosinc.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240705165316.127494-1-rbradford@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- disas/riscv.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/disas/riscv.c b/disas/riscv.c index XXXXXXX..XXXXXXX 100644 --- a/disas/riscv.c +++ b/disas/riscv.c @@ -XXX,XX +XXX,XX @@ typedef enum { rv_op_amomaxu_h = 943, rv_op_amocas_b = 944, rv_op_amocas_h = 945, + rv_op_wrs_sto = 946, + rv_op_wrs_nto = 947, } rv_op; /* register names */ @@ -XXX,XX +XXX,XX @@ const rv_opcode_data rvi_opcode_data[] = { { "amomaxu.h", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, { "amocas.b", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, { "amocas.h", rv_codec_r_a, rv_fmt_aqrl_rd_rs2_rs1, NULL, 0, 0, 0 }, + { "wrs.sto", rv_codec_none, rv_fmt_none, NULL, 0, 0, 0 }, + { "wrs.nto", rv_codec_none, rv_fmt_none, NULL, 0, 0, 0 }, }; /* CSR names */ @@ -XXX,XX +XXX,XX @@ static void decode_inst_opcode(rv_decode *dec, rv_isa isa) case 0: op = rv_op_ecall; break; case 32: op = rv_op_ebreak; break; case 64: op = rv_op_uret; break; + case 416: op = rv_op_wrs_nto; break; + case 928: op = rv_op_wrs_sto; break; } break; case 256: -- 2.45.2 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Two new regs added: ztso and zacas. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240709085431.455541-1-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/kvm/kvm-cpu.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/target/riscv/kvm/kvm-cpu.c b/target/riscv/kvm/kvm-cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm/kvm-cpu.c +++ b/target/riscv/kvm/kvm-cpu.c @@ -XXX,XX +XXX,XX @@ static KVMCPUConfig kvm_multi_ext_cfgs[] = { KVM_EXT_CFG("zihintntl", ext_zihintntl, KVM_RISCV_ISA_EXT_ZIHINTNTL), KVM_EXT_CFG("zihintpause", ext_zihintpause, KVM_RISCV_ISA_EXT_ZIHINTPAUSE), KVM_EXT_CFG("zihpm", ext_zihpm, KVM_RISCV_ISA_EXT_ZIHPM), + KVM_EXT_CFG("zacas", ext_zacas, KVM_RISCV_ISA_EXT_ZACAS), KVM_EXT_CFG("zfa", ext_zfa, KVM_RISCV_ISA_EXT_ZFA), KVM_EXT_CFG("zfh", ext_zfh, KVM_RISCV_ISA_EXT_ZFH), KVM_EXT_CFG("zfhmin", ext_zfhmin, KVM_RISCV_ISA_EXT_ZFHMIN), @@ -XXX,XX +XXX,XX @@ static KVMCPUConfig kvm_multi_ext_cfgs[] = { KVM_EXT_CFG("zksed", ext_zksed, KVM_RISCV_ISA_EXT_ZKSED), KVM_EXT_CFG("zksh", ext_zksh, KVM_RISCV_ISA_EXT_ZKSH), KVM_EXT_CFG("zkt", ext_zkt, KVM_RISCV_ISA_EXT_ZKT), + KVM_EXT_CFG("ztso", ext_ztso, KVM_RISCV_ISA_EXT_ZTSO), KVM_EXT_CFG("zvbb", ext_zvbb, KVM_RISCV_ISA_EXT_ZVBB), KVM_EXT_CFG("zvbc", ext_zvbc, KVM_RISCV_ISA_EXT_ZVBC), KVM_EXT_CFG("zvfh", ext_zvfh, KVM_RISCV_ISA_EXT_ZVFH), -- 2.45.2 From: Rajnesh Kanwal <rkanwal@rivosinc.com> Combining riscv_cpu_set_virt_enabled() and riscv_cpu_set_mode() functions. This is to make complete mode change information available through a single function. This allows to easily differentiate between HS->VS, VS->HS and VS->VS transitions when executing state update codes. For example: One use-case which inspired this change is to update mode-specific instruction and cycle counters which requires information of both prev mode and current mode. Signed-off-by: Rajnesh Kanwal <rkanwal@rivosinc.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Message-ID: <20240711-smcntrpmf_v7-v8-1-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 2 +- target/riscv/cpu_helper.c | 57 +++++++++++++++++++-------------------- target/riscv/op_helper.c | 17 +++++------- 3 files changed, 35 insertions(+), 41 deletions(-) diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.h +++ b/target/riscv/cpu.h @@ -XXX,XX +XXX,XX @@ void riscv_cpu_set_aia_ireg_rmw_fn(CPURISCVState *env, uint32_t priv, RISCVException smstateen_acc_ok(CPURISCVState *env, int index, uint64_t bit); #endif /* !CONFIG_USER_ONLY */ -void riscv_cpu_set_mode(CPURISCVState *env, target_ulong newpriv); +void riscv_cpu_set_mode(CPURISCVState *env, target_ulong newpriv, bool virt_en); void riscv_translate_init(void); G_NORETURN void riscv_raise_exception(CPURISCVState *env, diff --git a/target/riscv/cpu_helper.c b/target/riscv/cpu_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_helper.c +++ b/target/riscv/cpu_helper.c @@ -XXX,XX +XXX,XX @@ void riscv_cpu_set_geilen(CPURISCVState *env, target_ulong geilen) env->geilen = geilen; } -/* This function can only be called to set virt when RVH is enabled */ -void riscv_cpu_set_virt_enabled(CPURISCVState *env, bool enable) -{ - /* Flush the TLB on all virt mode changes. */ - if (env->virt_enabled != enable) { - tlb_flush(env_cpu(env)); - } - - env->virt_enabled = enable; - - if (enable) { - /* - * The guest external interrupts from an interrupt controller are - * delivered only when the Guest/VM is running (i.e. V=1). This means - * any guest external interrupt which is triggered while the Guest/VM - * is not running (i.e. V=0) will be missed on QEMU resulting in guest - * with sluggish response to serial console input and other I/O events. - * - * To solve this, we check and inject interrupt after setting V=1. - */ - riscv_cpu_update_mip(env, 0, 0); - } -} - int riscv_cpu_claim_interrupts(RISCVCPU *cpu, uint64_t interrupts) { CPURISCVState *env = &cpu->env; @@ -XXX,XX +XXX,XX @@ void riscv_cpu_set_aia_ireg_rmw_fn(CPURISCVState *env, uint32_t priv, } } -void riscv_cpu_set_mode(CPURISCVState *env, target_ulong newpriv) +void riscv_cpu_set_mode(CPURISCVState *env, target_ulong newpriv, bool virt_en) { g_assert(newpriv <= PRV_M && newpriv != PRV_RESERVED); @@ -XXX,XX +XXX,XX @@ void riscv_cpu_set_mode(CPURISCVState *env, target_ulong newpriv) * preemptive context switch. As a result, do both. */ env->load_res = -1; + + if (riscv_has_ext(env, RVH)) { + /* Flush the TLB on all virt mode changes. */ + if (env->virt_enabled != virt_en) { + tlb_flush(env_cpu(env)); + } + + env->virt_enabled = virt_en; + if (virt_en) { + /* + * The guest external interrupts from an interrupt controller are + * delivered only when the Guest/VM is running (i.e. V=1). This + * means any guest external interrupt which is triggered while the + * Guest/VM is not running (i.e. V=0) will be missed on QEMU + * resulting in guest with sluggish response to serial console + * input and other I/O events. + * + * To solve this, we check and inject interrupt after setting V=1. + */ + riscv_cpu_update_mip(env, 0, 0); + } + } } /* @@ -XXX,XX +XXX,XX @@ void riscv_cpu_do_interrupt(CPUState *cs) { RISCVCPU *cpu = RISCV_CPU(cs); CPURISCVState *env = &cpu->env; + bool virt = env->virt_enabled; bool write_gva = false; uint64_t s; @@ -XXX,XX +XXX,XX @@ void riscv_cpu_do_interrupt(CPUState *cs) htval = env->guest_phys_fault_addr; - riscv_cpu_set_virt_enabled(env, 0); + virt = false; } else { /* Trap into HS mode */ env->hstatus = set_field(env->hstatus, HSTATUS_SPV, false); @@ -XXX,XX +XXX,XX @@ void riscv_cpu_do_interrupt(CPUState *cs) env->htinst = tinst; env->pc = (env->stvec >> 2 << 2) + ((async && (env->stvec & 3) == 1) ? cause * 4 : 0); - riscv_cpu_set_mode(env, PRV_S); + riscv_cpu_set_mode(env, PRV_S, virt); } else { /* handle the trap in M-mode */ if (riscv_has_ext(env, RVH)) { @@ -XXX,XX +XXX,XX @@ void riscv_cpu_do_interrupt(CPUState *cs) mtval2 = env->guest_phys_fault_addr; /* Trapping to M mode, virt is disabled */ - riscv_cpu_set_virt_enabled(env, 0); + virt = false; } s = env->mstatus; @@ -XXX,XX +XXX,XX @@ void riscv_cpu_do_interrupt(CPUState *cs) env->mtinst = tinst; env->pc = (env->mtvec >> 2 << 2) + ((async && (env->mtvec & 3) == 1) ? cause * 4 : 0); - riscv_cpu_set_mode(env, PRV_M); + riscv_cpu_set_mode(env, PRV_M, virt); } /* diff --git a/target/riscv/op_helper.c b/target/riscv/op_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/op_helper.c +++ b/target/riscv/op_helper.c @@ -XXX,XX +XXX,XX @@ void helper_cbo_inval(CPURISCVState *env, target_ulong address) target_ulong helper_sret(CPURISCVState *env) { uint64_t mstatus; - target_ulong prev_priv, prev_virt; + target_ulong prev_priv, prev_virt = env->virt_enabled; if (!(env->priv >= PRV_S)) { riscv_raise_exception(env, RISCV_EXCP_ILLEGAL_INST, GETPC()); @@ -XXX,XX +XXX,XX @@ target_ulong helper_sret(CPURISCVState *env) if (prev_virt) { riscv_cpu_swap_hypervisor_regs(env); } - - riscv_cpu_set_virt_enabled(env, prev_virt); } - riscv_cpu_set_mode(env, prev_priv); + riscv_cpu_set_mode(env, prev_priv, prev_virt); return retpc; } @@ -XXX,XX +XXX,XX @@ target_ulong helper_mret(CPURISCVState *env) mstatus = set_field(mstatus, MSTATUS_MPRV, 0); } env->mstatus = mstatus; - riscv_cpu_set_mode(env, prev_priv); - - if (riscv_has_ext(env, RVH)) { - if (prev_virt) { - riscv_cpu_swap_hypervisor_regs(env); - } - riscv_cpu_set_virt_enabled(env, prev_virt); + if (riscv_has_ext(env, RVH) && prev_virt) { + riscv_cpu_swap_hypervisor_regs(env); } + riscv_cpu_set_mode(env, prev_priv, prev_virt); + return retpc; } -- 2.45.2 From: Atish Patra <atishp@rivosinc.com> mhpmeventhX CSRs are available for RV32. The predicate function should check that first before checking sscofpmf extension. Fixes: 14664483457b ("target/riscv: Add sscofpmf extension support") Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Atish Patra <atishp@rivosinc.com> Message-ID: <20240711-smcntrpmf_v7-v8-2-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/csr.c | 67 ++++++++++++++++++++++++++-------------------- 1 file changed, 38 insertions(+), 29 deletions(-) diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static RISCVException sscofpmf(CPURISCVState *env, int csrno) return RISCV_EXCP_NONE; } +static RISCVException sscofpmf_32(CPURISCVState *env, int csrno) +{ + if (riscv_cpu_mxl(env) != MXL_RV32) { + return RISCV_EXCP_ILLEGAL_INST; + } + + return sscofpmf(env, csrno); +} + static RISCVException any(CPURISCVState *env, int csrno) { return RISCV_EXCP_NONE; @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_MHPMEVENT31] = { "mhpmevent31", any, read_mhpmevent, write_mhpmevent }, - [CSR_MHPMEVENT3H] = { "mhpmevent3h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT3H] = { "mhpmevent3h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT4H] = { "mhpmevent4h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT4H] = { "mhpmevent4h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT5H] = { "mhpmevent5h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT5H] = { "mhpmevent5h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT6H] = { "mhpmevent6h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT6H] = { "mhpmevent6h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT7H] = { "mhpmevent7h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT7H] = { "mhpmevent7h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT8H] = { "mhpmevent8h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT8H] = { "mhpmevent8h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT9H] = { "mhpmevent9h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT9H] = { "mhpmevent9h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT10H] = { "mhpmevent10h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT10H] = { "mhpmevent10h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT11H] = { "mhpmevent11h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT11H] = { "mhpmevent11h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT12H] = { "mhpmevent12h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT12H] = { "mhpmevent12h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT13H] = { "mhpmevent13h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT13H] = { "mhpmevent13h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT14H] = { "mhpmevent14h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT14H] = { "mhpmevent14h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT15H] = { "mhpmevent15h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT15H] = { "mhpmevent15h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT16H] = { "mhpmevent16h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT16H] = { "mhpmevent16h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT17H] = { "mhpmevent17h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT17H] = { "mhpmevent17h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT18H] = { "mhpmevent18h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT18H] = { "mhpmevent18h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT19H] = { "mhpmevent19h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT19H] = { "mhpmevent19h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT20H] = { "mhpmevent20h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT20H] = { "mhpmevent20h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT21H] = { "mhpmevent21h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT21H] = { "mhpmevent21h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT22H] = { "mhpmevent22h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT22H] = { "mhpmevent22h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT23H] = { "mhpmevent23h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT23H] = { "mhpmevent23h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT24H] = { "mhpmevent24h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT24H] = { "mhpmevent24h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT25H] = { "mhpmevent25h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT25H] = { "mhpmevent25h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT26H] = { "mhpmevent26h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT26H] = { "mhpmevent26h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT27H] = { "mhpmevent27h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT27H] = { "mhpmevent27h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT28H] = { "mhpmevent28h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT28H] = { "mhpmevent28h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT29H] = { "mhpmevent29h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT29H] = { "mhpmevent29h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT30H] = { "mhpmevent30h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT30H] = { "mhpmevent30h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, - [CSR_MHPMEVENT31H] = { "mhpmevent31h", sscofpmf, read_mhpmeventh, + [CSR_MHPMEVENT31H] = { "mhpmevent31h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, -- 2.45.2 From: Kaiwen Xue <kaiwenx@rivosinc.com> This adds the properties for ISA extension smcntrpmf. Patches implementing it will follow. Signed-off-by: Kaiwen Xue <kaiwenx@rivosinc.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Signed-off-by: Atish Patra <atishp@rivosinc.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240711-smcntrpmf_v7-v8-3-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_cfg.h | 1 + target/riscv/cpu.c | 1 + 2 files changed, 2 insertions(+) diff --git a/target/riscv/cpu_cfg.h b/target/riscv/cpu_cfg.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_cfg.h +++ b/target/riscv/cpu_cfg.h @@ -XXX,XX +XXX,XX @@ struct RISCVCPUConfig { bool ext_ztso; bool ext_smstateen; bool ext_sstc; + bool ext_smcntrpmf; bool ext_svadu; bool ext_svinval; bool ext_svnapot; diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const RISCVIsaExtData isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(zhinx, PRIV_VERSION_1_12_0, ext_zhinx), ISA_EXT_DATA_ENTRY(zhinxmin, PRIV_VERSION_1_12_0, ext_zhinxmin), ISA_EXT_DATA_ENTRY(smaia, PRIV_VERSION_1_12_0, ext_smaia), + ISA_EXT_DATA_ENTRY(smcntrpmf, PRIV_VERSION_1_12_0, ext_smcntrpmf), ISA_EXT_DATA_ENTRY(smepmp, PRIV_VERSION_1_12_0, ext_smepmp), ISA_EXT_DATA_ENTRY(smstateen, PRIV_VERSION_1_12_0, ext_smstateen), ISA_EXT_DATA_ENTRY(ssaia, PRIV_VERSION_1_12_0, ext_ssaia), -- 2.45.2 From: Kaiwen Xue <kaiwenx@rivosinc.com> This adds the definitions for ISA extension smcntrpmf. Signed-off-by: Kaiwen Xue <kaiwenx@rivosinc.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Atish Patra <atishp@rivosinc.com> Message-ID: <20240711-smcntrpmf_v7-v8-4-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 6 ++++++ target/riscv/cpu_bits.h | 29 +++++++++++++++++++++++++++++ 2 files changed, 35 insertions(+) diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.h +++ b/target/riscv/cpu.h @@ -XXX,XX +XXX,XX @@ struct CPUArchState { uint32_t mcountinhibit; + /* PMU cycle & instret privilege mode filtering */ + target_ulong mcyclecfg; + target_ulong mcyclecfgh; + target_ulong minstretcfg; + target_ulong minstretcfgh; + /* PMU counter state */ PMUCTRState pmu_ctrs[RV_MAX_MHPMCOUNTERS]; diff --git a/target/riscv/cpu_bits.h b/target/riscv/cpu_bits.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_bits.h +++ b/target/riscv/cpu_bits.h @@ -XXX,XX +XXX,XX @@ /* Machine counter-inhibit register */ #define CSR_MCOUNTINHIBIT 0x320 +/* Machine counter configuration registers */ +#define CSR_MCYCLECFG 0x321 +#define CSR_MINSTRETCFG 0x322 + #define CSR_MHPMEVENT3 0x323 #define CSR_MHPMEVENT4 0x324 #define CSR_MHPMEVENT5 0x325 @@ -XXX,XX +XXX,XX @@ #define CSR_MHPMEVENT30 0x33e #define CSR_MHPMEVENT31 0x33f +#define CSR_MCYCLECFGH 0x721 +#define CSR_MINSTRETCFGH 0x722 + #define CSR_MHPMEVENT3H 0x723 #define CSR_MHPMEVENT4H 0x724 #define CSR_MHPMEVENT5H 0x725 @@ -XXX,XX +XXX,XX @@ typedef enum RISCVException { /* PMU related bits */ #define MIE_LCOFIE (1 << IRQ_PMU_OVF) +#define MCYCLECFG_BIT_MINH BIT_ULL(62) +#define MCYCLECFGH_BIT_MINH BIT(30) +#define MCYCLECFG_BIT_SINH BIT_ULL(61) +#define MCYCLECFGH_BIT_SINH BIT(29) +#define MCYCLECFG_BIT_UINH BIT_ULL(60) +#define MCYCLECFGH_BIT_UINH BIT(28) +#define MCYCLECFG_BIT_VSINH BIT_ULL(59) +#define MCYCLECFGH_BIT_VSINH BIT(27) +#define MCYCLECFG_BIT_VUINH BIT_ULL(58) +#define MCYCLECFGH_BIT_VUINH BIT(26) + +#define MINSTRETCFG_BIT_MINH BIT_ULL(62) +#define MINSTRETCFGH_BIT_MINH BIT(30) +#define MINSTRETCFG_BIT_SINH BIT_ULL(61) +#define MINSTRETCFGH_BIT_SINH BIT(29) +#define MINSTRETCFG_BIT_UINH BIT_ULL(60) +#define MINSTRETCFGH_BIT_UINH BIT(28) +#define MINSTRETCFG_BIT_VSINH BIT_ULL(59) +#define MINSTRETCFGH_BIT_VSINH BIT(27) +#define MINSTRETCFG_BIT_VUINH BIT_ULL(58) +#define MINSTRETCFGH_BIT_VUINH BIT(26) + #define MHPMEVENT_BIT_OF BIT_ULL(63) #define MHPMEVENTH_BIT_OF BIT(31) #define MHPMEVENT_BIT_MINH BIT_ULL(62) -- 2.45.2 From: Kaiwen Xue <kaiwenx@rivosinc.com> QEMU only calculates dummy cycles and instructions, so there is no actual means to stop the icount in QEMU. Hence this patch merely adds the functionality of accessing the cfg registers, and cause no actual effects on the counting of cycle and instret counters. Signed-off-by: Atish Patra <atishp@rivosinc.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Signed-off-by: Kaiwen Xue <kaiwenx@rivosinc.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240711-smcntrpmf_v7-v8-5-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_bits.h | 12 ++++ target/riscv/csr.c | 138 +++++++++++++++++++++++++++++++++++++++- 2 files changed, 149 insertions(+), 1 deletion(-) diff --git a/target/riscv/cpu_bits.h b/target/riscv/cpu_bits.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_bits.h +++ b/target/riscv/cpu_bits.h @@ -XXX,XX +XXX,XX @@ typedef enum RISCVException { #define MHPMEVENT_BIT_VUINH BIT_ULL(58) #define MHPMEVENTH_BIT_VUINH BIT(26) +#define MHPMEVENT_FILTER_MASK (MHPMEVENT_BIT_MINH | \ + MHPMEVENT_BIT_SINH | \ + MHPMEVENT_BIT_UINH | \ + MHPMEVENT_BIT_VSINH | \ + MHPMEVENT_BIT_VUINH) + +#define MHPMEVENTH_FILTER_MASK (MHPMEVENTH_BIT_MINH | \ + MHPMEVENTH_BIT_SINH | \ + MHPMEVENTH_BIT_UINH | \ + MHPMEVENTH_BIT_VSINH | \ + MHPMEVENTH_BIT_VUINH) + #define MHPMEVENT_SSCOF_MASK _ULL(0xFFFF000000000000) #define MHPMEVENT_IDX_MASK 0xFFFFF #define MHPMEVENT_SSCOF_RESVD 16 diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ #include "qemu/guest-random.h" #include "qapi/error.h" - /* CSR function table public API */ void riscv_get_csr_ops(int csrno, riscv_csr_operations *ops) { @@ -XXX,XX +XXX,XX @@ static RISCVException sscofpmf_32(CPURISCVState *env, int csrno) return sscofpmf(env, csrno); } +static RISCVException smcntrpmf(CPURISCVState *env, int csrno) +{ + if (!riscv_cpu_cfg(env)->ext_smcntrpmf) { + return RISCV_EXCP_ILLEGAL_INST; + } + + return RISCV_EXCP_NONE; +} + +static RISCVException smcntrpmf_32(CPURISCVState *env, int csrno) +{ + if (riscv_cpu_mxl(env) != MXL_RV32) { + return RISCV_EXCP_ILLEGAL_INST; + } + + return smcntrpmf(env, csrno); +} + static RISCVException any(CPURISCVState *env, int csrno) { return RISCV_EXCP_NONE; @@ -XXX,XX +XXX,XX @@ static RISCVException read_hpmcounterh(CPURISCVState *env, int csrno, #else /* CONFIG_USER_ONLY */ +static RISCVException read_mcyclecfg(CPURISCVState *env, int csrno, + target_ulong *val) +{ + *val = env->mcyclecfg; + return RISCV_EXCP_NONE; +} + +static RISCVException write_mcyclecfg(CPURISCVState *env, int csrno, + target_ulong val) +{ + uint64_t inh_avail_mask; + + if (riscv_cpu_mxl(env) == MXL_RV32) { + env->mcyclecfg = val; + } else { + /* Set xINH fields if priv mode supported */ + inh_avail_mask = ~MHPMEVENT_FILTER_MASK | MCYCLECFG_BIT_MINH; + inh_avail_mask |= riscv_has_ext(env, RVU) ? MCYCLECFG_BIT_UINH : 0; + inh_avail_mask |= riscv_has_ext(env, RVS) ? MCYCLECFG_BIT_SINH : 0; + inh_avail_mask |= (riscv_has_ext(env, RVH) && + riscv_has_ext(env, RVU)) ? MCYCLECFG_BIT_VUINH : 0; + inh_avail_mask |= (riscv_has_ext(env, RVH) && + riscv_has_ext(env, RVS)) ? MCYCLECFG_BIT_VSINH : 0; + env->mcyclecfg = val & inh_avail_mask; + } + + return RISCV_EXCP_NONE; +} + +static RISCVException read_mcyclecfgh(CPURISCVState *env, int csrno, + target_ulong *val) +{ + *val = env->mcyclecfgh; + return RISCV_EXCP_NONE; +} + +static RISCVException write_mcyclecfgh(CPURISCVState *env, int csrno, + target_ulong val) +{ + target_ulong inh_avail_mask = (target_ulong)(~MHPMEVENTH_FILTER_MASK | + MCYCLECFGH_BIT_MINH); + + /* Set xINH fields if priv mode supported */ + inh_avail_mask |= riscv_has_ext(env, RVU) ? MCYCLECFGH_BIT_UINH : 0; + inh_avail_mask |= riscv_has_ext(env, RVS) ? MCYCLECFGH_BIT_SINH : 0; + inh_avail_mask |= (riscv_has_ext(env, RVH) && + riscv_has_ext(env, RVU)) ? MCYCLECFGH_BIT_VUINH : 0; + inh_avail_mask |= (riscv_has_ext(env, RVH) && + riscv_has_ext(env, RVS)) ? MCYCLECFGH_BIT_VSINH : 0; + + env->mcyclecfgh = val & inh_avail_mask; + return RISCV_EXCP_NONE; +} + +static RISCVException read_minstretcfg(CPURISCVState *env, int csrno, + target_ulong *val) +{ + *val = env->minstretcfg; + return RISCV_EXCP_NONE; +} + +static RISCVException write_minstretcfg(CPURISCVState *env, int csrno, + target_ulong val) +{ + uint64_t inh_avail_mask; + + if (riscv_cpu_mxl(env) == MXL_RV32) { + env->minstretcfg = val; + } else { + inh_avail_mask = ~MHPMEVENT_FILTER_MASK | MINSTRETCFG_BIT_MINH; + inh_avail_mask |= riscv_has_ext(env, RVU) ? MINSTRETCFG_BIT_UINH : 0; + inh_avail_mask |= riscv_has_ext(env, RVS) ? MINSTRETCFG_BIT_SINH : 0; + inh_avail_mask |= (riscv_has_ext(env, RVH) && + riscv_has_ext(env, RVU)) ? MINSTRETCFG_BIT_VUINH : 0; + inh_avail_mask |= (riscv_has_ext(env, RVH) && + riscv_has_ext(env, RVS)) ? MINSTRETCFG_BIT_VSINH : 0; + env->minstretcfg = val & inh_avail_mask; + } + return RISCV_EXCP_NONE; +} + +static RISCVException read_minstretcfgh(CPURISCVState *env, int csrno, + target_ulong *val) +{ + *val = env->minstretcfgh; + return RISCV_EXCP_NONE; +} + +static RISCVException write_minstretcfgh(CPURISCVState *env, int csrno, + target_ulong val) +{ + target_ulong inh_avail_mask = (target_ulong)(~MHPMEVENTH_FILTER_MASK | + MINSTRETCFGH_BIT_MINH); + + inh_avail_mask |= riscv_has_ext(env, RVU) ? MINSTRETCFGH_BIT_UINH : 0; + inh_avail_mask |= riscv_has_ext(env, RVS) ? MINSTRETCFGH_BIT_SINH : 0; + inh_avail_mask |= (riscv_has_ext(env, RVH) && + riscv_has_ext(env, RVU)) ? MINSTRETCFGH_BIT_VUINH : 0; + inh_avail_mask |= (riscv_has_ext(env, RVH) && + riscv_has_ext(env, RVS)) ? MINSTRETCFGH_BIT_VSINH : 0; + + env->minstretcfgh = val & inh_avail_mask; + return RISCV_EXCP_NONE; +} + static RISCVException read_mhpmevent(CPURISCVState *env, int csrno, target_ulong *val) { @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { write_mcountinhibit, .min_priv_ver = PRIV_VERSION_1_11_0 }, + [CSR_MCYCLECFG] = { "mcyclecfg", smcntrpmf, read_mcyclecfg, + write_mcyclecfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_MINSTRETCFG] = { "minstretcfg", smcntrpmf, read_minstretcfg, + write_minstretcfg, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_MHPMEVENT3] = { "mhpmevent3", any, read_mhpmevent, write_mhpmevent }, [CSR_MHPMEVENT4] = { "mhpmevent4", any, read_mhpmevent, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_MHPMEVENT31] = { "mhpmevent31", any, read_mhpmevent, write_mhpmevent }, + [CSR_MCYCLECFGH] = { "mcyclecfgh", smcntrpmf_32, read_mcyclecfgh, + write_mcyclecfgh, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_MINSTRETCFGH] = { "minstretcfgh", smcntrpmf_32, read_minstretcfgh, + write_minstretcfgh, + .min_priv_ver = PRIV_VERSION_1_12_0 }, + [CSR_MHPMEVENT3H] = { "mhpmevent3h", sscofpmf_32, read_mhpmeventh, write_mhpmeventh, .min_priv_ver = PRIV_VERSION_1_12_0 }, -- 2.45.2 From: Atish Patra <atishp@rivosinc.com> Currently, the INH fields are set in mhpmevent uncoditionally without checking if a particular priv mode is supported or not. Suggested-by: Alistair Francis <alistair23@gmail.com> Signed-off-by: Atish Patra <atishp@rivosinc.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240711-smcntrpmf_v7-v8-6-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/csr.c | 29 +++++++++++++++++++++++++---- 1 file changed, 25 insertions(+), 4 deletions(-) diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static RISCVException write_mhpmevent(CPURISCVState *env, int csrno, { int evt_index = csrno - CSR_MCOUNTINHIBIT; uint64_t mhpmevt_val = val; - - env->mhpmevent_val[evt_index] = val; + uint64_t inh_avail_mask; if (riscv_cpu_mxl(env) == MXL_RV32) { + env->mhpmevent_val[evt_index] = val; mhpmevt_val = mhpmevt_val | ((uint64_t)env->mhpmeventh_val[evt_index] << 32); + } else { + inh_avail_mask = ~MHPMEVENT_FILTER_MASK | MHPMEVENT_BIT_MINH; + inh_avail_mask |= riscv_has_ext(env, RVU) ? MHPMEVENT_BIT_UINH : 0; + inh_avail_mask |= riscv_has_ext(env, RVS) ? MHPMEVENT_BIT_SINH : 0; + inh_avail_mask |= (riscv_has_ext(env, RVH) && + riscv_has_ext(env, RVU)) ? MHPMEVENT_BIT_VUINH : 0; + inh_avail_mask |= (riscv_has_ext(env, RVH) && + riscv_has_ext(env, RVS)) ? MHPMEVENT_BIT_VSINH : 0; + mhpmevt_val = val & inh_avail_mask; + env->mhpmevent_val[evt_index] = mhpmevt_val; } + riscv_pmu_update_event_map(env, mhpmevt_val, evt_index); return RISCV_EXCP_NONE; @@ -XXX,XX +XXX,XX @@ static RISCVException write_mhpmeventh(CPURISCVState *env, int csrno, target_ulong val) { int evt_index = csrno - CSR_MHPMEVENT3H + 3; - uint64_t mhpmevth_val = val; + uint64_t mhpmevth_val; uint64_t mhpmevt_val = env->mhpmevent_val[evt_index]; + target_ulong inh_avail_mask = (target_ulong)(~MHPMEVENTH_FILTER_MASK | + MHPMEVENTH_BIT_MINH); + + inh_avail_mask |= riscv_has_ext(env, RVU) ? MHPMEVENTH_BIT_UINH : 0; + inh_avail_mask |= riscv_has_ext(env, RVS) ? MHPMEVENTH_BIT_SINH : 0; + inh_avail_mask |= (riscv_has_ext(env, RVH) && + riscv_has_ext(env, RVU)) ? MHPMEVENTH_BIT_VUINH : 0; + inh_avail_mask |= (riscv_has_ext(env, RVH) && + riscv_has_ext(env, RVS)) ? MHPMEVENTH_BIT_VSINH : 0; + mhpmevth_val = val & inh_avail_mask; mhpmevt_val = mhpmevt_val | (mhpmevth_val << 32); - env->mhpmeventh_val[evt_index] = val; + env->mhpmeventh_val[evt_index] = mhpmevth_val; riscv_pmu_update_event_map(env, mhpmevt_val, evt_index); -- 2.45.2 From: Atish Patra <atishp@rivosinc.com> Privilege mode filtering can also be emulated for cycle/instret by tracking host_ticks/icount during each privilege mode switch. This patch implements that for both cycle/instret and mhpmcounters. The first one requires Smcntrpmf while the other one requires Sscofpmf to be enabled. The cycle/instret are still computed using host ticks when icount is not enabled. Otherwise, they are computed using raw icount which is more accurate in icount mode. Co-Developed-by: Rajnesh Kanwal <rkanwal@rivosinc.com> Signed-off-by: Rajnesh Kanwal <rkanwal@rivosinc.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Atish Patra <atishp@rivosinc.com> Message-ID: <20240711-smcntrpmf_v7-v8-7-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 11 ++++ target/riscv/pmu.h | 2 + target/riscv/cpu_helper.c | 9 ++- target/riscv/csr.c | 117 ++++++++++++++++++++++++++------------ target/riscv/pmu.c | 92 ++++++++++++++++++++++++++++++ 5 files changed, 194 insertions(+), 37 deletions(-) diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.h +++ b/target/riscv/cpu.h @@ -XXX,XX +XXX,XX @@ typedef struct PMUCTRState { target_ulong irq_overflow_left; } PMUCTRState; +typedef struct PMUFixedCtrState { + /* Track cycle and icount for each privilege mode */ + uint64_t counter[4]; + uint64_t counter_prev[4]; + /* Track cycle and icount for each privilege mode when V = 1*/ + uint64_t counter_virt[2]; + uint64_t counter_virt_prev[2]; +} PMUFixedCtrState; + struct CPUArchState { target_ulong gpr[32]; target_ulong gprh[32]; /* 64 top bits of the 128-bit registers */ @@ -XXX,XX +XXX,XX @@ struct CPUArchState { /* PMU event selector configured values for RV32 */ target_ulong mhpmeventh_val[RV_MAX_MHPMEVENTS]; + PMUFixedCtrState pmu_fixed_ctrs[2]; + target_ulong sscratch; target_ulong mscratch; diff --git a/target/riscv/pmu.h b/target/riscv/pmu.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/pmu.h +++ b/target/riscv/pmu.h @@ -XXX,XX +XXX,XX @@ int riscv_pmu_incr_ctr(RISCVCPU *cpu, enum riscv_pmu_event_idx event_idx); void riscv_pmu_generate_fdt_node(void *fdt, uint32_t cmask, char *pmu_name); int riscv_pmu_setup_timer(CPURISCVState *env, uint64_t value, uint32_t ctr_idx); +void riscv_pmu_update_fixed_ctrs(CPURISCVState *env, target_ulong newpriv, + bool new_virt); #endif /* RISCV_PMU_H */ diff --git a/target/riscv/cpu_helper.c b/target/riscv/cpu_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_helper.c +++ b/target/riscv/cpu_helper.c @@ -XXX,XX +XXX,XX @@ void riscv_cpu_set_mode(CPURISCVState *env, target_ulong newpriv, bool virt_en) { g_assert(newpriv <= PRV_M && newpriv != PRV_RESERVED); - if (icount_enabled() && newpriv != env->priv) { - riscv_itrigger_update_priv(env); + if (newpriv != env->priv || env->virt_enabled != virt_en) { + if (icount_enabled()) { + riscv_itrigger_update_priv(env); + } + + riscv_pmu_update_fixed_ctrs(env, newpriv, virt_en); } + /* tlb_flush is unnecessary as mode is contained in mmu_idx */ env->priv = newpriv; env->xl = cpu_recompute_xl(env); diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static RISCVException write_vcsr(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } +#if defined(CONFIG_USER_ONLY) /* User Timers and Counters */ -static target_ulong get_ticks(bool shift, bool instructions) +static target_ulong get_ticks(bool shift) { - int64_t val; - target_ulong result; - -#if !defined(CONFIG_USER_ONLY) - if (icount_enabled()) { - if (instructions) { - val = icount_get_raw(); - } else { - val = icount_get(); - } - } else { - val = cpu_get_host_ticks(); - } -#else - val = cpu_get_host_ticks(); -#endif - - if (shift) { - result = val >> 32; - } else { - result = val; - } + int64_t val = cpu_get_host_ticks(); + target_ulong result = shift ? val >> 32 : val; return result; } -#if defined(CONFIG_USER_ONLY) static RISCVException read_time(CPURISCVState *env, int csrno, target_ulong *val) { @@ -XXX,XX +XXX,XX @@ static RISCVException read_timeh(CPURISCVState *env, int csrno, static RISCVException read_hpmcounter(CPURISCVState *env, int csrno, target_ulong *val) { - *val = get_ticks(false, (csrno == CSR_INSTRET)); + *val = get_ticks(false); return RISCV_EXCP_NONE; } static RISCVException read_hpmcounterh(CPURISCVState *env, int csrno, target_ulong *val) { - *val = get_ticks(true, (csrno == CSR_INSTRETH)); + *val = get_ticks(true); return RISCV_EXCP_NONE; } @@ -XXX,XX +XXX,XX @@ static RISCVException write_mhpmeventh(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } +static target_ulong riscv_pmu_ctr_get_fixed_counters_val(CPURISCVState *env, + int counter_idx, + bool upper_half) +{ + int inst = riscv_pmu_ctr_monitor_instructions(env, counter_idx); + uint64_t *counter_arr_virt = env->pmu_fixed_ctrs[inst].counter_virt; + uint64_t *counter_arr = env->pmu_fixed_ctrs[inst].counter; + target_ulong result = 0; + uint64_t curr_val = 0; + uint64_t cfg_val = 0; + + if (counter_idx == 0) { + cfg_val = upper_half ? ((uint64_t)env->mcyclecfgh << 32) : + env->mcyclecfg; + } else if (counter_idx == 2) { + cfg_val = upper_half ? ((uint64_t)env->minstretcfgh << 32) : + env->minstretcfg; + } else { + cfg_val = upper_half ? + ((uint64_t)env->mhpmeventh_val[counter_idx] << 32) : + env->mhpmevent_val[counter_idx]; + cfg_val &= MHPMEVENT_FILTER_MASK; + } + + if (!cfg_val) { + if (icount_enabled()) { + curr_val = inst ? icount_get_raw() : icount_get(); + } else { + curr_val = cpu_get_host_ticks(); + } + + goto done; + } + + if (!(cfg_val & MCYCLECFG_BIT_MINH)) { + curr_val += counter_arr[PRV_M]; + } + + if (!(cfg_val & MCYCLECFG_BIT_SINH)) { + curr_val += counter_arr[PRV_S]; + } + + if (!(cfg_val & MCYCLECFG_BIT_UINH)) { + curr_val += counter_arr[PRV_U]; + } + + if (!(cfg_val & MCYCLECFG_BIT_VSINH)) { + curr_val += counter_arr_virt[PRV_S]; + } + + if (!(cfg_val & MCYCLECFG_BIT_VUINH)) { + curr_val += counter_arr_virt[PRV_U]; + } + +done: + if (riscv_cpu_mxl(env) == MXL_RV32) { + result = upper_half ? curr_val >> 32 : curr_val; + } else { + result = curr_val; + } + + return result; +} + static RISCVException write_mhpmcounter(CPURISCVState *env, int csrno, target_ulong val) { int ctr_idx = csrno - CSR_MCYCLE; PMUCTRState *counter = &env->pmu_ctrs[ctr_idx]; uint64_t mhpmctr_val = val; - bool instr = riscv_pmu_ctr_monitor_instructions(env, ctr_idx); counter->mhpmcounter_val = val; - if (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) || instr) { - counter->mhpmcounter_prev = get_ticks(false, instr); + if (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) || + riscv_pmu_ctr_monitor_instructions(env, ctr_idx)) { + counter->mhpmcounter_prev = riscv_pmu_ctr_get_fixed_counters_val(env, + ctr_idx, false); if (ctr_idx > 2) { if (riscv_cpu_mxl(env) == MXL_RV32) { mhpmctr_val = mhpmctr_val | @@ -XXX,XX +XXX,XX @@ static RISCVException write_mhpmcounterh(CPURISCVState *env, int csrno, PMUCTRState *counter = &env->pmu_ctrs[ctr_idx]; uint64_t mhpmctr_val = counter->mhpmcounter_val; uint64_t mhpmctrh_val = val; - bool instr = riscv_pmu_ctr_monitor_instructions(env, ctr_idx); counter->mhpmcounterh_val = val; mhpmctr_val = mhpmctr_val | (mhpmctrh_val << 32); - if (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) || instr) { - counter->mhpmcounterh_prev = get_ticks(true, instr); + if (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) || + riscv_pmu_ctr_monitor_instructions(env, ctr_idx)) { + counter->mhpmcounterh_prev = riscv_pmu_ctr_get_fixed_counters_val(env, + ctr_idx, true); if (ctr_idx > 2) { riscv_pmu_setup_timer(env, mhpmctr_val, ctr_idx); } @@ -XXX,XX +XXX,XX @@ static RISCVException riscv_pmu_read_ctr(CPURISCVState *env, target_ulong *val, counter->mhpmcounter_prev; target_ulong ctr_val = upper_half ? counter->mhpmcounterh_val : counter->mhpmcounter_val; - bool instr = riscv_pmu_ctr_monitor_instructions(env, ctr_idx); if (get_field(env->mcountinhibit, BIT(ctr_idx))) { /* @@ -XXX,XX +XXX,XX @@ static RISCVException riscv_pmu_read_ctr(CPURISCVState *env, target_ulong *val, * The kernel computes the perf delta by subtracting the current value from * the value it initialized previously (ctr_val). */ - if (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) || instr) { - *val = get_ticks(upper_half, instr) - ctr_prev + ctr_val; + if (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) || + riscv_pmu_ctr_monitor_instructions(env, ctr_idx)) { + *val = riscv_pmu_ctr_get_fixed_counters_val(env, ctr_idx, upper_half) - + ctr_prev + ctr_val; } else { *val = ctr_val; } diff --git a/target/riscv/pmu.c b/target/riscv/pmu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/pmu.c +++ b/target/riscv/pmu.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "qemu/log.h" #include "qemu/error-report.h" +#include "qemu/timer.h" #include "cpu.h" #include "pmu.h" #include "sysemu/cpu-timers.h" @@ -XXX,XX +XXX,XX @@ static int riscv_pmu_incr_ctr_rv64(RISCVCPU *cpu, uint32_t ctr_idx) return 0; } +/* + * Information needed to update counters: + * new_priv, new_virt: To correctly save starting snapshot for the newly + * started mode. Look at array being indexed with newprv. + * old_priv, old_virt: To correctly select previous snapshot for old priv + * and compute delta. Also to select correct counter + * to inc. Look at arrays being indexed with env->priv. + * + * To avoid the complexity of calling this function, we assume that + * env->priv and env->virt_enabled contain old priv and old virt and + * new priv and new virt values are passed in as arguments. + */ +static void riscv_pmu_icount_update_priv(CPURISCVState *env, + target_ulong newpriv, bool new_virt) +{ + uint64_t *snapshot_prev, *snapshot_new; + uint64_t current_icount; + uint64_t *counter_arr; + uint64_t delta; + + if (icount_enabled()) { + current_icount = icount_get_raw(); + } else { + current_icount = cpu_get_host_ticks(); + } + + if (env->virt_enabled) { + counter_arr = env->pmu_fixed_ctrs[1].counter_virt; + snapshot_prev = env->pmu_fixed_ctrs[1].counter_virt_prev; + } else { + counter_arr = env->pmu_fixed_ctrs[1].counter; + snapshot_prev = env->pmu_fixed_ctrs[1].counter_prev; + } + + if (new_virt) { + snapshot_new = env->pmu_fixed_ctrs[1].counter_virt_prev; + } else { + snapshot_new = env->pmu_fixed_ctrs[1].counter_prev; + } + + /* + * new_priv can be same as env->priv. So we need to calculate + * delta first before updating snapshot_new[new_priv]. + */ + delta = current_icount - snapshot_prev[env->priv]; + snapshot_new[newpriv] = current_icount; + + counter_arr[env->priv] += delta; +} + +static void riscv_pmu_cycle_update_priv(CPURISCVState *env, + target_ulong newpriv, bool new_virt) +{ + uint64_t *snapshot_prev, *snapshot_new; + uint64_t current_ticks; + uint64_t *counter_arr; + uint64_t delta; + + if (icount_enabled()) { + current_ticks = icount_get(); + } else { + current_ticks = cpu_get_host_ticks(); + } + + if (env->virt_enabled) { + counter_arr = env->pmu_fixed_ctrs[0].counter_virt; + snapshot_prev = env->pmu_fixed_ctrs[0].counter_virt_prev; + } else { + counter_arr = env->pmu_fixed_ctrs[0].counter; + snapshot_prev = env->pmu_fixed_ctrs[0].counter_prev; + } + + if (new_virt) { + snapshot_new = env->pmu_fixed_ctrs[0].counter_virt_prev; + } else { + snapshot_new = env->pmu_fixed_ctrs[0].counter_prev; + } + + delta = current_ticks - snapshot_prev[env->priv]; + snapshot_new[newpriv] = current_ticks; + + counter_arr[env->priv] += delta; +} + +void riscv_pmu_update_fixed_ctrs(CPURISCVState *env, target_ulong newpriv, + bool new_virt) +{ + riscv_pmu_cycle_update_priv(env, newpriv, new_virt); + riscv_pmu_icount_update_priv(env, newpriv, new_virt); +} + int riscv_pmu_incr_ctr(RISCVCPU *cpu, enum riscv_pmu_event_idx event_idx) { uint32_t ctr_idx; -- 2.45.2 From: Atish Patra <atishp@rivosinc.com> Currently, if a counter monitoring cycle/instret is stopped via mcountinhibit we just update the state while the value is saved during the next read. This is not accurate as the read may happen many cycles after the counter is stopped. Ideally, the read should return the value saved when the counter is stopped. Thus, save the value of the counter during the inhibit update operation and return that value during the read if corresponding bit in mcountihibit is set. Acked-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Signed-off-by: Atish Patra <atishp@rivosinc.com> Message-ID: <20240711-smcntrpmf_v7-v8-8-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 1 - target/riscv/csr.c | 34 ++++++++++++++++++++++------------ target/riscv/machine.c | 5 ++--- 3 files changed, 24 insertions(+), 16 deletions(-) diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.h +++ b/target/riscv/cpu.h @@ -XXX,XX +XXX,XX @@ typedef struct PMUCTRState { target_ulong mhpmcounter_prev; /* Snapshort value of a counter in RV32 */ target_ulong mhpmcounterh_prev; - bool started; /* Value beyond UINT32_MAX/UINT64_MAX before overflow interrupt trigger */ target_ulong irq_overflow_left; } PMUCTRState; diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static RISCVException riscv_pmu_read_ctr(CPURISCVState *env, target_ulong *val, if (get_field(env->mcountinhibit, BIT(ctr_idx))) { /* - * Counter should not increment if inhibit bit is set. We can't really - * stop the icount counting. Just return the counter value written by - * the supervisor to indicate that counter was not incremented. + * Counter should not increment if inhibit bit is set. Just return the + * current counter value. */ - if (!counter->started) { - *val = ctr_val; - return RISCV_EXCP_NONE; - } else { - /* Mark that the counter has been stopped */ - counter->started = false; - } + *val = ctr_val; + return RISCV_EXCP_NONE; } /* @@ -XXX,XX +XXX,XX @@ static RISCVException write_mcountinhibit(CPURISCVState *env, int csrno, /* Check if any other counter is also monitoring cycles/instructions */ for (cidx = 0; cidx < RV_MAX_MHPMCOUNTERS; cidx++) { - if (!get_field(env->mcountinhibit, BIT(cidx))) { counter = &env->pmu_ctrs[cidx]; - counter->started = true; + if (get_field(env->mcountinhibit, BIT(cidx)) && (val & BIT(cidx))) { + /* + * Update the counter value for cycle/instret as we can't stop the + * host ticks. But we should show the current value at this moment. + */ + if (riscv_pmu_ctr_monitor_cycles(env, cidx) || + riscv_pmu_ctr_monitor_instructions(env, cidx)) { + counter->mhpmcounter_val = + riscv_pmu_ctr_get_fixed_counters_val(env, cidx, false) - + counter->mhpmcounter_prev + + counter->mhpmcounter_val; + if (riscv_cpu_mxl(env) == MXL_RV32) { + counter->mhpmcounterh_val = + riscv_pmu_ctr_get_fixed_counters_val(env, cidx, true) - + counter->mhpmcounterh_prev + + counter->mhpmcounterh_val; + } + } } } diff --git a/target/riscv/machine.c b/target/riscv/machine.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/machine.c +++ b/target/riscv/machine.c @@ -XXX,XX +XXX,XX @@ static bool pmu_needed(void *opaque) static const VMStateDescription vmstate_pmu_ctr_state = { .name = "cpu/pmu", - .version_id = 1, - .minimum_version_id = 1, + .version_id = 2, + .minimum_version_id = 2, .needed = pmu_needed, .fields = (const VMStateField[]) { VMSTATE_UINTTL(mhpmcounter_val, PMUCTRState), VMSTATE_UINTTL(mhpmcounterh_val, PMUCTRState), VMSTATE_UINTTL(mhpmcounter_prev, PMUCTRState), VMSTATE_UINTTL(mhpmcounterh_prev, PMUCTRState), - VMSTATE_BOOL(started, PMUCTRState), VMSTATE_END_OF_LIST() } }; -- 2.45.2 From: Atish Patra <atishp@rivosinc.com> scounteren/hcountern are also WARL registers similar to mcountern. Only set the bits for the available counters during the write to preserve the WARL behavior. Signed-off-by: Atish Patra <atishp@rivosinc.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240711-smcntrpmf_v7-v8-9-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/csr.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static RISCVException read_scounteren(CPURISCVState *env, int csrno, static RISCVException write_scounteren(CPURISCVState *env, int csrno, target_ulong val) { - env->scounteren = val; + RISCVCPU *cpu = env_archcpu(env); + + /* WARL register - disable unavailable counters */ + env->scounteren = val & (cpu->pmu_avail_ctrs | COUNTEREN_CY | COUNTEREN_TM | + COUNTEREN_IR); return RISCV_EXCP_NONE; } @@ -XXX,XX +XXX,XX @@ static RISCVException read_hcounteren(CPURISCVState *env, int csrno, static RISCVException write_hcounteren(CPURISCVState *env, int csrno, target_ulong val) { - env->hcounteren = val; + RISCVCPU *cpu = env_archcpu(env); + + /* WARL register - disable unavailable counters */ + env->hcounteren = val & (cpu->pmu_avail_ctrs | COUNTEREN_CY | COUNTEREN_TM | + COUNTEREN_IR); return RISCV_EXCP_NONE; } -- 2.45.2 From: Rajnesh Kanwal <rkanwal@rivosinc.com> Currently we start timer counter from write_mhpmcounter path only without checking for mcountinhibit bit. This changes adds mcountinhibit check and also programs the counter from write_mcountinhibit as well. When a counter is stopped using mcountinhibit we simply update the value of the counter based on current host ticks and save it for future reads. We don't need to disable running timer as pmu_timer_trigger_irq will discard the interrupt if the counter has been inhibited. Signed-off-by: Rajnesh Kanwal <rkanwal@rivosinc.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Message-ID: <20240711-smcntrpmf_v7-v8-10-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/csr.c | 75 ++++++++++++++++++++++++++++++++-------------- target/riscv/pmu.c | 3 +- 2 files changed, 54 insertions(+), 24 deletions(-) diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static RISCVException write_mhpmcounter(CPURISCVState *env, int csrno, uint64_t mhpmctr_val = val; counter->mhpmcounter_val = val; - if (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) || - riscv_pmu_ctr_monitor_instructions(env, ctr_idx)) { + if (!get_field(env->mcountinhibit, BIT(ctr_idx)) && + (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) || + riscv_pmu_ctr_monitor_instructions(env, ctr_idx))) { counter->mhpmcounter_prev = riscv_pmu_ctr_get_fixed_counters_val(env, ctr_idx, false); if (ctr_idx > 2) { @@ -XXX,XX +XXX,XX @@ static RISCVException write_mhpmcounterh(CPURISCVState *env, int csrno, counter->mhpmcounterh_val = val; mhpmctr_val = mhpmctr_val | (mhpmctrh_val << 32); - if (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) || - riscv_pmu_ctr_monitor_instructions(env, ctr_idx)) { + if (!get_field(env->mcountinhibit, BIT(ctr_idx)) && + (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) || + riscv_pmu_ctr_monitor_instructions(env, ctr_idx))) { counter->mhpmcounterh_prev = riscv_pmu_ctr_get_fixed_counters_val(env, ctr_idx, true); if (ctr_idx > 2) { @@ -XXX,XX +XXX,XX @@ static RISCVException write_mcountinhibit(CPURISCVState *env, int csrno, int cidx; PMUCTRState *counter; RISCVCPU *cpu = env_archcpu(env); + uint32_t present_ctrs = cpu->pmu_avail_ctrs | COUNTEREN_CY | COUNTEREN_IR; + target_ulong updated_ctrs = (env->mcountinhibit ^ val) & present_ctrs; + uint64_t mhpmctr_val, prev_count, curr_count; /* WARL register - disable unavailable counters; TM bit is always 0 */ - env->mcountinhibit = - val & (cpu->pmu_avail_ctrs | COUNTEREN_CY | COUNTEREN_IR); + env->mcountinhibit = val & present_ctrs; /* Check if any other counter is also monitoring cycles/instructions */ for (cidx = 0; cidx < RV_MAX_MHPMCOUNTERS; cidx++) { - counter = &env->pmu_ctrs[cidx]; - if (get_field(env->mcountinhibit, BIT(cidx)) && (val & BIT(cidx))) { - /* - * Update the counter value for cycle/instret as we can't stop the - * host ticks. But we should show the current value at this moment. - */ - if (riscv_pmu_ctr_monitor_cycles(env, cidx) || - riscv_pmu_ctr_monitor_instructions(env, cidx)) { - counter->mhpmcounter_val = - riscv_pmu_ctr_get_fixed_counters_val(env, cidx, false) - - counter->mhpmcounter_prev + - counter->mhpmcounter_val; + if (!(updated_ctrs & BIT(cidx)) || + (!riscv_pmu_ctr_monitor_cycles(env, cidx) && + !riscv_pmu_ctr_monitor_instructions(env, cidx))) { + continue; + } + + counter = &env->pmu_ctrs[cidx]; + + if (!get_field(env->mcountinhibit, BIT(cidx))) { + counter->mhpmcounter_prev = + riscv_pmu_ctr_get_fixed_counters_val(env, cidx, false); + if (riscv_cpu_mxl(env) == MXL_RV32) { + counter->mhpmcounterh_prev = + riscv_pmu_ctr_get_fixed_counters_val(env, cidx, true); + } + + if (cidx > 2) { + mhpmctr_val = counter->mhpmcounter_val; if (riscv_cpu_mxl(env) == MXL_RV32) { - counter->mhpmcounterh_val = - riscv_pmu_ctr_get_fixed_counters_val(env, cidx, true) - - counter->mhpmcounterh_prev + - counter->mhpmcounterh_val; + mhpmctr_val = mhpmctr_val | + ((uint64_t)counter->mhpmcounterh_val << 32); } + riscv_pmu_setup_timer(env, mhpmctr_val, cidx); + } + } else { + curr_count = riscv_pmu_ctr_get_fixed_counters_val(env, cidx, false); + + mhpmctr_val = counter->mhpmcounter_val; + prev_count = counter->mhpmcounter_prev; + if (riscv_cpu_mxl(env) == MXL_RV32) { + uint64_t tmp = + riscv_pmu_ctr_get_fixed_counters_val(env, cidx, true); + + curr_count = curr_count | (tmp << 32); + mhpmctr_val = mhpmctr_val | + ((uint64_t)counter->mhpmcounterh_val << 32); + prev_count = prev_count | + ((uint64_t)counter->mhpmcounterh_prev << 32); + } + + /* Adjust the counter for later reads. */ + mhpmctr_val = curr_count - prev_count + mhpmctr_val; + counter->mhpmcounter_val = mhpmctr_val; + if (riscv_cpu_mxl(env) == MXL_RV32) { + counter->mhpmcounterh_val = mhpmctr_val >> 32; } } } diff --git a/target/riscv/pmu.c b/target/riscv/pmu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/pmu.c +++ b/target/riscv/pmu.c @@ -XXX,XX +XXX,XX @@ int riscv_pmu_incr_ctr(RISCVCPU *cpu, enum riscv_pmu_event_idx event_idx) } ctr_idx = GPOINTER_TO_UINT(value); - if (!riscv_pmu_counter_enabled(cpu, ctr_idx) || - get_field(env->mcountinhibit, BIT(ctr_idx))) { + if (!riscv_pmu_counter_enabled(cpu, ctr_idx)) { return -1; } -- 2.45.2 From: Rajnesh Kanwal <rkanwal@rivosinc.com> In case of programmable counters configured to count inst/cycles we often end-up with counter not incrementing at all from kernel's perspective. For example: - Kernel configures hpm3 to count instructions and sets hpmcounter to -10000 and all modes except U mode are inhibited. - In QEMU we configure a timer to expire after ~10000 instructions. - Problem is, it's often the case that kernel might not even schedule Umode task and we hit the timer callback in QEMU. - In the timer callback we inject the interrupt into kernel, kernel runs the handler and reads hpmcounter3 value. - Given QEMU maintains individual counters to count for each privilege mode, and given umode never ran, the umode counter didn't increment and QEMU returns same value as was programmed by the kernel when starting the counter. - Kernel checks for overflow using previous and current value of the counter and reprograms the counter given there wasn't an overflow as per the counter value. (Which itself is a problem. We have QEMU telling kernel that counter3 overflowed but the counter value returned by QEMU doesn't seem to reflect that.). This change makes sure that timer is reprogrammed from the handler if the counter didn't overflow based on the counter value. Second, this change makes sure that whenever the counter is read, it's value is updated to reflect the latest count. Signed-off-by: Rajnesh Kanwal <rkanwal@rivosinc.com> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Message-ID: <20240711-smcntrpmf_v7-v8-11-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/pmu.h | 2 ++ target/riscv/csr.c | 5 ++++- target/riscv/pmu.c | 30 +++++++++++++++++++++++++++--- 3 files changed, 33 insertions(+), 4 deletions(-) diff --git a/target/riscv/pmu.h b/target/riscv/pmu.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/pmu.h +++ b/target/riscv/pmu.h @@ -XXX,XX +XXX,XX @@ int riscv_pmu_setup_timer(CPURISCVState *env, uint64_t value, uint32_t ctr_idx); void riscv_pmu_update_fixed_ctrs(CPURISCVState *env, target_ulong newpriv, bool new_virt); +RISCVException riscv_pmu_read_ctr(CPURISCVState *env, target_ulong *val, + bool upper_half, uint32_t ctr_idx); #endif /* RISCV_PMU_H */ diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static target_ulong riscv_pmu_ctr_get_fixed_counters_val(CPURISCVState *env, goto done; } + /* Update counter before reading. */ + riscv_pmu_update_fixed_ctrs(env, env->priv, env->virt_enabled); + if (!(cfg_val & MCYCLECFG_BIT_MINH)) { curr_val += counter_arr[PRV_M]; } @@ -XXX,XX +XXX,XX @@ static RISCVException write_mhpmcounterh(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } -static RISCVException riscv_pmu_read_ctr(CPURISCVState *env, target_ulong *val, +RISCVException riscv_pmu_read_ctr(CPURISCVState *env, target_ulong *val, bool upper_half, uint32_t ctr_idx) { PMUCTRState *counter = &env->pmu_ctrs[ctr_idx]; diff --git a/target/riscv/pmu.c b/target/riscv/pmu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/pmu.c +++ b/target/riscv/pmu.c @@ -XXX,XX +XXX,XX @@ static void pmu_timer_trigger_irq(RISCVCPU *cpu, target_ulong *mhpmevent_val; uint64_t of_bit_mask; int64_t irq_trigger_at; + uint64_t curr_ctr_val, curr_ctrh_val; + uint64_t ctr_val; if (evt_idx != RISCV_PMU_EVENT_HW_CPU_CYCLES && evt_idx != RISCV_PMU_EVENT_HW_INSTRUCTIONS) { @@ -XXX,XX +XXX,XX @@ static void pmu_timer_trigger_irq(RISCVCPU *cpu, return; } + riscv_pmu_read_ctr(env, (target_ulong *)&curr_ctr_val, false, ctr_idx); + ctr_val = counter->mhpmcounter_val; + if (riscv_cpu_mxl(env) == MXL_RV32) { + riscv_pmu_read_ctr(env, (target_ulong *)&curr_ctrh_val, true, ctr_idx); + curr_ctr_val = curr_ctr_val | (curr_ctrh_val << 32); + ctr_val = ctr_val | + ((uint64_t)counter->mhpmcounterh_val << 32); + } + + /* + * We can not accommodate for inhibited modes when setting up timer. Check + * if the counter has actually overflowed or not by comparing current + * counter value (accommodated for inhibited modes) with software written + * counter value. + */ + if (curr_ctr_val >= ctr_val) { + riscv_pmu_setup_timer(env, curr_ctr_val, ctr_idx); + return; + } + if (cpu->pmu_avail_ctrs & BIT(ctr_idx)) { /* Generate interrupt only if OF bit is clear */ if (!(*mhpmevent_val & of_bit_mask)) { @@ -XXX,XX +XXX,XX @@ void riscv_pmu_timer_cb(void *priv) int riscv_pmu_setup_timer(CPURISCVState *env, uint64_t value, uint32_t ctr_idx) { - uint64_t overflow_delta, overflow_at; + uint64_t overflow_delta, overflow_at, curr_ns; int64_t overflow_ns, overflow_left = 0; RISCVCPU *cpu = env_archcpu(env); PMUCTRState *counter = &env->pmu_ctrs[ctr_idx]; @@ -XXX,XX +XXX,XX @@ int riscv_pmu_setup_timer(CPURISCVState *env, uint64_t value, uint32_t ctr_idx) } else { return -1; } - overflow_at = (uint64_t)qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL) + - overflow_ns; + curr_ns = (uint64_t)qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL); + overflow_at = curr_ns + overflow_ns; + if (overflow_at <= curr_ns) + overflow_at = UINT64_MAX; if (overflow_at > INT64_MAX) { overflow_left += overflow_at - INT64_MAX; -- 2.45.2 From: Atish Patra <atishp@rivosinc.com> The timer is setup function is invoked in both hpmcounter write and mcountinhibit write path. If the OF bit set, the LCOFI interrupt is disabled. There is no benefitting in setting up the qemu timer until LCOFI is cleared to indicate that interrupts can be fired again. Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Signed-off-by: Atish Patra <atishp@rivosinc.com> Message-ID: <20240711-smcntrpmf_v7-v8-12-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/pmu.c | 56 ++++++++++++++++++++++++++++++++++++---------- 1 file changed, 44 insertions(+), 12 deletions(-) diff --git a/target/riscv/pmu.c b/target/riscv/pmu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/pmu.c +++ b/target/riscv/pmu.c @@ -XXX,XX +XXX,XX @@ int riscv_pmu_update_event_map(CPURISCVState *env, uint64_t value, return 0; } +static bool pmu_hpmevent_is_of_set(CPURISCVState *env, uint32_t ctr_idx) +{ + target_ulong mhpmevent_val; + uint64_t of_bit_mask; + + if (riscv_cpu_mxl(env) == MXL_RV32) { + mhpmevent_val = env->mhpmeventh_val[ctr_idx]; + of_bit_mask = MHPMEVENTH_BIT_OF; + } else { + mhpmevent_val = env->mhpmevent_val[ctr_idx]; + of_bit_mask = MHPMEVENT_BIT_OF; + } + + return get_field(mhpmevent_val, of_bit_mask); +} + +static bool pmu_hpmevent_set_of_if_clear(CPURISCVState *env, uint32_t ctr_idx) +{ + target_ulong *mhpmevent_val; + uint64_t of_bit_mask; + + if (riscv_cpu_mxl(env) == MXL_RV32) { + mhpmevent_val = &env->mhpmeventh_val[ctr_idx]; + of_bit_mask = MHPMEVENTH_BIT_OF; + } else { + mhpmevent_val = &env->mhpmevent_val[ctr_idx]; + of_bit_mask = MHPMEVENT_BIT_OF; + } + + if (!get_field(*mhpmevent_val, of_bit_mask)) { + *mhpmevent_val |= of_bit_mask; + return true; + } + + return false; +} + static void pmu_timer_trigger_irq(RISCVCPU *cpu, enum riscv_pmu_event_idx evt_idx) { uint32_t ctr_idx; CPURISCVState *env = &cpu->env; PMUCTRState *counter; - target_ulong *mhpmevent_val; - uint64_t of_bit_mask; int64_t irq_trigger_at; uint64_t curr_ctr_val, curr_ctrh_val; uint64_t ctr_val; @@ -XXX,XX +XXX,XX @@ static void pmu_timer_trigger_irq(RISCVCPU *cpu, return; } - if (riscv_cpu_mxl(env) == MXL_RV32) { - mhpmevent_val = &env->mhpmeventh_val[ctr_idx]; - of_bit_mask = MHPMEVENTH_BIT_OF; - } else { - mhpmevent_val = &env->mhpmevent_val[ctr_idx]; - of_bit_mask = MHPMEVENT_BIT_OF; + /* Generate interrupt only if OF bit is clear */ + if (pmu_hpmevent_is_of_set(env, ctr_idx)) { + return; } counter = &env->pmu_ctrs[ctr_idx]; @@ -XXX,XX +XXX,XX @@ static void pmu_timer_trigger_irq(RISCVCPU *cpu, } if (cpu->pmu_avail_ctrs & BIT(ctr_idx)) { - /* Generate interrupt only if OF bit is clear */ - if (!(*mhpmevent_val & of_bit_mask)) { - *mhpmevent_val |= of_bit_mask; + if (pmu_hpmevent_set_of_if_clear(env, ctr_idx)) { riscv_cpu_update_mip(env, MIP_LCOFIP, BOOL_TO_MASK(1)); } } @@ -XXX,XX +XXX,XX @@ int riscv_pmu_setup_timer(CPURISCVState *env, uint64_t value, uint32_t ctr_idx) RISCVCPU *cpu = env_archcpu(env); PMUCTRState *counter = &env->pmu_ctrs[ctr_idx]; - if (!riscv_pmu_counter_valid(cpu, ctr_idx) || !cpu->cfg.ext_sscofpmf) { + /* No need to setup a timer if LCOFI is disabled when OF is set */ + if (!riscv_pmu_counter_valid(cpu, ctr_idx) || !cpu->cfg.ext_sscofpmf || + pmu_hpmevent_is_of_set(env, ctr_idx)) { return -1; } -- 2.45.2 From: Atish Patra <atishp@rivosinc.com> Create a new config for Smcntrpmf extension so that it can be enabled/ disabled from the qemu commandline. Signed-off-by: Atish Patra <atishp@rivosinc.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240711-smcntrpmf_v7-v8-13-b7c38ae7b263@rivosinc.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 1 + 1 file changed, 1 insertion(+) diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const char *riscv_get_misa_ext_description(uint32_t bit) const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { /* Defaults for standard extensions */ MULTI_EXT_CFG_BOOL("sscofpmf", ext_sscofpmf, false), + MULTI_EXT_CFG_BOOL("smcntrpmf", ext_smcntrpmf, false), MULTI_EXT_CFG_BOOL("zifencei", ext_zifencei, true), MULTI_EXT_CFG_BOOL("zicsr", ext_zicsr, true), MULTI_EXT_CFG_BOOL("zihintntl", ext_zihintntl, true), -- 2.45.2 From: Yu-Ming Chang <yumin686@andestech.com> Both CSRRS and CSRRC always read the addressed CSR and cause any read side effects regardless of rs1 and rd fields. Note that if rs1 specifies a register holding a zero value other than x0, the instruction will still attempt to write the unmodified value back to the CSR and will cause any attendant side effects. So if CSRRS or CSRRC tries to write a read-only CSR with rs1 which specifies a register holding a zero value, an illegal instruction exception should be raised. Signed-off-by: Yu-Ming Chang <yumin686@andestech.com> Signed-off-by: Alvin Chang <alvinga@andestech.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <172100444279.18077.6893072378718059541-0@git.sr.ht> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 4 +++ target/riscv/csr.c | 57 ++++++++++++++++++++++++++++++++++++---- target/riscv/op_helper.c | 6 ++--- 3 files changed, 58 insertions(+), 9 deletions(-) diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.h +++ b/target/riscv/cpu.h @@ -XXX,XX +XXX,XX @@ void cpu_get_tb_cpu_state(CPURISCVState *env, vaddr *pc, void riscv_cpu_update_mask(CPURISCVState *env); bool riscv_cpu_is_32bit(RISCVCPU *cpu); +RISCVException riscv_csrr(CPURISCVState *env, int csrno, + target_ulong *ret_value); RISCVException riscv_csrrw(CPURISCVState *env, int csrno, target_ulong *ret_value, target_ulong new_value, target_ulong write_mask); @@ -XXX,XX +XXX,XX @@ typedef RISCVException (*riscv_csr_op_fn)(CPURISCVState *env, int csrno, target_ulong new_value, target_ulong write_mask); +RISCVException riscv_csrr_i128(CPURISCVState *env, int csrno, + Int128 *ret_value); RISCVException riscv_csrrw_i128(CPURISCVState *env, int csrno, Int128 *ret_value, Int128 new_value, Int128 write_mask); diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_seed(CPURISCVState *env, int csrno, static inline RISCVException riscv_csrrw_check(CPURISCVState *env, int csrno, - bool write_mask) + bool write) { /* check privileges and return RISCV_EXCP_ILLEGAL_INST if check fails */ bool read_only = get_field(csrno, 0xC00) == 3; @@ -XXX,XX +XXX,XX @@ static inline RISCVException riscv_csrrw_check(CPURISCVState *env, } /* read / write check */ - if (write_mask && read_only) { + if (write && read_only) { return RISCV_EXCP_ILLEGAL_INST; } @@ -XXX,XX +XXX,XX @@ static RISCVException riscv_csrrw_do64(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } +RISCVException riscv_csrr(CPURISCVState *env, int csrno, + target_ulong *ret_value) +{ + RISCVException ret = riscv_csrrw_check(env, csrno, false); + if (ret != RISCV_EXCP_NONE) { + return ret; + } + + return riscv_csrrw_do64(env, csrno, ret_value, 0, 0); +} + RISCVException riscv_csrrw(CPURISCVState *env, int csrno, target_ulong *ret_value, target_ulong new_value, target_ulong write_mask) { - RISCVException ret = riscv_csrrw_check(env, csrno, write_mask); + RISCVException ret = riscv_csrrw_check(env, csrno, true); if (ret != RISCV_EXCP_NONE) { return ret; } @@ -XXX,XX +XXX,XX @@ static RISCVException riscv_csrrw_do128(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } +RISCVException riscv_csrr_i128(CPURISCVState *env, int csrno, + Int128 *ret_value) +{ + RISCVException ret; + + ret = riscv_csrrw_check(env, csrno, false); + if (ret != RISCV_EXCP_NONE) { + return ret; + } + + if (csr_ops[csrno].read128) { + return riscv_csrrw_do128(env, csrno, ret_value, + int128_zero(), int128_zero()); + } + + /* + * Fall back to 64-bit version for now, if the 128-bit alternative isn't + * at all defined. + * Note, some CSRs don't need to extend to MXLEN (64 upper bits non + * significant), for those, this fallback is correctly handling the + * accesses + */ + target_ulong old_value; + ret = riscv_csrrw_do64(env, csrno, &old_value, + (target_ulong)0, + (target_ulong)0); + if (ret == RISCV_EXCP_NONE && ret_value) { + *ret_value = int128_make64(old_value); + } + return ret; +} + RISCVException riscv_csrrw_i128(CPURISCVState *env, int csrno, Int128 *ret_value, Int128 new_value, Int128 write_mask) { RISCVException ret; - ret = riscv_csrrw_check(env, csrno, int128_nz(write_mask)); + ret = riscv_csrrw_check(env, csrno, true); if (ret != RISCV_EXCP_NONE) { return ret; } @@ -XXX,XX +XXX,XX @@ RISCVException riscv_csrrw_debug(CPURISCVState *env, int csrno, #if !defined(CONFIG_USER_ONLY) env->debugger = true; #endif - ret = riscv_csrrw(env, csrno, ret_value, new_value, write_mask); + if (!write_mask) { + ret = riscv_csrr(env, csrno, ret_value); + } else { + ret = riscv_csrrw(env, csrno, ret_value, new_value, write_mask); + } #if !defined(CONFIG_USER_ONLY) env->debugger = false; #endif diff --git a/target/riscv/op_helper.c b/target/riscv/op_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/op_helper.c +++ b/target/riscv/op_helper.c @@ -XXX,XX +XXX,XX @@ target_ulong helper_csrr(CPURISCVState *env, int csr) } target_ulong val = 0; - RISCVException ret = riscv_csrrw(env, csr, &val, 0, 0); + RISCVException ret = riscv_csrr(env, csr, &val); if (ret != RISCV_EXCP_NONE) { riscv_raise_exception(env, ret, GETPC()); @@ -XXX,XX +XXX,XX @@ target_ulong helper_csrrw(CPURISCVState *env, int csr, target_ulong helper_csrr_i128(CPURISCVState *env, int csr) { Int128 rv = int128_zero(); - RISCVException ret = riscv_csrrw_i128(env, csr, &rv, - int128_zero(), - int128_zero()); + RISCVException ret = riscv_csrr_i128(env, csr, &rv); if (ret != RISCV_EXCP_NONE) { riscv_raise_exception(env, ret, GETPC()); -- 2.45.2 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Commit b1f1e9dcfa renamed 'riscv,delegate' to 'riscv,delegation' since it is the correct name as per dt-bindings, and the absence of the correct name will result in validation fails when dumping the dtb and using dt-validate. But this change has a side-effect: every other firmware available that is AIA capable is using 'riscv,delegate', and it will fault/misbehave if this property isn't present. The property was added back in QEMU 7.0, meaning we have 2 years of firmware development using the wrong property. Re-introducing 'riscv,delegate' while keeping 'riscv,delegation' allows older firmwares to keep booting with the 'virt' machine. 'riscv,delegate' is then marked for future deprecation with its use being discouraged from now on. Cc: Conor Dooley <conor@kernel.org> Cc: Anup Patel <apatel@ventanamicro.com> Fixes: b1f1e9dcfa ("hw/riscv/virt.c: aplic DT: rename prop to 'riscv, delegation'") Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240715090455.145888-1-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- docs/about/deprecated.rst | 11 +++++++++++ hw/riscv/virt.c | 9 +++++++++ 2 files changed, 20 insertions(+) diff --git a/docs/about/deprecated.rst b/docs/about/deprecated.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/about/deprecated.rst +++ b/docs/about/deprecated.rst @@ -XXX,XX +XXX,XX @@ versions, aliases will point to newer CPU model versions depending on the machine type, so management software must resolve CPU model aliases before starting a virtual machine. +RISC-V "virt" board "riscv,delegate" DT property (since 9.1) +'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''' + +The "riscv,delegate" DT property was added in QEMU 7.0 as part of +the AIA APLIC support. The property changed name during the review +process in Linux and the correct name ended up being +"riscv,delegation". Changing the DT property name will break all +available firmwares that are using the current (wrong) name. The +property is kept as is in 9.1, together with "riscv,delegation", to +give more time for firmware developers to change their code. + Migration --------- diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -XXX,XX +XXX,XX @@ static void create_fdt_one_aplic(RISCVVirtState *s, int socket, qemu_fdt_setprop_cells(ms->fdt, aplic_name, "riscv,delegation", aplic_child_phandle, 0x1, VIRT_IRQCHIP_NUM_SOURCES); + /* + * DEPRECATED_9.1: Compat property kept temporarily + * to allow old firmwares to work with AIA. Do *not* + * use 'riscv,delegate' in new code: use + * 'riscv,delegation' instead. + */ + qemu_fdt_setprop_cells(ms->fdt, aplic_name, "riscv,delegate", + aplic_child_phandle, 0x1, + VIRT_IRQCHIP_NUM_SOURCES); } riscv_socket_fdt_write_id(ms, aplic_name, socket); -- 2.45.2 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Update OpenSBI and the pre-built opensbi32 and opensbi64 images to v1.5. The following commits were included in v1.5: 455de67 include: Bump-up version to 1.5 23b7bad lib: sbi: check incoming dbtr shmem address 0e45b63 docs: Fix wrong filename caae2f7 lib: sbi: fwft: return SBI_EINVAL rather than SBI_ERR_INVALID_PARAM e8717d1 lib: sbi: fwft: check feature value to be exactly 1 or 0 ecef14d lib: sbi: implement SBI FWFT extension e9ee967 lib: sbi: fwft: add support for SBI_FWFT_PTE_AD_HW_UPDATING c97a1d5 lib: sbi: fwft: add support for SBI_FWFT_MISALIGNED_EXC_DELEG aa5a859 lib: sbi: add support for firmware features extension 53844c9 lib: sbi: Add support for Svade and Svadu extensions 52dcf35 platform: generic: andes: Add support for RV32 to set up PMA f09f164 platform: generic: andes: Refine Andes PMA related code 7830e98 lib: serial: fix RX path in litex-uart 62e178a lib: utils/reset: Try initializing all reset devices in dt 3a94a32 sbi: sbi_domain_context: Fix trap context for domain context switching a73ff04 lib: utils/reset: Fix fdt_reset to search for more dt nodes b5c984b lib: utils/reset: Skip initialize reset when dt is not enabled 86bbe6c lib: utils/serial: Fix fdt_serial to match more dt nodes 179e00a lib: utils/serial: Skip initialize serial when dt is not enabled b1c7c75 lib: utils/irqchip: Skip initialize irqchip when dt is not enabled 5e3ad7d lib: utils/timer: Skip initialize timer when dt is not enabled c5be0e1 lib: utils/ipi: Skip initialize ipi when dt is not enabled df3db6a lib: utils/fdt: Fix DT property for APLIC delegation d962db2 lib: utils/gpio: respect flag GPIO_FLAG_ACTIVE_LOW ae5ef18 lib: sbi: sse: handle missing writable attributes 858754a lib: utils/irqchip: Add sanity checks in imsic_get_data() and imsic_get_target_file() 96f0a2e firmware: Bring back FW_TEXT_START as an optional parameter e3a30a2 lib: utils/irqchip: Check before initializing imsic 2bed4c1 platform: generic: thead: add Sophgo CV18XX/SG200X series 533067d lib: sbi: Put event after use in sbi_sse_exit() loop ea9cf6a utils/reset: Add SG2042 hwmon MCU support. 1cb792d lib: sbi: simplify inline function in sbi_dtbr.c 7b37da3 lib: sbi: fix return type of sbi_dbtr_shmem_disabled e065c3c lib: sbi: Fixed memory permission check in sbi_dbtr_setup_shmem 7f54527 lib: sbi: fix DBTR_SHMEM_MAKE_PHYS for RV64 744f495 lib: sbi: Removal unnecessary check dbtr_thishart_state_ptr 4953bd7 lib: sbi: fix hart_shmem_base 019a8e6 platform: generic: thead: add Canaan Kendryte K230 33e21c9 platform: generic: thead: separate T-Head PMU Errata 2b93ce0 platform: andes: Change all occurrences of andes45 to andes f68b3ae platform: andes: Rename files with the prefix andes45 17e8291 sbi: sbi_domain_context: Add spinlock for updating domain assigned_harts 1d89a9d lib: sbi: Refine the settings for switching to Virtual Supervisor Mode. 033104d lib: sbi: sse: check handler entry to belong to supervisor mode bd00765 lib: sbi: sse: use PRV_S instead of hardcoded value for mode ce3c82c lib: sbi: sse: call enable callback before sending IPI d528dbf lib: sbi: sse: remove superfluous sbi_list_empty() check 22ff750 lib: sbi: sse: simplify 32bits overflow check 7aa80ea lib: sbi: sse: rename sse_hart_unlock() to sse_enabled_event_unlock() c21c99d lib: sbi: sse: fix typos, comments and spacing errors 7b1ed96 lib: tests: Add test for spinlocks 7bdf41a lib: tests: Add test for atomic_t f6243d9 lib: tests: Add test suite init function d4d2582 firmware: remove FW_TEXT_START 73344d4 lib: utils: check correct value in fdt_node_offset_by_compatible 37e1544 lib: sbi: sse_event_get() may return NULL 68bc031 lib: sbi: Add missing sscrind and sscfg extensions in sbi_hart_ext[] a7c5c2c Makefile: Remove unnecessary dependencies 268feab Makefile: Respect manual changes to .config 29ecda9 sbi: sbi_domain_context: Check privilege spec version before accessing S-mode CSRs 7862c24 lib: sbi: Wakeup non-coldboot HARTs early in the coldboot path beb0cd1 lib: sbi: Simplify wait_for_coldboot() implementation f5375bc platform: generic: allwinner: Optimize current hart scratch access b94396c lib: utils/timer: Optimize current hart scratch access 5c9a735 include: sbi: Support byteorder macros in assembly 06fc453 lib: sbi: Add SSE support for PMU events 09ad214 lib: sbi: Implement SBI SSE extension c8cdf01 lib: sbi: Add support for Supervisor Software Events extension 76d7e9b firmware: remove copy-base relocation 5186da6 platform: generic: allwinner: sun20i-d1: Remove duplicated CSR definitions 3b2f89e docs: writing_tests: Make docs correspond the latest changes f7d0050 lib: sbi: Extend sbi_trap_error() to dump state in a nested trap 5b11f16 lib: sbi: Pass trap context pointer to sbi_ecall_handler() 43d346c lib: sbi: Remove regs parameter from trap irq handling functions d84e7eb lib: sbi: Remove regs paramter of sbi_irqchip_process() f414cf9 lib: sbi: Simplify parameters of sbi_illegal_insn_handler() fea33a9 lib: sbi: Simplify parameters of misaligned and access fault handlers abea949 lib: sbi: Introduce trap context 60ffc15 include: sbi: Add trap_context pointer in struct sbi_scratch ebb697a lib: sbi: Remove sbi_trap_exit() and related code 2e85178 lib: sbi: Remove epc from struct sbi_trap_info 86224ec docs/writing_tests: Update tests paths 5c992a1 lib: tests: Move tests to a separate directory 81e3ba7 lib: sbi: call platform load/store emulators ddf3b64 include: sbi: add emulate_load/store handler to platform ops 4c11265 lib: sbi: abstract out insn decoding to unify mem fault handlers 9221fe5 lib: sbi: change prototype of sbi_misaligned_load/store_handler a17600c lib: sbi: change prototype of sbi_trap_redirect 2471cf2 include: sbi: rename sbi_misaligned_ldst.h to sbi_trap_ldst.h c0a6320 lib: sbi: rename sbi_misaligned_ldst.c to sbi_trap_ldst.c e11025c lib: sbi: Add initial domain context management support 87d8fe7 lib: tests: Add sbi_console test e5f53fd lib: tests: Add a test for sbi_bitmap 874fcef lib: Add SBIUnit testing macros and functions b9e4de0 docs: Add documentation about tests and SBIUnit 526b9ce firmware: fw_base.S: fix _reset_regs 8151105 firmware: fw_base.S: Remove _relocate_lottery 187397f firmware: fw_dynamic.S: Remove _bad_dynamic_info b27b7c6 firmware: fw_base: Simplified setup trap handler fdf5589 firmware: fw_base.S: Simplify address get 748bef1 lib: sbi_misaligned_ldst: Add handling of C.LHU/C.LH and C.SH bc36678 platform: andes: Drop andes_pmu_setup() 6bb6b61 lib: sbi: Add support for smcsrind and smcdeleg 322b598 lib: sbi_hsm: Restor hart state to stop when fails to start 96a35db docs/firmware: document new options for jump and payload firmwares 2cff7f3 platform: Apply relocatable address f056939 firmware: Add relocatable FW_PAYLOAD_FDT_ADDR 7227cdd firmware: Add relocatable FW_JUMP_ADDR and FW_JUMP_FDT_ADDR 741e941 platform: starfive: call starfive_jh7110_inst_init() in pm_reset_init() 3edf044 platform: starfive: return error if needed devices are not present 80ae046 platform: starfive: rename "stf,axp15060-regulator" -> "x-powers,axp15060" 5335340 platform: starfive: remove redundant compatibility check in pmic_ops 4d8569d platform: starfive: get I2C offset address from clocks property 034af1f platform: starfive: correct system clock device tree node 88273fe lib: sbi_pmu: Before using we should ensure PMU init done 46c8c65 docs: move documentation of system suspend test. 8df836d platform: generic: Parse system suspend test from config node. 23e7e48 docs: Add OpenSBI DT configuration guide. 67ce5a7 platform: generic: Add support for specify coldboot harts in DT 9c8b18e firmware: fw_base.S: remove _runtime_offset 4c6b7cb firmware: fw_base.S: Improve loading u32 92e8aff firmware: always create dynsym section d1dad07 Makefile: check for --exclude-libs 4a76f79 Makefile: don't pass -mstrict-align if not supported 21caaa3 fw_base.S: Fix comment errors 1ec353d lib: sbi: Use mask to check the free bit during trigger allocation bb90a9e lib: sbi: Print number of debug triggers found 76a2a15 lib: sbi: Implement SBI debug trigger extension fa87ec9 include: sbi: Add SBI debug trigger extension related defines 97f234f lib: sbi: Introduce the SBI debug triggers extension support 40dac6b lib: sbi: Detect support of debug triggers 2499769 include: sbi: Introduce debug trigger register encodings 20ca19a include: sbi: Add TINFO debug trigger CSR b752099 include: sbi: Introduce common endianess conversion macro Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20240715171521.179517-1-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- .../opensbi-riscv32-generic-fw_dynamic.bin | Bin 267416 -> 268312 bytes .../opensbi-riscv64-generic-fw_dynamic.bin | Bin 270808 -> 272504 bytes roms/opensbi | 2 +- 3 files changed, 1 insertion(+), 1 deletion(-) diff --git a/pc-bios/opensbi-riscv32-generic-fw_dynamic.bin b/pc-bios/opensbi-riscv32-generic-fw_dynamic.bin index XXXXXXX..XXXXXXX 100644 Binary files a/pc-bios/opensbi-riscv32-generic-fw_dynamic.bin and b/pc-bios/opensbi-riscv32-generic-fw_dynamic.bin differ diff --git a/pc-bios/opensbi-riscv64-generic-fw_dynamic.bin b/pc-bios/opensbi-riscv64-generic-fw_dynamic.bin index XXXXXXX..XXXXXXX 100644 Binary files a/pc-bios/opensbi-riscv64-generic-fw_dynamic.bin and b/pc-bios/opensbi-riscv64-generic-fw_dynamic.bin differ diff --git a/roms/opensbi b/roms/opensbi index XXXXXXX..XXXXXXX 160000 --- a/roms/opensbi +++ b/roms/opensbi @@ -1 +1 @@ -Subproject commit a2b255b88918715173942f2c5e1f97ac9e90c877 +Subproject commit 455de672dd7c2aa1992df54dfb08dc11abbc1b1a -- 2.45.2