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 From: Alistair Francis <alistair.francis@wdc.com> The following changes since commit ad38520bdeb2b1e0b487db317f29119e94c1c88d: Merge remote-tracking branch 'remotes/stefanha-gitlab/tags/block-pull-request' into staging (2022-02-15 19:30:33 +0000) are available in the Git repository at: git@github.com:alistair23/qemu.git tags/pull-riscv-to-apply-20220216 for you to fetch changes up to 7035b8420fa52e8c94cf4317c0f88c1b73ced28d: docs/system: riscv: Update description of CPU (2022-02-16 12:25:52 +1000) ---------------------------------------------------------------- Fourth RISC-V PR for QEMU 7.0 * Remove old Ibex PLIC header file * Allow writing 8 bytes with generic loader * Fixes for RV128 * Refactor RISC-V CPU configs * Initial support for XVentanaCondOps custom extension * Fix for vill field in vtype * Fix trap cause for RV32 HS-mode CSR access from RV64 HS-mode * Support for svnapot, svinval and svpbmt extensions ---------------------------------------------------------------- Anup Patel (18): target/riscv: Fix trap cause for RV32 HS-mode CSR access from RV64 HS-mode target/riscv: Implement SGEIP bit in hip and hie CSRs target/riscv: Implement hgeie and hgeip CSRs target/riscv: Improve delivery of guest external interrupts target/riscv: Allow setting CPU feature from machine/device emulation target/riscv: Add AIA cpu feature target/riscv: Add defines for AIA CSRs target/riscv: Allow AIA device emulation to set ireg rmw callback target/riscv: Implement AIA local interrupt priorities target/riscv: Implement AIA CSRs for 64 local interrupts on RV32 target/riscv: Implement AIA hvictl and hviprioX CSRs target/riscv: Implement AIA interrupt filtering CSRs target/riscv: Implement AIA mtopi, stopi, and vstopi CSRs target/riscv: Implement AIA xiselect and xireg CSRs target/riscv: Implement AIA IMSIC interface CSRs hw/riscv: virt: Use AIA INTC compatible string when available target/riscv: Allow users to force enable AIA CSRs in HART hw/intc: Add RISC-V AIA APLIC device emulation Frédéric Pétrot (1): target/riscv: correct "code should not be reached" for x-rv128 Guo Ren (1): target/riscv: Ignore reserved bits in PTE for RV64 LIU Zhiwei (1): target/riscv: Fix vill field write in vtype Petr Tesarik (1): Allow setting up to 8 bytes with the generic loader Philipp Tomsich (7): target/riscv: refactor (anonymous struct) RISCVCPU.cfg into 'struct RISCVCPUConfig' target/riscv: riscv_tr_init_disas_context: copy pointer-to-cfg into cfg_ptr target/riscv: access configuration through cfg_ptr in DisasContext target/riscv: access cfg structure through DisasContext target/riscv: iterate over a table of decoders target/riscv: Add XVentanaCondOps custom extension target/riscv: add a MAINTAINERS entry for XVentanaCondOps Weiwei Li (4): target/riscv: add PTE_A/PTE_D/PTE_U bits check for inner PTE target/riscv: add support for svnapot extension target/riscv: add support for svinval extension target/riscv: add support for svpbmt extension Wilfred Mallawa (1): include: hw: remove ibex_plic.h Yu Li (1): docs/system: riscv: Update description of CPU docs/system/riscv/virt.rst | 6 +- include/hw/intc/ibex_plic.h | 67 - include/hw/intc/riscv_aplic.h | 79 ++ target/riscv/cpu.h | 169 ++- target/riscv/cpu_bits.h | 129 ++ target/riscv/XVentanaCondOps.decode | 25 + target/riscv/insn32.decode | 7 + hw/core/generic-loader.c | 2 +- hw/intc/riscv_aplic.c | 978 +++++++++++++++ hw/riscv/virt.c | 13 +- target/riscv/cpu.c | 113 +- target/riscv/cpu_helper.c | 377 +++++- target/riscv/csr.c | 1282 ++++++++++++++++++-- target/riscv/gdbstub.c | 3 + target/riscv/machine.c | 24 +- target/riscv/translate.c | 61 +- target/riscv/vector_helper.c | 1 + target/riscv/insn_trans/trans_rvb.c.inc | 8 +- target/riscv/insn_trans/trans_rvi.c.inc | 2 +- target/riscv/insn_trans/trans_rvv.c.inc | 146 ++- target/riscv/insn_trans/trans_rvzfh.c.inc | 4 +- target/riscv/insn_trans/trans_svinval.c.inc | 75 ++ .../riscv/insn_trans/trans_xventanacondops.c.inc | 39 + MAINTAINERS | 7 + hw/intc/Kconfig | 3 + hw/intc/meson.build | 1 + target/riscv/meson.build | 1 + 27 files changed, 3252 insertions(+), 370 deletions(-) delete mode 100644 include/hw/intc/ibex_plic.h create mode 100644 include/hw/intc/riscv_aplic.h create mode 100644 target/riscv/XVentanaCondOps.decode create mode 100644 hw/intc/riscv_aplic.c create mode 100644 target/riscv/insn_trans/trans_svinval.c.inc create mode 100644 target/riscv/insn_trans/trans_xventanacondops.c.inc From: Wilfred Mallawa <wilfred.mallawa@wdc.com> This patch removes the left-over/unused `ibex_plic.h` file. Previously used by opentitan, which now follows the RISC-V standard and uses the SiFivePlicState. Fixes: 434e7e021 ("hw/intc: Remove the Ibex PLIC") Signed-off-by: Wilfred Mallawa <wilfred.mallawa@wdc.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220121055005.3159846-1-alistair.francis@opensource.wdc.com Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- include/hw/intc/ibex_plic.h | 67 ------------------------------------- 1 file changed, 67 deletions(-) delete mode 100644 include/hw/intc/ibex_plic.h diff --git a/include/hw/intc/ibex_plic.h b/include/hw/intc/ibex_plic.h deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/include/hw/intc/ibex_plic.h +++ /dev/null @@ -XXX,XX +XXX,XX @@ -/* - * QEMU RISC-V lowRISC Ibex PLIC - * - * Copyright (c) 2020 Western Digital - * - * 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/>. - */ - -#ifndef HW_IBEX_PLIC_H -#define HW_IBEX_PLIC_H - -#include "hw/sysbus.h" -#include "qom/object.h" - -#define TYPE_IBEX_PLIC "ibex-plic" -OBJECT_DECLARE_SIMPLE_TYPE(IbexPlicState, IBEX_PLIC) - -struct IbexPlicState { - /*< private >*/ - SysBusDevice parent_obj; - - /*< public >*/ - MemoryRegion mmio; - - uint32_t *pending; - uint32_t *hidden_pending; - uint32_t *claimed; - uint32_t *source; - uint32_t *priority; - uint32_t *enable; - uint32_t threshold; - uint32_t claim; - - /* config */ - uint32_t num_cpus; - uint32_t num_sources; - - uint32_t pending_base; - uint32_t pending_num; - - uint32_t source_base; - uint32_t source_num; - - uint32_t priority_base; - uint32_t priority_num; - - uint32_t enable_base; - uint32_t enable_num; - - uint32_t threshold_base; - - uint32_t claim_base; - - qemu_irq *external_irqs; -}; - -#endif /* HW_IBEX_PLIC_H */ -- 2.34.1 From: Petr Tesarik <ptesarik@suse.com> The documentation for the generic loader says that "the maximum size of the data is 8 bytes". However, attempts to set data-len=8 trigger the following assertion failure: ../hw/core/generic-loader.c:59: generic_loader_reset: Assertion `s->data_len < sizeof(s->data)' failed. The type of s->data is uint64_t (i.e. 8 bytes long), so I believe this assert should use <= instead of <. Fixes: e481a1f63c93 ("generic-loader: Add a generic loader") Signed-off-by: Petr Tesarik <ptesarik@suse.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-id: 20220120092715.7805-1-ptesarik@suse.com Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/core/generic-loader.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/core/generic-loader.c b/hw/core/generic-loader.c index XXXXXXX..XXXXXXX 100644 --- a/hw/core/generic-loader.c +++ b/hw/core/generic-loader.c @@ -XXX,XX +XXX,XX @@ static void generic_loader_reset(void *opaque) } if (s->data_len) { - assert(s->data_len < sizeof(s->data)); + assert(s->data_len <= sizeof(s->data)); dma_memory_write(s->cpu->as, s->addr, &s->data, s->data_len, MEMTXATTRS_UNSPECIFIED); } -- 2.34.1 From: Frédéric Pétrot <frederic.petrot@univ-grenoble-alpes.fr> The addition of uxl support in gdbstub adds a few checks on the maximum register length, but omitted MXL_RV128, an experimental feature. This patch makes rv128 react as rv64, as previously. Signed-off-by: Frédéric Pétrot <frederic.petrot@univ-grenoble-alpes.fr> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-id: 20220124202456.420258-1-frederic.petrot@univ-grenoble-alpes.fr Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 3 +-- target/riscv/gdbstub.c | 3 +++ 2 files changed, 4 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 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp) switch (env->misa_mxl_max) { #ifdef TARGET_RISCV64 case MXL_RV64: - cc->gdb_core_xml_file = "riscv-64bit-cpu.xml"; - break; case MXL_RV128: + cc->gdb_core_xml_file = "riscv-64bit-cpu.xml"; break; #endif case MXL_RV32: diff --git a/target/riscv/gdbstub.c b/target/riscv/gdbstub.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/gdbstub.c +++ b/target/riscv/gdbstub.c @@ -XXX,XX +XXX,XX @@ int riscv_cpu_gdb_read_register(CPUState *cs, GByteArray *mem_buf, int n) case MXL_RV32: return gdb_get_reg32(mem_buf, tmp); case MXL_RV64: + case MXL_RV128: return gdb_get_reg64(mem_buf, tmp); default: g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ int riscv_cpu_gdb_write_register(CPUState *cs, uint8_t *mem_buf, int n) length = 4; break; case MXL_RV64: + case MXL_RV128: if (env->xl < MXL_RV64) { tmp = (int32_t)ldq_p(mem_buf); } else { @@ -XXX,XX +XXX,XX @@ void riscv_cpu_register_gdb_regs_for_features(CPUState *cs) 1, "riscv-32bit-virtual.xml", 0); break; case MXL_RV64: + case MXL_RV128: gdb_register_coprocessor(cs, riscv_gdb_get_virtual, riscv_gdb_set_virtual, 1, "riscv-64bit-virtual.xml", 0); -- 2.34.1 From: Philipp Tomsich <philipp.tomsich@vrull.eu> Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Suggested-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20220202005249.3566542-2-philipp.tomsich@vrull.eu> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 78 ++++++++++++++++++++++++---------------------- 1 file changed, 41 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 @@ struct RISCVCPUClass { DeviceReset parent_reset; }; +struct RISCVCPUConfig { + bool ext_i; + bool ext_e; + bool ext_g; + bool ext_m; + bool ext_a; + bool ext_f; + bool ext_d; + bool ext_c; + bool ext_s; + bool ext_u; + bool ext_h; + bool ext_j; + bool ext_v; + bool ext_zba; + bool ext_zbb; + bool ext_zbc; + bool ext_zbs; + bool ext_counters; + bool ext_ifencei; + bool ext_icsr; + bool ext_zfh; + bool ext_zfhmin; + bool ext_zve32f; + bool ext_zve64f; + + char *priv_spec; + char *user_spec; + char *bext_spec; + char *vext_spec; + uint16_t vlen; + uint16_t elen; + bool mmu; + bool pmp; + bool epmp; + uint64_t resetvec; +}; + +typedef struct RISCVCPUConfig RISCVCPUConfig; + /** * RISCVCPU: * @env: #CPURISCVState @@ -XXX,XX +XXX,XX @@ struct RISCVCPU { char *dyn_vreg_xml; /* Configuration Settings */ - struct { - bool ext_i; - bool ext_e; - bool ext_g; - bool ext_m; - bool ext_a; - bool ext_f; - bool ext_d; - bool ext_c; - bool ext_s; - bool ext_u; - bool ext_h; - bool ext_j; - bool ext_v; - bool ext_zba; - bool ext_zbb; - bool ext_zbc; - bool ext_zbs; - bool ext_counters; - bool ext_ifencei; - bool ext_icsr; - bool ext_zfh; - bool ext_zfhmin; - bool ext_zve32f; - bool ext_zve64f; - - char *priv_spec; - char *user_spec; - char *bext_spec; - char *vext_spec; - uint16_t vlen; - uint16_t elen; - bool mmu; - bool pmp; - bool epmp; - uint64_t resetvec; - } cfg; + RISCVCPUConfig cfg; }; static inline int riscv_has_ext(CPURISCVState *env, target_ulong ext) -- 2.34.1 From: Philipp Tomsich <philipp.tomsich@vrull.eu> As the number of extensions is growing, copying them individiually into the DisasContext will scale less and less... instead we populate a pointer to the RISCVCPUConfig structure in the DisasContext. This adds an extra indirection when checking for the availability of an extension (compared to copying the fields into DisasContext). While not a performance problem today, we can always (shallow) copy the entire structure into the DisasContext (instead of putting a pointer to it) if this is ever deemed necessary. Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Suggested-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20220202005249.3566542-3-philipp.tomsich@vrull.eu> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/translate.c | 2 ++ 1 file changed, 2 insertions(+) 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 @@ typedef struct DisasContext { int frm; RISCVMXL ol; bool virt_enabled; + const RISCVCPUConfig *cfg_ptr; bool ext_ifencei; bool ext_zfh; bool ext_zfhmin; @@ -XXX,XX +XXX,XX @@ static void riscv_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs) #endif ctx->misa_ext = env->misa_ext; ctx->frm = -1; /* unknown rounding mode */ + ctx->cfg_ptr = &(cpu->cfg); ctx->ext_ifencei = cpu->cfg.ext_ifencei; ctx->ext_zfh = cpu->cfg.ext_zfh; ctx->ext_zfhmin = cpu->cfg.ext_zfhmin; -- 2.34.1 From: Philipp Tomsich <philipp.tomsich@vrull.eu> The implementation in trans_{rvi,rvv,rvzfh}.c.inc accesses the shallow copies (in DisasContext) of some of the elements available in the RISCVCPUConfig structure. This commit redirects accesses to use the cfg_ptr copied into DisasContext and removes the shallow copies. Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Suggested-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20220202005249.3566542-4-philipp.tomsich@vrull.eu> [ Changes by AF: - Fixup checkpatch failures ] Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/translate.c | 14 --- target/riscv/insn_trans/trans_rvi.c.inc | 2 +- target/riscv/insn_trans/trans_rvv.c.inc | 146 ++++++++++++++-------- target/riscv/insn_trans/trans_rvzfh.c.inc | 4 +- 4 files changed, 97 insertions(+), 69 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 @@ typedef struct DisasContext { RISCVMXL ol; bool virt_enabled; const RISCVCPUConfig *cfg_ptr; - bool ext_ifencei; - bool ext_zfh; - bool ext_zfhmin; - bool ext_zve32f; - bool ext_zve64f; bool hlsx; /* vector extension */ bool vill; @@ -XXX,XX +XXX,XX @@ typedef struct DisasContext { */ int8_t lmul; uint8_t sew; - uint16_t vlen; - uint16_t elen; target_ulong vstart; bool vl_eq_vlmax; uint8_t ntemp; @@ -XXX,XX +XXX,XX @@ static void riscv_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs) ctx->misa_ext = env->misa_ext; ctx->frm = -1; /* unknown rounding mode */ ctx->cfg_ptr = &(cpu->cfg); - ctx->ext_ifencei = cpu->cfg.ext_ifencei; - ctx->ext_zfh = cpu->cfg.ext_zfh; - ctx->ext_zfhmin = cpu->cfg.ext_zfhmin; - ctx->ext_zve32f = cpu->cfg.ext_zve32f; - ctx->ext_zve64f = cpu->cfg.ext_zve64f; - ctx->vlen = cpu->cfg.vlen; - ctx->elen = cpu->cfg.elen; ctx->mstatus_hs_fs = FIELD_EX32(tb_flags, TB_FLAGS, MSTATUS_HS_FS); ctx->mstatus_hs_vs = FIELD_EX32(tb_flags, TB_FLAGS, MSTATUS_HS_VS); ctx->hlsx = FIELD_EX32(tb_flags, TB_FLAGS, HLSX); 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 trans_fence(DisasContext *ctx, arg_fence *a) static bool trans_fence_i(DisasContext *ctx, arg_fence_i *a) { - if (!ctx->ext_ifencei) { + if (!ctx->cfg_ptr->ext_ifencei) { return false; } 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 require_zve32f(DisasContext *s) } /* Zve32f doesn't support FP64. (Section 18.2) */ - return s->ext_zve32f ? s->sew <= MO_32 : true; + return s->cfg_ptr->ext_zve32f ? s->sew <= MO_32 : true; } static bool require_scale_zve32f(DisasContext *s) @@ -XXX,XX +XXX,XX @@ static bool require_scale_zve32f(DisasContext *s) } /* Zve32f doesn't support FP64. (Section 18.2) */ - return s->ext_zve64f ? s->sew <= MO_16 : true; + return s->cfg_ptr->ext_zve64f ? s->sew <= MO_16 : true; } static bool require_zve64f(DisasContext *s) @@ -XXX,XX +XXX,XX @@ static bool require_zve64f(DisasContext *s) } /* Zve64f doesn't support FP64. (Section 18.2) */ - return s->ext_zve64f ? s->sew <= MO_32 : true; + return s->cfg_ptr->ext_zve64f ? s->sew <= MO_32 : true; } static bool require_scale_zve64f(DisasContext *s) @@ -XXX,XX +XXX,XX @@ static bool require_scale_zve64f(DisasContext *s) } /* Zve64f doesn't support FP64. (Section 18.2) */ - return s->ext_zve64f ? s->sew <= MO_16 : true; + return s->cfg_ptr->ext_zve64f ? s->sew <= MO_16 : true; } /* Destination vector register group cannot overlap source mask register. */ @@ -XXX,XX +XXX,XX @@ static bool do_vsetvl(DisasContext *s, int rd, int rs1, TCGv s2) TCGv s1, dst; if (!require_rvv(s) || - !(has_ext(s, RVV) || s->ext_zve32f || s->ext_zve64f)) { + !(has_ext(s, RVV) || s->cfg_ptr->ext_zve32f || + s->cfg_ptr->ext_zve64f)) { return false; } @@ -XXX,XX +XXX,XX @@ static bool do_vsetivli(DisasContext *s, int rd, TCGv s1, TCGv s2) TCGv dst; if (!require_rvv(s) || - !(has_ext(s, RVV) || s->ext_zve32f || s->ext_zve64f)) { + !(has_ext(s, RVV) || s->cfg_ptr->ext_zve32f || + s->cfg_ptr->ext_zve64f)) { return false; } @@ -XXX,XX +XXX,XX @@ static bool trans_vsetivli(DisasContext *s, arg_vsetivli *a) /* vector register offset from env */ static uint32_t vreg_ofs(DisasContext *s, int reg) { - return offsetof(CPURISCVState, vreg) + reg * s->vlen / 8; + return offsetof(CPURISCVState, vreg) + reg * s->cfg_ptr->vlen / 8; } /* check functions */ @@ -XXX,XX +XXX,XX @@ static bool vext_check_st_index(DisasContext *s, int vd, int vs2, int nf, * when XLEN=32. (Section 18.2) */ if (get_xl(s) == MXL_RV32) { - ret &= (!has_ext(s, RVV) && s->ext_zve64f ? eew != MO_64 : true); + ret &= (!has_ext(s, RVV) && + s->cfg_ptr->ext_zve64f ? eew != MO_64 : true); } return ret; @@ -XXX,XX +XXX,XX @@ static bool vext_wide_check_common(DisasContext *s, int vd, int vm) { return (s->lmul <= 2) && (s->sew < MO_64) && - ((s->sew + 1) <= (s->elen >> 4)) && + ((s->sew + 1) <= (s->cfg_ptr->elen >> 4)) && require_align(vd, s->lmul + 1) && require_vm(vm, vd); } @@ -XXX,XX +XXX,XX @@ static bool vext_narrow_check_common(DisasContext *s, int vd, int vs2, { return (s->lmul <= 2) && (s->sew < MO_64) && - ((s->sew + 1) <= (s->elen >> 4)) && + ((s->sew + 1) <= (s->cfg_ptr->elen >> 4)) && require_align(vs2, s->lmul + 1) && require_align(vd, s->lmul) && require_vm(vm, vd); @@ -XXX,XX +XXX,XX @@ static bool ldst_us_trans(uint32_t vd, uint32_t rs1, uint32_t data, * The first part is vlen in bytes, encoded in maxsz of simd_desc. * The second part is lmul, encoded in data of simd_desc. */ - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd)); tcg_gen_addi_ptr(mask, cpu_env, vreg_ofs(s, 0)); @@ -XXX,XX +XXX,XX @@ static bool ldst_stride_trans(uint32_t vd, uint32_t rs1, uint32_t rs2, mask = tcg_temp_new_ptr(); base = get_gpr(s, rs1, EXT_NONE); stride = get_gpr(s, rs2, EXT_NONE); - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd)); tcg_gen_addi_ptr(mask, cpu_env, vreg_ofs(s, 0)); @@ -XXX,XX +XXX,XX @@ static bool ldst_index_trans(uint32_t vd, uint32_t rs1, uint32_t vs2, mask = tcg_temp_new_ptr(); index = tcg_temp_new_ptr(); base = get_gpr(s, rs1, EXT_NONE); - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd)); tcg_gen_addi_ptr(index, cpu_env, vreg_ofs(s, vs2)); @@ -XXX,XX +XXX,XX @@ static bool ldff_trans(uint32_t vd, uint32_t rs1, uint32_t data, dest = tcg_temp_new_ptr(); mask = tcg_temp_new_ptr(); base = get_gpr(s, rs1, EXT_NONE); - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd)); tcg_gen_addi_ptr(mask, cpu_env, vreg_ofs(s, 0)); @@ -XXX,XX +XXX,XX @@ static bool ldst_whole_trans(uint32_t vd, uint32_t rs1, uint32_t nf, uint32_t data = FIELD_DP32(0, VDATA, NF, nf); dest = tcg_temp_new_ptr(); - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); base = get_gpr(s, rs1, EXT_NONE); tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd)); @@ -XXX,XX +XXX,XX @@ GEN_LDST_WHOLE_TRANS(vs8r_v, 8, true) static inline uint32_t MAXSZ(DisasContext *s) { int scale = s->lmul - 3; - return scale < 0 ? s->vlen >> -scale : s->vlen << scale; + return scale < 0 ? s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale; } static bool opivv_check(DisasContext *s, arg_rmrr *a) @@ -XXX,XX +XXX,XX @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn, data = FIELD_DP32(data, VDATA, LMUL, s->lmul); tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2), - cpu_env, s->vlen / 8, s->vlen / 8, data, fn); + cpu_env, s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data, fn); } mark_vs_dirty(s); gen_set_label(over); @@ -XXX,XX +XXX,XX @@ static bool opivx_trans(uint32_t vd, uint32_t rs1, uint32_t vs2, uint32_t vm, data = FIELD_DP32(data, VDATA, VM, vm); data = FIELD_DP32(data, VDATA, LMUL, s->lmul); - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd)); tcg_gen_addi_ptr(src2, cpu_env, vreg_ofs(s, vs2)); @@ -XXX,XX +XXX,XX @@ static bool opivi_trans(uint32_t vd, uint32_t imm, uint32_t vs2, uint32_t vm, data = FIELD_DP32(data, VDATA, VM, vm); data = FIELD_DP32(data, VDATA, LMUL, s->lmul); - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd)); tcg_gen_addi_ptr(src2, cpu_env, vreg_ofs(s, vs2)); @@ -XXX,XX +XXX,XX @@ static bool do_opivv_widen(DisasContext *s, arg_rmrr *a, tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2), - cpu_env, s->vlen / 8, s->vlen / 8, + cpu_env, s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data, fn); mark_vs_dirty(s); gen_set_label(over); @@ -XXX,XX +XXX,XX @@ static bool do_opiwv_widen(DisasContext *s, arg_rmrr *a, tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2), - cpu_env, s->vlen / 8, s->vlen / 8, data, fn); + cpu_env, s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data, fn); mark_vs_dirty(s); gen_set_label(over); return true; @@ -XXX,XX +XXX,XX @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a) \ tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), \ vreg_ofs(s, a->rs1), \ vreg_ofs(s, a->rs2), cpu_env, \ - s->vlen / 8, s->vlen / 8, data, \ + s->cfg_ptr->vlen / 8, \ + s->cfg_ptr->vlen / 8, data, \ fns[s->sew]); \ mark_vs_dirty(s); \ gen_set_label(over); \ @@ -XXX,XX +XXX,XX @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a) \ tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), \ vreg_ofs(s, a->rs1), \ vreg_ofs(s, a->rs2), cpu_env, \ - s->vlen / 8, s->vlen / 8, data, \ + s->cfg_ptr->vlen / 8, \ + s->cfg_ptr->vlen / 8, data, \ fns[s->sew]); \ mark_vs_dirty(s); \ gen_set_label(over); \ @@ -XXX,XX +XXX,XX @@ static bool vmulh_vv_check(DisasContext *s, arg_rmrr *a) * are not included for EEW=64 in Zve64*. (Section 18.2) */ return opivv_check(s, a) && - (!has_ext(s, RVV) && s->ext_zve64f ? s->sew != MO_64 : true); + (!has_ext(s, RVV) && + s->cfg_ptr->ext_zve64f ? s->sew != MO_64 : true); } static bool vmulh_vx_check(DisasContext *s, arg_rmrr *a) @@ -XXX,XX +XXX,XX @@ static bool vmulh_vx_check(DisasContext *s, arg_rmrr *a) * are not included for EEW=64 in Zve64*. (Section 18.2) */ return opivx_check(s, a) && - (!has_ext(s, RVV) && s->ext_zve64f ? s->sew != MO_64 : true); + (!has_ext(s, RVV) && + s->cfg_ptr->ext_zve64f ? s->sew != MO_64 : true); } GEN_OPIVV_GVEC_TRANS(vmul_vv, mul) @@ -XXX,XX +XXX,XX @@ static bool trans_vmv_v_v(DisasContext *s, arg_vmv_v_v *a) tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over); tcg_gen_gvec_2_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, a->rs1), - cpu_env, s->vlen / 8, s->vlen / 8, data, + cpu_env, s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data, fns[s->sew]); gen_set_label(over); } @@ -XXX,XX +XXX,XX @@ static bool trans_vmv_v_x(DisasContext *s, arg_vmv_v_x *a) }; tcg_gen_ext_tl_i64(s1_i64, s1); - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, a->rd)); fns[s->sew](dest, s1_i64, cpu_env, desc); @@ -XXX,XX +XXX,XX @@ static bool trans_vmv_v_i(DisasContext *s, arg_vmv_v_i *a) s1 = tcg_constant_i64(simm); dest = tcg_temp_new_ptr(); - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, a->rd)); fns[s->sew](dest, s1, cpu_env, desc); @@ -XXX,XX +XXX,XX @@ static bool vsmul_vv_check(DisasContext *s, arg_rmrr *a) * for EEW=64 in Zve64*. (Section 18.2) */ return opivv_check(s, a) && - (!has_ext(s, RVV) && s->ext_zve64f ? s->sew != MO_64 : true); + (!has_ext(s, RVV) && + s->cfg_ptr->ext_zve64f ? s->sew != MO_64 : true); } static bool vsmul_vx_check(DisasContext *s, arg_rmrr *a) @@ -XXX,XX +XXX,XX @@ static bool vsmul_vx_check(DisasContext *s, arg_rmrr *a) * for EEW=64 in Zve64*. (Section 18.2) */ return opivx_check(s, a) && - (!has_ext(s, RVV) && s->ext_zve64f ? s->sew != MO_64 : true); + (!has_ext(s, RVV) && + s->cfg_ptr->ext_zve64f ? s->sew != MO_64 : true); } GEN_OPIVV_TRANS(vsmul_vv, vsmul_vv_check) @@ -XXX,XX +XXX,XX @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a) \ tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), \ vreg_ofs(s, a->rs1), \ vreg_ofs(s, a->rs2), cpu_env, \ - s->vlen / 8, s->vlen / 8, data, \ + s->cfg_ptr->vlen / 8, \ + s->cfg_ptr->vlen / 8, data, \ fns[s->sew - 1]); \ mark_vs_dirty(s); \ gen_set_label(over); \ @@ -XXX,XX +XXX,XX @@ static bool opfvf_trans(uint32_t vd, uint32_t rs1, uint32_t vs2, dest = tcg_temp_new_ptr(); mask = tcg_temp_new_ptr(); src2 = tcg_temp_new_ptr(); - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd)); tcg_gen_addi_ptr(src2, cpu_env, vreg_ofs(s, vs2)); @@ -XXX,XX +XXX,XX @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a) \ tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), \ vreg_ofs(s, a->rs1), \ vreg_ofs(s, a->rs2), cpu_env, \ - s->vlen / 8, s->vlen / 8, data, \ + s->cfg_ptr->vlen / 8, \ + s->cfg_ptr->vlen / 8, data, \ fns[s->sew - 1]); \ mark_vs_dirty(s); \ gen_set_label(over); \ @@ -XXX,XX +XXX,XX @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a) \ tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), \ vreg_ofs(s, a->rs1), \ vreg_ofs(s, a->rs2), cpu_env, \ - s->vlen / 8, s->vlen / 8, data, \ + s->cfg_ptr->vlen / 8, \ + s->cfg_ptr->vlen / 8, data, \ fns[s->sew - 1]); \ mark_vs_dirty(s); \ gen_set_label(over); \ @@ -XXX,XX +XXX,XX @@ static bool do_opfv(DisasContext *s, arg_rmr *a, data = FIELD_DP32(data, VDATA, LMUL, s->lmul); tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), vreg_ofs(s, a->rs2), cpu_env, - s->vlen / 8, s->vlen / 8, data, fn); + s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data, fn); mark_vs_dirty(s); gen_set_label(over); return true; @@ -XXX,XX +XXX,XX @@ static bool trans_vfmv_v_f(DisasContext *s, arg_vfmv_v_f *a) do_nanbox(s, t1, cpu_fpr[a->rs1]); dest = tcg_temp_new_ptr(); - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, a->rd)); fns[s->sew - 1](dest, t1, cpu_env, desc); @@ -XXX,XX +XXX,XX @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a) \ data = FIELD_DP32(data, VDATA, LMUL, s->lmul); \ tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), \ vreg_ofs(s, a->rs2), cpu_env, \ - s->vlen / 8, s->vlen / 8, data, \ + s->cfg_ptr->vlen / 8, \ + s->cfg_ptr->vlen / 8, data, \ fns[s->sew - 1]); \ mark_vs_dirty(s); \ gen_set_label(over); \ @@ -XXX,XX +XXX,XX @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a) \ data = FIELD_DP32(data, VDATA, VM, a->vm); \ tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), \ vreg_ofs(s, a->rs2), cpu_env, \ - s->vlen / 8, s->vlen / 8, data, \ + s->cfg_ptr->vlen / 8, \ + s->cfg_ptr->vlen / 8, data, \ fns[s->sew]); \ mark_vs_dirty(s); \ gen_set_label(over); \ @@ -XXX,XX +XXX,XX @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a) \ data = FIELD_DP32(data, VDATA, LMUL, s->lmul); \ tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), \ vreg_ofs(s, a->rs2), cpu_env, \ - s->vlen / 8, s->vlen / 8, data, \ + s->cfg_ptr->vlen / 8, \ + s->cfg_ptr->vlen / 8, data, \ fns[s->sew - 1]); \ mark_vs_dirty(s); \ gen_set_label(over); \ @@ -XXX,XX +XXX,XX @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a) \ data = FIELD_DP32(data, VDATA, VM, a->vm); \ tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), \ vreg_ofs(s, a->rs2), cpu_env, \ - s->vlen / 8, s->vlen / 8, data, \ + s->cfg_ptr->vlen / 8, \ + s->cfg_ptr->vlen / 8, data, \ fns[s->sew]); \ mark_vs_dirty(s); \ gen_set_label(over); \ @@ -XXX,XX +XXX,XX @@ GEN_OPIVV_TRANS(vredxor_vs, reduction_check) static bool reduction_widen_check(DisasContext *s, arg_rmrr *a) { return reduction_check(s, a) && (s->sew < MO_64) && - ((s->sew + 1) <= (s->elen >> 4)); + ((s->sew + 1) <= (s->cfg_ptr->elen >> 4)); } GEN_OPIVV_WIDEN_TRANS(vwredsum_vs, reduction_widen_check) @@ -XXX,XX +XXX,XX @@ static bool trans_##NAME(DisasContext *s, arg_r *a) \ tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), \ vreg_ofs(s, a->rs1), \ vreg_ofs(s, a->rs2), cpu_env, \ - s->vlen / 8, s->vlen / 8, data, fn); \ + s->cfg_ptr->vlen / 8, \ + s->cfg_ptr->vlen / 8, data, fn); \ mark_vs_dirty(s); \ gen_set_label(over); \ return true; \ @@ -XXX,XX +XXX,XX @@ static bool trans_vcpop_m(DisasContext *s, arg_rmr *a) mask = tcg_temp_new_ptr(); src2 = tcg_temp_new_ptr(); dst = dest_gpr(s, a->rd); - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); tcg_gen_addi_ptr(src2, cpu_env, vreg_ofs(s, a->rs2)); tcg_gen_addi_ptr(mask, cpu_env, vreg_ofs(s, 0)); @@ -XXX,XX +XXX,XX @@ static bool trans_vfirst_m(DisasContext *s, arg_rmr *a) mask = tcg_temp_new_ptr(); src2 = tcg_temp_new_ptr(); dst = dest_gpr(s, a->rd); - desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data)); + desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data)); tcg_gen_addi_ptr(src2, cpu_env, vreg_ofs(s, a->rs2)); tcg_gen_addi_ptr(mask, cpu_env, vreg_ofs(s, 0)); @@ -XXX,XX +XXX,XX @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a) \ data = FIELD_DP32(data, VDATA, LMUL, s->lmul); \ tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), \ vreg_ofs(s, 0), vreg_ofs(s, a->rs2), \ - cpu_env, s->vlen / 8, s->vlen / 8, \ + cpu_env, s->cfg_ptr->vlen / 8, \ + s->cfg_ptr->vlen / 8, \ data, fn); \ mark_vs_dirty(s); \ gen_set_label(over); \ @@ -XXX,XX +XXX,XX @@ static bool trans_viota_m(DisasContext *s, arg_viota_m *a) }; tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), vreg_ofs(s, a->rs2), cpu_env, - s->vlen / 8, s->vlen / 8, data, fns[s->sew]); + s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data, fns[s->sew]); mark_vs_dirty(s); gen_set_label(over); return true; @@ -XXX,XX +XXX,XX @@ static bool trans_vid_v(DisasContext *s, arg_vid_v *a) gen_helper_vid_v_w, gen_helper_vid_v_d, }; tcg_gen_gvec_2_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), - cpu_env, s->vlen / 8, s->vlen / 8, + cpu_env, s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data, fns[s->sew]); mark_vs_dirty(s); gen_set_label(over); @@ -XXX,XX +XXX,XX @@ static bool trans_vrgather_vx(DisasContext *s, arg_rmrr *a) if (a->vm && s->vl_eq_vlmax) { int scale = s->lmul - (s->sew + 3); - int vlmax = scale < 0 ? s->vlen >> -scale : s->vlen << scale; + int vlmax = scale < 0 ? + s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale; TCGv_i64 dest = tcg_temp_new_i64(); if (a->rs1 == 0) { @@ -XXX,XX +XXX,XX @@ static bool trans_vrgather_vi(DisasContext *s, arg_rmrr *a) if (a->vm && s->vl_eq_vlmax) { int scale = s->lmul - (s->sew + 3); - int vlmax = scale < 0 ? s->vlen >> -scale : s->vlen << scale; + int vlmax = scale < 0 ? + s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale; if (a->rs1 >= vlmax) { tcg_gen_gvec_dup_imm(MO_64, vreg_ofs(s, a->rd), MAXSZ(s), MAXSZ(s), 0); @@ -XXX,XX +XXX,XX @@ static bool trans_vcompress_vm(DisasContext *s, arg_r *a) data = FIELD_DP32(data, VDATA, LMUL, s->lmul); tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2), - cpu_env, s->vlen / 8, s->vlen / 8, data, + cpu_env, s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data, fns[s->sew]); mark_vs_dirty(s); gen_set_label(over); @@ -XXX,XX +XXX,XX @@ static bool trans_##NAME(DisasContext *s, arg_##NAME * a) \ if (require_rvv(s) && \ QEMU_IS_ALIGNED(a->rd, LEN) && \ QEMU_IS_ALIGNED(a->rs2, LEN)) { \ - uint32_t maxsz = (s->vlen >> 3) * LEN; \ + uint32_t maxsz = (s->cfg_ptr->vlen >> 3) * LEN; \ if (s->vstart == 0) { \ /* EEW = 8 */ \ tcg_gen_gvec_mov(MO_8, vreg_ofs(s, a->rd), \ @@ -XXX,XX +XXX,XX @@ static bool int_ext_op(DisasContext *s, arg_rmr *a, uint8_t seq) tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), vreg_ofs(s, a->rs2), cpu_env, - s->vlen / 8, s->vlen / 8, data, fn); + s->cfg_ptr->vlen / 8, + s->cfg_ptr->vlen / 8, data, fn); mark_vs_dirty(s); gen_set_label(over); diff --git a/target/riscv/insn_trans/trans_rvzfh.c.inc b/target/riscv/insn_trans/trans_rvzfh.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn_trans/trans_rvzfh.c.inc +++ b/target/riscv/insn_trans/trans_rvzfh.c.inc @@ -XXX,XX +XXX,XX @@ */ #define REQUIRE_ZFH(ctx) do { \ - if (!ctx->ext_zfh) { \ + if (!ctx->cfg_ptr->ext_zfh) { \ return false; \ } \ } while (0) #define REQUIRE_ZFH_OR_ZFHMIN(ctx) do { \ - if (!(ctx->ext_zfh || ctx->ext_zfhmin)) { \ + if (!(ctx->cfg_ptr->ext_zfh || ctx->cfg_ptr->ext_zfhmin)) { \ return false; \ } \ } while (0) -- 2.34.1 From: Philipp Tomsich <philipp.tomsich@vrull.eu> The Zb[abcs] support code still uses the RISCV_CPU macros to access the configuration information (i.e., check whether an extension is available/enabled). Now that we provide this information directly from DisasContext, we can access this directly via the cfg_ptr field. Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Suggested-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20220202005249.3566542-5-philipp.tomsich@vrull.eu> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/insn_trans/trans_rvb.c.inc | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/target/riscv/insn_trans/trans_rvb.c.inc b/target/riscv/insn_trans/trans_rvb.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/insn_trans/trans_rvb.c.inc +++ b/target/riscv/insn_trans/trans_rvb.c.inc @@ -XXX,XX +XXX,XX @@ */ #define REQUIRE_ZBA(ctx) do { \ - if (!RISCV_CPU(ctx->cs)->cfg.ext_zba) { \ + if (ctx->cfg_ptr->ext_zba) { \ return false; \ } \ } while (0) #define REQUIRE_ZBB(ctx) do { \ - if (!RISCV_CPU(ctx->cs)->cfg.ext_zbb) { \ + if (ctx->cfg_ptr->ext_zbb) { \ return false; \ } \ } while (0) #define REQUIRE_ZBC(ctx) do { \ - if (!RISCV_CPU(ctx->cs)->cfg.ext_zbc) { \ + if (ctx->cfg_ptr->ext_zbc) { \ return false; \ } \ } while (0) #define REQUIRE_ZBS(ctx) do { \ - if (!RISCV_CPU(ctx->cs)->cfg.ext_zbs) { \ + if (ctx->cfg_ptr->ext_zbs) { \ return false; \ } \ } while (0) -- 2.34.1 From: Philipp Tomsich <philipp.tomsich@vrull.eu> To split up the decoder into multiple functions (both to support vendor-specific opcodes in separate files and to simplify maintenance of orthogonal extensions), this changes decode_op to iterate over a table of decoders predicated on guard functions. This commit only adds the new structure and the table, allowing for the easy addition of additional decoders in the future. Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-Id: <20220202005249.3566542-6-philipp.tomsich@vrull.eu> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/translate.c | 32 +++++++++++++++++++++++++++----- 1 file changed, 27 insertions(+), 5 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 inline bool has_ext(DisasContext *ctx, uint32_t ext) return ctx->misa_ext & ext; } +static bool always_true_p(DisasContext *ctx __attribute__((__unused__))) +{ + return true; +} + #ifdef TARGET_RISCV32 #define get_xl(ctx) MXL_RV32 #elif defined(CONFIG_USER_ONLY) @@ -XXX,XX +XXX,XX @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc) static void decode_opc(CPURISCVState *env, DisasContext *ctx, uint16_t opcode) { - /* check for compressed insn */ + /* + * A table with predicate (i.e., guard) functions and decoder functions + * that are tested in-order until a decoder matches onto the opcode. + */ + static const struct { + bool (*guard_func)(DisasContext *); + bool (*decode_func)(DisasContext *, uint32_t); + } decoders[] = { + { always_true_p, decode_insn32 }, + }; + + /* Check for compressed insn */ if (extract16(opcode, 0, 2) != 3) { if (!has_ext(ctx, RVC)) { gen_exception_illegal(ctx); } else { ctx->opcode = opcode; ctx->pc_succ_insn = ctx->base.pc_next + 2; - if (!decode_insn16(ctx, opcode)) { - gen_exception_illegal(ctx); + if (decode_insn16(ctx, opcode)) { + return; } } } else { @@ -XXX,XX +XXX,XX @@ static void decode_opc(CPURISCVState *env, DisasContext *ctx, uint16_t opcode) ctx->base.pc_next + 2)); ctx->opcode = opcode32; ctx->pc_succ_insn = ctx->base.pc_next + 4; - if (!decode_insn32(ctx, opcode32)) { - gen_exception_illegal(ctx); + + for (size_t i = 0; i < ARRAY_SIZE(decoders); ++i) { + if (decoders[i].guard_func(ctx) && + decoders[i].decode_func(ctx, opcode32)) { + return; + } } } + + gen_exception_illegal(ctx); } static void riscv_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs) -- 2.34.1 From: Philipp Tomsich <philipp.tomsich@vrull.eu> This adds the decoder and translation for the XVentanaCondOps custom extension (vendor-defined by Ventana Micro Systems), which is documented at https://github.com/ventanamicro/ventana-custom-extensions/releases/download/v1.0.0/ventana-custom-extensions-v1.0.0.pdf This commit then also adds a guard-function (has_XVentanaCondOps_p) and the decoder function to the table of decoders, enabling the support for the XVentanaCondOps extension. Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-Id: <20220202005249.3566542-7-philipp.tomsich@vrull.eu> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 3 ++ target/riscv/XVentanaCondOps.decode | 25 ++++++++++++ target/riscv/cpu.c | 3 ++ target/riscv/translate.c | 12 ++++++ .../insn_trans/trans_xventanacondops.c.inc | 39 +++++++++++++++++++ target/riscv/meson.build | 1 + 6 files changed, 83 insertions(+) create mode 100644 target/riscv/XVentanaCondOps.decode create mode 100644 target/riscv/insn_trans/trans_xventanacondops.c.inc 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 RISCVCPUConfig { bool ext_zve32f; bool ext_zve64f; + /* Vendor-specific custom extensions */ + bool ext_XVentanaCondOps; + char *priv_spec; char *user_spec; char *bext_spec; diff --git a/target/riscv/XVentanaCondOps.decode b/target/riscv/XVentanaCondOps.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/riscv/XVentanaCondOps.decode @@ -XXX,XX +XXX,XX @@ +# +# RISC-V translation routines for the XVentanaCondOps extension +# +# Copyright (c) 2022 Dr. Philipp Tomsich, philipp.tomsich@vrull.eu +# +# SPDX-License-Identifier: LGPL-2.1-or-later +# +# Reference: VTx-family custom instructions +# Custom ISA extensions for Ventana Micro Systems RISC-V cores +# (https://github.com/ventanamicro/ventana-custom-extensions/releases/download/v1.0.0/ventana-custom-extensions-v1.0.0.pdf) + +# Fields +%rs2 20:5 +%rs1 15:5 +%rd 7:5 + +# Argument sets +&r rd rs1 rs2 !extern + +# Formats +@r ....... ..... ..... ... ..... ....... &r %rs2 %rs1 %rd + +# *** RV64 Custom-3 Extension *** +vt_maskc 0000000 ..... ..... 110 ..... 1111011 @r +vt_maskcn 0000000 ..... ..... 111 ..... 1111011 @r 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 Property riscv_cpu_properties[] = { DEFINE_PROP_BOOL("zbc", RISCVCPU, cfg.ext_zbc, true), DEFINE_PROP_BOOL("zbs", RISCVCPU, cfg.ext_zbs, true), + /* Vendor-specific custom extensions */ + DEFINE_PROP_BOOL("xventanacondops", RISCVCPU, cfg.ext_XVentanaCondOps, false), + /* These are experimental so mark with 'x-' */ DEFINE_PROP_BOOL("x-j", RISCVCPU, cfg.ext_j, false), /* ePMP 0.9.3 */ 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 always_true_p(DisasContext *ctx __attribute__((__unused__))) return true; } +#define MATERIALISE_EXT_PREDICATE(ext) \ + static bool has_ ## ext ## _p(DisasContext *ctx) \ + { \ + return ctx->cfg_ptr->ext_ ## ext ; \ + } + +MATERIALISE_EXT_PREDICATE(XVentanaCondOps); + #ifdef TARGET_RISCV32 #define get_xl(ctx) MXL_RV32 #elif defined(CONFIG_USER_ONLY) @@ -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_rvzfh.c.inc" #include "insn_trans/trans_privileged.c.inc" +#include "insn_trans/trans_xventanacondops.c.inc" /* Include the auto-generated decoder for 16 bit insn */ #include "decode-insn16.c.inc" +/* Include decoders for factored-out extensions */ +#include "decode-XVentanaCondOps.c.inc" static void decode_opc(CPURISCVState *env, DisasContext *ctx, uint16_t opcode) { @@ -XXX,XX +XXX,XX @@ static void decode_opc(CPURISCVState *env, DisasContext *ctx, uint16_t opcode) bool (*decode_func)(DisasContext *, uint32_t); } decoders[] = { { always_true_p, decode_insn32 }, + { has_XVentanaCondOps_p, decode_XVentanaCodeOps }, }; /* Check for compressed insn */ diff --git a/target/riscv/insn_trans/trans_xventanacondops.c.inc b/target/riscv/insn_trans/trans_xventanacondops.c.inc new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/riscv/insn_trans/trans_xventanacondops.c.inc @@ -XXX,XX +XXX,XX @@ +/* + * RISC-V translation routines for the XVentanaCondOps extension. + * + * Copyright (c) 2021-2022 VRULL GmbH. + * + * 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/>. + */ + +static bool gen_vt_condmask(DisasContext *ctx, arg_r *a, TCGCond cond) +{ + TCGv dest = dest_gpr(ctx, a->rd); + TCGv src1 = get_gpr(ctx, a->rs1, EXT_NONE); + TCGv src2 = get_gpr(ctx, a->rs2, EXT_NONE); + + tcg_gen_movcond_tl(cond, dest, src2, ctx->zero, src1, ctx->zero); + + gen_set_gpr(ctx, a->rd, dest); + return true; +} + +static bool trans_vt_maskc(DisasContext *ctx, arg_r *a) +{ + return gen_vt_condmask(ctx, a, TCG_COND_NE); +} + +static bool trans_vt_maskcn(DisasContext *ctx, arg_r *a) +{ + return gen_vt_condmask(ctx, a, TCG_COND_EQ); +} diff --git a/target/riscv/meson.build b/target/riscv/meson.build index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/meson.build +++ b/target/riscv/meson.build @@ -XXX,XX +XXX,XX @@ dir = meson.current_source_dir() gen = [ decodetree.process('insn16.decode', extra_args: ['--static-decode=decode_insn16', '--insnwidth=16']), decodetree.process('insn32.decode', extra_args: '--static-decode=decode_insn32'), + decodetree.process('XVentanaCondOps.decode', extra_args: '--static-decode=decode_XVentanaCodeOps'), ] riscv_ss = ss.source_set() -- 2.34.1 From: Philipp Tomsich <philipp.tomsich@vrull.eu> The XVentanaCondOps extension is supported by VRULL on behalf of the Ventana Micro. Add myself as a point-of-contact. Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-Id: <20220202005249.3566542-8-philipp.tomsich@vrull.eu> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- MAINTAINERS | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ F: include/hw/riscv/ F: linux-user/host/riscv32/ F: linux-user/host/riscv64/ +RISC-V XVentanaCondOps extension +M: Philipp Tomsich <philipp.tomsich@vrull.eu> +L: qemu-riscv@nongnu.org +S: Supported +F: target/riscv/XVentanaCondOps.decode +F: target/riscv/insn_trans/trans_xventanacondops.c.inc + RENESAS RX CPUs R: Yoshinori Sato <ysato@users.sourceforge.jp> S: Orphan -- 2.34.1 From: LIU Zhiwei <zhiwei_liu@c-sky.com> The guest should be able to set the vill bit as part of vsetvl. Currently we may set env->vill to 1 in the vsetvl helper, but there is nowhere that we set it to 0, so once it transitions to 1 it's stuck there until the system is reset. Signed-off-by: LIU Zhiwei <zhiwei_liu@c-sky.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-Id: <20220201064601.41143-1-zhiwei_liu@c-sky.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/vector_helper.c | 1 + 1 file changed, 1 insertion(+) 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 @@ target_ulong HELPER(vsetvl)(CPURISCVState *env, target_ulong s1, env->vl = vl; env->vtype = s2; env->vstart = 0; + env->vill = 0; return vl; } -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> We should be returning illegal instruction trap when RV64 HS-mode tries to access RV32 HS-mode CSR. Fixes: d6f20dacea51 ("target/riscv: Fix 32-bit HS mode access permissions") Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-2-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/csr.c | 2 +- 1 file changed, 1 insertion(+), 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 hmode(CPURISCVState *env, int csrno) static RISCVException hmode32(CPURISCVState *env, int csrno) { if (riscv_cpu_mxl(env) != MXL_RV32) { - if (riscv_cpu_virt_enabled(env)) { + if (!riscv_cpu_virt_enabled(env)) { return RISCV_EXCP_ILLEGAL_INST; } else { return RISCV_EXCP_VIRT_INSTRUCTION_FAULT; -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> A hypervisor can optionally take guest external interrupts using SGEIP bit of hip and hie CSRs. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-3-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_bits.h | 3 +++ target/riscv/cpu.c | 3 ++- target/riscv/csr.c | 18 +++++++++++------- 3 files changed, 16 insertions(+), 8 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 @@ typedef enum RISCVException { #define IRQ_S_EXT 9 #define IRQ_VS_EXT 10 #define IRQ_M_EXT 11 +#define IRQ_S_GEXT 12 +#define IRQ_LOCAL_MAX 16 /* mip masks */ #define MIP_USIP (1 << IRQ_U_SOFT) @@ -XXX,XX +XXX,XX @@ typedef enum RISCVException { #define MIP_SEIP (1 << IRQ_S_EXT) #define MIP_VSEIP (1 << IRQ_VS_EXT) #define MIP_MEIP (1 << IRQ_M_EXT) +#define MIP_SGEIP (1 << IRQ_S_GEXT) /* sip masks */ #define SIP_SSIP MIP_SSIP 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_reset(DeviceState *dev) } } env->mcause = 0; + env->miclaim = MIP_SGEIP; env->pc = env->resetvec; env->two_stage_lookup = false; /* mmte is supposed to have pm.current hardwired to 1 */ @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_init(Object *obj) cpu_set_cpustate_pointers(cpu); #ifndef CONFIG_USER_ONLY - qdev_init_gpio_in(DEVICE(cpu), riscv_cpu_set_irq, 12); + qdev_init_gpio_in(DEVICE(cpu), riscv_cpu_set_irq, IRQ_LOCAL_MAX); #endif /* CONFIG_USER_ONLY */ } 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_timeh(CPURISCVState *env, int csrno, #define M_MODE_INTERRUPTS (MIP_MSIP | MIP_MTIP | MIP_MEIP) #define S_MODE_INTERRUPTS (MIP_SSIP | MIP_STIP | MIP_SEIP) #define VS_MODE_INTERRUPTS (MIP_VSSIP | MIP_VSTIP | MIP_VSEIP) +#define HS_MODE_INTERRUPTS (MIP_SGEIP | VS_MODE_INTERRUPTS) static const target_ulong delegable_ints = S_MODE_INTERRUPTS | VS_MODE_INTERRUPTS; static const target_ulong vs_delegable_ints = VS_MODE_INTERRUPTS; static const target_ulong all_ints = M_MODE_INTERRUPTS | S_MODE_INTERRUPTS | - VS_MODE_INTERRUPTS; + HS_MODE_INTERRUPTS; #define DELEGABLE_EXCPS ((1ULL << (RISCV_EXCP_INST_ADDR_MIS)) | \ (1ULL << (RISCV_EXCP_INST_ACCESS_FAULT)) | \ (1ULL << (RISCV_EXCP_ILLEGAL_INST)) | \ @@ -XXX,XX +XXX,XX @@ static RISCVException write_mideleg(CPURISCVState *env, int csrno, { env->mideleg = (env->mideleg & ~delegable_ints) | (val & delegable_ints); if (riscv_has_ext(env, RVH)) { - env->mideleg |= VS_MODE_INTERRUPTS; + env->mideleg |= HS_MODE_INTERRUPTS; } return RISCV_EXCP_NONE; } @@ -XXX,XX +XXX,XX @@ static RISCVException write_mie(CPURISCVState *env, int csrno, target_ulong val) { env->mie = (env->mie & ~all_ints) | (val & all_ints); + if (!riscv_has_ext(env, RVH)) { + env->mie &= ~MIP_SGEIP; + } return RISCV_EXCP_NONE; } @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_sip(CPURISCVState *env, int csrno, } if (ret_value) { - *ret_value &= env->mideleg; + *ret_value &= env->mideleg & S_MODE_INTERRUPTS; } return ret; } @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_hvip(CPURISCVState *env, int csrno, write_mask & hvip_writable_mask); if (ret_value) { - *ret_value &= hvip_writable_mask; + *ret_value &= VS_MODE_INTERRUPTS; } return ret; } @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_hip(CPURISCVState *env, int csrno, write_mask & hip_writable_mask); if (ret_value) { - *ret_value &= hip_writable_mask; + *ret_value &= HS_MODE_INTERRUPTS; } return ret; } @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_hip(CPURISCVState *env, int csrno, static RISCVException read_hie(CPURISCVState *env, int csrno, target_ulong *val) { - *val = env->mie & VS_MODE_INTERRUPTS; + *val = env->mie & HS_MODE_INTERRUPTS; return RISCV_EXCP_NONE; } static RISCVException write_hie(CPURISCVState *env, int csrno, target_ulong val) { - target_ulong newval = (env->mie & ~VS_MODE_INTERRUPTS) | (val & VS_MODE_INTERRUPTS); + target_ulong newval = (env->mie & ~HS_MODE_INTERRUPTS) | (val & HS_MODE_INTERRUPTS); return write_mie(env, CSR_MIE, newval); } -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> The hgeie and hgeip CSRs are required for emulating an external interrupt controller capable of injecting virtual external interrupt to Guest/VM running at VS-level. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-4-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 5 +++ target/riscv/cpu_bits.h | 1 + target/riscv/cpu.c | 67 +++++++++++++++++++++++++++------------ target/riscv/cpu_helper.c | 37 +++++++++++++++++++-- target/riscv/csr.c | 43 +++++++++++++++++-------- target/riscv/machine.c | 6 ++-- 6 files changed, 121 insertions(+), 38 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 @@ struct CPURISCVState { target_ulong priv; /* This contains QEMU specific information about the virt state. */ target_ulong virt; + target_ulong geilen; target_ulong resetvec; target_ulong mhartid; @@ -XXX,XX +XXX,XX @@ struct CPURISCVState { target_ulong htval; target_ulong htinst; target_ulong hgatp; + target_ulong hgeie; + target_ulong hgeip; uint64_t htimedelta; /* Upper 64-bits of 128-bit CSRs */ @@ -XXX,XX +XXX,XX @@ int riscv_cpu_write_elf32_note(WriteCoreDumpFunction f, CPUState *cs, int riscv_cpu_gdb_read_register(CPUState *cpu, GByteArray *buf, int reg); int riscv_cpu_gdb_write_register(CPUState *cpu, uint8_t *buf, int reg); bool riscv_cpu_fp_enabled(CPURISCVState *env); +target_ulong riscv_cpu_get_geilen(CPURISCVState *env); +void riscv_cpu_set_geilen(CPURISCVState *env, target_ulong geilen); bool riscv_cpu_vector_enabled(CPURISCVState *env); bool riscv_cpu_virt_enabled(CPURISCVState *env); void riscv_cpu_set_virt_enabled(CPURISCVState *env, bool enable); 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 IRQ_M_EXT 11 #define IRQ_S_GEXT 12 #define IRQ_LOCAL_MAX 16 +#define IRQ_LOCAL_GUEST_MAX (TARGET_LONG_BITS - 1) /* mip masks */ #define MIP_USIP (1 << IRQ_U_SOFT) 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_realize(DeviceState *dev, Error **errp) static void riscv_cpu_set_irq(void *opaque, int irq, int level) { RISCVCPU *cpu = RISCV_CPU(opaque); + CPURISCVState *env = &cpu->env; - switch (irq) { - case IRQ_U_SOFT: - case IRQ_S_SOFT: - case IRQ_VS_SOFT: - case IRQ_M_SOFT: - case IRQ_U_TIMER: - case IRQ_S_TIMER: - case IRQ_VS_TIMER: - case IRQ_M_TIMER: - case IRQ_U_EXT: - case IRQ_S_EXT: - case IRQ_VS_EXT: - case IRQ_M_EXT: - if (kvm_enabled()) { - kvm_riscv_set_irq(cpu, irq, level); - } else { - riscv_cpu_update_mip(cpu, 1 << irq, BOOL_TO_MASK(level)); + if (irq < IRQ_LOCAL_MAX) { + switch (irq) { + case IRQ_U_SOFT: + case IRQ_S_SOFT: + case IRQ_VS_SOFT: + case IRQ_M_SOFT: + case IRQ_U_TIMER: + case IRQ_S_TIMER: + case IRQ_VS_TIMER: + case IRQ_M_TIMER: + case IRQ_U_EXT: + case IRQ_S_EXT: + case IRQ_VS_EXT: + case IRQ_M_EXT: + if (kvm_enabled()) { + kvm_riscv_set_irq(cpu, irq, level); + } else { + riscv_cpu_update_mip(cpu, 1 << irq, BOOL_TO_MASK(level)); + } + break; + default: + g_assert_not_reached(); } - break; - default: + } else if (irq < (IRQ_LOCAL_MAX + IRQ_LOCAL_GUEST_MAX)) { + /* Require H-extension for handling guest local interrupts */ + if (!riscv_has_ext(env, RVH)) { + g_assert_not_reached(); + } + + /* Compute bit position in HGEIP CSR */ + irq = irq - IRQ_LOCAL_MAX + 1; + if (env->geilen < irq) { + g_assert_not_reached(); + } + + /* Update HGEIP CSR */ + env->hgeip &= ~((target_ulong)1 << irq); + if (level) { + env->hgeip |= (target_ulong)1 << irq; + } + + /* Update mip.SGEIP bit */ + riscv_cpu_update_mip(cpu, MIP_SGEIP, + BOOL_TO_MASK(!!(env->hgeie & env->hgeip))); + } else { g_assert_not_reached(); } } @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_init(Object *obj) cpu_set_cpustate_pointers(cpu); #ifndef CONFIG_USER_ONLY - qdev_init_gpio_in(DEVICE(cpu), riscv_cpu_set_irq, IRQ_LOCAL_MAX); + qdev_init_gpio_in(DEVICE(cpu), riscv_cpu_set_irq, + IRQ_LOCAL_MAX + IRQ_LOCAL_GUEST_MAX); #endif /* CONFIG_USER_ONLY */ } 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 riscv_cpu_local_irq_pending(CPURISCVState *env) target_ulong mstatus_mie = get_field(env->mstatus, MSTATUS_MIE); target_ulong mstatus_sie = get_field(env->mstatus, MSTATUS_SIE); - target_ulong pending = env->mip & env->mie; + target_ulong vsgemask = + (target_ulong)1 << get_field(env->hstatus, HSTATUS_VGEIN); + target_ulong vsgein = (env->hgeip & vsgemask) ? MIP_VSEIP : 0; + + target_ulong pending = (env->mip | vsgein) & env->mie; target_ulong mie = env->priv < PRV_M || (env->priv == PRV_M && mstatus_mie); @@ -XXX,XX +XXX,XX @@ void riscv_cpu_swap_hypervisor_regs(CPURISCVState *env) } } +target_ulong riscv_cpu_get_geilen(CPURISCVState *env) +{ + if (!riscv_has_ext(env, RVH)) { + return 0; + } + + return env->geilen; +} + +void riscv_cpu_set_geilen(CPURISCVState *env, target_ulong geilen) +{ + if (!riscv_has_ext(env, RVH)) { + return; + } + + if (geilen > (TARGET_LONG_BITS - 1)) { + return; + } + + env->geilen = geilen; +} + bool riscv_cpu_virt_enabled(CPURISCVState *env) { if (!riscv_has_ext(env, RVH)) { @@ -XXX,XX +XXX,XX @@ uint32_t riscv_cpu_update_mip(RISCVCPU *cpu, uint32_t mask, uint32_t value) { CPURISCVState *env = &cpu->env; CPUState *cs = CPU(cpu); - uint32_t old = env->mip; + uint32_t gein, vsgein = 0, old = env->mip; bool locked = false; + if (riscv_cpu_virt_enabled(env)) { + gein = get_field(env->hstatus, HSTATUS_VGEIN); + vsgein = (env->hgeip & (1ULL << gein)) ? MIP_VSEIP : 0; + } + if (!qemu_mutex_iothread_locked()) { locked = true; qemu_mutex_lock_iothread(); @@ -XXX,XX +XXX,XX @@ uint32_t riscv_cpu_update_mip(RISCVCPU *cpu, uint32_t mask, uint32_t value) env->mip = (env->mip & ~mask) | (value & mask); - if (env->mip) { + if (env->mip | vsgein) { cpu_interrupt(cs, CPU_INTERRUPT_HARD); } else { cpu_reset_interrupt(cs, CPU_INTERRUPT_HARD); 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_mip(CPURISCVState *env, int csrno, RISCVCPU *cpu = env_archcpu(env); /* Allow software control of delegable interrupts not claimed by hardware */ target_ulong mask = write_mask & delegable_ints & ~env->miclaim; - uint32_t old_mip; + uint32_t gin, old_mip; if (mask) { old_mip = riscv_cpu_update_mip(cpu, mask, (new_value & mask)); @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_mip(CPURISCVState *env, int csrno, old_mip = env->mip; } + if (csrno != CSR_HVIP) { + gin = get_field(env->hstatus, HSTATUS_VGEIN); + old_mip |= (env->hgeip & ((target_ulong)1 << gin)) ? MIP_VSEIP : 0; + } + if (ret_value) { *ret_value = old_mip; } @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_vsip(CPURISCVState *env, int csrno, target_ulong new_value, target_ulong write_mask) { /* Shift the S bits to their VS bit location in mip */ - int ret = rmw_mip(env, 0, ret_value, new_value << 1, + int ret = rmw_mip(env, csrno, ret_value, new_value << 1, (write_mask << 1) & vsip_writable_mask & env->hideleg); if (ret_value) { @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_sip(CPURISCVState *env, int csrno, if (riscv_cpu_virt_enabled(env)) { ret = rmw_vsip(env, CSR_VSIP, ret_value, new_value, write_mask); } else { - ret = rmw_mip(env, CSR_MSTATUS, ret_value, new_value, + ret = rmw_mip(env, csrno, ret_value, new_value, write_mask & env->mideleg & sip_writable_mask); } @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_hvip(CPURISCVState *env, int csrno, target_ulong *ret_value, target_ulong new_value, target_ulong write_mask) { - int ret = rmw_mip(env, 0, ret_value, new_value, + int ret = rmw_mip(env, csrno, ret_value, new_value, write_mask & hvip_writable_mask); if (ret_value) { @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_hip(CPURISCVState *env, int csrno, target_ulong *ret_value, target_ulong new_value, target_ulong write_mask) { - int ret = rmw_mip(env, 0, ret_value, new_value, + int ret = rmw_mip(env, csrno, ret_value, new_value, write_mask & hip_writable_mask); if (ret_value) { @@ -XXX,XX +XXX,XX @@ static RISCVException write_hcounteren(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } -static RISCVException write_hgeie(CPURISCVState *env, int csrno, - target_ulong val) +static RISCVException read_hgeie(CPURISCVState *env, int csrno, + target_ulong *val) { if (val) { - qemu_log_mask(LOG_UNIMP, "No support for a non-zero GEILEN."); + *val = env->hgeie; } return RISCV_EXCP_NONE; } +static RISCVException write_hgeie(CPURISCVState *env, int csrno, + target_ulong val) +{ + /* Only GEILEN:1 bits implemented and BIT0 is never implemented */ + val &= ((((target_ulong)1) << env->geilen) - 1) << 1; + env->hgeie = val; + /* Update mip.SGEIP bit */ + riscv_cpu_update_mip(env_archcpu(env), MIP_SGEIP, + BOOL_TO_MASK(!!(env->hgeie & env->hgeip))); + return RISCV_EXCP_NONE; +} + static RISCVException read_htval(CPURISCVState *env, int csrno, target_ulong *val) { @@ -XXX,XX +XXX,XX @@ static RISCVException write_htinst(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } -static RISCVException write_hgeip(CPURISCVState *env, int csrno, - target_ulong val) +static RISCVException read_hgeip(CPURISCVState *env, int csrno, + target_ulong *val) { if (val) { - qemu_log_mask(LOG_UNIMP, "No support for a non-zero GEILEN."); + *val = env->hgeip; } return RISCV_EXCP_NONE; } @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_HIP] = { "hip", hmode, NULL, NULL, rmw_hip }, [CSR_HIE] = { "hie", hmode, read_hie, write_hie }, [CSR_HCOUNTEREN] = { "hcounteren", hmode, read_hcounteren, write_hcounteren }, - [CSR_HGEIE] = { "hgeie", hmode, read_zero, write_hgeie }, + [CSR_HGEIE] = { "hgeie", hmode, read_hgeie, write_hgeie }, [CSR_HTVAL] = { "htval", hmode, read_htval, write_htval }, [CSR_HTINST] = { "htinst", hmode, read_htinst, write_htinst }, - [CSR_HGEIP] = { "hgeip", hmode, read_zero, write_hgeip }, + [CSR_HGEIP] = { "hgeip", hmode, read_hgeip, NULL }, [CSR_HGATP] = { "hgatp", hmode, read_hgatp, write_hgatp }, [CSR_HTIMEDELTA] = { "htimedelta", hmode, read_htimedelta, write_htimedelta }, [CSR_HTIMEDELTAH] = { "htimedeltah", hmode32, read_htimedeltah, write_htimedeltah }, 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 hyper_needed(void *opaque) static const VMStateDescription vmstate_hyper = { .name = "cpu/hyper", - .version_id = 1, - .minimum_version_id = 1, + .version_id = 2, + .minimum_version_id = 2, .needed = hyper_needed, .fields = (VMStateField[]) { VMSTATE_UINTTL(env.hstatus, RISCVCPU), @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_hyper = { VMSTATE_UINTTL(env.htval, RISCVCPU), VMSTATE_UINTTL(env.htinst, RISCVCPU), VMSTATE_UINTTL(env.hgatp, RISCVCPU), + VMSTATE_UINTTL(env.hgeie, RISCVCPU), + VMSTATE_UINTTL(env.hgeip, RISCVCPU), VMSTATE_UINT64(env.htimedelta, RISCVCPU), VMSTATE_UINT64(env.vsstatus, RISCVCPU), -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> 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. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-5-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_helper.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) 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_virt_enabled(CPURISCVState *env, bool enable) } env->virt = set_field(env->virt, VIRT_ONOFF, 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_archcpu(env), 0, 0); + } } bool riscv_cpu_two_stage_lookup(int mmu_idx) -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> The machine or device emulation should be able to force set certain CPU features because: 1) We can have certain CPU features which are in-general optional but implemented by RISC-V CPUs on the machine. 2) We can have devices which require a certain CPU feature. For example, AIA IMSIC devices expect AIA CSRs implemented by RISC-V CPUs. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-6-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 5 +++++ target/riscv/cpu.c | 11 +++-------- 2 files changed, 8 insertions(+), 8 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 @@ static inline bool riscv_feature(CPURISCVState *env, int feature) return env->features & (1ULL << feature); } +static inline void riscv_set_feature(CPURISCVState *env, int feature) +{ + env->features |= (1ULL << feature); +} + #include "cpu_user.h" extern const char * const riscv_int_regnames[]; 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 set_vext_version(CPURISCVState *env, int vext_ver) env->vext_ver = vext_ver; } -static void set_feature(CPURISCVState *env, int feature) -{ - env->features |= (1ULL << feature); -} - static void set_resetvec(CPURISCVState *env, target_ulong resetvec) { #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp) } if (cpu->cfg.mmu) { - set_feature(env, RISCV_FEATURE_MMU); + riscv_set_feature(env, RISCV_FEATURE_MMU); } if (cpu->cfg.pmp) { - set_feature(env, RISCV_FEATURE_PMP); + riscv_set_feature(env, RISCV_FEATURE_PMP); /* * Enhanced PMP should only be available * on harts with PMP support */ if (cpu->cfg.epmp) { - set_feature(env, RISCV_FEATURE_EPMP); + riscv_set_feature(env, RISCV_FEATURE_EPMP); } } -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> We define a CPU feature for AIA CSR support in RISC-V CPUs which can be set by machine/device emulation. The RISC-V CSR emulation will also check this feature for emulating AIA CSRs. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-7-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) 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 @@ enum { RISCV_FEATURE_MMU, RISCV_FEATURE_PMP, RISCV_FEATURE_EPMP, - RISCV_FEATURE_MISA + RISCV_FEATURE_MISA, + RISCV_FEATURE_AIA }; #define PRIV_VERSION_1_10_0 0x00011000 -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> The RISC-V AIA specification extends RISC-V local interrupts and introduces new CSRs. This patch adds defines for the new AIA CSRs. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-8-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_bits.h | 119 ++++++++++++++++++++++++++++++++++++++++ 1 file changed, 119 insertions(+) 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_MTVAL 0x343 #define CSR_MIP 0x344 +/* Machine-Level Window to Indirectly Accessed Registers (AIA) */ +#define CSR_MISELECT 0x350 +#define CSR_MIREG 0x351 + +/* Machine-Level Interrupts (AIA) */ +#define CSR_MTOPI 0xfb0 + +/* Machine-Level IMSIC Interface (AIA) */ +#define CSR_MSETEIPNUM 0x358 +#define CSR_MCLREIPNUM 0x359 +#define CSR_MSETEIENUM 0x35a +#define CSR_MCLREIENUM 0x35b +#define CSR_MTOPEI 0x35c + +/* Virtual Interrupts for Supervisor Level (AIA) */ +#define CSR_MVIEN 0x308 +#define CSR_MVIP 0x309 + +/* Machine-Level High-Half CSRs (AIA) */ +#define CSR_MIDELEGH 0x313 +#define CSR_MIEH 0x314 +#define CSR_MVIENH 0x318 +#define CSR_MVIPH 0x319 +#define CSR_MIPH 0x354 + /* Supervisor Trap Setup */ #define CSR_SSTATUS 0x100 #define CSR_SEDELEG 0x102 @@ -XXX,XX +XXX,XX @@ #define CSR_SPTBR 0x180 #define CSR_SATP 0x180 +/* Supervisor-Level Window to Indirectly Accessed Registers (AIA) */ +#define CSR_SISELECT 0x150 +#define CSR_SIREG 0x151 + +/* Supervisor-Level Interrupts (AIA) */ +#define CSR_STOPI 0xdb0 + +/* Supervisor-Level IMSIC Interface (AIA) */ +#define CSR_SSETEIPNUM 0x158 +#define CSR_SCLREIPNUM 0x159 +#define CSR_SSETEIENUM 0x15a +#define CSR_SCLREIENUM 0x15b +#define CSR_STOPEI 0x15c + +/* Supervisor-Level High-Half CSRs (AIA) */ +#define CSR_SIEH 0x114 +#define CSR_SIPH 0x154 + /* Hpervisor CSRs */ #define CSR_HSTATUS 0x600 #define CSR_HEDELEG 0x602 @@ -XXX,XX +XXX,XX @@ #define CSR_MTINST 0x34a #define CSR_MTVAL2 0x34b +/* Virtual Interrupts and Interrupt Priorities (H-extension with AIA) */ +#define CSR_HVIEN 0x608 +#define CSR_HVICTL 0x609 +#define CSR_HVIPRIO1 0x646 +#define CSR_HVIPRIO2 0x647 + +/* VS-Level Window to Indirectly Accessed Registers (H-extension with AIA) */ +#define CSR_VSISELECT 0x250 +#define CSR_VSIREG 0x251 + +/* VS-Level Interrupts (H-extension with AIA) */ +#define CSR_VSTOPI 0xeb0 + +/* VS-Level IMSIC Interface (H-extension with AIA) */ +#define CSR_VSSETEIPNUM 0x258 +#define CSR_VSCLREIPNUM 0x259 +#define CSR_VSSETEIENUM 0x25a +#define CSR_VSCLREIENUM 0x25b +#define CSR_VSTOPEI 0x25c + +/* Hypervisor and VS-Level High-Half CSRs (H-extension with AIA) */ +#define CSR_HIDELEGH 0x613 +#define CSR_HVIENH 0x618 +#define CSR_HVIPH 0x655 +#define CSR_HVIPRIO1H 0x656 +#define CSR_HVIPRIO2H 0x657 +#define CSR_VSIEH 0x214 +#define CSR_VSIPH 0x254 + /* Enhanced Physical Memory Protection (ePMP) */ #define CSR_MSECCFG 0x747 #define CSR_MSECCFGH 0x757 @@ -XXX,XX +XXX,XX @@ typedef enum RISCVException { #define UMTE_U_PM_INSN U_PM_INSN #define UMTE_MASK (UMTE_U_PM_ENABLE | MMTE_U_PM_CURRENT | UMTE_U_PM_INSN) +/* MISELECT, SISELECT, and VSISELECT bits (AIA) */ +#define ISELECT_IPRIO0 0x30 +#define ISELECT_IPRIO15 0x3f +#define ISELECT_IMSIC_EIDELIVERY 0x70 +#define ISELECT_IMSIC_EITHRESHOLD 0x72 +#define ISELECT_IMSIC_EIP0 0x80 +#define ISELECT_IMSIC_EIP63 0xbf +#define ISELECT_IMSIC_EIE0 0xc0 +#define ISELECT_IMSIC_EIE63 0xff +#define ISELECT_IMSIC_FIRST ISELECT_IMSIC_EIDELIVERY +#define ISELECT_IMSIC_LAST ISELECT_IMSIC_EIE63 +#define ISELECT_MASK 0x1ff + +/* Dummy [M|S|VS]ISELECT value for emulating [M|S|VS]TOPEI CSRs */ +#define ISELECT_IMSIC_TOPEI (ISELECT_MASK + 1) + +/* IMSIC bits (AIA) */ +#define IMSIC_TOPEI_IID_SHIFT 16 +#define IMSIC_TOPEI_IID_MASK 0x7ff +#define IMSIC_TOPEI_IPRIO_MASK 0x7ff +#define IMSIC_EIPx_BITS 32 +#define IMSIC_EIEx_BITS 32 + +/* MTOPI and STOPI bits (AIA) */ +#define TOPI_IID_SHIFT 16 +#define TOPI_IID_MASK 0xfff +#define TOPI_IPRIO_MASK 0xff + +/* Interrupt priority bits (AIA) */ +#define IPRIO_IRQ_BITS 8 +#define IPRIO_MMAXIPRIO 255 +#define IPRIO_DEFAULT_UPPER 4 +#define IPRIO_DEFAULT_MIDDLE (IPRIO_DEFAULT_UPPER + 24) +#define IPRIO_DEFAULT_M IPRIO_DEFAULT_MIDDLE +#define IPRIO_DEFAULT_S (IPRIO_DEFAULT_M + 3) +#define IPRIO_DEFAULT_SGEXT (IPRIO_DEFAULT_S + 3) +#define IPRIO_DEFAULT_VS (IPRIO_DEFAULT_SGEXT + 1) +#define IPRIO_DEFAULT_LOWER (IPRIO_DEFAULT_VS + 3) + +/* HVICTL bits (AIA) */ +#define HVICTL_VTI 0x40000000 +#define HVICTL_IID 0x0fff0000 +#define HVICTL_IPRIOM 0x00000100 +#define HVICTL_IPRIO 0x000000ff +#define HVICTL_VALID_MASK \ + (HVICTL_VTI | HVICTL_IID | HVICTL_IPRIOM | HVICTL_IPRIO) + #endif -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> The AIA device emulation (such as AIA IMSIC) should be able to set (or provide) AIA ireg read-modify-write callback for each privilege level of a RISC-V HART. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-9-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 23 +++++++++++++++++++++++ target/riscv/cpu_helper.c | 14 ++++++++++++++ 2 files changed, 37 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 CPURISCVState { uint64_t (*rdtime_fn)(uint32_t); uint32_t rdtime_fn_arg; + /* machine specific AIA ireg read-modify-write callback */ +#define AIA_MAKE_IREG(__isel, __priv, __virt, __vgein, __xlen) \ + ((((__xlen) & 0xff) << 24) | \ + (((__vgein) & 0x3f) << 20) | \ + (((__virt) & 0x1) << 18) | \ + (((__priv) & 0x3) << 16) | \ + (__isel & 0xffff)) +#define AIA_IREG_ISEL(__ireg) ((__ireg) & 0xffff) +#define AIA_IREG_PRIV(__ireg) (((__ireg) >> 16) & 0x3) +#define AIA_IREG_VIRT(__ireg) (((__ireg) >> 18) & 0x1) +#define AIA_IREG_VGEIN(__ireg) (((__ireg) >> 20) & 0x3f) +#define AIA_IREG_XLEN(__ireg) (((__ireg) >> 24) & 0xff) + int (*aia_ireg_rmw_fn[4])(void *arg, target_ulong reg, + target_ulong *val, target_ulong new_val, target_ulong write_mask); + void *aia_ireg_rmw_fn_arg[4]; + /* True if in debugger mode. */ bool debugger; @@ -XXX,XX +XXX,XX @@ uint32_t riscv_cpu_update_mip(RISCVCPU *cpu, uint32_t mask, uint32_t value); #define BOOL_TO_MASK(x) (-!!(x)) /* helper for riscv_cpu_update_mip value */ void riscv_cpu_set_rdtime_fn(CPURISCVState *env, uint64_t (*fn)(uint32_t), uint32_t arg); +void riscv_cpu_set_aia_ireg_rmw_fn(CPURISCVState *env, uint32_t priv, + int (*rmw_fn)(void *arg, + target_ulong reg, + target_ulong *val, + target_ulong new_val, + target_ulong write_mask), + void *rmw_fn_arg); #endif void riscv_cpu_set_mode(CPURISCVState *env, target_ulong newpriv); 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_rdtime_fn(CPURISCVState *env, uint64_t (*fn)(uint32_t), env->rdtime_fn_arg = arg; } +void riscv_cpu_set_aia_ireg_rmw_fn(CPURISCVState *env, uint32_t priv, + int (*rmw_fn)(void *arg, + target_ulong reg, + target_ulong *val, + target_ulong new_val, + target_ulong write_mask), + void *rmw_fn_arg) +{ + if (priv <= PRV_M) { + env->aia_ireg_rmw_fn[priv] = rmw_fn; + env->aia_ireg_rmw_fn_arg[priv] = rmw_fn_arg; + } +} + void riscv_cpu_set_mode(CPURISCVState *env, target_ulong newpriv) { if (newpriv > PRV_M) { -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> The AIA spec defines programmable 8-bit priority for each local interrupt at M-level, S-level and VS-level so we extend local interrupt processing to consider AIA interrupt priorities. The AIA CSRs which help software configure local interrupt priorities will be added by subsequent patches. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-id: 20220204174700.534953-10-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 12 ++ target/riscv/cpu.c | 19 +++ target/riscv/cpu_helper.c | 281 +++++++++++++++++++++++++++++++++++--- target/riscv/machine.c | 3 + 4 files changed, 294 insertions(+), 21 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 @@ struct CPURISCVState { target_ulong mcause; target_ulong mtval; /* since: priv-1.10.0 */ + /* Machine and Supervisor interrupt priorities */ + uint8_t miprio[64]; + uint8_t siprio[64]; + /* Hypervisor CSRs */ target_ulong hstatus; target_ulong hedeleg; @@ -XXX,XX +XXX,XX @@ struct CPURISCVState { target_ulong hgeip; uint64_t htimedelta; + /* Hypervisor controlled virtual interrupt priorities */ + uint8_t hviprio[64]; + /* Upper 64-bits of 128-bit CSRs */ uint64_t mscratchh; uint64_t sscratchh; @@ -XXX,XX +XXX,XX @@ int riscv_cpu_write_elf32_note(WriteCoreDumpFunction f, CPUState *cs, int cpuid, void *opaque); int riscv_cpu_gdb_read_register(CPUState *cpu, GByteArray *buf, int reg); int riscv_cpu_gdb_write_register(CPUState *cpu, uint8_t *buf, int reg); +int riscv_cpu_hviprio_index2irq(int index, int *out_irq, int *out_rdzero); +uint8_t riscv_cpu_default_priority(int irq); +int riscv_cpu_mirq_pending(CPURISCVState *env); +int riscv_cpu_sirq_pending(CPURISCVState *env); +int riscv_cpu_vsirq_pending(CPURISCVState *env); bool riscv_cpu_fp_enabled(CPURISCVState *env); target_ulong riscv_cpu_get_geilen(CPURISCVState *env); void riscv_cpu_set_geilen(CPURISCVState *env, target_ulong geilen); 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 @@ void restore_state_to_opc(CPURISCVState *env, TranslationBlock *tb, static void riscv_cpu_reset(DeviceState *dev) { +#ifndef CONFIG_USER_ONLY + uint8_t iprio; + int i, irq, rdzero; +#endif CPUState *cs = CPU(dev); RISCVCPU *cpu = RISCV_CPU(cs); RISCVCPUClass *mcc = RISCV_CPU_GET_CLASS(cpu); @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_reset(DeviceState *dev) env->miclaim = MIP_SGEIP; env->pc = env->resetvec; env->two_stage_lookup = false; + + /* Initialized default priorities of local interrupts. */ + for (i = 0; i < ARRAY_SIZE(env->miprio); i++) { + iprio = riscv_cpu_default_priority(i); + env->miprio[i] = (i == IRQ_M_EXT) ? 0 : iprio; + env->siprio[i] = (i == IRQ_S_EXT) ? 0 : iprio; + env->hviprio[i] = 0; + } + i = 0; + while (!riscv_cpu_hviprio_index2irq(i, &irq, &rdzero)) { + if (!rdzero) { + env->hviprio[irq] = env->miprio[irq]; + } + i++; + } /* mmte is supposed to have pm.current hardwired to 1 */ env->mmte |= (PM_EXT_INITIAL | MMTE_M_PM_CURRENT); #endif 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_update_mask(CPURISCVState *env) } #ifndef CONFIG_USER_ONLY -static int riscv_cpu_local_irq_pending(CPURISCVState *env) + +/* + * The HS-mode is allowed to configure priority only for the + * following VS-mode local interrupts: + * + * 0 (Reserved interrupt, reads as zero) + * 1 Supervisor software interrupt + * 4 (Reserved interrupt, reads as zero) + * 5 Supervisor timer interrupt + * 8 (Reserved interrupt, reads as zero) + * 13 (Reserved interrupt) + * 14 " + * 15 " + * 16 " + * 18 Debug/trace interrupt + * 20 (Reserved interrupt) + * 22 " + * 24 " + * 26 " + * 28 " + * 30 (Reserved for standard reporting of bus or system errors) + */ + +static const int hviprio_index2irq[] = { + 0, 1, 4, 5, 8, 13, 14, 15, 16, 18, 20, 22, 24, 26, 28, 30 }; +static const int hviprio_index2rdzero[] = { + 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }; + +int riscv_cpu_hviprio_index2irq(int index, int *out_irq, int *out_rdzero) +{ + if (index < 0 || ARRAY_SIZE(hviprio_index2irq) <= index) { + return -EINVAL; + } + + if (out_irq) { + *out_irq = hviprio_index2irq[index]; + } + + if (out_rdzero) { + *out_rdzero = hviprio_index2rdzero[index]; + } + + return 0; +} + +/* + * Default priorities of local interrupts are defined in the + * RISC-V Advanced Interrupt Architecture specification. + * + * ---------------------------------------------------------------- + * Default | + * Priority | Major Interrupt Numbers + * ---------------------------------------------------------------- + * Highest | 63 (3f), 62 (3e), 31 (1f), 30 (1e), 61 (3d), 60 (3c), + * | 59 (3b), 58 (3a), 29 (1d), 28 (1c), 57 (39), 56 (38), + * | 55 (37), 54 (36), 27 (1b), 26 (1a), 53 (35), 52 (34), + * | 51 (33), 50 (32), 25 (19), 24 (18), 49 (31), 48 (30) + * | + * | 11 (0b), 3 (03), 7 (07) + * | 9 (09), 1 (01), 5 (05) + * | 12 (0c) + * | 10 (0a), 2 (02), 6 (06) + * | + * | 47 (2f), 46 (2e), 23 (17), 22 (16), 45 (2d), 44 (2c), + * | 43 (2b), 42 (2a), 21 (15), 20 (14), 41 (29), 40 (28), + * | 39 (27), 38 (26), 19 (13), 18 (12), 37 (25), 36 (24), + * Lowest | 35 (23), 34 (22), 17 (11), 16 (10), 33 (21), 32 (20) + * ---------------------------------------------------------------- + */ +static const uint8_t default_iprio[64] = { + [63] = IPRIO_DEFAULT_UPPER, + [62] = IPRIO_DEFAULT_UPPER + 1, + [31] = IPRIO_DEFAULT_UPPER + 2, + [30] = IPRIO_DEFAULT_UPPER + 3, + [61] = IPRIO_DEFAULT_UPPER + 4, + [60] = IPRIO_DEFAULT_UPPER + 5, + + [59] = IPRIO_DEFAULT_UPPER + 6, + [58] = IPRIO_DEFAULT_UPPER + 7, + [29] = IPRIO_DEFAULT_UPPER + 8, + [28] = IPRIO_DEFAULT_UPPER + 9, + [57] = IPRIO_DEFAULT_UPPER + 10, + [56] = IPRIO_DEFAULT_UPPER + 11, + + [55] = IPRIO_DEFAULT_UPPER + 12, + [54] = IPRIO_DEFAULT_UPPER + 13, + [27] = IPRIO_DEFAULT_UPPER + 14, + [26] = IPRIO_DEFAULT_UPPER + 15, + [53] = IPRIO_DEFAULT_UPPER + 16, + [52] = IPRIO_DEFAULT_UPPER + 17, + + [51] = IPRIO_DEFAULT_UPPER + 18, + [50] = IPRIO_DEFAULT_UPPER + 19, + [25] = IPRIO_DEFAULT_UPPER + 20, + [24] = IPRIO_DEFAULT_UPPER + 21, + [49] = IPRIO_DEFAULT_UPPER + 22, + [48] = IPRIO_DEFAULT_UPPER + 23, + + [11] = IPRIO_DEFAULT_M, + [3] = IPRIO_DEFAULT_M + 1, + [7] = IPRIO_DEFAULT_M + 2, + + [9] = IPRIO_DEFAULT_S, + [1] = IPRIO_DEFAULT_S + 1, + [5] = IPRIO_DEFAULT_S + 2, + + [12] = IPRIO_DEFAULT_SGEXT, + + [10] = IPRIO_DEFAULT_VS, + [2] = IPRIO_DEFAULT_VS + 1, + [6] = IPRIO_DEFAULT_VS + 2, + + [47] = IPRIO_DEFAULT_LOWER, + [46] = IPRIO_DEFAULT_LOWER + 1, + [23] = IPRIO_DEFAULT_LOWER + 2, + [22] = IPRIO_DEFAULT_LOWER + 3, + [45] = IPRIO_DEFAULT_LOWER + 4, + [44] = IPRIO_DEFAULT_LOWER + 5, + + [43] = IPRIO_DEFAULT_LOWER + 6, + [42] = IPRIO_DEFAULT_LOWER + 7, + [21] = IPRIO_DEFAULT_LOWER + 8, + [20] = IPRIO_DEFAULT_LOWER + 9, + [41] = IPRIO_DEFAULT_LOWER + 10, + [40] = IPRIO_DEFAULT_LOWER + 11, + + [39] = IPRIO_DEFAULT_LOWER + 12, + [38] = IPRIO_DEFAULT_LOWER + 13, + [19] = IPRIO_DEFAULT_LOWER + 14, + [18] = IPRIO_DEFAULT_LOWER + 15, + [37] = IPRIO_DEFAULT_LOWER + 16, + [36] = IPRIO_DEFAULT_LOWER + 17, + + [35] = IPRIO_DEFAULT_LOWER + 18, + [34] = IPRIO_DEFAULT_LOWER + 19, + [17] = IPRIO_DEFAULT_LOWER + 20, + [16] = IPRIO_DEFAULT_LOWER + 21, + [33] = IPRIO_DEFAULT_LOWER + 22, + [32] = IPRIO_DEFAULT_LOWER + 23, +}; + +uint8_t riscv_cpu_default_priority(int irq) { - target_ulong virt_enabled = riscv_cpu_virt_enabled(env); + if (irq < 0 || irq > 63) { + return IPRIO_MMAXIPRIO; + } + + return default_iprio[irq] ? default_iprio[irq] : IPRIO_MMAXIPRIO; +}; - target_ulong mstatus_mie = get_field(env->mstatus, MSTATUS_MIE); - target_ulong mstatus_sie = get_field(env->mstatus, MSTATUS_SIE); +static int riscv_cpu_pending_to_irq(CPURISCVState *env, + int extirq, unsigned int extirq_def_prio, + uint64_t pending, uint8_t *iprio) +{ + int irq, best_irq = RISCV_EXCP_NONE; + unsigned int prio, best_prio = UINT_MAX; - target_ulong vsgemask = - (target_ulong)1 << get_field(env->hstatus, HSTATUS_VGEIN); - target_ulong vsgein = (env->hgeip & vsgemask) ? MIP_VSEIP : 0; + if (!pending) { + return RISCV_EXCP_NONE; + } - target_ulong pending = (env->mip | vsgein) & env->mie; + irq = ctz64(pending); + if (!riscv_feature(env, RISCV_FEATURE_AIA)) { + return irq; + } - target_ulong mie = env->priv < PRV_M || - (env->priv == PRV_M && mstatus_mie); - target_ulong sie = env->priv < PRV_S || - (env->priv == PRV_S && mstatus_sie); - target_ulong hsie = virt_enabled || sie; - target_ulong vsie = virt_enabled && sie; + pending = pending >> irq; + while (pending) { + prio = iprio[irq]; + if (!prio) { + if (irq == extirq) { + prio = extirq_def_prio; + } else { + prio = (riscv_cpu_default_priority(irq) < extirq_def_prio) ? + 1 : IPRIO_MMAXIPRIO; + } + } + if ((pending & 0x1) && (prio <= best_prio)) { + best_irq = irq; + best_prio = prio; + } + irq++; + pending = pending >> 1; + } - target_ulong irqs = - (pending & ~env->mideleg & -mie) | - (pending & env->mideleg & ~env->hideleg & -hsie) | - (pending & env->mideleg & env->hideleg & -vsie); + return best_irq; +} - if (irqs) { - return ctz64(irqs); /* since non-zero */ +static uint64_t riscv_cpu_all_pending(CPURISCVState *env) +{ + uint32_t gein = get_field(env->hstatus, HSTATUS_VGEIN); + uint64_t vsgein = (env->hgeip & (1ULL << gein)) ? MIP_VSEIP : 0; + + return (env->mip | vsgein) & env->mie; +} + +int riscv_cpu_mirq_pending(CPURISCVState *env) +{ + uint64_t irqs = riscv_cpu_all_pending(env) & ~env->mideleg & + ~(MIP_SGEIP | MIP_VSSIP | MIP_VSTIP | MIP_VSEIP); + + return riscv_cpu_pending_to_irq(env, IRQ_M_EXT, IPRIO_DEFAULT_M, + irqs, env->miprio); +} + +int riscv_cpu_sirq_pending(CPURISCVState *env) +{ + uint64_t irqs = riscv_cpu_all_pending(env) & env->mideleg & + ~(MIP_VSSIP | MIP_VSTIP | MIP_VSEIP); + + return riscv_cpu_pending_to_irq(env, IRQ_S_EXT, IPRIO_DEFAULT_S, + irqs, env->siprio); +} + +int riscv_cpu_vsirq_pending(CPURISCVState *env) +{ + uint64_t irqs = riscv_cpu_all_pending(env) & env->mideleg & + (MIP_VSSIP | MIP_VSTIP | MIP_VSEIP); + + return riscv_cpu_pending_to_irq(env, IRQ_S_EXT, IPRIO_DEFAULT_S, + irqs >> 1, env->hviprio); +} + +static int riscv_cpu_local_irq_pending(CPURISCVState *env) +{ + int virq; + uint64_t irqs, pending, mie, hsie, vsie; + + /* Determine interrupt enable state of all privilege modes */ + if (riscv_cpu_virt_enabled(env)) { + mie = 1; + hsie = 1; + vsie = (env->priv < PRV_S) || + (env->priv == PRV_S && get_field(env->mstatus, MSTATUS_SIE)); } else { - return RISCV_EXCP_NONE; /* indicates no pending interrupt */ + mie = (env->priv < PRV_M) || + (env->priv == PRV_M && get_field(env->mstatus, MSTATUS_MIE)); + hsie = (env->priv < PRV_S) || + (env->priv == PRV_S && get_field(env->mstatus, MSTATUS_SIE)); + vsie = 0; + } + + /* Determine all pending interrupts */ + pending = riscv_cpu_all_pending(env); + + /* Check M-mode interrupts */ + irqs = pending & ~env->mideleg & -mie; + if (irqs) { + return riscv_cpu_pending_to_irq(env, IRQ_M_EXT, IPRIO_DEFAULT_M, + irqs, env->miprio); } + + /* Check HS-mode interrupts */ + irqs = pending & env->mideleg & ~env->hideleg & -hsie; + if (irqs) { + return riscv_cpu_pending_to_irq(env, IRQ_S_EXT, IPRIO_DEFAULT_S, + irqs, env->siprio); + } + + /* Check VS-mode interrupts */ + irqs = pending & env->mideleg & env->hideleg & -vsie; + if (irqs) { + virq = riscv_cpu_pending_to_irq(env, IRQ_S_EXT, IPRIO_DEFAULT_S, + irqs >> 1, env->hviprio); + return (virq <= 0) ? virq : virq + 1; + } + + /* Indicate no pending interrupt */ + return RISCV_EXCP_NONE; } bool riscv_cpu_exec_interrupt(CPUState *cs, int interrupt_request) 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 const VMStateDescription vmstate_hyper = { VMSTATE_UINTTL(env.hgeie, RISCVCPU), VMSTATE_UINTTL(env.hgeip, RISCVCPU), VMSTATE_UINT64(env.htimedelta, RISCVCPU), + VMSTATE_UINT8_ARRAY(env.hviprio, RISCVCPU, 64), VMSTATE_UINT64(env.vsstatus, RISCVCPU), VMSTATE_UINTTL(env.vstvec, RISCVCPU), @@ -XXX,XX +XXX,XX @@ const VMStateDescription vmstate_riscv_cpu = { .fields = (VMStateField[]) { VMSTATE_UINTTL_ARRAY(env.gpr, RISCVCPU, 32), VMSTATE_UINT64_ARRAY(env.fpr, RISCVCPU, 32), + VMSTATE_UINT8_ARRAY(env.miprio, RISCVCPU, 64), + VMSTATE_UINT8_ARRAY(env.siprio, RISCVCPU, 64), VMSTATE_UINTTL(env.pc, RISCVCPU), VMSTATE_UINTTL(env.load_res, RISCVCPU), VMSTATE_UINTTL(env.load_val, RISCVCPU), -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> The AIA specification adds new CSRs for RV32 so that RISC-V hart can support 64 local interrupts on both RV32 and RV64. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-11-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 14 +- target/riscv/cpu_helper.c | 10 +- target/riscv/csr.c | 560 +++++++++++++++++++++++++++++++------- target/riscv/machine.c | 10 +- 4 files changed, 474 insertions(+), 120 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 @@ struct CPURISCVState { */ uint64_t mstatus; - target_ulong mip; + uint64_t mip; - uint32_t miclaim; + uint64_t miclaim; - target_ulong mie; - target_ulong mideleg; + uint64_t mie; + uint64_t mideleg; target_ulong satp; /* since: priv-1.10.0 */ target_ulong stval; @@ -XXX,XX +XXX,XX @@ struct CPURISCVState { /* Hypervisor CSRs */ target_ulong hstatus; target_ulong hedeleg; - target_ulong hideleg; + uint64_t hideleg; target_ulong hcounteren; target_ulong htval; target_ulong htinst; @@ -XXX,XX +XXX,XX @@ void riscv_cpu_list(void); #ifndef CONFIG_USER_ONLY bool riscv_cpu_exec_interrupt(CPUState *cs, int interrupt_request); void riscv_cpu_swap_hypervisor_regs(CPURISCVState *env); -int riscv_cpu_claim_interrupts(RISCVCPU *cpu, uint32_t interrupts); -uint32_t riscv_cpu_update_mip(RISCVCPU *cpu, uint32_t mask, uint32_t value); +int riscv_cpu_claim_interrupts(RISCVCPU *cpu, uint64_t interrupts); +uint64_t riscv_cpu_update_mip(RISCVCPU *cpu, uint64_t mask, uint64_t value); #define BOOL_TO_MASK(x) (-!!(x)) /* helper for riscv_cpu_update_mip value */ void riscv_cpu_set_rdtime_fn(CPURISCVState *env, uint64_t (*fn)(uint32_t), uint32_t arg); 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 @@ bool riscv_cpu_two_stage_lookup(int mmu_idx) return mmu_idx & TB_FLAGS_PRIV_HYP_ACCESS_MASK; } -int riscv_cpu_claim_interrupts(RISCVCPU *cpu, uint32_t interrupts) +int riscv_cpu_claim_interrupts(RISCVCPU *cpu, uint64_t interrupts) { CPURISCVState *env = &cpu->env; if (env->miclaim & interrupts) { @@ -XXX,XX +XXX,XX @@ int riscv_cpu_claim_interrupts(RISCVCPU *cpu, uint32_t interrupts) } } -uint32_t riscv_cpu_update_mip(RISCVCPU *cpu, uint32_t mask, uint32_t value) +uint64_t riscv_cpu_update_mip(RISCVCPU *cpu, uint64_t mask, uint64_t value) { CPURISCVState *env = &cpu->env; CPUState *cs = CPU(cpu); - uint32_t gein, vsgein = 0, old = env->mip; + uint64_t gein, vsgein = 0, old = env->mip; bool locked = false; if (riscv_cpu_virt_enabled(env)) { @@ -XXX,XX +XXX,XX @@ void riscv_cpu_do_interrupt(CPUState *cs) */ bool async = !!(cs->exception_index & RISCV_EXCP_INT_FLAG); target_ulong cause = cs->exception_index & RISCV_EXCP_INT_MASK; - target_ulong deleg = async ? env->mideleg : env->medeleg; + uint64_t deleg = async ? env->mideleg : env->medeleg; target_ulong tval = 0; target_ulong htval = 0; target_ulong mtval2 = 0; @@ -XXX,XX +XXX,XX @@ void riscv_cpu_do_interrupt(CPUState *cs) cause < TARGET_LONG_BITS && ((deleg >> cause) & 1)) { /* handle the trap in S-mode */ if (riscv_has_ext(env, RVH)) { - target_ulong hdeleg = async ? env->hideleg : env->hedeleg; + uint64_t hdeleg = async ? env->hideleg : env->hedeleg; if (riscv_cpu_virt_enabled(env) && ((hdeleg >> cause) & 1)) { /* Trap to VS mode */ 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 any32(CPURISCVState *env, int csrno) } +static int aia_any32(CPURISCVState *env, int csrno) +{ + if (!riscv_feature(env, RISCV_FEATURE_AIA)) { + return RISCV_EXCP_ILLEGAL_INST; + } + + return any32(env, csrno); +} + static RISCVException smode(CPURISCVState *env, int csrno) { if (riscv_has_ext(env, RVS)) { @@ -XXX,XX +XXX,XX @@ static RISCVException smode(CPURISCVState *env, int csrno) return RISCV_EXCP_ILLEGAL_INST; } +static int smode32(CPURISCVState *env, int csrno) +{ + if (riscv_cpu_mxl(env) != MXL_RV32) { + return RISCV_EXCP_ILLEGAL_INST; + } + + return smode(env, csrno); +} + +static int aia_smode32(CPURISCVState *env, int csrno) +{ + if (!riscv_feature(env, RISCV_FEATURE_AIA)) { + return RISCV_EXCP_ILLEGAL_INST; + } + + return smode32(env, csrno); +} + static RISCVException hmode(CPURISCVState *env, int csrno) { if (riscv_has_ext(env, RVS) && @@ -XXX,XX +XXX,XX @@ static RISCVException pointer_masking(CPURISCVState *env, int csrno) return RISCV_EXCP_ILLEGAL_INST; } +static int aia_hmode32(CPURISCVState *env, int csrno) +{ + if (!riscv_feature(env, RISCV_FEATURE_AIA)) { + return RISCV_EXCP_ILLEGAL_INST; + } + + return hmode32(env, csrno); +} + static RISCVException pmp(CPURISCVState *env, int csrno) { if (riscv_feature(env, RISCV_FEATURE_PMP)) { @@ -XXX,XX +XXX,XX @@ static RISCVException read_timeh(CPURISCVState *env, int csrno, /* Machine constants */ -#define M_MODE_INTERRUPTS (MIP_MSIP | MIP_MTIP | MIP_MEIP) -#define S_MODE_INTERRUPTS (MIP_SSIP | MIP_STIP | MIP_SEIP) -#define VS_MODE_INTERRUPTS (MIP_VSSIP | MIP_VSTIP | MIP_VSEIP) -#define HS_MODE_INTERRUPTS (MIP_SGEIP | VS_MODE_INTERRUPTS) +#define M_MODE_INTERRUPTS ((uint64_t)(MIP_MSIP | MIP_MTIP | MIP_MEIP)) +#define S_MODE_INTERRUPTS ((uint64_t)(MIP_SSIP | MIP_STIP | MIP_SEIP)) +#define VS_MODE_INTERRUPTS ((uint64_t)(MIP_VSSIP | MIP_VSTIP | MIP_VSEIP)) +#define HS_MODE_INTERRUPTS ((uint64_t)(MIP_SGEIP | VS_MODE_INTERRUPTS)) -static const target_ulong delegable_ints = S_MODE_INTERRUPTS | +static const uint64_t delegable_ints = S_MODE_INTERRUPTS | VS_MODE_INTERRUPTS; -static const target_ulong vs_delegable_ints = VS_MODE_INTERRUPTS; -static const target_ulong all_ints = M_MODE_INTERRUPTS | S_MODE_INTERRUPTS | +static const uint64_t vs_delegable_ints = VS_MODE_INTERRUPTS; +static const uint64_t all_ints = M_MODE_INTERRUPTS | S_MODE_INTERRUPTS | HS_MODE_INTERRUPTS; #define DELEGABLE_EXCPS ((1ULL << (RISCV_EXCP_INST_ADDR_MIS)) | \ (1ULL << (RISCV_EXCP_INST_ACCESS_FAULT)) | \ @@ -XXX,XX +XXX,XX @@ static RISCVException write_medeleg(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } -static RISCVException read_mideleg(CPURISCVState *env, int csrno, - target_ulong *val) +static RISCVException rmw_mideleg64(CPURISCVState *env, int csrno, + uint64_t *ret_val, + uint64_t new_val, uint64_t wr_mask) { - *val = env->mideleg; - return RISCV_EXCP_NONE; -} + uint64_t mask = wr_mask & delegable_ints; + + if (ret_val) { + *ret_val = env->mideleg; + } + + env->mideleg = (env->mideleg & ~mask) | (new_val & mask); -static RISCVException write_mideleg(CPURISCVState *env, int csrno, - target_ulong val) -{ - env->mideleg = (env->mideleg & ~delegable_ints) | (val & delegable_ints); if (riscv_has_ext(env, RVH)) { env->mideleg |= HS_MODE_INTERRUPTS; } + return RISCV_EXCP_NONE; } -static RISCVException read_mie(CPURISCVState *env, int csrno, - target_ulong *val) +static RISCVException rmw_mideleg(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) { - *val = env->mie; - return RISCV_EXCP_NONE; + uint64_t rval; + RISCVException ret; + + ret = rmw_mideleg64(env, csrno, &rval, new_val, wr_mask); + if (ret_val) { + *ret_val = rval; + } + + return ret; } -static RISCVException write_mie(CPURISCVState *env, int csrno, - target_ulong val) +static RISCVException rmw_midelegh(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, + target_ulong wr_mask) { - env->mie = (env->mie & ~all_ints) | (val & all_ints); + uint64_t rval; + RISCVException ret; + + ret = rmw_mideleg64(env, csrno, &rval, + ((uint64_t)new_val) << 32, ((uint64_t)wr_mask) << 32); + if (ret_val) { + *ret_val = rval >> 32; + } + + return ret; +} + +static RISCVException rmw_mie64(CPURISCVState *env, int csrno, + uint64_t *ret_val, + uint64_t new_val, uint64_t wr_mask) +{ + uint64_t mask = wr_mask & all_ints; + + if (ret_val) { + *ret_val = env->mie; + } + + env->mie = (env->mie & ~mask) | (new_val & mask); + if (!riscv_has_ext(env, RVH)) { - env->mie &= ~MIP_SGEIP; + env->mie &= ~((uint64_t)MIP_SGEIP); } + return RISCV_EXCP_NONE; } +static RISCVException rmw_mie(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) +{ + uint64_t rval; + RISCVException ret; + + ret = rmw_mie64(env, csrno, &rval, new_val, wr_mask); + if (ret_val) { + *ret_val = rval; + } + + return ret; +} + +static RISCVException rmw_mieh(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) +{ + uint64_t rval; + RISCVException ret; + + ret = rmw_mie64(env, csrno, &rval, + ((uint64_t)new_val) << 32, ((uint64_t)wr_mask) << 32); + if (ret_val) { + *ret_val = rval >> 32; + } + + return ret; +} + static RISCVException read_mtvec(CPURISCVState *env, int csrno, target_ulong *val) { @@ -XXX,XX +XXX,XX @@ static RISCVException write_mtval(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } -static RISCVException rmw_mip(CPURISCVState *env, int csrno, - target_ulong *ret_value, - target_ulong new_value, target_ulong write_mask) +static RISCVException rmw_mip64(CPURISCVState *env, int csrno, + uint64_t *ret_val, + uint64_t new_val, uint64_t wr_mask) { RISCVCPU *cpu = env_archcpu(env); /* Allow software control of delegable interrupts not claimed by hardware */ - target_ulong mask = write_mask & delegable_ints & ~env->miclaim; - uint32_t gin, old_mip; + uint64_t old_mip, mask = wr_mask & delegable_ints & ~env->miclaim; + uint32_t gin; if (mask) { - old_mip = riscv_cpu_update_mip(cpu, mask, (new_value & mask)); + old_mip = riscv_cpu_update_mip(cpu, mask, (new_val & mask)); } else { old_mip = env->mip; } @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_mip(CPURISCVState *env, int csrno, old_mip |= (env->hgeip & ((target_ulong)1 << gin)) ? MIP_VSEIP : 0; } - if (ret_value) { - *ret_value = old_mip; + if (ret_val) { + *ret_val = old_mip; } return RISCV_EXCP_NONE; } +static RISCVException rmw_mip(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) +{ + uint64_t rval; + RISCVException ret; + + ret = rmw_mip64(env, csrno, &rval, new_val, wr_mask); + if (ret_val) { + *ret_val = rval; + } + + return ret; +} + +static RISCVException rmw_miph(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) +{ + uint64_t rval; + RISCVException ret; + + ret = rmw_mip64(env, csrno, &rval, + ((uint64_t)new_val) << 32, ((uint64_t)wr_mask) << 32); + if (ret_val) { + *ret_val = rval >> 32; + } + + return ret; +} + /* Supervisor Trap Setup */ static RISCVException read_sstatus_i128(CPURISCVState *env, int csrno, Int128 *val) @@ -XXX,XX +XXX,XX @@ static RISCVException write_sstatus(CPURISCVState *env, int csrno, return write_mstatus(env, CSR_MSTATUS, newval); } -static RISCVException read_vsie(CPURISCVState *env, int csrno, - target_ulong *val) +static RISCVException rmw_vsie64(CPURISCVState *env, int csrno, + uint64_t *ret_val, + uint64_t new_val, uint64_t wr_mask) { - /* Shift the VS bits to their S bit location in vsie */ - *val = (env->mie & env->hideleg & VS_MODE_INTERRUPTS) >> 1; - return RISCV_EXCP_NONE; + RISCVException ret; + uint64_t rval, vsbits, mask = env->hideleg & VS_MODE_INTERRUPTS; + + /* Bring VS-level bits to correct position */ + vsbits = new_val & (VS_MODE_INTERRUPTS >> 1); + new_val &= ~(VS_MODE_INTERRUPTS >> 1); + new_val |= vsbits << 1; + vsbits = wr_mask & (VS_MODE_INTERRUPTS >> 1); + wr_mask &= ~(VS_MODE_INTERRUPTS >> 1); + wr_mask |= vsbits << 1; + + ret = rmw_mie64(env, csrno, &rval, new_val, wr_mask & mask); + if (ret_val) { + rval &= mask; + vsbits = rval & VS_MODE_INTERRUPTS; + rval &= ~VS_MODE_INTERRUPTS; + *ret_val = rval | (vsbits >> 1); + } + + return ret; } -static RISCVException read_sie(CPURISCVState *env, int csrno, - target_ulong *val) +static RISCVException rmw_vsie(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) { - if (riscv_cpu_virt_enabled(env)) { - read_vsie(env, CSR_VSIE, val); - } else { - *val = env->mie & env->mideleg; + uint64_t rval; + RISCVException ret; + + ret = rmw_vsie64(env, csrno, &rval, new_val, wr_mask); + if (ret_val) { + *ret_val = rval; } - return RISCV_EXCP_NONE; + + return ret; } -static RISCVException write_vsie(CPURISCVState *env, int csrno, - target_ulong val) +static RISCVException rmw_vsieh(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) { - /* Shift the S bits to their VS bit location in mie */ - target_ulong newval = (env->mie & ~VS_MODE_INTERRUPTS) | - ((val << 1) & env->hideleg & VS_MODE_INTERRUPTS); - return write_mie(env, CSR_MIE, newval); + uint64_t rval; + RISCVException ret; + + ret = rmw_vsie64(env, csrno, &rval, + ((uint64_t)new_val) << 32, ((uint64_t)wr_mask) << 32); + if (ret_val) { + *ret_val = rval >> 32; + } + + return ret; } -static int write_sie(CPURISCVState *env, int csrno, target_ulong val) +static RISCVException rmw_sie64(CPURISCVState *env, int csrno, + uint64_t *ret_val, + uint64_t new_val, uint64_t wr_mask) { + RISCVException ret; + uint64_t mask = env->mideleg & S_MODE_INTERRUPTS; + if (riscv_cpu_virt_enabled(env)) { - write_vsie(env, CSR_VSIE, val); + ret = rmw_vsie64(env, CSR_VSIE, ret_val, new_val, wr_mask); } else { - target_ulong newval = (env->mie & ~S_MODE_INTERRUPTS) | - (val & S_MODE_INTERRUPTS); - write_mie(env, CSR_MIE, newval); + ret = rmw_mie64(env, csrno, ret_val, new_val, wr_mask & mask); } - return RISCV_EXCP_NONE; + if (ret_val) { + *ret_val &= mask; + } + + return ret; +} + +static RISCVException rmw_sie(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) +{ + uint64_t rval; + RISCVException ret; + + ret = rmw_sie64(env, csrno, &rval, new_val, wr_mask); + if (ret_val) { + *ret_val = rval; + } + + return ret; +} + +static RISCVException rmw_sieh(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) +{ + uint64_t rval; + RISCVException ret; + + ret = rmw_sie64(env, csrno, &rval, + ((uint64_t)new_val) << 32, ((uint64_t)wr_mask) << 32); + if (ret_val) { + *ret_val = rval >> 32; + } + + return ret; } static RISCVException read_stvec(CPURISCVState *env, int csrno, @@ -XXX,XX +XXX,XX @@ static RISCVException write_stval(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } +static RISCVException rmw_vsip64(CPURISCVState *env, int csrno, + uint64_t *ret_val, + uint64_t new_val, uint64_t wr_mask) +{ + RISCVException ret; + uint64_t rval, vsbits, mask = env->hideleg & vsip_writable_mask; + + /* Bring VS-level bits to correct position */ + vsbits = new_val & (VS_MODE_INTERRUPTS >> 1); + new_val &= ~(VS_MODE_INTERRUPTS >> 1); + new_val |= vsbits << 1; + vsbits = wr_mask & (VS_MODE_INTERRUPTS >> 1); + wr_mask &= ~(VS_MODE_INTERRUPTS >> 1); + wr_mask |= vsbits << 1; + + ret = rmw_mip64(env, csrno, &rval, new_val, wr_mask & mask); + if (ret_val) { + rval &= mask; + vsbits = rval & VS_MODE_INTERRUPTS; + rval &= ~VS_MODE_INTERRUPTS; + *ret_val = rval | (vsbits >> 1); + } + + return ret; +} + static RISCVException rmw_vsip(CPURISCVState *env, int csrno, - target_ulong *ret_value, - target_ulong new_value, target_ulong write_mask) + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) { - /* Shift the S bits to their VS bit location in mip */ - int ret = rmw_mip(env, csrno, ret_value, new_value << 1, - (write_mask << 1) & vsip_writable_mask & env->hideleg); + uint64_t rval; + RISCVException ret; - if (ret_value) { - *ret_value &= VS_MODE_INTERRUPTS; - /* Shift the VS bits to their S bit location in vsip */ - *ret_value >>= 1; + ret = rmw_vsip64(env, csrno, &rval, new_val, wr_mask); + if (ret_val) { + *ret_val = rval; } + return ret; } -static RISCVException rmw_sip(CPURISCVState *env, int csrno, - target_ulong *ret_value, - target_ulong new_value, target_ulong write_mask) +static RISCVException rmw_vsiph(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) { - int ret; + uint64_t rval; + RISCVException ret; + + ret = rmw_vsip64(env, csrno, &rval, + ((uint64_t)new_val) << 32, ((uint64_t)wr_mask) << 32); + if (ret_val) { + *ret_val = rval >> 32; + } + + return ret; +} + +static RISCVException rmw_sip64(CPURISCVState *env, int csrno, + uint64_t *ret_val, + uint64_t new_val, uint64_t wr_mask) +{ + RISCVException ret; + uint64_t mask = env->mideleg & sip_writable_mask; if (riscv_cpu_virt_enabled(env)) { - ret = rmw_vsip(env, CSR_VSIP, ret_value, new_value, write_mask); + ret = rmw_vsip64(env, CSR_VSIP, ret_val, new_val, wr_mask); } else { - ret = rmw_mip(env, csrno, ret_value, new_value, - write_mask & env->mideleg & sip_writable_mask); + ret = rmw_mip64(env, csrno, ret_val, new_val, wr_mask & mask); } - if (ret_value) { - *ret_value &= env->mideleg & S_MODE_INTERRUPTS; + if (ret_val) { + *ret_val &= env->mideleg & S_MODE_INTERRUPTS; + } + + return ret; +} + +static RISCVException rmw_sip(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) +{ + uint64_t rval; + RISCVException ret; + + ret = rmw_sip64(env, csrno, &rval, new_val, wr_mask); + if (ret_val) { + *ret_val = rval; } + + return ret; +} + +static RISCVException rmw_siph(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) +{ + uint64_t rval; + RISCVException ret; + + ret = rmw_sip64(env, csrno, &rval, + ((uint64_t)new_val) << 32, ((uint64_t)wr_mask) << 32); + if (ret_val) { + *ret_val = rval >> 32; + } + return ret; } @@ -XXX,XX +XXX,XX @@ static RISCVException write_hedeleg(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } -static RISCVException read_hideleg(CPURISCVState *env, int csrno, - target_ulong *val) +static RISCVException rmw_hideleg64(CPURISCVState *env, int csrno, + uint64_t *ret_val, + uint64_t new_val, uint64_t wr_mask) { - *val = env->hideleg; + uint64_t mask = wr_mask & vs_delegable_ints; + + if (ret_val) { + *ret_val = env->hideleg & vs_delegable_ints; + } + + env->hideleg = (env->hideleg & ~mask) | (new_val & mask); return RISCV_EXCP_NONE; } -static RISCVException write_hideleg(CPURISCVState *env, int csrno, - target_ulong val) +static RISCVException rmw_hideleg(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) { - env->hideleg = val & vs_delegable_ints; - return RISCV_EXCP_NONE; + uint64_t rval; + RISCVException ret; + + ret = rmw_hideleg64(env, csrno, &rval, new_val, wr_mask); + if (ret_val) { + *ret_val = rval; + } + + return ret; +} + +static RISCVException rmw_hidelegh(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) +{ + uint64_t rval; + RISCVException ret; + + ret = rmw_hideleg64(env, csrno, &rval, + ((uint64_t)new_val) << 32, ((uint64_t)wr_mask) << 32); + if (ret_val) { + *ret_val = rval >> 32; + } + + return ret; +} + +static RISCVException rmw_hvip64(CPURISCVState *env, int csrno, + uint64_t *ret_val, + uint64_t new_val, uint64_t wr_mask) +{ + RISCVException ret; + + ret = rmw_mip64(env, csrno, ret_val, new_val, + wr_mask & hvip_writable_mask); + if (ret_val) { + *ret_val &= VS_MODE_INTERRUPTS; + } + + return ret; } static RISCVException rmw_hvip(CPURISCVState *env, int csrno, - target_ulong *ret_value, - target_ulong new_value, target_ulong write_mask) + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) { - int ret = rmw_mip(env, csrno, ret_value, new_value, - write_mask & hvip_writable_mask); + uint64_t rval; + RISCVException ret; - if (ret_value) { - *ret_value &= VS_MODE_INTERRUPTS; + ret = rmw_hvip64(env, csrno, &rval, new_val, wr_mask); + if (ret_val) { + *ret_val = rval; + } + + return ret; +} + +static RISCVException rmw_hviph(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) +{ + uint64_t rval; + RISCVException ret; + + ret = rmw_hvip64(env, csrno, &rval, + ((uint64_t)new_val) << 32, ((uint64_t)wr_mask) << 32); + if (ret_val) { + *ret_val = rval >> 32; } + return ret; } @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_hip(CPURISCVState *env, int csrno, return ret; } -static RISCVException read_hie(CPURISCVState *env, int csrno, - target_ulong *val) +static RISCVException rmw_hie(CPURISCVState *env, int csrno, + target_ulong *ret_val, + target_ulong new_val, target_ulong wr_mask) { - *val = env->mie & HS_MODE_INTERRUPTS; - return RISCV_EXCP_NONE; -} + uint64_t rval; + RISCVException ret; -static RISCVException write_hie(CPURISCVState *env, int csrno, - target_ulong val) -{ - target_ulong newval = (env->mie & ~HS_MODE_INTERRUPTS) | (val & HS_MODE_INTERRUPTS); - return write_mie(env, CSR_MIE, newval); + ret = rmw_mie64(env, csrno, &rval, new_val, wr_mask & HS_MODE_INTERRUPTS); + if (ret_val) { + *ret_val = rval & HS_MODE_INTERRUPTS; + } + + return ret; } static RISCVException read_hcounteren(CPURISCVState *env, int csrno, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { read_mstatus_i128 }, [CSR_MISA] = { "misa", any, read_misa, write_misa, NULL, read_misa_i128 }, - [CSR_MIDELEG] = { "mideleg", any, read_mideleg, write_mideleg }, + [CSR_MIDELEG] = { "mideleg", any, NULL, NULL, rmw_mideleg }, [CSR_MEDELEG] = { "medeleg", any, read_medeleg, write_medeleg }, - [CSR_MIE] = { "mie", any, read_mie, write_mie }, + [CSR_MIE] = { "mie", any, NULL, NULL, rmw_mie }, [CSR_MTVEC] = { "mtvec", any, read_mtvec, write_mtvec }, [CSR_MCOUNTEREN] = { "mcounteren", any, read_mcounteren, write_mcounteren }, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_MTVAL] = { "mtval", any, read_mtval, write_mtval }, [CSR_MIP] = { "mip", any, NULL, NULL, rmw_mip }, + /* Machine-Level High-Half CSRs (AIA) */ + [CSR_MIDELEGH] = { "midelegh", aia_any32, NULL, NULL, rmw_midelegh }, + [CSR_MIEH] = { "mieh", aia_any32, NULL, NULL, rmw_mieh }, + [CSR_MIPH] = { "miph", aia_any32, NULL, NULL, rmw_miph }, + /* Supervisor Trap Setup */ [CSR_SSTATUS] = { "sstatus", smode, read_sstatus, write_sstatus, NULL, read_sstatus_i128 }, - [CSR_SIE] = { "sie", smode, read_sie, write_sie }, + [CSR_SIE] = { "sie", smode, NULL, NULL, rmw_sie }, [CSR_STVEC] = { "stvec", smode, read_stvec, write_stvec }, [CSR_SCOUNTEREN] = { "scounteren", smode, read_scounteren, write_scounteren }, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { /* Supervisor Protection and Translation */ [CSR_SATP] = { "satp", smode, read_satp, write_satp }, + /* Supervisor-Level High-Half CSRs (AIA) */ + [CSR_SIEH] = { "sieh", aia_smode32, NULL, NULL, rmw_sieh }, + [CSR_SIPH] = { "siph", aia_smode32, NULL, NULL, rmw_siph }, + [CSR_HSTATUS] = { "hstatus", hmode, read_hstatus, write_hstatus }, [CSR_HEDELEG] = { "hedeleg", hmode, read_hedeleg, write_hedeleg }, - [CSR_HIDELEG] = { "hideleg", hmode, read_hideleg, write_hideleg }, + [CSR_HIDELEG] = { "hideleg", hmode, NULL, NULL, rmw_hideleg }, [CSR_HVIP] = { "hvip", hmode, NULL, NULL, rmw_hvip }, [CSR_HIP] = { "hip", hmode, NULL, NULL, rmw_hip }, - [CSR_HIE] = { "hie", hmode, read_hie, write_hie }, + [CSR_HIE] = { "hie", hmode, NULL, NULL, rmw_hie }, [CSR_HCOUNTEREN] = { "hcounteren", hmode, read_hcounteren, write_hcounteren }, [CSR_HGEIE] = { "hgeie", hmode, read_hgeie, write_hgeie }, [CSR_HTVAL] = { "htval", hmode, read_htval, write_htval }, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_VSSTATUS] = { "vsstatus", hmode, read_vsstatus, write_vsstatus }, [CSR_VSIP] = { "vsip", hmode, NULL, NULL, rmw_vsip }, - [CSR_VSIE] = { "vsie", hmode, read_vsie, write_vsie }, + [CSR_VSIE] = { "vsie", hmode, NULL, NULL, rmw_vsie }, [CSR_VSTVEC] = { "vstvec", hmode, read_vstvec, write_vstvec }, [CSR_VSSCRATCH] = { "vsscratch", hmode, read_vsscratch, write_vsscratch }, [CSR_VSEPC] = { "vsepc", hmode, read_vsepc, write_vsepc }, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_MTVAL2] = { "mtval2", hmode, read_mtval2, write_mtval2 }, [CSR_MTINST] = { "mtinst", hmode, read_mtinst, write_mtinst }, + /* Hypervisor and VS-Level High-Half CSRs (H-extension with AIA) */ + [CSR_HIDELEGH] = { "hidelegh", aia_hmode32, NULL, NULL, rmw_hidelegh }, + [CSR_HVIPH] = { "hviph", aia_hmode32, NULL, NULL, rmw_hviph }, + [CSR_VSIEH] = { "vsieh", aia_hmode32, NULL, NULL, rmw_vsieh }, + [CSR_VSIPH] = { "vsiph", aia_hmode32, NULL, NULL, rmw_vsiph }, + /* Physical Memory Protection */ [CSR_MSECCFG] = { "mseccfg", epmp, read_mseccfg, write_mseccfg }, [CSR_PMPCFG0] = { "pmpcfg0", pmp, read_pmpcfg, write_pmpcfg }, 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 const VMStateDescription vmstate_hyper = { .fields = (VMStateField[]) { VMSTATE_UINTTL(env.hstatus, RISCVCPU), VMSTATE_UINTTL(env.hedeleg, RISCVCPU), - VMSTATE_UINTTL(env.hideleg, RISCVCPU), + VMSTATE_UINT64(env.hideleg, RISCVCPU), VMSTATE_UINTTL(env.hcounteren, RISCVCPU), VMSTATE_UINTTL(env.htval, RISCVCPU), VMSTATE_UINTTL(env.htinst, RISCVCPU), @@ -XXX,XX +XXX,XX @@ const VMStateDescription vmstate_riscv_cpu = { VMSTATE_UINTTL(env.resetvec, RISCVCPU), VMSTATE_UINTTL(env.mhartid, RISCVCPU), VMSTATE_UINT64(env.mstatus, RISCVCPU), - VMSTATE_UINTTL(env.mip, RISCVCPU), - VMSTATE_UINT32(env.miclaim, RISCVCPU), - VMSTATE_UINTTL(env.mie, RISCVCPU), - VMSTATE_UINTTL(env.mideleg, RISCVCPU), + VMSTATE_UINT64(env.mip, RISCVCPU), + VMSTATE_UINT64(env.miclaim, RISCVCPU), + VMSTATE_UINT64(env.mie, RISCVCPU), + VMSTATE_UINT64(env.mideleg, RISCVCPU), VMSTATE_UINTTL(env.satp, RISCVCPU), VMSTATE_UINTTL(env.stval, RISCVCPU), VMSTATE_UINTTL(env.medeleg, RISCVCPU), -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> The AIA hvictl and hviprioX CSRs allow hypervisor to control interrupts visible at VS-level. This patch implements AIA hvictl and hviprioX CSRs. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-12-anup@brainfault.org [ Changes by AF: - Fix possible unintilised variable error in rmw_sie() ] Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 2 + target/riscv/csr.c | 128 ++++++++++++++++++++++++++++++++++++++++- target/riscv/machine.c | 2 + 3 files changed, 131 insertions(+), 1 deletion(-) 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 CPURISCVState { uint64_t htimedelta; /* Hypervisor controlled virtual interrupt priorities */ + target_ulong hvictl; uint8_t hviprio[64]; /* Upper 64-bits of 128-bit CSRs */ @@ -XXX,XX +XXX,XX @@ static inline RISCVMXL riscv_cpu_mxl(CPURISCVState *env) return env->misa_mxl; } #endif +#define riscv_cpu_mxl_bits(env) (1UL << (4 + riscv_cpu_mxl(env))) #if defined(TARGET_RISCV32) #define cpu_recompute_xl(env) ((void)(env), MXL_RV32) 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 pointer_masking(CPURISCVState *env, int csrno) return RISCV_EXCP_ILLEGAL_INST; } +static int aia_hmode(CPURISCVState *env, int csrno) +{ + if (!riscv_feature(env, RISCV_FEATURE_AIA)) { + return RISCV_EXCP_ILLEGAL_INST; + } + + return hmode(env, csrno); +} + static int aia_hmode32(CPURISCVState *env, int csrno) { if (!riscv_feature(env, RISCV_FEATURE_AIA)) { @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_sie64(CPURISCVState *env, int csrno, uint64_t mask = env->mideleg & S_MODE_INTERRUPTS; if (riscv_cpu_virt_enabled(env)) { + if (env->hvictl & HVICTL_VTI) { + return RISCV_EXCP_VIRT_INSTRUCTION_FAULT; + } ret = rmw_vsie64(env, CSR_VSIE, ret_val, new_val, wr_mask); } else { ret = rmw_mie64(env, csrno, ret_val, new_val, wr_mask & mask); @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_sie(CPURISCVState *env, int csrno, RISCVException ret; ret = rmw_sie64(env, csrno, &rval, new_val, wr_mask); - if (ret_val) { + if (ret == RISCV_EXCP_NONE && ret_val) { *ret_val = rval; } @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_sip64(CPURISCVState *env, int csrno, uint64_t mask = env->mideleg & sip_writable_mask; if (riscv_cpu_virt_enabled(env)) { + if (env->hvictl & HVICTL_VTI) { + return RISCV_EXCP_VIRT_INSTRUCTION_FAULT; + } ret = rmw_vsip64(env, CSR_VSIP, ret_val, new_val, wr_mask); } else { ret = rmw_mip64(env, csrno, ret_val, new_val, wr_mask & mask); @@ -XXX,XX +XXX,XX @@ static RISCVException write_htimedeltah(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } +static int read_hvictl(CPURISCVState *env, int csrno, target_ulong *val) +{ + *val = env->hvictl; + return RISCV_EXCP_NONE; +} + +static int write_hvictl(CPURISCVState *env, int csrno, target_ulong val) +{ + env->hvictl = val & HVICTL_VALID_MASK; + return RISCV_EXCP_NONE; +} + +static int read_hvipriox(CPURISCVState *env, int first_index, + uint8_t *iprio, target_ulong *val) +{ + int i, irq, rdzero, num_irqs = 4 * (riscv_cpu_mxl_bits(env) / 32); + + /* First index has to be a multiple of number of irqs per register */ + if (first_index % num_irqs) { + return (riscv_cpu_virt_enabled(env)) ? + RISCV_EXCP_VIRT_INSTRUCTION_FAULT : RISCV_EXCP_ILLEGAL_INST; + } + + /* Fill-up return value */ + *val = 0; + for (i = 0; i < num_irqs; i++) { + if (riscv_cpu_hviprio_index2irq(first_index + i, &irq, &rdzero)) { + continue; + } + if (rdzero) { + continue; + } + *val |= ((target_ulong)iprio[irq]) << (i * 8); + } + + return RISCV_EXCP_NONE; +} + +static int write_hvipriox(CPURISCVState *env, int first_index, + uint8_t *iprio, target_ulong val) +{ + int i, irq, rdzero, num_irqs = 4 * (riscv_cpu_mxl_bits(env) / 32); + + /* First index has to be a multiple of number of irqs per register */ + if (first_index % num_irqs) { + return (riscv_cpu_virt_enabled(env)) ? + RISCV_EXCP_VIRT_INSTRUCTION_FAULT : RISCV_EXCP_ILLEGAL_INST; + } + + /* Fill-up priority arrary */ + for (i = 0; i < num_irqs; i++) { + if (riscv_cpu_hviprio_index2irq(first_index + i, &irq, &rdzero)) { + continue; + } + if (rdzero) { + iprio[irq] = 0; + } else { + iprio[irq] = (val >> (i * 8)) & 0xff; + } + } + + return RISCV_EXCP_NONE; +} + +static int read_hviprio1(CPURISCVState *env, int csrno, target_ulong *val) +{ + return read_hvipriox(env, 0, env->hviprio, val); +} + +static int write_hviprio1(CPURISCVState *env, int csrno, target_ulong val) +{ + return write_hvipriox(env, 0, env->hviprio, val); +} + +static int read_hviprio1h(CPURISCVState *env, int csrno, target_ulong *val) +{ + return read_hvipriox(env, 4, env->hviprio, val); +} + +static int write_hviprio1h(CPURISCVState *env, int csrno, target_ulong val) +{ + return write_hvipriox(env, 4, env->hviprio, val); +} + +static int read_hviprio2(CPURISCVState *env, int csrno, target_ulong *val) +{ + return read_hvipriox(env, 8, env->hviprio, val); +} + +static int write_hviprio2(CPURISCVState *env, int csrno, target_ulong val) +{ + return write_hvipriox(env, 8, env->hviprio, val); +} + +static int read_hviprio2h(CPURISCVState *env, int csrno, target_ulong *val) +{ + return read_hvipriox(env, 12, env->hviprio, val); +} + +static int write_hviprio2h(CPURISCVState *env, int csrno, target_ulong val) +{ + return write_hvipriox(env, 12, env->hviprio, val); +} + /* Virtual CSR Registers */ static RISCVException read_vsstatus(CPURISCVState *env, int csrno, target_ulong *val) @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_MTVAL2] = { "mtval2", hmode, read_mtval2, write_mtval2 }, [CSR_MTINST] = { "mtinst", hmode, read_mtinst, write_mtinst }, + /* Virtual Interrupts and Interrupt Priorities (H-extension with AIA) */ + [CSR_HVICTL] = { "hvictl", aia_hmode, read_hvictl, write_hvictl }, + [CSR_HVIPRIO1] = { "hviprio1", aia_hmode, read_hviprio1, write_hviprio1 }, + [CSR_HVIPRIO2] = { "hviprio2", aia_hmode, read_hviprio2, write_hviprio2 }, + /* Hypervisor and VS-Level High-Half CSRs (H-extension with AIA) */ [CSR_HIDELEGH] = { "hidelegh", aia_hmode32, NULL, NULL, rmw_hidelegh }, [CSR_HVIPH] = { "hviph", aia_hmode32, NULL, NULL, rmw_hviph }, + [CSR_HVIPRIO1H] = { "hviprio1h", aia_hmode32, read_hviprio1h, write_hviprio1h }, + [CSR_HVIPRIO2H] = { "hviprio2h", aia_hmode32, read_hviprio2h, write_hviprio2h }, [CSR_VSIEH] = { "vsieh", aia_hmode32, NULL, NULL, rmw_vsieh }, [CSR_VSIPH] = { "vsiph", aia_hmode32, NULL, NULL, rmw_vsiph }, 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 const VMStateDescription vmstate_hyper = { VMSTATE_UINTTL(env.hgeie, RISCVCPU), VMSTATE_UINTTL(env.hgeip, RISCVCPU), VMSTATE_UINT64(env.htimedelta, RISCVCPU), + + VMSTATE_UINTTL(env.hvictl, RISCVCPU), VMSTATE_UINT8_ARRAY(env.hviprio, RISCVCPU, 64), VMSTATE_UINT64(env.vsstatus, RISCVCPU), -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> The AIA specificaiton adds interrupt filtering support for M-mode and HS-mode. Using AIA interrupt filtering M-mode and H-mode can take local interrupt 13 or above and selectively inject same local interrupt to lower privilege modes. At the moment, we don't have any local interrupts above 12 so we add dummy implementation (i.e. read zero and ignore write) of AIA interrupt filtering CSRs. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-13-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/csr.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) 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 any32(CPURISCVState *env, int csrno) } +static int aia_any(CPURISCVState *env, int csrno) +{ + if (!riscv_feature(env, RISCV_FEATURE_AIA)) { + return RISCV_EXCP_ILLEGAL_INST; + } + + return any(env, csrno); +} + static int aia_any32(CPURISCVState *env, int csrno) { if (!riscv_feature(env, RISCV_FEATURE_AIA)) { @@ -XXX,XX +XXX,XX @@ static RISCVException read_zero(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } +static RISCVException write_ignore(CPURISCVState *env, int csrno, + target_ulong val) +{ + return RISCV_EXCP_NONE; +} + static RISCVException read_mhartid(CPURISCVState *env, int csrno, target_ulong *val) { @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_MTVAL] = { "mtval", any, read_mtval, write_mtval }, [CSR_MIP] = { "mip", any, NULL, NULL, rmw_mip }, + /* Virtual Interrupts for Supervisor Level (AIA) */ + [CSR_MVIEN] = { "mvien", aia_any, read_zero, write_ignore }, + [CSR_MVIP] = { "mvip", aia_any, read_zero, write_ignore }, + /* Machine-Level High-Half CSRs (AIA) */ [CSR_MIDELEGH] = { "midelegh", aia_any32, NULL, NULL, rmw_midelegh }, [CSR_MIEH] = { "mieh", aia_any32, NULL, NULL, rmw_mieh }, + [CSR_MVIENH] = { "mvienh", aia_any32, read_zero, write_ignore }, + [CSR_MVIPH] = { "mviph", aia_any32, read_zero, write_ignore }, [CSR_MIPH] = { "miph", aia_any32, NULL, NULL, rmw_miph }, /* Supervisor Trap Setup */ @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_MTINST] = { "mtinst", hmode, read_mtinst, write_mtinst }, /* Virtual Interrupts and Interrupt Priorities (H-extension with AIA) */ + [CSR_HVIEN] = { "hvien", aia_hmode, read_zero, write_ignore }, [CSR_HVICTL] = { "hvictl", aia_hmode, read_hvictl, write_hvictl }, [CSR_HVIPRIO1] = { "hviprio1", aia_hmode, read_hviprio1, write_hviprio1 }, [CSR_HVIPRIO2] = { "hviprio2", aia_hmode, read_hviprio2, write_hviprio2 }, /* Hypervisor and VS-Level High-Half CSRs (H-extension with AIA) */ [CSR_HIDELEGH] = { "hidelegh", aia_hmode32, NULL, NULL, rmw_hidelegh }, + [CSR_HVIENH] = { "hvienh", aia_hmode32, read_zero, write_ignore }, [CSR_HVIPH] = { "hviph", aia_hmode32, NULL, NULL, rmw_hviph }, [CSR_HVIPRIO1H] = { "hviprio1h", aia_hmode32, read_hviprio1h, write_hviprio1h }, [CSR_HVIPRIO2H] = { "hviprio2h", aia_hmode32, read_hviprio2h, write_hviprio2h }, -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> The AIA specification introduces new [m|s|vs]topi CSRs for reporting pending local IRQ number and associated IRQ priority. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-14-anup@brainfault.org [ Changed by AF: - Fixup indentation ] Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/csr.c | 156 +++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 156 insertions(+) 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 int smode32(CPURISCVState *env, int csrno) return smode(env, csrno); } +static int aia_smode(CPURISCVState *env, int csrno) +{ + if (!riscv_feature(env, RISCV_FEATURE_AIA)) { + return RISCV_EXCP_ILLEGAL_INST; + } + + return smode(env, csrno); +} + static int aia_smode32(CPURISCVState *env, int csrno) { if (!riscv_feature(env, RISCV_FEATURE_AIA)) { @@ -XXX,XX +XXX,XX @@ static RISCVException read_timeh(CPURISCVState *env, int csrno, #define VS_MODE_INTERRUPTS ((uint64_t)(MIP_VSSIP | MIP_VSTIP | MIP_VSEIP)) #define HS_MODE_INTERRUPTS ((uint64_t)(MIP_SGEIP | VS_MODE_INTERRUPTS)) +#define VSTOPI_NUM_SRCS 5 + static const uint64_t delegable_ints = S_MODE_INTERRUPTS | VS_MODE_INTERRUPTS; static const uint64_t vs_delegable_ints = VS_MODE_INTERRUPTS; @@ -XXX,XX +XXX,XX @@ static RISCVException rmw_mieh(CPURISCVState *env, int csrno, return ret; } +static int read_mtopi(CPURISCVState *env, int csrno, target_ulong *val) +{ + int irq; + uint8_t iprio; + + irq = riscv_cpu_mirq_pending(env); + if (irq <= 0 || irq > 63) { + *val = 0; + } else { + iprio = env->miprio[irq]; + if (!iprio) { + if (riscv_cpu_default_priority(irq) > IPRIO_DEFAULT_M) { + iprio = IPRIO_MMAXIPRIO; + } + } + *val = (irq & TOPI_IID_MASK) << TOPI_IID_SHIFT; + *val |= iprio; + } + + return RISCV_EXCP_NONE; +} + static RISCVException read_mtvec(CPURISCVState *env, int csrno, target_ulong *val) { @@ -XXX,XX +XXX,XX @@ static RISCVException write_satp(CPURISCVState *env, int csrno, return RISCV_EXCP_NONE; } +static int read_vstopi(CPURISCVState *env, int csrno, target_ulong *val) +{ + int irq, ret; + target_ulong topei; + uint64_t vseip, vsgein; + uint32_t iid, iprio, hviid, hviprio, gein; + uint32_t s, scount = 0, siid[VSTOPI_NUM_SRCS], siprio[VSTOPI_NUM_SRCS]; + + gein = get_field(env->hstatus, HSTATUS_VGEIN); + hviid = get_field(env->hvictl, HVICTL_IID); + hviprio = get_field(env->hvictl, HVICTL_IPRIO); + + if (gein) { + vsgein = (env->hgeip & (1ULL << gein)) ? MIP_VSEIP : 0; + vseip = env->mie & (env->mip | vsgein) & MIP_VSEIP; + if (gein <= env->geilen && vseip) { + siid[scount] = IRQ_S_EXT; + siprio[scount] = IPRIO_MMAXIPRIO + 1; + if (env->aia_ireg_rmw_fn[PRV_S]) { + /* + * Call machine specific IMSIC register emulation for + * reading TOPEI. + */ + ret = env->aia_ireg_rmw_fn[PRV_S]( + env->aia_ireg_rmw_fn_arg[PRV_S], + AIA_MAKE_IREG(ISELECT_IMSIC_TOPEI, PRV_S, true, gein, + riscv_cpu_mxl_bits(env)), + &topei, 0, 0); + if (!ret && topei) { + siprio[scount] = topei & IMSIC_TOPEI_IPRIO_MASK; + } + } + scount++; + } + } else { + if (hviid == IRQ_S_EXT && hviprio) { + siid[scount] = IRQ_S_EXT; + siprio[scount] = hviprio; + scount++; + } + } + + if (env->hvictl & HVICTL_VTI) { + if (hviid != IRQ_S_EXT) { + siid[scount] = hviid; + siprio[scount] = hviprio; + scount++; + } + } else { + irq = riscv_cpu_vsirq_pending(env); + if (irq != IRQ_S_EXT && 0 < irq && irq <= 63) { + siid[scount] = irq; + siprio[scount] = env->hviprio[irq]; + scount++; + } + } + + iid = 0; + iprio = UINT_MAX; + for (s = 0; s < scount; s++) { + if (siprio[s] < iprio) { + iid = siid[s]; + iprio = siprio[s]; + } + } + + if (iid) { + if (env->hvictl & HVICTL_IPRIOM) { + if (iprio > IPRIO_MMAXIPRIO) { + iprio = IPRIO_MMAXIPRIO; + } + if (!iprio) { + if (riscv_cpu_default_priority(iid) > IPRIO_DEFAULT_S) { + iprio = IPRIO_MMAXIPRIO; + } + } + } else { + iprio = 1; + } + } else { + iprio = 0; + } + + *val = (iid & TOPI_IID_MASK) << TOPI_IID_SHIFT; + *val |= iprio; + return RISCV_EXCP_NONE; +} + +static int read_stopi(CPURISCVState *env, int csrno, target_ulong *val) +{ + int irq; + uint8_t iprio; + + if (riscv_cpu_virt_enabled(env)) { + return read_vstopi(env, CSR_VSTOPI, val); + } + + irq = riscv_cpu_sirq_pending(env); + if (irq <= 0 || irq > 63) { + *val = 0; + } else { + iprio = env->siprio[irq]; + if (!iprio) { + if (riscv_cpu_default_priority(irq) > IPRIO_DEFAULT_S) { + iprio = IPRIO_MMAXIPRIO; + } + } + *val = (irq & TOPI_IID_MASK) << TOPI_IID_SHIFT; + *val |= iprio; + } + + return RISCV_EXCP_NONE; +} + /* Hypervisor Extensions */ static RISCVException read_hstatus(CPURISCVState *env, int csrno, target_ulong *val) @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_MTVAL] = { "mtval", any, read_mtval, write_mtval }, [CSR_MIP] = { "mip", any, NULL, NULL, rmw_mip }, + /* Machine-Level Interrupts (AIA) */ + [CSR_MTOPI] = { "mtopi", aia_any, read_mtopi }, + /* Virtual Interrupts for Supervisor Level (AIA) */ [CSR_MVIEN] = { "mvien", aia_any, read_zero, write_ignore }, [CSR_MVIP] = { "mvip", aia_any, read_zero, write_ignore }, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { /* Supervisor Protection and Translation */ [CSR_SATP] = { "satp", smode, read_satp, write_satp }, + /* Supervisor-Level Interrupts (AIA) */ + [CSR_STOPI] = { "stopi", aia_smode, read_stopi }, + /* Supervisor-Level High-Half CSRs (AIA) */ [CSR_SIEH] = { "sieh", aia_smode32, NULL, NULL, rmw_sieh }, [CSR_SIPH] = { "siph", aia_smode32, NULL, NULL, rmw_siph }, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_HVIPRIO1] = { "hviprio1", aia_hmode, read_hviprio1, write_hviprio1 }, [CSR_HVIPRIO2] = { "hviprio2", aia_hmode, read_hviprio2, write_hviprio2 }, + /* VS-Level Interrupts (H-extension with AIA) */ + [CSR_VSTOPI] = { "vstopi", aia_hmode, read_vstopi }, + /* Hypervisor and VS-Level High-Half CSRs (H-extension with AIA) */ [CSR_HIDELEGH] = { "hidelegh", aia_hmode32, NULL, NULL, rmw_hidelegh }, [CSR_HVIENH] = { "hvienh", aia_hmode32, read_zero, write_ignore }, -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> The AIA specification defines [m|s|vs]iselect and [m|s|vs]ireg CSRs which allow indirect access to interrupt priority arrays and per-HART IMSIC registers. This patch implements AIA xiselect and xireg CSRs. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-15-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 7 ++ target/riscv/csr.c | 177 +++++++++++++++++++++++++++++++++++++++++ target/riscv/machine.c | 3 + 3 files changed, 187 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 CPURISCVState { uint8_t miprio[64]; uint8_t siprio[64]; + /* AIA CSRs */ + target_ulong miselect; + target_ulong siselect; + /* Hypervisor CSRs */ target_ulong hstatus; target_ulong hedeleg; @@ -XXX,XX +XXX,XX @@ struct CPURISCVState { target_ulong vstval; target_ulong vsatp; + /* AIA VS-mode CSRs */ + target_ulong vsiselect; + target_ulong mtval2; target_ulong mtinst; 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 int read_mtopi(CPURISCVState *env, int csrno, target_ulong *val) return RISCV_EXCP_NONE; } +static int aia_xlate_vs_csrno(CPURISCVState *env, int csrno) +{ + if (!riscv_cpu_virt_enabled(env)) { + return csrno; + } + + switch (csrno) { + case CSR_SISELECT: + return CSR_VSISELECT; + case CSR_SIREG: + return CSR_VSIREG; + default: + return csrno; + }; +} + +static int rmw_xiselect(CPURISCVState *env, int csrno, target_ulong *val, + target_ulong new_val, target_ulong wr_mask) +{ + target_ulong *iselect; + + /* Translate CSR number for VS-mode */ + csrno = aia_xlate_vs_csrno(env, csrno); + + /* Find the iselect CSR based on CSR number */ + switch (csrno) { + case CSR_MISELECT: + iselect = &env->miselect; + break; + case CSR_SISELECT: + iselect = &env->siselect; + break; + case CSR_VSISELECT: + iselect = &env->vsiselect; + break; + default: + return RISCV_EXCP_ILLEGAL_INST; + }; + + if (val) { + *val = *iselect; + } + + wr_mask &= ISELECT_MASK; + if (wr_mask) { + *iselect = (*iselect & ~wr_mask) | (new_val & wr_mask); + } + + return RISCV_EXCP_NONE; +} + +static int rmw_iprio(target_ulong xlen, + target_ulong iselect, uint8_t *iprio, + target_ulong *val, target_ulong new_val, + target_ulong wr_mask, int ext_irq_no) +{ + int i, firq, nirqs; + target_ulong old_val; + + if (iselect < ISELECT_IPRIO0 || ISELECT_IPRIO15 < iselect) { + return -EINVAL; + } + if (xlen != 32 && iselect & 0x1) { + return -EINVAL; + } + + nirqs = 4 * (xlen / 32); + firq = ((iselect - ISELECT_IPRIO0) / (xlen / 32)) * (nirqs); + + old_val = 0; + for (i = 0; i < nirqs; i++) { + old_val |= ((target_ulong)iprio[firq + i]) << (IPRIO_IRQ_BITS * i); + } + + if (val) { + *val = old_val; + } + + if (wr_mask) { + new_val = (old_val & ~wr_mask) | (new_val & wr_mask); + for (i = 0; i < nirqs; i++) { + /* + * M-level and S-level external IRQ priority always read-only + * zero. This means default priority order is always preferred + * for M-level and S-level external IRQs. + */ + if ((firq + i) == ext_irq_no) { + continue; + } + iprio[firq + i] = (new_val >> (IPRIO_IRQ_BITS * i)) & 0xff; + } + } + + return 0; +} + +static int rmw_xireg(CPURISCVState *env, int csrno, target_ulong *val, + target_ulong new_val, target_ulong wr_mask) +{ + bool virt; + uint8_t *iprio; + int ret = -EINVAL; + target_ulong priv, isel, vgein; + + /* Translate CSR number for VS-mode */ + csrno = aia_xlate_vs_csrno(env, csrno); + + /* Decode register details from CSR number */ + virt = false; + switch (csrno) { + case CSR_MIREG: + iprio = env->miprio; + isel = env->miselect; + priv = PRV_M; + break; + case CSR_SIREG: + iprio = env->siprio; + isel = env->siselect; + priv = PRV_S; + break; + case CSR_VSIREG: + iprio = env->hviprio; + isel = env->vsiselect; + priv = PRV_S; + virt = true; + break; + default: + goto done; + }; + + /* Find the selected guest interrupt file */ + vgein = (virt) ? get_field(env->hstatus, HSTATUS_VGEIN) : 0; + + if (ISELECT_IPRIO0 <= isel && isel <= ISELECT_IPRIO15) { + /* Local interrupt priority registers not available for VS-mode */ + if (!virt) { + ret = rmw_iprio(riscv_cpu_mxl_bits(env), + isel, iprio, val, new_val, wr_mask, + (priv == PRV_M) ? IRQ_M_EXT : IRQ_S_EXT); + } + } else if (ISELECT_IMSIC_FIRST <= isel && isel <= ISELECT_IMSIC_LAST) { + /* IMSIC registers only available when machine implements it. */ + if (env->aia_ireg_rmw_fn[priv]) { + /* Selected guest interrupt file should not be zero */ + if (virt && (!vgein || env->geilen < vgein)) { + goto done; + } + /* Call machine specific IMSIC register emulation */ + ret = env->aia_ireg_rmw_fn[priv](env->aia_ireg_rmw_fn_arg[priv], + AIA_MAKE_IREG(isel, priv, virt, vgein, + riscv_cpu_mxl_bits(env)), + val, new_val, wr_mask); + } + } + +done: + if (ret) { + return (riscv_cpu_virt_enabled(env) && virt) ? + RISCV_EXCP_VIRT_INSTRUCTION_FAULT : RISCV_EXCP_ILLEGAL_INST; + } + return RISCV_EXCP_NONE; +} + static RISCVException read_mtvec(CPURISCVState *env, int csrno, target_ulong *val) { @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_MTVAL] = { "mtval", any, read_mtval, write_mtval }, [CSR_MIP] = { "mip", any, NULL, NULL, rmw_mip }, + /* Machine-Level Window to Indirectly Accessed Registers (AIA) */ + [CSR_MISELECT] = { "miselect", aia_any, NULL, NULL, rmw_xiselect }, + [CSR_MIREG] = { "mireg", aia_any, NULL, NULL, rmw_xireg }, + /* Machine-Level Interrupts (AIA) */ [CSR_MTOPI] = { "mtopi", aia_any, read_mtopi }, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { /* Supervisor Protection and Translation */ [CSR_SATP] = { "satp", smode, read_satp, write_satp }, + /* Supervisor-Level Window to Indirectly Accessed Registers (AIA) */ + [CSR_SISELECT] = { "siselect", aia_smode, NULL, NULL, rmw_xiselect }, + [CSR_SIREG] = { "sireg", aia_smode, NULL, NULL, rmw_xireg }, + /* Supervisor-Level Interrupts (AIA) */ [CSR_STOPI] = { "stopi", aia_smode, read_stopi }, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { [CSR_HVIPRIO1] = { "hviprio1", aia_hmode, read_hviprio1, write_hviprio1 }, [CSR_HVIPRIO2] = { "hviprio2", aia_hmode, read_hviprio2, write_hviprio2 }, + /* + * VS-Level Window to Indirectly Accessed Registers (H-extension with AIA) + */ + [CSR_VSISELECT] = { "vsiselect", aia_hmode, NULL, NULL, rmw_xiselect }, + [CSR_VSIREG] = { "vsireg", aia_hmode, NULL, NULL, rmw_xireg }, + /* VS-Level Interrupts (H-extension with AIA) */ [CSR_VSTOPI] = { "vstopi", aia_hmode, read_vstopi }, 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 const VMStateDescription vmstate_hyper = { VMSTATE_UINTTL(env.vscause, RISCVCPU), VMSTATE_UINTTL(env.vstval, RISCVCPU), VMSTATE_UINTTL(env.vsatp, RISCVCPU), + VMSTATE_UINTTL(env.vsiselect, RISCVCPU), VMSTATE_UINTTL(env.mtval2, RISCVCPU), VMSTATE_UINTTL(env.mtinst, RISCVCPU), @@ -XXX,XX +XXX,XX @@ const VMStateDescription vmstate_riscv_cpu = { VMSTATE_UINTTL(env.mepc, RISCVCPU), VMSTATE_UINTTL(env.mcause, RISCVCPU), VMSTATE_UINTTL(env.mtval, RISCVCPU), + VMSTATE_UINTTL(env.miselect, RISCVCPU), + VMSTATE_UINTTL(env.siselect, RISCVCPU), VMSTATE_UINTTL(env.scounteren, RISCVCPU), VMSTATE_UINTTL(env.mcounteren, RISCVCPU), VMSTATE_UINTTL(env.sscratch, RISCVCPU), -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> The AIA specification defines IMSIC interface CSRs for easy access to the per-HART IMSIC registers without using indirect xiselect and xireg CSRs. This patch implements the AIA IMSIC interface CSRs. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-16-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/csr.c | 203 +++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 203 insertions(+) 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 int aia_xlate_vs_csrno(CPURISCVState *env, int csrno) return CSR_VSISELECT; case CSR_SIREG: return CSR_VSIREG; + case CSR_SSETEIPNUM: + return CSR_VSSETEIPNUM; + case CSR_SCLREIPNUM: + return CSR_VSCLREIPNUM; + case CSR_SSETEIENUM: + return CSR_VSSETEIENUM; + case CSR_SCLREIENUM: + return CSR_VSCLREIENUM; + case CSR_STOPEI: + return CSR_VSTOPEI; default: return csrno; }; @@ -XXX,XX +XXX,XX @@ done: return RISCV_EXCP_NONE; } +static int rmw_xsetclreinum(CPURISCVState *env, int csrno, target_ulong *val, + target_ulong new_val, target_ulong wr_mask) +{ + int ret = -EINVAL; + bool set, pend, virt; + target_ulong priv, isel, vgein, xlen, nval, wmask; + + /* Translate CSR number for VS-mode */ + csrno = aia_xlate_vs_csrno(env, csrno); + + /* Decode register details from CSR number */ + virt = set = pend = false; + switch (csrno) { + case CSR_MSETEIPNUM: + priv = PRV_M; + set = true; + pend = true; + break; + case CSR_MCLREIPNUM: + priv = PRV_M; + pend = true; + break; + case CSR_MSETEIENUM: + priv = PRV_M; + set = true; + break; + case CSR_MCLREIENUM: + priv = PRV_M; + break; + case CSR_SSETEIPNUM: + priv = PRV_S; + set = true; + pend = true; + break; + case CSR_SCLREIPNUM: + priv = PRV_S; + pend = true; + break; + case CSR_SSETEIENUM: + priv = PRV_S; + set = true; + break; + case CSR_SCLREIENUM: + priv = PRV_S; + break; + case CSR_VSSETEIPNUM: + priv = PRV_S; + virt = true; + set = true; + pend = true; + break; + case CSR_VSCLREIPNUM: + priv = PRV_S; + virt = true; + pend = true; + break; + case CSR_VSSETEIENUM: + priv = PRV_S; + virt = true; + set = true; + break; + case CSR_VSCLREIENUM: + priv = PRV_S; + virt = true; + break; + default: + goto done; + }; + + /* IMSIC CSRs only available when machine implements IMSIC. */ + if (!env->aia_ireg_rmw_fn[priv]) { + goto done; + } + + /* Find the selected guest interrupt file */ + vgein = (virt) ? get_field(env->hstatus, HSTATUS_VGEIN) : 0; + + /* Selected guest interrupt file should be valid */ + if (virt && (!vgein || env->geilen < vgein)) { + goto done; + } + + /* Set/Clear CSRs always read zero */ + if (val) { + *val = 0; + } + + if (wr_mask) { + /* Get interrupt number */ + new_val &= wr_mask; + + /* Find target interrupt pending/enable register */ + xlen = riscv_cpu_mxl_bits(env); + isel = (new_val / xlen); + isel *= (xlen / IMSIC_EIPx_BITS); + isel += (pend) ? ISELECT_IMSIC_EIP0 : ISELECT_IMSIC_EIE0; + + /* Find the interrupt bit to be set/clear */ + wmask = ((target_ulong)1) << (new_val % xlen); + nval = (set) ? wmask : 0; + + /* Call machine specific IMSIC register emulation */ + ret = env->aia_ireg_rmw_fn[priv](env->aia_ireg_rmw_fn_arg[priv], + AIA_MAKE_IREG(isel, priv, virt, + vgein, xlen), + NULL, nval, wmask); + } else { + ret = 0; + } + +done: + if (ret) { + return (riscv_cpu_virt_enabled(env) && virt) ? + RISCV_EXCP_VIRT_INSTRUCTION_FAULT : RISCV_EXCP_ILLEGAL_INST; + } + return RISCV_EXCP_NONE; +} + +static int rmw_xtopei(CPURISCVState *env, int csrno, target_ulong *val, + target_ulong new_val, target_ulong wr_mask) +{ + bool virt; + int ret = -EINVAL; + target_ulong priv, vgein; + + /* Translate CSR number for VS-mode */ + csrno = aia_xlate_vs_csrno(env, csrno); + + /* Decode register details from CSR number */ + virt = false; + switch (csrno) { + case CSR_MTOPEI: + priv = PRV_M; + break; + case CSR_STOPEI: + priv = PRV_S; + break; + case CSR_VSTOPEI: + priv = PRV_S; + virt = true; + break; + default: + goto done; + }; + + /* IMSIC CSRs only available when machine implements IMSIC. */ + if (!env->aia_ireg_rmw_fn[priv]) { + goto done; + } + + /* Find the selected guest interrupt file */ + vgein = (virt) ? get_field(env->hstatus, HSTATUS_VGEIN) : 0; + + /* Selected guest interrupt file should be valid */ + if (virt && (!vgein || env->geilen < vgein)) { + goto done; + } + + /* Call machine specific IMSIC register emulation for TOPEI */ + ret = env->aia_ireg_rmw_fn[priv](env->aia_ireg_rmw_fn_arg[priv], + AIA_MAKE_IREG(ISELECT_IMSIC_TOPEI, priv, virt, vgein, + riscv_cpu_mxl_bits(env)), + val, new_val, wr_mask); + +done: + if (ret) { + return (riscv_cpu_virt_enabled(env) && virt) ? + RISCV_EXCP_VIRT_INSTRUCTION_FAULT : RISCV_EXCP_ILLEGAL_INST; + } + return RISCV_EXCP_NONE; +} + static RISCVException read_mtvec(CPURISCVState *env, int csrno, target_ulong *val) { @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { /* Machine-Level Interrupts (AIA) */ [CSR_MTOPI] = { "mtopi", aia_any, read_mtopi }, + /* Machine-Level IMSIC Interface (AIA) */ + [CSR_MSETEIPNUM] = { "mseteipnum", aia_any, NULL, NULL, rmw_xsetclreinum }, + [CSR_MCLREIPNUM] = { "mclreipnum", aia_any, NULL, NULL, rmw_xsetclreinum }, + [CSR_MSETEIENUM] = { "mseteienum", aia_any, NULL, NULL, rmw_xsetclreinum }, + [CSR_MCLREIENUM] = { "mclreienum", aia_any, NULL, NULL, rmw_xsetclreinum }, + [CSR_MTOPEI] = { "mtopei", aia_any, NULL, NULL, rmw_xtopei }, + /* Virtual Interrupts for Supervisor Level (AIA) */ [CSR_MVIEN] = { "mvien", aia_any, read_zero, write_ignore }, [CSR_MVIP] = { "mvip", aia_any, read_zero, write_ignore }, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { /* Supervisor-Level Interrupts (AIA) */ [CSR_STOPI] = { "stopi", aia_smode, read_stopi }, + /* Supervisor-Level IMSIC Interface (AIA) */ + [CSR_SSETEIPNUM] = { "sseteipnum", aia_smode, NULL, NULL, rmw_xsetclreinum }, + [CSR_SCLREIPNUM] = { "sclreipnum", aia_smode, NULL, NULL, rmw_xsetclreinum }, + [CSR_SSETEIENUM] = { "sseteienum", aia_smode, NULL, NULL, rmw_xsetclreinum }, + [CSR_SCLREIENUM] = { "sclreienum", aia_smode, NULL, NULL, rmw_xsetclreinum }, + [CSR_STOPEI] = { "stopei", aia_smode, NULL, NULL, rmw_xtopei }, + /* Supervisor-Level High-Half CSRs (AIA) */ [CSR_SIEH] = { "sieh", aia_smode32, NULL, NULL, rmw_sieh }, [CSR_SIPH] = { "siph", aia_smode32, NULL, NULL, rmw_siph }, @@ -XXX,XX +XXX,XX @@ riscv_csr_operations csr_ops[CSR_TABLE_SIZE] = { /* VS-Level Interrupts (H-extension with AIA) */ [CSR_VSTOPI] = { "vstopi", aia_hmode, read_vstopi }, + /* VS-Level IMSIC Interface (H-extension with AIA) */ + [CSR_VSSETEIPNUM] = { "vsseteipnum", aia_hmode, NULL, NULL, rmw_xsetclreinum }, + [CSR_VSCLREIPNUM] = { "vsclreipnum", aia_hmode, NULL, NULL, rmw_xsetclreinum }, + [CSR_VSSETEIENUM] = { "vsseteienum", aia_hmode, NULL, NULL, rmw_xsetclreinum }, + [CSR_VSCLREIENUM] = { "vsclreienum", aia_hmode, NULL, NULL, rmw_xsetclreinum }, + [CSR_VSTOPEI] = { "vstopei", aia_hmode, NULL, NULL, rmw_xtopei }, + /* Hypervisor and VS-Level High-Half CSRs (H-extension with AIA) */ [CSR_HIDELEGH] = { "hidelegh", aia_hmode32, NULL, NULL, rmw_hidelegh }, [CSR_HVIENH] = { "hvienh", aia_hmode32, read_zero, write_ignore }, -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> We should use the AIA INTC compatible string in the CPU INTC DT nodes when the CPUs support AIA feature. This will allow Linux INTC driver to use AIA local interrupt CSRs. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-17-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/riscv/virt.c | 13 +++++++++++-- 1 file changed, 11 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_cpus(RISCVVirtState *s, int socket, qemu_fdt_add_subnode(mc->fdt, intc_name); qemu_fdt_setprop_cell(mc->fdt, intc_name, "phandle", intc_phandles[cpu]); - qemu_fdt_setprop_string(mc->fdt, intc_name, "compatible", - "riscv,cpu-intc"); + if (riscv_feature(&s->soc[socket].harts[cpu].env, + RISCV_FEATURE_AIA)) { + static const char * const compat[2] = { + "riscv,cpu-intc-aia", "riscv,cpu-intc" + }; + qemu_fdt_setprop_string_array(mc->fdt, intc_name, "compatible", + (char **)&compat, ARRAY_SIZE(compat)); + } else { + qemu_fdt_setprop_string(mc->fdt, intc_name, "compatible", + "riscv,cpu-intc"); + } qemu_fdt_setprop(mc->fdt, intc_name, "interrupt-controller", NULL, 0); qemu_fdt_setprop_cell(mc->fdt, intc_name, "#interrupt-cells", 1); -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> We add "x-aia" command-line option for RISC-V HART using which allows users to force enable CPU AIA CSRs without changing the interrupt controller available in RISC-V machine. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-18-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 1 + target/riscv/cpu.c | 5 +++++ 2 files changed, 6 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 RISCVCPUConfig { bool mmu; bool pmp; bool epmp; + bool aia; uint64_t resetvec; }; 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_realize(DeviceState *dev, Error **errp) } } + if (cpu->cfg.aia) { + riscv_set_feature(env, RISCV_FEATURE_AIA); + } + set_resetvec(env, cpu->cfg.resetvec); /* Validate that MISA_MXL is set properly. */ @@ -XXX,XX +XXX,XX @@ static Property riscv_cpu_properties[] = { DEFINE_PROP_BOOL("x-j", RISCVCPU, cfg.ext_j, false), /* ePMP 0.9.3 */ DEFINE_PROP_BOOL("x-epmp", RISCVCPU, cfg.epmp, false), + DEFINE_PROP_BOOL("x-aia", RISCVCPU, cfg.aia, false), DEFINE_PROP_UINT64("resetvec", RISCVCPU, cfg.resetvec, DEFAULT_RSTVEC), DEFINE_PROP_END_OF_LIST(), -- 2.34.1 From: Anup Patel <anup.patel@wdc.com> The RISC-V AIA (Advanced Interrupt Architecture) defines a new interrupt controller for wired interrupts called APLIC (Advanced Platform Level Interrupt Controller). The APLIC is capabable of forwarding wired interupts to RISC-V HARTs directly or as MSIs (Message Signaled Interupts). This patch adds device emulation for RISC-V AIA APLIC. Signed-off-by: Anup Patel <anup.patel@wdc.com> Signed-off-by: Anup Patel <anup@brainfault.org> Reviewed-by: Frank Chang <frank.chang@sifive.com> Message-id: 20220204174700.534953-19-anup@brainfault.org Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- include/hw/intc/riscv_aplic.h | 79 +++ hw/intc/riscv_aplic.c | 978 ++++++++++++++++++++++++++++++++++ hw/intc/Kconfig | 3 + hw/intc/meson.build | 1 + 4 files changed, 1061 insertions(+) create mode 100644 include/hw/intc/riscv_aplic.h create mode 100644 hw/intc/riscv_aplic.c diff --git a/include/hw/intc/riscv_aplic.h b/include/hw/intc/riscv_aplic.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/intc/riscv_aplic.h @@ -XXX,XX +XXX,XX @@ +/* + * RISC-V APLIC (Advanced Platform Level Interrupt Controller) interface + * + * Copyright (c) 2021 Western Digital Corporation or its affiliates. + * + * 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/>. + */ + +#ifndef HW_RISCV_APLIC_H +#define HW_RISCV_APLIC_H + +#include "hw/sysbus.h" +#include "qom/object.h" + +#define TYPE_RISCV_APLIC "riscv.aplic" + +typedef struct RISCVAPLICState RISCVAPLICState; +DECLARE_INSTANCE_CHECKER(RISCVAPLICState, RISCV_APLIC, TYPE_RISCV_APLIC) + +#define APLIC_MIN_SIZE 0x4000 +#define APLIC_SIZE_ALIGN(__x) (((__x) + (APLIC_MIN_SIZE - 1)) & \ + ~(APLIC_MIN_SIZE - 1)) +#define APLIC_SIZE(__num_harts) (APLIC_MIN_SIZE + \ + APLIC_SIZE_ALIGN(32 * (__num_harts))) + +struct RISCVAPLICState { + /*< private >*/ + SysBusDevice parent_obj; + qemu_irq *external_irqs; + + /*< public >*/ + MemoryRegion mmio; + uint32_t bitfield_words; + uint32_t domaincfg; + uint32_t mmsicfgaddr; + uint32_t mmsicfgaddrH; + uint32_t smsicfgaddr; + uint32_t smsicfgaddrH; + uint32_t genmsi; + uint32_t *sourcecfg; + uint32_t *state; + uint32_t *target; + uint32_t *idelivery; + uint32_t *iforce; + uint32_t *ithreshold; + + /* topology */ +#define QEMU_APLIC_MAX_CHILDREN 16 + struct RISCVAPLICState *parent; + struct RISCVAPLICState *children[QEMU_APLIC_MAX_CHILDREN]; + uint16_t num_children; + + /* config */ + uint32_t aperture_size; + uint32_t hartid_base; + uint32_t num_harts; + uint32_t iprio_mask; + uint32_t num_irqs; + bool msimode; + bool mmode; +}; + +void riscv_aplic_add_child(DeviceState *parent, DeviceState *child); + +DeviceState *riscv_aplic_create(hwaddr addr, hwaddr size, + uint32_t hartid_base, uint32_t num_harts, uint32_t num_sources, + uint32_t iprio_bits, bool msimode, bool mmode, DeviceState *parent); + +#endif diff --git a/hw/intc/riscv_aplic.c b/hw/intc/riscv_aplic.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/intc/riscv_aplic.c @@ -XXX,XX +XXX,XX @@ +/* + * RISC-V APLIC (Advanced Platform Level Interrupt Controller) + * + * Copyright (c) 2021 Western Digital Corporation or its affiliates. + * + * 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 "qemu/log.h" +#include "qemu/module.h" +#include "qemu/error-report.h" +#include "qemu/bswap.h" +#include "exec/address-spaces.h" +#include "hw/sysbus.h" +#include "hw/pci/msi.h" +#include "hw/boards.h" +#include "hw/qdev-properties.h" +#include "hw/intc/riscv_aplic.h" +#include "hw/irq.h" +#include "target/riscv/cpu.h" +#include "sysemu/sysemu.h" +#include "migration/vmstate.h" + +#define APLIC_MAX_IDC (1UL << 14) +#define APLIC_MAX_SOURCE 1024 +#define APLIC_MIN_IPRIO_BITS 1 +#define APLIC_MAX_IPRIO_BITS 8 +#define APLIC_MAX_CHILDREN 1024 + +#define APLIC_DOMAINCFG 0x0000 +#define APLIC_DOMAINCFG_RDONLY 0x80000000 +#define APLIC_DOMAINCFG_IE (1 << 8) +#define APLIC_DOMAINCFG_DM (1 << 2) +#define APLIC_DOMAINCFG_BE (1 << 0) + +#define APLIC_SOURCECFG_BASE 0x0004 +#define APLIC_SOURCECFG_D (1 << 10) +#define APLIC_SOURCECFG_CHILDIDX_MASK 0x000003ff +#define APLIC_SOURCECFG_SM_MASK 0x00000007 +#define APLIC_SOURCECFG_SM_INACTIVE 0x0 +#define APLIC_SOURCECFG_SM_DETACH 0x1 +#define APLIC_SOURCECFG_SM_EDGE_RISE 0x4 +#define APLIC_SOURCECFG_SM_EDGE_FALL 0x5 +#define APLIC_SOURCECFG_SM_LEVEL_HIGH 0x6 +#define APLIC_SOURCECFG_SM_LEVEL_LOW 0x7 + +#define APLIC_MMSICFGADDR 0x1bc0 +#define APLIC_MMSICFGADDRH 0x1bc4 +#define APLIC_SMSICFGADDR 0x1bc8 +#define APLIC_SMSICFGADDRH 0x1bcc + +#define APLIC_xMSICFGADDRH_L (1UL << 31) +#define APLIC_xMSICFGADDRH_HHXS_MASK 0x1f +#define APLIC_xMSICFGADDRH_HHXS_SHIFT 24 +#define APLIC_xMSICFGADDRH_LHXS_MASK 0x7 +#define APLIC_xMSICFGADDRH_LHXS_SHIFT 20 +#define APLIC_xMSICFGADDRH_HHXW_MASK 0x7 +#define APLIC_xMSICFGADDRH_HHXW_SHIFT 16 +#define APLIC_xMSICFGADDRH_LHXW_MASK 0xf +#define APLIC_xMSICFGADDRH_LHXW_SHIFT 12 +#define APLIC_xMSICFGADDRH_BAPPN_MASK 0xfff + +#define APLIC_xMSICFGADDR_PPN_SHIFT 12 + +#define APLIC_xMSICFGADDR_PPN_HART(__lhxs) \ + ((1UL << (__lhxs)) - 1) + +#define APLIC_xMSICFGADDR_PPN_LHX_MASK(__lhxw) \ + ((1UL << (__lhxw)) - 1) +#define APLIC_xMSICFGADDR_PPN_LHX_SHIFT(__lhxs) \ + ((__lhxs)) +#define APLIC_xMSICFGADDR_PPN_LHX(__lhxw, __lhxs) \ + (APLIC_xMSICFGADDR_PPN_LHX_MASK(__lhxw) << \ + APLIC_xMSICFGADDR_PPN_LHX_SHIFT(__lhxs)) + +#define APLIC_xMSICFGADDR_PPN_HHX_MASK(__hhxw) \ + ((1UL << (__hhxw)) - 1) +#define APLIC_xMSICFGADDR_PPN_HHX_SHIFT(__hhxs) \ + ((__hhxs) + APLIC_xMSICFGADDR_PPN_SHIFT) +#define APLIC_xMSICFGADDR_PPN_HHX(__hhxw, __hhxs) \ + (APLIC_xMSICFGADDR_PPN_HHX_MASK(__hhxw) << \ + APLIC_xMSICFGADDR_PPN_HHX_SHIFT(__hhxs)) + +#define APLIC_xMSICFGADDRH_VALID_MASK \ + (APLIC_xMSICFGADDRH_L | \ + (APLIC_xMSICFGADDRH_HHXS_MASK << APLIC_xMSICFGADDRH_HHXS_SHIFT) | \ + (APLIC_xMSICFGADDRH_LHXS_MASK << APLIC_xMSICFGADDRH_LHXS_SHIFT) | \ + (APLIC_xMSICFGADDRH_HHXW_MASK << APLIC_xMSICFGADDRH_HHXW_SHIFT) | \ + (APLIC_xMSICFGADDRH_LHXW_MASK << APLIC_xMSICFGADDRH_LHXW_SHIFT) | \ + APLIC_xMSICFGADDRH_BAPPN_MASK) + +#define APLIC_SETIP_BASE 0x1c00 +#define APLIC_SETIPNUM 0x1cdc + +#define APLIC_CLRIP_BASE 0x1d00 +#define APLIC_CLRIPNUM 0x1ddc + +#define APLIC_SETIE_BASE 0x1e00 +#define APLIC_SETIENUM 0x1edc + +#define APLIC_CLRIE_BASE 0x1f00 +#define APLIC_CLRIENUM 0x1fdc + +#define APLIC_SETIPNUM_LE 0x2000 +#define APLIC_SETIPNUM_BE 0x2004 + +#define APLIC_ISTATE_PENDING (1U << 0) +#define APLIC_ISTATE_ENABLED (1U << 1) +#define APLIC_ISTATE_ENPEND (APLIC_ISTATE_ENABLED | \ + APLIC_ISTATE_PENDING) +#define APLIC_ISTATE_INPUT (1U << 8) + +#define APLIC_GENMSI 0x3000 + +#define APLIC_TARGET_BASE 0x3004 +#define APLIC_TARGET_HART_IDX_SHIFT 18 +#define APLIC_TARGET_HART_IDX_MASK 0x3fff +#define APLIC_TARGET_GUEST_IDX_SHIFT 12 +#define APLIC_TARGET_GUEST_IDX_MASK 0x3f +#define APLIC_TARGET_IPRIO_MASK 0xff +#define APLIC_TARGET_EIID_MASK 0x7ff + +#define APLIC_IDC_BASE 0x4000 +#define APLIC_IDC_SIZE 32 + +#define APLIC_IDC_IDELIVERY 0x00 + +#define APLIC_IDC_IFORCE 0x04 + +#define APLIC_IDC_ITHRESHOLD 0x08 + +#define APLIC_IDC_TOPI 0x18 +#define APLIC_IDC_TOPI_ID_SHIFT 16 +#define APLIC_IDC_TOPI_ID_MASK 0x3ff +#define APLIC_IDC_TOPI_PRIO_MASK 0xff + +#define APLIC_IDC_CLAIMI 0x1c + +static uint32_t riscv_aplic_read_input_word(RISCVAPLICState *aplic, + uint32_t word) +{ + uint32_t i, irq, ret = 0; + + for (i = 0; i < 32; i++) { + irq = word * 32 + i; + if (!irq || aplic->num_irqs <= irq) { + continue; + } + + ret |= ((aplic->state[irq] & APLIC_ISTATE_INPUT) ? 1 : 0) << i; + } + + return ret; +} + +static uint32_t riscv_aplic_read_pending_word(RISCVAPLICState *aplic, + uint32_t word) +{ + uint32_t i, irq, ret = 0; + + for (i = 0; i < 32; i++) { + irq = word * 32 + i; + if (!irq || aplic->num_irqs <= irq) { + continue; + } + + ret |= ((aplic->state[irq] & APLIC_ISTATE_PENDING) ? 1 : 0) << i; + } + + return ret; +} + +static void riscv_aplic_set_pending_raw(RISCVAPLICState *aplic, + uint32_t irq, bool pending) +{ + if (pending) { + aplic->state[irq] |= APLIC_ISTATE_PENDING; + } else { + aplic->state[irq] &= ~APLIC_ISTATE_PENDING; + } +} + +static void riscv_aplic_set_pending(RISCVAPLICState *aplic, + uint32_t irq, bool pending) +{ + uint32_t sourcecfg, sm; + + if ((irq <= 0) || (aplic->num_irqs <= irq)) { + return; + } + + sourcecfg = aplic->sourcecfg[irq]; + if (sourcecfg & APLIC_SOURCECFG_D) { + return; + } + + sm = sourcecfg & APLIC_SOURCECFG_SM_MASK; + if ((sm == APLIC_SOURCECFG_SM_INACTIVE) || + ((!aplic->msimode || (aplic->msimode && !pending)) && + ((sm == APLIC_SOURCECFG_SM_LEVEL_HIGH) || + (sm == APLIC_SOURCECFG_SM_LEVEL_LOW)))) { + return; + } + + riscv_aplic_set_pending_raw(aplic, irq, pending); +} + +static void riscv_aplic_set_pending_word(RISCVAPLICState *aplic, + uint32_t word, uint32_t value, + bool pending) +{ + uint32_t i, irq; + + for (i = 0; i < 32; i++) { + irq = word * 32 + i; + if (!irq || aplic->num_irqs <= irq) { + continue; + } + + if (value & (1U << i)) { + riscv_aplic_set_pending(aplic, irq, pending); + } + } +} + +static uint32_t riscv_aplic_read_enabled_word(RISCVAPLICState *aplic, + int word) +{ + uint32_t i, irq, ret = 0; + + for (i = 0; i < 32; i++) { + irq = word * 32 + i; + if (!irq || aplic->num_irqs <= irq) { + continue; + } + + ret |= ((aplic->state[irq] & APLIC_ISTATE_ENABLED) ? 1 : 0) << i; + } + + return ret; +} + +static void riscv_aplic_set_enabled_raw(RISCVAPLICState *aplic, + uint32_t irq, bool enabled) +{ + if (enabled) { + aplic->state[irq] |= APLIC_ISTATE_ENABLED; + } else { + aplic->state[irq] &= ~APLIC_ISTATE_ENABLED; + } +} + +static void riscv_aplic_set_enabled(RISCVAPLICState *aplic, + uint32_t irq, bool enabled) +{ + uint32_t sourcecfg, sm; + + if ((irq <= 0) || (aplic->num_irqs <= irq)) { + return; + } + + sourcecfg = aplic->sourcecfg[irq]; + if (sourcecfg & APLIC_SOURCECFG_D) { + return; + } + + sm = sourcecfg & APLIC_SOURCECFG_SM_MASK; + if (sm == APLIC_SOURCECFG_SM_INACTIVE) { + return; + } + + riscv_aplic_set_enabled_raw(aplic, irq, enabled); +} + +static void riscv_aplic_set_enabled_word(RISCVAPLICState *aplic, + uint32_t word, uint32_t value, + bool enabled) +{ + uint32_t i, irq; + + for (i = 0; i < 32; i++) { + irq = word * 32 + i; + if (!irq || aplic->num_irqs <= irq) { + continue; + } + + if (value & (1U << i)) { + riscv_aplic_set_enabled(aplic, irq, enabled); + } + } +} + +static void riscv_aplic_msi_send(RISCVAPLICState *aplic, + uint32_t hart_idx, uint32_t guest_idx, + uint32_t eiid) +{ + uint64_t addr; + MemTxResult result; + RISCVAPLICState *aplic_m; + uint32_t lhxs, lhxw, hhxs, hhxw, group_idx, msicfgaddr, msicfgaddrH; + + aplic_m = aplic; + while (aplic_m && !aplic_m->mmode) { + aplic_m = aplic_m->parent; + } + if (!aplic_m) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: m-level APLIC not found\n", + __func__); + return; + } + + if (aplic->mmode) { + msicfgaddr = aplic_m->mmsicfgaddr; + msicfgaddrH = aplic_m->mmsicfgaddrH; + } else { + msicfgaddr = aplic_m->smsicfgaddr; + msicfgaddrH = aplic_m->smsicfgaddrH; + } + + lhxs = (msicfgaddrH >> APLIC_xMSICFGADDRH_LHXS_SHIFT) & + APLIC_xMSICFGADDRH_LHXS_MASK; + lhxw = (msicfgaddrH >> APLIC_xMSICFGADDRH_LHXW_SHIFT) & + APLIC_xMSICFGADDRH_LHXW_MASK; + hhxs = (msicfgaddrH >> APLIC_xMSICFGADDRH_HHXS_SHIFT) & + APLIC_xMSICFGADDRH_HHXS_MASK; + hhxw = (msicfgaddrH >> APLIC_xMSICFGADDRH_HHXW_SHIFT) & + APLIC_xMSICFGADDRH_HHXW_MASK; + + group_idx = hart_idx >> lhxw; + hart_idx &= APLIC_xMSICFGADDR_PPN_LHX_MASK(lhxw); + + addr = msicfgaddr; + addr |= ((uint64_t)(msicfgaddrH & APLIC_xMSICFGADDRH_BAPPN_MASK)) << 32; + addr |= ((uint64_t)(group_idx & APLIC_xMSICFGADDR_PPN_HHX_MASK(hhxw))) << + APLIC_xMSICFGADDR_PPN_HHX_SHIFT(hhxs); + addr |= ((uint64_t)(hart_idx & APLIC_xMSICFGADDR_PPN_LHX_MASK(lhxw))) << + APLIC_xMSICFGADDR_PPN_LHX_SHIFT(lhxs); + addr |= (uint64_t)(guest_idx & APLIC_xMSICFGADDR_PPN_HART(lhxs)); + addr <<= APLIC_xMSICFGADDR_PPN_SHIFT; + + address_space_stl_le(&address_space_memory, addr, + eiid, MEMTXATTRS_UNSPECIFIED, &result); + if (result != MEMTX_OK) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: MSI write failed for " + "hart_index=%d guest_index=%d eiid=%d\n", + __func__, hart_idx, guest_idx, eiid); + } +} + +static void riscv_aplic_msi_irq_update(RISCVAPLICState *aplic, uint32_t irq) +{ + uint32_t hart_idx, guest_idx, eiid; + + if (!aplic->msimode || (aplic->num_irqs <= irq) || + !(aplic->domaincfg & APLIC_DOMAINCFG_IE)) { + return; + } + + if ((aplic->state[irq] & APLIC_ISTATE_ENPEND) != APLIC_ISTATE_ENPEND) { + return; + } + + riscv_aplic_set_pending_raw(aplic, irq, false); + + hart_idx = aplic->target[irq] >> APLIC_TARGET_HART_IDX_SHIFT; + hart_idx &= APLIC_TARGET_HART_IDX_MASK; + if (aplic->mmode) { + /* M-level APLIC ignores guest_index */ + guest_idx = 0; + } else { + guest_idx = aplic->target[irq] >> APLIC_TARGET_GUEST_IDX_SHIFT; + guest_idx &= APLIC_TARGET_GUEST_IDX_MASK; + } + eiid = aplic->target[irq] & APLIC_TARGET_EIID_MASK; + riscv_aplic_msi_send(aplic, hart_idx, guest_idx, eiid); +} + +static uint32_t riscv_aplic_idc_topi(RISCVAPLICState *aplic, uint32_t idc) +{ + uint32_t best_irq, best_iprio; + uint32_t irq, iprio, ihartidx, ithres; + + if (aplic->num_harts <= idc) { + return 0; + } + + ithres = aplic->ithreshold[idc]; + best_irq = best_iprio = UINT32_MAX; + for (irq = 1; irq < aplic->num_irqs; irq++) { + if ((aplic->state[irq] & APLIC_ISTATE_ENPEND) != + APLIC_ISTATE_ENPEND) { + continue; + } + + ihartidx = aplic->target[irq] >> APLIC_TARGET_HART_IDX_SHIFT; + ihartidx &= APLIC_TARGET_HART_IDX_MASK; + if (ihartidx != idc) { + continue; + } + + iprio = aplic->target[irq] & aplic->iprio_mask; + if (ithres && iprio >= ithres) { + continue; + } + + if (iprio < best_iprio) { + best_irq = irq; + best_iprio = iprio; + } + } + + if (best_irq < aplic->num_irqs && best_iprio <= aplic->iprio_mask) { + return (best_irq << APLIC_IDC_TOPI_ID_SHIFT) | best_iprio; + } + + return 0; +} + +static void riscv_aplic_idc_update(RISCVAPLICState *aplic, uint32_t idc) +{ + uint32_t topi; + + if (aplic->msimode || aplic->num_harts <= idc) { + return; + } + + topi = riscv_aplic_idc_topi(aplic, idc); + if ((aplic->domaincfg & APLIC_DOMAINCFG_IE) && + aplic->idelivery[idc] && + (aplic->iforce[idc] || topi)) { + qemu_irq_raise(aplic->external_irqs[idc]); + } else { + qemu_irq_lower(aplic->external_irqs[idc]); + } +} + +static uint32_t riscv_aplic_idc_claimi(RISCVAPLICState *aplic, uint32_t idc) +{ + uint32_t irq, state, sm, topi = riscv_aplic_idc_topi(aplic, idc); + + if (!topi) { + aplic->iforce[idc] = 0; + return 0; + } + + irq = (topi >> APLIC_IDC_TOPI_ID_SHIFT) & APLIC_IDC_TOPI_ID_MASK; + sm = aplic->sourcecfg[irq] & APLIC_SOURCECFG_SM_MASK; + state = aplic->state[irq]; + riscv_aplic_set_pending_raw(aplic, irq, false); + if ((sm == APLIC_SOURCECFG_SM_LEVEL_HIGH) && + (state & APLIC_ISTATE_INPUT)) { + riscv_aplic_set_pending_raw(aplic, irq, true); + } else if ((sm == APLIC_SOURCECFG_SM_LEVEL_LOW) && + !(state & APLIC_ISTATE_INPUT)) { + riscv_aplic_set_pending_raw(aplic, irq, true); + } + riscv_aplic_idc_update(aplic, idc); + + return topi; +} + +static void riscv_aplic_request(void *opaque, int irq, int level) +{ + bool update = false; + RISCVAPLICState *aplic = opaque; + uint32_t sourcecfg, childidx, state, idc; + + assert((0 < irq) && (irq < aplic->num_irqs)); + + sourcecfg = aplic->sourcecfg[irq]; + if (sourcecfg & APLIC_SOURCECFG_D) { + childidx = sourcecfg & APLIC_SOURCECFG_CHILDIDX_MASK; + if (childidx < aplic->num_children) { + riscv_aplic_request(aplic->children[childidx], irq, level); + } + return; + } + + state = aplic->state[irq]; + switch (sourcecfg & APLIC_SOURCECFG_SM_MASK) { + case APLIC_SOURCECFG_SM_EDGE_RISE: + if ((level > 0) && !(state & APLIC_ISTATE_INPUT) && + !(state & APLIC_ISTATE_PENDING)) { + riscv_aplic_set_pending_raw(aplic, irq, true); + update = true; + } + break; + case APLIC_SOURCECFG_SM_EDGE_FALL: + if ((level <= 0) && (state & APLIC_ISTATE_INPUT) && + !(state & APLIC_ISTATE_PENDING)) { + riscv_aplic_set_pending_raw(aplic, irq, true); + update = true; + } + break; + case APLIC_SOURCECFG_SM_LEVEL_HIGH: + if ((level > 0) && !(state & APLIC_ISTATE_PENDING)) { + riscv_aplic_set_pending_raw(aplic, irq, true); + update = true; + } + break; + case APLIC_SOURCECFG_SM_LEVEL_LOW: + if ((level <= 0) && !(state & APLIC_ISTATE_PENDING)) { + riscv_aplic_set_pending_raw(aplic, irq, true); + update = true; + } + break; + default: + break; + } + + if (level <= 0) { + aplic->state[irq] &= ~APLIC_ISTATE_INPUT; + } else { + aplic->state[irq] |= APLIC_ISTATE_INPUT; + } + + if (update) { + if (aplic->msimode) { + riscv_aplic_msi_irq_update(aplic, irq); + } else { + idc = aplic->target[irq] >> APLIC_TARGET_HART_IDX_SHIFT; + idc &= APLIC_TARGET_HART_IDX_MASK; + riscv_aplic_idc_update(aplic, idc); + } + } +} + +static uint64_t riscv_aplic_read(void *opaque, hwaddr addr, unsigned size) +{ + uint32_t irq, word, idc; + RISCVAPLICState *aplic = opaque; + + /* Reads must be 4 byte words */ + if ((addr & 0x3) != 0) { + goto err; + } + + if (addr == APLIC_DOMAINCFG) { + return APLIC_DOMAINCFG_RDONLY | aplic->domaincfg | + (aplic->msimode ? APLIC_DOMAINCFG_DM : 0); + } else if ((APLIC_SOURCECFG_BASE <= addr) && + (addr < (APLIC_SOURCECFG_BASE + (aplic->num_irqs - 1) * 4))) { + irq = ((addr - APLIC_SOURCECFG_BASE) >> 2) + 1; + return aplic->sourcecfg[irq]; + } else if (aplic->mmode && aplic->msimode && + (addr == APLIC_MMSICFGADDR)) { + return aplic->mmsicfgaddr; + } else if (aplic->mmode && aplic->msimode && + (addr == APLIC_MMSICFGADDRH)) { + return aplic->mmsicfgaddrH; + } else if (aplic->mmode && aplic->msimode && + (addr == APLIC_SMSICFGADDR)) { + /* + * Registers SMSICFGADDR and SMSICFGADDRH are implemented only if: + * (a) the interrupt domain is at machine level + * (b) the domain's harts implement supervisor mode + * (c) the domain has one or more child supervisor-level domains + * that support MSI delivery mode (domaincfg.DM is not read- + * only zero in at least one of the supervisor-level child + * domains). + */ + return (aplic->num_children) ? aplic->smsicfgaddr : 0; + } else if (aplic->mmode && aplic->msimode && + (addr == APLIC_SMSICFGADDRH)) { + return (aplic->num_children) ? aplic->smsicfgaddrH : 0; + } else if ((APLIC_SETIP_BASE <= addr) && + (addr < (APLIC_SETIP_BASE + aplic->bitfield_words * 4))) { + word = (addr - APLIC_SETIP_BASE) >> 2; + return riscv_aplic_read_pending_word(aplic, word); + } else if (addr == APLIC_SETIPNUM) { + return 0; + } else if ((APLIC_CLRIP_BASE <= addr) && + (addr < (APLIC_CLRIP_BASE + aplic->bitfield_words * 4))) { + word = (addr - APLIC_CLRIP_BASE) >> 2; + return riscv_aplic_read_input_word(aplic, word); + } else if (addr == APLIC_CLRIPNUM) { + return 0; + } else if ((APLIC_SETIE_BASE <= addr) && + (addr < (APLIC_SETIE_BASE + aplic->bitfield_words * 4))) { + word = (addr - APLIC_SETIE_BASE) >> 2; + return riscv_aplic_read_enabled_word(aplic, word); + } else if (addr == APLIC_SETIENUM) { + return 0; + } else if ((APLIC_CLRIE_BASE <= addr) && + (addr < (APLIC_CLRIE_BASE + aplic->bitfield_words * 4))) { + return 0; + } else if (addr == APLIC_CLRIENUM) { + return 0; + } else if (addr == APLIC_SETIPNUM_LE) { + return 0; + } else if (addr == APLIC_SETIPNUM_BE) { + return 0; + } else if (addr == APLIC_GENMSI) { + return (aplic->msimode) ? aplic->genmsi : 0; + } else if ((APLIC_TARGET_BASE <= addr) && + (addr < (APLIC_TARGET_BASE + (aplic->num_irqs - 1) * 4))) { + irq = ((addr - APLIC_TARGET_BASE) >> 2) + 1; + return aplic->target[irq]; + } else if (!aplic->msimode && (APLIC_IDC_BASE <= addr) && + (addr < (APLIC_IDC_BASE + aplic->num_harts * APLIC_IDC_SIZE))) { + idc = (addr - APLIC_IDC_BASE) / APLIC_IDC_SIZE; + switch (addr - (APLIC_IDC_BASE + idc * APLIC_IDC_SIZE)) { + case APLIC_IDC_IDELIVERY: + return aplic->idelivery[idc]; + case APLIC_IDC_IFORCE: + return aplic->iforce[idc]; + case APLIC_IDC_ITHRESHOLD: + return aplic->ithreshold[idc]; + case APLIC_IDC_TOPI: + return riscv_aplic_idc_topi(aplic, idc); + case APLIC_IDC_CLAIMI: + return riscv_aplic_idc_claimi(aplic, idc); + default: + goto err; + }; + } + +err: + qemu_log_mask(LOG_GUEST_ERROR, + "%s: Invalid register read 0x%" HWADDR_PRIx "\n", + __func__, addr); + return 0; +} + +static void riscv_aplic_write(void *opaque, hwaddr addr, uint64_t value, + unsigned size) +{ + RISCVAPLICState *aplic = opaque; + uint32_t irq, word, idc = UINT32_MAX; + + /* Writes must be 4 byte words */ + if ((addr & 0x3) != 0) { + goto err; + } + + if (addr == APLIC_DOMAINCFG) { + /* Only IE bit writeable at the moment */ + value &= APLIC_DOMAINCFG_IE; + aplic->domaincfg = value; + } else if ((APLIC_SOURCECFG_BASE <= addr) && + (addr < (APLIC_SOURCECFG_BASE + (aplic->num_irqs - 1) * 4))) { + irq = ((addr - APLIC_SOURCECFG_BASE) >> 2) + 1; + if (!aplic->num_children && (value & APLIC_SOURCECFG_D)) { + value = 0; + } + if (value & APLIC_SOURCECFG_D) { + value &= (APLIC_SOURCECFG_D | APLIC_SOURCECFG_CHILDIDX_MASK); + } else { + value &= (APLIC_SOURCECFG_D | APLIC_SOURCECFG_SM_MASK); + } + aplic->sourcecfg[irq] = value; + if ((aplic->sourcecfg[irq] & APLIC_SOURCECFG_D) || + (aplic->sourcecfg[irq] == 0)) { + riscv_aplic_set_pending_raw(aplic, irq, false); + riscv_aplic_set_enabled_raw(aplic, irq, false); + } + } else if (aplic->mmode && aplic->msimode && + (addr == APLIC_MMSICFGADDR)) { + if (!(aplic->mmsicfgaddrH & APLIC_xMSICFGADDRH_L)) { + aplic->mmsicfgaddr = value; + } + } else if (aplic->mmode && aplic->msimode && + (addr == APLIC_MMSICFGADDRH)) { + if (!(aplic->mmsicfgaddrH & APLIC_xMSICFGADDRH_L)) { + aplic->mmsicfgaddrH = value & APLIC_xMSICFGADDRH_VALID_MASK; + } + } else if (aplic->mmode && aplic->msimode && + (addr == APLIC_SMSICFGADDR)) { + /* + * Registers SMSICFGADDR and SMSICFGADDRH are implemented only if: + * (a) the interrupt domain is at machine level + * (b) the domain's harts implement supervisor mode + * (c) the domain has one or more child supervisor-level domains + * that support MSI delivery mode (domaincfg.DM is not read- + * only zero in at least one of the supervisor-level child + * domains). + */ + if (aplic->num_children && + !(aplic->smsicfgaddrH & APLIC_xMSICFGADDRH_L)) { + aplic->smsicfgaddr = value; + } + } else if (aplic->mmode && aplic->msimode && + (addr == APLIC_SMSICFGADDRH)) { + if (aplic->num_children && + !(aplic->smsicfgaddrH & APLIC_xMSICFGADDRH_L)) { + aplic->smsicfgaddrH = value & APLIC_xMSICFGADDRH_VALID_MASK; + } + } else if ((APLIC_SETIP_BASE <= addr) && + (addr < (APLIC_SETIP_BASE + aplic->bitfield_words * 4))) { + word = (addr - APLIC_SETIP_BASE) >> 2; + riscv_aplic_set_pending_word(aplic, word, value, true); + } else if (addr == APLIC_SETIPNUM) { + riscv_aplic_set_pending(aplic, value, true); + } else if ((APLIC_CLRIP_BASE <= addr) && + (addr < (APLIC_CLRIP_BASE + aplic->bitfield_words * 4))) { + word = (addr - APLIC_CLRIP_BASE) >> 2; + riscv_aplic_set_pending_word(aplic, word, value, false); + } else if (addr == APLIC_CLRIPNUM) { + riscv_aplic_set_pending(aplic, value, false); + } else if ((APLIC_SETIE_BASE <= addr) && + (addr < (APLIC_SETIE_BASE + aplic->bitfield_words * 4))) { + word = (addr - APLIC_SETIE_BASE) >> 2; + riscv_aplic_set_enabled_word(aplic, word, value, true); + } else if (addr == APLIC_SETIENUM) { + riscv_aplic_set_enabled(aplic, value, true); + } else if ((APLIC_CLRIE_BASE <= addr) && + (addr < (APLIC_CLRIE_BASE + aplic->bitfield_words * 4))) { + word = (addr - APLIC_CLRIE_BASE) >> 2; + riscv_aplic_set_enabled_word(aplic, word, value, false); + } else if (addr == APLIC_CLRIENUM) { + riscv_aplic_set_enabled(aplic, value, false); + } else if (addr == APLIC_SETIPNUM_LE) { + riscv_aplic_set_pending(aplic, value, true); + } else if (addr == APLIC_SETIPNUM_BE) { + riscv_aplic_set_pending(aplic, bswap32(value), true); + } else if (addr == APLIC_GENMSI) { + if (aplic->msimode) { + aplic->genmsi = value & ~(APLIC_TARGET_GUEST_IDX_MASK << + APLIC_TARGET_GUEST_IDX_SHIFT); + riscv_aplic_msi_send(aplic, + value >> APLIC_TARGET_HART_IDX_SHIFT, + 0, + value & APLIC_TARGET_EIID_MASK); + } + } else if ((APLIC_TARGET_BASE <= addr) && + (addr < (APLIC_TARGET_BASE + (aplic->num_irqs - 1) * 4))) { + irq = ((addr - APLIC_TARGET_BASE) >> 2) + 1; + if (aplic->msimode) { + aplic->target[irq] = value; + } else { + aplic->target[irq] = (value & ~APLIC_TARGET_IPRIO_MASK) | + ((value & aplic->iprio_mask) ? + (value & aplic->iprio_mask) : 1); + } + } else if (!aplic->msimode && (APLIC_IDC_BASE <= addr) && + (addr < (APLIC_IDC_BASE + aplic->num_harts * APLIC_IDC_SIZE))) { + idc = (addr - APLIC_IDC_BASE) / APLIC_IDC_SIZE; + switch (addr - (APLIC_IDC_BASE + idc * APLIC_IDC_SIZE)) { + case APLIC_IDC_IDELIVERY: + aplic->idelivery[idc] = value & 0x1; + break; + case APLIC_IDC_IFORCE: + aplic->iforce[idc] = value & 0x1; + break; + case APLIC_IDC_ITHRESHOLD: + aplic->ithreshold[idc] = value & aplic->iprio_mask; + break; + default: + goto err; + }; + } else { + goto err; + } + + if (aplic->msimode) { + for (irq = 1; irq < aplic->num_irqs; irq++) { + riscv_aplic_msi_irq_update(aplic, irq); + } + } else { + if (idc == UINT32_MAX) { + for (idc = 0; idc < aplic->num_harts; idc++) { + riscv_aplic_idc_update(aplic, idc); + } + } else { + riscv_aplic_idc_update(aplic, idc); + } + } + + return; + +err: + qemu_log_mask(LOG_GUEST_ERROR, + "%s: Invalid register write 0x%" HWADDR_PRIx "\n", + __func__, addr); +} + +static const MemoryRegionOps riscv_aplic_ops = { + .read = riscv_aplic_read, + .write = riscv_aplic_write, + .endianness = DEVICE_LITTLE_ENDIAN, + .valid = { + .min_access_size = 4, + .max_access_size = 4 + } +}; + +static void riscv_aplic_realize(DeviceState *dev, Error **errp) +{ + uint32_t i; + RISCVAPLICState *aplic = RISCV_APLIC(dev); + + aplic->bitfield_words = (aplic->num_irqs + 31) >> 5; + aplic->sourcecfg = g_new0(uint32_t, aplic->num_irqs); + aplic->state = g_new(uint32_t, aplic->num_irqs); + aplic->target = g_new0(uint32_t, aplic->num_irqs); + if (!aplic->msimode) { + for (i = 0; i < aplic->num_irqs; i++) { + aplic->target[i] = 1; + } + } + aplic->idelivery = g_new0(uint32_t, aplic->num_harts); + aplic->iforce = g_new0(uint32_t, aplic->num_harts); + aplic->ithreshold = g_new0(uint32_t, aplic->num_harts); + + memory_region_init_io(&aplic->mmio, OBJECT(dev), &riscv_aplic_ops, aplic, + TYPE_RISCV_APLIC, aplic->aperture_size); + sysbus_init_mmio(SYS_BUS_DEVICE(dev), &aplic->mmio); + + /* + * Only root APLICs have hardware IRQ lines. All non-root APLICs + * have IRQ lines delegated by their parent APLIC. + */ + if (!aplic->parent) { + qdev_init_gpio_in(dev, riscv_aplic_request, aplic->num_irqs); + } + + /* Create output IRQ lines for non-MSI mode */ + if (!aplic->msimode) { + aplic->external_irqs = g_malloc(sizeof(qemu_irq) * aplic->num_harts); + qdev_init_gpio_out(dev, aplic->external_irqs, aplic->num_harts); + + /* Claim the CPU interrupt to be triggered by this APLIC */ + for (i = 0; i < aplic->num_harts; i++) { + RISCVCPU *cpu = RISCV_CPU(qemu_get_cpu(aplic->hartid_base + i)); + if (riscv_cpu_claim_interrupts(cpu, + (aplic->mmode) ? MIP_MEIP : MIP_SEIP) < 0) { + error_report("%s already claimed", + (aplic->mmode) ? "MEIP" : "SEIP"); + exit(1); + } + } + } + + msi_nonbroken = true; +} + +static Property riscv_aplic_properties[] = { + DEFINE_PROP_UINT32("aperture-size", RISCVAPLICState, aperture_size, 0), + DEFINE_PROP_UINT32("hartid-base", RISCVAPLICState, hartid_base, 0), + DEFINE_PROP_UINT32("num-harts", RISCVAPLICState, num_harts, 0), + DEFINE_PROP_UINT32("iprio-mask", RISCVAPLICState, iprio_mask, 0), + DEFINE_PROP_UINT32("num-irqs", RISCVAPLICState, num_irqs, 0), + DEFINE_PROP_BOOL("msimode", RISCVAPLICState, msimode, 0), + DEFINE_PROP_BOOL("mmode", RISCVAPLICState, mmode, 0), + DEFINE_PROP_END_OF_LIST(), +}; + +static const VMStateDescription vmstate_riscv_aplic = { + .name = "riscv_aplic", + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32(domaincfg, RISCVAPLICState), + VMSTATE_UINT32(mmsicfgaddr, RISCVAPLICState), + VMSTATE_UINT32(mmsicfgaddrH, RISCVAPLICState), + VMSTATE_UINT32(smsicfgaddr, RISCVAPLICState), + VMSTATE_UINT32(smsicfgaddrH, RISCVAPLICState), + VMSTATE_UINT32(genmsi, RISCVAPLICState), + VMSTATE_VARRAY_UINT32(sourcecfg, RISCVAPLICState, + num_irqs, 0, + vmstate_info_uint32, uint32_t), + VMSTATE_VARRAY_UINT32(state, RISCVAPLICState, + num_irqs, 0, + vmstate_info_uint32, uint32_t), + VMSTATE_VARRAY_UINT32(target, RISCVAPLICState, + num_irqs, 0, + vmstate_info_uint32, uint32_t), + VMSTATE_VARRAY_UINT32(idelivery, RISCVAPLICState, + num_harts, 0, + vmstate_info_uint32, uint32_t), + VMSTATE_VARRAY_UINT32(iforce, RISCVAPLICState, + num_harts, 0, + vmstate_info_uint32, uint32_t), + VMSTATE_VARRAY_UINT32(ithreshold, RISCVAPLICState, + num_harts, 0, + vmstate_info_uint32, uint32_t), + VMSTATE_END_OF_LIST() + } +}; + +static void riscv_aplic_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + device_class_set_props(dc, riscv_aplic_properties); + dc->realize = riscv_aplic_realize; + dc->vmsd = &vmstate_riscv_aplic; +} + +static const TypeInfo riscv_aplic_info = { + .name = TYPE_RISCV_APLIC, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(RISCVAPLICState), + .class_init = riscv_aplic_class_init, +}; + +static void riscv_aplic_register_types(void) +{ + type_register_static(&riscv_aplic_info); +} + +type_init(riscv_aplic_register_types) + +/* + * Add a APLIC device to another APLIC device as child for + * interrupt delegation. + */ +void riscv_aplic_add_child(DeviceState *parent, DeviceState *child) +{ + RISCVAPLICState *caplic, *paplic; + + assert(parent && child); + caplic = RISCV_APLIC(child); + paplic = RISCV_APLIC(parent); + + assert(paplic->num_irqs == caplic->num_irqs); + assert(paplic->num_children <= QEMU_APLIC_MAX_CHILDREN); + + caplic->parent = paplic; + paplic->children[paplic->num_children] = caplic; + paplic->num_children++; +} + +/* + * Create APLIC device. + */ +DeviceState *riscv_aplic_create(hwaddr addr, hwaddr size, + uint32_t hartid_base, uint32_t num_harts, uint32_t num_sources, + uint32_t iprio_bits, bool msimode, bool mmode, DeviceState *parent) +{ + DeviceState *dev = qdev_new(TYPE_RISCV_APLIC); + uint32_t i; + + assert(num_harts < APLIC_MAX_IDC); + assert((APLIC_IDC_BASE + (num_harts * APLIC_IDC_SIZE)) <= size); + assert(num_sources < APLIC_MAX_SOURCE); + assert(APLIC_MIN_IPRIO_BITS <= iprio_bits); + assert(iprio_bits <= APLIC_MAX_IPRIO_BITS); + + qdev_prop_set_uint32(dev, "aperture-size", size); + qdev_prop_set_uint32(dev, "hartid-base", hartid_base); + qdev_prop_set_uint32(dev, "num-harts", num_harts); + qdev_prop_set_uint32(dev, "iprio-mask", ((1U << iprio_bits) - 1)); + qdev_prop_set_uint32(dev, "num-irqs", num_sources + 1); + qdev_prop_set_bit(dev, "msimode", msimode); + qdev_prop_set_bit(dev, "mmode", mmode); + + sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(dev), 0, addr); + + if (parent) { + riscv_aplic_add_child(parent, dev); + } + + if (!msimode) { + for (i = 0; i < num_harts; i++) { + CPUState *cpu = qemu_get_cpu(hartid_base + i); + + qdev_connect_gpio_out_named(dev, NULL, i, + qdev_get_gpio_in(DEVICE(cpu), + (mmode) ? IRQ_M_EXT : IRQ_S_EXT)); + } + } + + return dev; +} diff --git a/hw/intc/Kconfig b/hw/intc/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/Kconfig +++ b/hw/intc/Kconfig @@ -XXX,XX +XXX,XX @@ config LOONGSON_LIOINTC config RISCV_ACLINT bool +config RISCV_APLIC + bool + config SIFIVE_PLIC bool diff --git a/hw/intc/meson.build b/hw/intc/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/meson.build +++ b/hw/intc/meson.build @@ -XXX,XX +XXX,XX @@ specific_ss.add(when: 'CONFIG_S390_FLIC', if_true: files('s390_flic.c')) specific_ss.add(when: 'CONFIG_S390_FLIC_KVM', if_true: files('s390_flic_kvm.c')) specific_ss.add(when: 'CONFIG_SH_INTC', if_true: files('sh_intc.c')) specific_ss.add(when: 'CONFIG_RISCV_ACLINT', if_true: files('riscv_aclint.c')) +specific_ss.add(when: 'CONFIG_RISCV_APLIC', if_true: files('riscv_aplic.c')) specific_ss.add(when: 'CONFIG_SIFIVE_PLIC', if_true: files('sifive_plic.c')) specific_ss.add(when: 'CONFIG_XICS', if_true: files('xics.c')) specific_ss.add(when: ['CONFIG_KVM', 'CONFIG_XICS'], -- 2.34.1 From: Guo Ren <ren_guo@c-sky.com> Highest bits of PTE has been used for svpbmt, ref: [1], [2], so we need to ignore them. They cannot be a part of ppn. 1: The RISC-V Instruction Set Manual, Volume II: Privileged Architecture 4.4 Sv39: Page-Based 39-bit Virtual-Memory System 4.5 Sv48: Page-Based 48-bit Virtual-Memory System 2: https://github.com/riscv/virtual-memory/blob/main/specs/663-Svpbmt-diff.pdf Signed-off-by: Guo Ren <ren_guo@c-sky.com> Reviewed-by: Liu Zhiwei <zhiwei_liu@c-sky.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Cc: Bin Meng <bmeng.cn@gmail.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-Id: <20220204022658.18097-2-liweiwei@iscas.ac.cn> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 15 +++++++++++++++ target/riscv/cpu_bits.h | 3 +++ target/riscv/cpu_helper.c | 13 ++++++++++++- 3 files changed, 30 insertions(+), 1 deletion(-) 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 RISCVCPUConfig { bool ext_counters; bool ext_ifencei; bool ext_icsr; + bool ext_svnapot; + bool ext_svpbmt; bool ext_zfh; bool ext_zfhmin; bool ext_zve32f; @@ -XXX,XX +XXX,XX @@ static inline int riscv_cpu_xlen(CPURISCVState *env) return 16 << env->xl; } +#ifdef TARGET_RISCV32 +#define riscv_cpu_sxl(env) ((void)(env), MXL_RV32) +#else +static inline RISCVMXL riscv_cpu_sxl(CPURISCVState *env) +{ +#ifdef CONFIG_USER_ONLY + return env->misa_mxl; +#else + return get_field(env->mstatus, MSTATUS64_SXL); +#endif +} +#endif + /* * Encode LMUL to lmul as follows: * LMUL vlmul lmul 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 { /* Page table PPN shift amount */ #define PTE_PPN_SHIFT 10 +/* Page table PPN mask */ +#define PTE_PPN_MASK 0x3FFFFFFFFFFC00ULL + /* Leaf page shift amount */ #define PGSHIFT 12 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, MemTxAttrs attrs = MEMTXATTRS_UNSPECIFIED; int mode = mmu_idx & TB_FLAGS_PRIV_MMU_MASK; bool use_background = false; + hwaddr ppn; + RISCVCPU *cpu = env_archcpu(env); /* * Check if we should use the background registers for the two @@ -XXX,XX +XXX,XX @@ restart: return TRANSLATE_FAIL; } - hwaddr ppn = pte >> PTE_PPN_SHIFT; + if (riscv_cpu_sxl(env) == MXL_RV32) { + ppn = pte >> PTE_PPN_SHIFT; + } else if (cpu->cfg.ext_svpbmt || cpu->cfg.ext_svnapot) { + ppn = (pte & (target_ulong)PTE_PPN_MASK) >> PTE_PPN_SHIFT; + } else { + ppn = pte >> PTE_PPN_SHIFT; + if ((pte & ~(target_ulong)PTE_PPN_MASK) >> PTE_PPN_SHIFT) { + return TRANSLATE_FAIL; + } + } if (!(pte & PTE_V)) { /* Invalid PTE */ -- 2.34.1 From: Weiwei Li <liweiwei@iscas.ac.cn> For non-leaf PTEs, the D, A, and U bits are reserved for future standard use. Signed-off-by: Weiwei Li <liweiwei@iscas.ac.cn> Signed-off-by: Junqiang Wang <wangjunqiang@iscas.ac.cn> Reviewed-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-Id: <20220204022658.18097-3-liweiwei@iscas.ac.cn> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_helper.c | 3 +++ 1 file changed, 3 insertions(+) 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 @@ restart: return TRANSLATE_FAIL; } else if (!(pte & (PTE_R | PTE_W | PTE_X))) { /* Inner PTE, continue walking */ + if (pte & (PTE_D | PTE_A | PTE_U)) { + return TRANSLATE_FAIL; + } base = ppn << PGSHIFT; } else if ((pte & (PTE_R | PTE_W | PTE_X)) == PTE_W) { /* Reserved leaf PTE flags: PTE_W */ -- 2.34.1 From: Weiwei Li <liweiwei@iscas.ac.cn> - add PTE_N bit - add PTE_N bit check for inner PTE - update address translation to support 64KiB continuous region (napot_bits = 4) Signed-off-by: Weiwei Li <liweiwei@iscas.ac.cn> Signed-off-by: Junqiang Wang <wangjunqiang@iscas.ac.cn> Reviewed-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-Id: <20220204022658.18097-4-liweiwei@iscas.ac.cn> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_bits.h | 1 + target/riscv/cpu.c | 2 ++ target/riscv/cpu_helper.c | 18 +++++++++++++++--- 3 files changed, 18 insertions(+), 3 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 @@ typedef enum { #define PTE_A 0x040 /* Accessed */ #define PTE_D 0x080 /* Dirty */ #define PTE_SOFT 0x300 /* Reserved for Software */ +#define PTE_N 0x8000000000000000ULL /* NAPOT translation */ /* Page table PPN shift amount */ #define PTE_PPN_SHIFT 10 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 Property riscv_cpu_properties[] = { DEFINE_PROP_UINT16("vlen", RISCVCPU, cfg.vlen, 128), DEFINE_PROP_UINT16("elen", RISCVCPU, cfg.elen, 64), + DEFINE_PROP_BOOL("svnapot", RISCVCPU, cfg.ext_svnapot, false), + DEFINE_PROP_BOOL("zba", RISCVCPU, cfg.ext_zba, true), DEFINE_PROP_BOOL("zbb", RISCVCPU, cfg.ext_zbb, true), DEFINE_PROP_BOOL("zbc", RISCVCPU, cfg.ext_zbc, 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 use_background = false; hwaddr ppn; RISCVCPU *cpu = env_archcpu(env); + int napot_bits = 0; + target_ulong napot_mask; /* * Check if we should use the background registers for the two @@ -XXX,XX +XXX,XX @@ restart: return TRANSLATE_FAIL; } else if (!(pte & (PTE_R | PTE_W | PTE_X))) { /* Inner PTE, continue walking */ - if (pte & (PTE_D | PTE_A | PTE_U)) { + if (pte & (PTE_D | PTE_A | PTE_U | PTE_N)) { return TRANSLATE_FAIL; } base = ppn << PGSHIFT; @@ -XXX,XX +XXX,XX @@ restart: /* for superpage mappings, make a fake leaf PTE for the TLB's benefit. */ target_ulong vpn = addr >> PGSHIFT; - *physical = ((ppn | (vpn & ((1L << ptshift) - 1))) << PGSHIFT) | - (addr & ~TARGET_PAGE_MASK); + + if (cpu->cfg.ext_svnapot && (pte & PTE_N)) { + napot_bits = ctzl(ppn) + 1; + if ((i != (levels - 1)) || (napot_bits != 4)) { + return TRANSLATE_FAIL; + } + } + + napot_mask = (1 << napot_bits) - 1; + *physical = (((ppn & ~napot_mask) | (vpn & napot_mask) | + (vpn & (((target_ulong)1 << ptshift) - 1)) + ) << PGSHIFT) | (addr & ~TARGET_PAGE_MASK); /* set permissions on the TLB entry */ if ((pte & PTE_R) || ((pte & PTE_X) && mxr)) { -- 2.34.1 From: Weiwei Li <liweiwei@iscas.ac.cn> - sinval.vma, hinval.vvma and hinval.gvma do the same as sfence.vma, hfence.vvma and hfence.gvma except extension check - do nothing other than extension check for sfence.w.inval and sfence.inval.ir Signed-off-by: Weiwei Li <liweiwei@iscas.ac.cn> Signed-off-by: Junqiang Wang <wangjunqiang@iscas.ac.cn> Reviewed-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-Id: <20220204022658.18097-5-liweiwei@iscas.ac.cn> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 1 + target/riscv/insn32.decode | 7 ++ target/riscv/cpu.c | 1 + target/riscv/translate.c | 1 + target/riscv/insn_trans/trans_svinval.c.inc | 75 +++++++++++++++++++++ 5 files changed, 85 insertions(+) create mode 100644 target/riscv/insn_trans/trans_svinval.c.inc 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 RISCVCPUConfig { bool ext_counters; bool ext_ifencei; bool ext_icsr; + bool ext_svinval; bool ext_svnapot; bool ext_svpbmt; bool ext_zfh; 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 @@ fcvt_l_h 1100010 00010 ..... ... ..... 1010011 @r2_rm fcvt_lu_h 1100010 00011 ..... ... ..... 1010011 @r2_rm fcvt_h_l 1101010 00010 ..... ... ..... 1010011 @r2_rm fcvt_h_lu 1101010 00011 ..... ... ..... 1010011 @r2_rm + +# *** Svinval Standard Extension *** +sinval_vma 0001011 ..... ..... 000 00000 1110011 @sfence_vma +sfence_w_inval 0001100 00000 00000 000 00000 1110011 +sfence_inval_ir 0001100 00001 00000 000 00000 1110011 +hinval_vvma 0010011 ..... ..... 000 00000 1110011 @hfence_vvma +hinval_gvma 0110011 ..... ..... 000 00000 1110011 @hfence_gvma 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 Property riscv_cpu_properties[] = { DEFINE_PROP_UINT16("vlen", RISCVCPU, cfg.vlen, 128), DEFINE_PROP_UINT16("elen", RISCVCPU, cfg.elen, 64), + DEFINE_PROP_BOOL("svinval", RISCVCPU, cfg.ext_svinval, false), DEFINE_PROP_BOOL("svnapot", RISCVCPU, cfg.ext_svnapot, false), DEFINE_PROP_BOOL("zba", RISCVCPU, cfg.ext_zba, true), 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_rvb.c.inc" #include "insn_trans/trans_rvzfh.c.inc" #include "insn_trans/trans_privileged.c.inc" +#include "insn_trans/trans_svinval.c.inc" #include "insn_trans/trans_xventanacondops.c.inc" /* Include the auto-generated decoder for 16 bit insn */ diff --git a/target/riscv/insn_trans/trans_svinval.c.inc b/target/riscv/insn_trans/trans_svinval.c.inc new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/riscv/insn_trans/trans_svinval.c.inc @@ -XXX,XX +XXX,XX @@ +/* + * RISC-V translation routines for the Svinval Standard Instruction Set. + * + * Copyright (c) 2020-2022 PLCT lab + * + * 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_SVINVAL(ctx) do { \ + if (!ctx->cfg_ptr->ext_svinval) { \ + return false; \ + } \ +} while (0) + +static bool trans_sinval_vma(DisasContext *ctx, arg_sinval_vma *a) +{ + REQUIRE_SVINVAL(ctx); + /* Do the same as sfence.vma currently */ + REQUIRE_EXT(ctx, RVS); +#ifndef CONFIG_USER_ONLY + gen_helper_tlb_flush(cpu_env); + return true; +#endif + return false; +} + +static bool trans_sfence_w_inval(DisasContext *ctx, arg_sfence_w_inval *a) +{ + REQUIRE_SVINVAL(ctx); + REQUIRE_EXT(ctx, RVS); + /* Do nothing currently */ + return true; +} + +static bool trans_sfence_inval_ir(DisasContext *ctx, arg_sfence_inval_ir *a) +{ + REQUIRE_SVINVAL(ctx); + REQUIRE_EXT(ctx, RVS); + /* Do nothing currently */ + return true; +} + +static bool trans_hinval_vvma(DisasContext *ctx, arg_hinval_vvma *a) +{ + REQUIRE_SVINVAL(ctx); + /* Do the same as hfence.vvma currently */ + REQUIRE_EXT(ctx, RVH); +#ifndef CONFIG_USER_ONLY + gen_helper_hyp_tlb_flush(cpu_env); + return true; +#endif + return false; +} + +static bool trans_hinval_gvma(DisasContext *ctx, arg_hinval_gvma *a) +{ + REQUIRE_SVINVAL(ctx); + /* Do the same as hfence.gvma currently */ + REQUIRE_EXT(ctx, RVH); +#ifndef CONFIG_USER_ONLY + gen_helper_hyp_gvma_tlb_flush(cpu_env); + return true; +#endif + return false; +} -- 2.34.1 From: Weiwei Li <liweiwei@iscas.ac.cn> - add PTE_PBMT bits: It uses two PTE bits, but otherwise has no effect on QEMU, since QEMU is sequentially consistent and doesn't model PMAs currently - add PTE_PBMT bit check for inner PTE Signed-off-by: Weiwei Li <liweiwei@iscas.ac.cn> Signed-off-by: Junqiang Wang <wangjunqiang@iscas.ac.cn> Reviewed-by: Anup Patel <anup@brainfault.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-Id: <20220204022658.18097-6-liweiwei@iscas.ac.cn> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_bits.h | 2 ++ target/riscv/cpu.c | 1 + target/riscv/cpu_helper.c | 4 +++- 3 files changed, 6 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 { #define PTE_A 0x040 /* Accessed */ #define PTE_D 0x080 /* Dirty */ #define PTE_SOFT 0x300 /* Reserved for Software */ +#define PTE_PBMT 0x6000000000000000ULL /* Page-based memory types */ #define PTE_N 0x8000000000000000ULL /* NAPOT translation */ +#define PTE_ATTR (PTE_N | PTE_PBMT) /* All attributes bits */ /* Page table PPN shift amount */ #define PTE_PPN_SHIFT 10 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 Property riscv_cpu_properties[] = { DEFINE_PROP_BOOL("svinval", RISCVCPU, cfg.ext_svinval, false), DEFINE_PROP_BOOL("svnapot", RISCVCPU, cfg.ext_svnapot, false), + DEFINE_PROP_BOOL("svpbmt", RISCVCPU, cfg.ext_svpbmt, false), DEFINE_PROP_BOOL("zba", RISCVCPU, cfg.ext_zba, true), DEFINE_PROP_BOOL("zbb", RISCVCPU, cfg.ext_zbb, 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 @@ restart: if (!(pte & PTE_V)) { /* Invalid PTE */ return TRANSLATE_FAIL; + } else if (!cpu->cfg.ext_svpbmt && (pte & PTE_PBMT)) { + return TRANSLATE_FAIL; } else if (!(pte & (PTE_R | PTE_W | PTE_X))) { /* Inner PTE, continue walking */ - if (pte & (PTE_D | PTE_A | PTE_U | PTE_N)) { + if (pte & (PTE_D | PTE_A | PTE_U | PTE_ATTR)) { return TRANSLATE_FAIL; } base = ppn << PGSHIFT; -- 2.34.1 From: Yu Li <liyu.yukiteru@bytedance.com> Since the hypervisor extension been non experimental and enabled for default CPU, the previous command is no longer available and the option `x-h=true` or `h=true` is also no longer required. Signed-off-by: Yu Li <liyu.yukiteru@bytedance.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-Id: <9040401e-8f87-ef4a-d840-6703f08d068c@bytedance.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- docs/system/riscv/virt.rst | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/docs/system/riscv/virt.rst b/docs/system/riscv/virt.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/riscv/virt.rst +++ b/docs/system/riscv/virt.rst @@ -XXX,XX +XXX,XX @@ The ``virt`` machine supports the following devices: * 1 generic PCIe host bridge * The fw_cfg device that allows a guest to obtain data from QEMU -Note that the default CPU is a generic RV32GC/RV64GC. Optional extensions -can be enabled via command line parameters, e.g.: ``-cpu rv64,x-h=true`` -enables the hypervisor extension for RV64. +The hypervisor extension has been enabled for the default CPU, so virtual +machines with hypervisor extension can simply be used without explicitly +declaring. Hardware configuration information ---------------------------------- -- 2.34.1