The following changes since commit a51e5124a655b3dad80b36b18547cb1eca2c5eb2: Merge tag 'pull-omnibus-111023-1' of https://gitlab.com/stsquad/qemu into staging (2023-10-11 09:43:10 -0400) are available in the Git repository at: https://github.com/alistair23/qemu.git tags/pull-riscv-to-apply-20231012-1 for you to fetch changes up to 837570cef237b634eb4c245363470deebea7089d: target/riscv: Fix vfwmaccbf16.vf (2023-10-12 12:50:13 +1000) ---------------------------------------------------------------- Second RISC-V PR for 8.2 * Add support for the max CPU * Detect user choice in TCG * Clear CSR values at reset and sync MPSTATE with host * Fix the typo of inverted order of pmpaddr13 and pmpaddr14 * Split TCG/KVM accelerators from cpu.c * Add extension properties for all cpus * Replace GDB exit calls with proper shutdown * Support KVM_GET_REG_LIST * Remove RVG warning * Use env_archcpu for better performance * Deprecate capital 'Z' CPU properties * Fix vfwmaccbf16.vf ---------------------------------------------------------------- Alvin Chang (1): disas/riscv: Fix the typo of inverted order of pmpaddr13 and pmpaddr14 Clément Chigot (5): softmmu: add means to pass an exit code when requesting a shutdown softmmu: pass the main loop status to gdb "Wxx" packet hw/misc/sifive_test.c: replace exit calls with proper shutdown hw/char: riscv_htif: replace exit calls with proper shutdown gdbstub: replace exit calls with proper shutdown for softmmu Daniel Henrique Barboza (45): target/riscv/cpu.c: split CPU options from riscv_cpu_extensions[] target/riscv/cpu.c: skip 'bool' check when filtering KVM props target/riscv/cpu.c: split kvm prop handling to its own helper target/riscv: add DEFINE_PROP_END_OF_LIST() to riscv_cpu_options[] target/riscv/cpu.c: split non-ratified exts from riscv_cpu_extensions[] target/riscv/cpu.c: split vendor exts from riscv_cpu_extensions[] target/riscv/cpu.c: add riscv_cpu_add_qdev_prop_array() target/riscv/cpu.c: add riscv_cpu_add_kvm_unavail_prop_array() target/riscv/cpu.c: limit cfg->vext_spec log message target/riscv: add 'max' CPU type avocado, risc-v: add tuxboot tests for 'max' CPU target/riscv: deprecate the 'any' CPU type target/riscv/cpu.c: use offset in isa_ext_is_enabled/update_enabled target/riscv: make CPUCFG() macro public target/riscv/cpu.c: introduce cpu_cfg_ext_auto_update() target/riscv/cpu.c: use cpu_cfg_ext_auto_update() during realize() target/riscv/cpu.c: introduce RISCVCPUMultiExtConfig target/riscv: use isa_ext_update_enabled() in init_max_cpu_extensions() target/riscv/cpu.c: honor user choice in cpu_cfg_ext_auto_update() target/riscv/cpu.c: consider user option with RVG target/riscv: introduce TCG AccelCPUClass target/riscv: move riscv_cpu_realize_tcg() to TCG::cpu_realizefn() target/riscv: move riscv_cpu_validate_set_extensions() to tcg-cpu.c target/riscv: move riscv_tcg_ops to tcg-cpu.c target/riscv/cpu.c: add .instance_post_init() target/riscv: move 'host' CPU declaration to kvm.c target/riscv/cpu.c: mark extensions arrays as 'const' target/riscv: move riscv_cpu_add_kvm_properties() to kvm.c target/riscv: make riscv_add_satp_mode_properties() public target/riscv: remove kvm-stub.c target/riscv: introduce KVM AccelCPUClass target/riscv: move KVM only files to kvm subdir target/riscv/kvm: do not use riscv_cpu_add_misa_properties() target/riscv/cpu.c: export set_misa() target/riscv/tcg: introduce tcg_cpu_instance_init() target/riscv/cpu.c: make misa_ext_cfgs[] 'const' target/riscv/tcg: move riscv_cpu_add_misa_properties() to tcg-cpu.c target/riscv/cpu.c: export isa_edata_arr[] target/riscv/cpu: move priv spec functions to tcg-cpu.c target/riscv: add riscv_cpu_get_name() target/riscv/tcg-cpu.c: add extension properties for all cpus target/riscv/kvm: improve 'init_multiext_cfg' error msg target/riscv/kvm: support KVM_GET_REG_LIST target/riscv/tcg: remove RVG warning target/riscv: deprecate capital 'Z' CPU properties Max Chou (1): target/riscv: Fix vfwmaccbf16.vf Richard W.M. Jones (1): target/riscv: Use env_archcpu for better performance liguang.zhang (1): target/riscv: Clear CSR values at reset and sync MPSTATE with host docs/about/deprecated.rst | 35 ++ include/gdbstub/syscalls.h | 9 + include/sysemu/runstate.h | 2 + include/sysemu/sysemu.h | 2 +- target/riscv/cpu-qom.h | 1 + target/riscv/cpu.h | 34 +- target/riscv/{ => kvm}/kvm_riscv.h | 2 +- target/riscv/tcg/tcg-cpu.h | 27 + disas/riscv.c | 4 +- gdbstub/gdbstub.c | 5 +- gdbstub/system.c | 6 + gdbstub/user.c | 6 + hw/char/riscv_htif.c | 5 +- hw/intc/riscv_aplic.c | 2 +- hw/misc/sifive_test.c | 9 +- hw/riscv/virt.c | 2 +- system/main.c | 2 +- system/runstate.c | 16 +- target/riscv/cpu.c | 1030 +++++++-------------------------- target/riscv/cpu_helper.c | 3 +- target/riscv/csr.c | 1 + target/riscv/kvm-stub.c | 30 - target/riscv/{kvm.c => kvm/kvm-cpu.c} | 272 ++++++++- target/riscv/tcg/tcg-cpu.c | 949 ++++++++++++++++++++++++++++++ target/riscv/vector_helper.c | 2 +- target/riscv/kvm/meson.build | 1 + target/riscv/meson.build | 4 +- target/riscv/tcg/meson.build | 2 + tests/avocado/tuxrun_baselines.py | 32 + 29 files changed, 1604 insertions(+), 891 deletions(-) rename target/riscv/{ => kvm}/kvm_riscv.h (95%) create mode 100644 target/riscv/tcg/tcg-cpu.h delete mode 100644 target/riscv/kvm-stub.c rename target/riscv/{kvm.c => kvm/kvm-cpu.c} (81%) create mode 100644 target/riscv/tcg/tcg-cpu.c create mode 100644 target/riscv/kvm/meson.build create mode 100644 target/riscv/tcg/meson.build From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> We'll add a new CPU type that will enable a considerable amount of extensions. To make it easier for us we'll do a few cleanups in our existing riscv_cpu_extensions[] array. Start by splitting all CPU non-boolean options from it. Create a new riscv_cpu_options[] array for them. Add all these properties in riscv_cpu_add_user_properties() as it is already being done today. 'mmu' and 'pmp' aren't really extensions in the usual way we think about RISC-V extensions. These are closer to CPU features/options, so move both to riscv_cpu_options[] too. In the near future we'll need to match all extensions with all entries in isa_edata_arr[], and so it happens that both 'mmu' and 'pmp' do not have a riscv,isa string (thus, no priv spec version restriction). This further emphasizes the point that these are more a CPU option than an extension. No functional changes made. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Weiwei Li <liweiwei@iscas.ac.cn> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-2-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 33 +++++++++++++++++++++++---------- 1 file changed, 23 insertions(+), 10 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_add_misa_properties(Object *cpu_obj) static Property riscv_cpu_extensions[] = { /* Defaults for standard extensions */ - DEFINE_PROP_UINT8("pmu-num", RISCVCPU, cfg.pmu_num, 16), DEFINE_PROP_BOOL("sscofpmf", RISCVCPU, cfg.ext_sscofpmf, false), DEFINE_PROP_BOOL("Zifencei", RISCVCPU, cfg.ext_ifencei, true), DEFINE_PROP_BOOL("Zicsr", RISCVCPU, cfg.ext_icsr, true), @@ -XXX,XX +XXX,XX @@ static Property riscv_cpu_extensions[] = { DEFINE_PROP_BOOL("Zve32f", RISCVCPU, cfg.ext_zve32f, false), DEFINE_PROP_BOOL("Zve64f", RISCVCPU, cfg.ext_zve64f, false), DEFINE_PROP_BOOL("Zve64d", RISCVCPU, cfg.ext_zve64d, false), - DEFINE_PROP_BOOL("mmu", RISCVCPU, cfg.mmu, true), - DEFINE_PROP_BOOL("pmp", RISCVCPU, cfg.pmp, true), DEFINE_PROP_BOOL("sstc", RISCVCPU, cfg.ext_sstc, true), - DEFINE_PROP_STRING("priv_spec", RISCVCPU, cfg.priv_spec), - DEFINE_PROP_STRING("vext_spec", RISCVCPU, cfg.vext_spec), - DEFINE_PROP_UINT16("vlen", RISCVCPU, cfg.vlen, 128), - DEFINE_PROP_UINT16("elen", RISCVCPU, cfg.elen, 64), - DEFINE_PROP_BOOL("smstateen", RISCVCPU, cfg.ext_smstateen, false), DEFINE_PROP_BOOL("svadu", RISCVCPU, cfg.ext_svadu, true), DEFINE_PROP_BOOL("svinval", RISCVCPU, cfg.ext_svinval, false), @@ -XXX,XX +XXX,XX @@ static Property riscv_cpu_extensions[] = { DEFINE_PROP_BOOL("zhinxmin", RISCVCPU, cfg.ext_zhinxmin, false), DEFINE_PROP_BOOL("zicbom", RISCVCPU, cfg.ext_icbom, true), - DEFINE_PROP_UINT16("cbom_blocksize", RISCVCPU, cfg.cbom_blocksize, 64), DEFINE_PROP_BOOL("zicboz", RISCVCPU, cfg.ext_icboz, true), - DEFINE_PROP_UINT16("cboz_blocksize", RISCVCPU, cfg.cboz_blocksize, 64), DEFINE_PROP_BOOL("zmmul", RISCVCPU, cfg.ext_zmmul, false), @@ -XXX,XX +XXX,XX @@ static Property riscv_cpu_extensions[] = { DEFINE_PROP_END_OF_LIST(), }; +static Property riscv_cpu_options[] = { + DEFINE_PROP_UINT8("pmu-num", RISCVCPU, cfg.pmu_num, 16), + + DEFINE_PROP_BOOL("mmu", RISCVCPU, cfg.mmu, true), + DEFINE_PROP_BOOL("pmp", RISCVCPU, cfg.pmp, true), + + DEFINE_PROP_STRING("priv_spec", RISCVCPU, cfg.priv_spec), + DEFINE_PROP_STRING("vext_spec", RISCVCPU, cfg.vext_spec), + + DEFINE_PROP_UINT16("vlen", RISCVCPU, cfg.vlen, 128), + DEFINE_PROP_UINT16("elen", RISCVCPU, cfg.elen, 64), + + DEFINE_PROP_UINT16("cbom_blocksize", RISCVCPU, cfg.cbom_blocksize, 64), + DEFINE_PROP_UINT16("cboz_blocksize", RISCVCPU, cfg.cboz_blocksize, 64), +}; #ifndef CONFIG_USER_ONLY static void cpu_set_cfg_unavailable(Object *obj, Visitor *v, @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_user_properties(Object *obj) #endif qdev_property_add_static(dev, prop); } + + for (int i = 0; i < ARRAY_SIZE(riscv_cpu_options); i++) { + /* Check if KVM created the property already */ + if (object_property_find(obj, riscv_cpu_options[i].name)) { + continue; + } + qdev_property_add_static(dev, &riscv_cpu_options[i]); + } } static Property riscv_cpu_properties[] = { -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> After the introduction of riscv_cpu_options[] all properties in riscv_cpu_extensions[] are booleans. This check is now obsolete. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-3-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 14 ++++---------- 1 file changed, 4 insertions(+), 10 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_add_user_properties(Object *obj) * Set the default to disabled for every extension * unknown to KVM and error out if the user attempts * to enable any of them. - * - * We're giving a pass for non-bool properties since they're - * not related to the availability of extensions and can be - * safely ignored as is. */ - if (prop->info == &qdev_prop_bool) { - object_property_add(obj, prop->name, "bool", - NULL, cpu_set_cfg_unavailable, - NULL, (void *)prop->name); - continue; - } + object_property_add(obj, prop->name, "bool", + NULL, cpu_set_cfg_unavailable, + NULL, (void *)prop->name); + continue; } #endif qdev_property_add_static(dev, prop); -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Future patches will split the existing Property arrays even further, and the existing code in riscv_cpu_add_user_properties() will start to scale bad with it because it's dealing with KVM constraints mixed in with TCG constraints. We're going to pay a high price to share a couple of common lines of code between the two. Create a new kvm_riscv_cpu_add_kvm_properties() helper that will be forked from riscv_cpu_add_user_properties() if we're running KVM. The helper includes all properties that a KVM CPU will add. The rest of riscv_cpu_add_user_properties() body will then be relieved from having to deal with KVM constraints. The helper was declared in kvm_stubs.h, while being implemented in cpu.c, to allow '--enable-debug' builds to work. The compiler won't remove the kvm_riscv_cpu_add_kvm_properties() reference when 'kvm_enabled()' is false if we end up with an unused function. Even though being a KVM only helper we can't implement it in kvm.c due to its many dependencies inside cpu.c, so make it public in kvm_riscv.h and keep its implementation in cpu.c for now. We'll move it to kvm.c in the near future. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-4-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/kvm_riscv.h | 3 ++ target/riscv/cpu.c | 65 +++++++++++++++++++++++++--------------- 2 files changed, 44 insertions(+), 24 deletions(-) diff --git a/target/riscv/kvm_riscv.h b/target/riscv/kvm_riscv.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm_riscv.h +++ b/target/riscv/kvm_riscv.h @@ -XXX,XX +XXX,XX @@ #ifndef QEMU_KVM_RISCV_H #define QEMU_KVM_RISCV_H +/* Temporarily implemented in cpu.c */ +void kvm_riscv_cpu_add_kvm_properties(Object *obj); + void kvm_riscv_init_user_properties(Object *cpu_obj); void kvm_riscv_reset_vcpu(RISCVCPU *cpu); void kvm_riscv_set_irq(RISCVCPU *cpu, int irq, int level); 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_options[] = { DEFINE_PROP_UINT16("cboz_blocksize", RISCVCPU, cfg.cboz_blocksize, 64), }; -#ifndef CONFIG_USER_ONLY +#ifdef CONFIG_KVM static void cpu_set_cfg_unavailable(Object *obj, Visitor *v, const char *name, void *opaque, Error **errp) @@ -XXX,XX +XXX,XX @@ static void cpu_set_cfg_unavailable(Object *obj, Visitor *v, propname); } } + +static void riscv_cpu_add_kvm_unavail_prop(Object *obj, const char *prop_name) +{ + /* Check if KVM created the property already */ + if (object_property_find(obj, prop_name)) { + return; + } + + /* + * Set the default to disabled for every extension + * unknown to KVM and error out if the user attempts + * to enable any of them. + */ + object_property_add(obj, prop_name, "bool", + NULL, cpu_set_cfg_unavailable, + NULL, (void *)prop_name); +} + +void kvm_riscv_cpu_add_kvm_properties(Object *obj) +{ + Property *prop; + DeviceState *dev = DEVICE(obj); + + kvm_riscv_init_user_properties(obj); + riscv_cpu_add_misa_properties(obj); + + for (prop = riscv_cpu_extensions; prop && prop->name; prop++) { + riscv_cpu_add_kvm_unavail_prop(obj, prop->name); + } + + for (int i = 0; i < ARRAY_SIZE(riscv_cpu_options); i++) { + /* Check if KVM created the property already */ + if (object_property_find(obj, riscv_cpu_options[i].name)) { + continue; + } + qdev_property_add_static(dev, &riscv_cpu_options[i]); + } +} #endif /* @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_user_properties(Object *obj) riscv_add_satp_mode_properties(obj); if (kvm_enabled()) { - kvm_riscv_init_user_properties(obj); + kvm_riscv_cpu_add_kvm_properties(obj); + return; } #endif riscv_cpu_add_misa_properties(obj); for (prop = riscv_cpu_extensions; prop && prop->name; prop++) { -#ifndef CONFIG_USER_ONLY - if (kvm_enabled()) { - /* Check if KVM created the property already */ - if (object_property_find(obj, prop->name)) { - continue; - } - - /* - * Set the default to disabled for every extension - * unknown to KVM and error out if the user attempts - * to enable any of them. - */ - object_property_add(obj, prop->name, "bool", - NULL, cpu_set_cfg_unavailable, - NULL, (void *)prop->name); - continue; - } -#endif qdev_property_add_static(dev, prop); } for (int i = 0; i < ARRAY_SIZE(riscv_cpu_options); i++) { - /* Check if KVM created the property already */ - if (object_property_find(obj, riscv_cpu_options[i].name)) { - continue; - } qdev_property_add_static(dev, &riscv_cpu_options[i]); } } -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Add DEFINE_PROP_END_OF_LIST() and eliminate the ARRAY_SIZE() usage when iterating in the riscv_cpu_options[] array, making it similar to what we already do when working with riscv_cpu_extensions[]. We also have a more sophisticated motivation behind this change. In the future we might need to export riscv_cpu_options[] to other files, and ARRAY_LIST() doesn't work properly in that case because the array size isn't exposed to the header file. Here's a future sight of what we would deal with: ./target/riscv/kvm.c:1057:5: error: nested extern declaration of 'riscv_cpu_add_misa_properties' [-Werror=nested-externs] n file included from ../target/riscv/kvm.c:19: home/danielhb/work/qemu/include/qemu/osdep.h:473:31: error: invalid application of 'sizeof' to incomplete type 'const RISCVCPUMultiExtConfig[]' 473 | #define ARRAY_SIZE(x) ((sizeof(x) / sizeof((x)[0])) + \ | ^ ./target/riscv/kvm.c:1047:29: note: in expansion of macro 'ARRAY_SIZE' 1047 | for (int i = 0; i < ARRAY_SIZE(_array); i++) { \ | ^~~~~~~~~~ ./target/riscv/kvm.c:1059:5: note: in expansion of macro 'ADD_UNAVAIL_KVM_PROP_ARRAY' 1059 | ADD_UNAVAIL_KVM_PROP_ARRAY(obj, riscv_cpu_extensions); | ^~~~~~~~~~~~~~~~~~~~~~~~~~ home/danielhb/work/qemu/include/qemu/osdep.h:473:31: error: invalid application of 'sizeof' to incomplete type 'const RISCVCPUMultiExtConfig[]' 473 | #define ARRAY_SIZE(x) ((sizeof(x) / sizeof((x)[0])) + \ | ^ ./target/riscv/kvm.c:1047:29: note: in expansion of macro 'ARRAY_SIZE' 1047 | for (int i = 0; i < ARRAY_SIZE(_array); i++) { \ Homogenize the present and change the future by using DEFINE_PROP_END_OF_LIST() in riscv_cpu_options[]. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-ID: <20230912132423.268494-5-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 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 Property riscv_cpu_options[] = { DEFINE_PROP_UINT16("cbom_blocksize", RISCVCPU, cfg.cbom_blocksize, 64), DEFINE_PROP_UINT16("cboz_blocksize", RISCVCPU, cfg.cboz_blocksize, 64), + + DEFINE_PROP_END_OF_LIST(), }; #ifdef CONFIG_KVM @@ -XXX,XX +XXX,XX @@ void kvm_riscv_cpu_add_kvm_properties(Object *obj) riscv_cpu_add_kvm_unavail_prop(obj, prop->name); } - for (int i = 0; i < ARRAY_SIZE(riscv_cpu_options); i++) { + for (prop = riscv_cpu_options; prop && prop->name; prop++) { /* Check if KVM created the property already */ - if (object_property_find(obj, riscv_cpu_options[i].name)) { + if (object_property_find(obj, prop->name)) { continue; } - qdev_property_add_static(dev, &riscv_cpu_options[i]); + qdev_property_add_static(dev, prop); } } #endif @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_user_properties(Object *obj) qdev_property_add_static(dev, prop); } - for (int i = 0; i < ARRAY_SIZE(riscv_cpu_options); i++) { - qdev_property_add_static(dev, &riscv_cpu_options[i]); + for (prop = riscv_cpu_options; prop && prop->name; prop++) { + qdev_property_add_static(dev, prop); } } -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Create a new riscv_cpu_experimental_exts[] to store the non-ratified extensions properties. Once they are ratified we'll move them back to riscv_cpu_extensions[]. riscv_cpu_add_user_properties() and riscv_cpu_add_kvm_properties() are changed to keep adding non-ratified properties to users. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-6-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 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 @@ static Property riscv_cpu_extensions[] = { DEFINE_PROP_BOOL("xtheadsync", RISCVCPU, cfg.ext_xtheadsync, false), DEFINE_PROP_BOOL("xventanacondops", RISCVCPU, cfg.ext_XVentanaCondOps, false), - /* These are experimental so mark with 'x-' */ + DEFINE_PROP_END_OF_LIST(), +}; +/* These are experimental so mark with 'x-' */ +static Property riscv_cpu_experimental_exts[] = { /* ePMP 0.9.3 */ DEFINE_PROP_BOOL("x-epmp", RISCVCPU, cfg.epmp, false), DEFINE_PROP_BOOL("x-smaia", RISCVCPU, cfg.ext_smaia, false), @@ -XXX,XX +XXX,XX @@ void kvm_riscv_cpu_add_kvm_properties(Object *obj) riscv_cpu_add_kvm_unavail_prop(obj, prop->name); } + for (prop = riscv_cpu_experimental_exts; prop && prop->name; prop++) { + riscv_cpu_add_kvm_unavail_prop(obj, prop->name); + } + for (prop = riscv_cpu_options; prop && prop->name; prop++) { /* Check if KVM created the property already */ if (object_property_find(obj, prop->name)) { @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_user_properties(Object *obj) for (prop = riscv_cpu_options; prop && prop->name; prop++) { qdev_property_add_static(dev, prop); } + + for (prop = riscv_cpu_experimental_exts; prop && prop->name; prop++) { + qdev_property_add_static(dev, prop); + } } static Property riscv_cpu_properties[] = { -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Our goal is to make riscv_cpu_extensions[] hold only ratified, non-vendor extensions. Create a new riscv_cpu_vendor_exts[] array for them, changing riscv_cpu_add_user_properties() and riscv_cpu_add_kvm_properties() accordingly. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-7-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 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 @@ static Property riscv_cpu_extensions[] = { DEFINE_PROP_BOOL("zcmt", RISCVCPU, cfg.ext_zcmt, false), DEFINE_PROP_BOOL("zicond", RISCVCPU, cfg.ext_zicond, false), - /* Vendor-specific custom extensions */ + DEFINE_PROP_END_OF_LIST(), +}; + +static Property riscv_cpu_vendor_exts[] = { DEFINE_PROP_BOOL("xtheadba", RISCVCPU, cfg.ext_xtheadba, false), DEFINE_PROP_BOOL("xtheadbb", RISCVCPU, cfg.ext_xtheadbb, false), DEFINE_PROP_BOOL("xtheadbs", RISCVCPU, cfg.ext_xtheadbs, false), @@ -XXX,XX +XXX,XX @@ void kvm_riscv_cpu_add_kvm_properties(Object *obj) riscv_cpu_add_kvm_unavail_prop(obj, prop->name); } + for (prop = riscv_cpu_vendor_exts; prop && prop->name; prop++) { + riscv_cpu_add_kvm_unavail_prop(obj, prop->name); + } + for (prop = riscv_cpu_experimental_exts; prop && prop->name; prop++) { riscv_cpu_add_kvm_unavail_prop(obj, prop->name); } @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_user_properties(Object *obj) qdev_property_add_static(dev, prop); } + for (prop = riscv_cpu_vendor_exts; prop && prop->name; prop++) { + qdev_property_add_static(dev, prop); + } + for (prop = riscv_cpu_experimental_exts; prop && prop->name; prop++) { qdev_property_add_static(dev, prop); } -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> The code inside riscv_cpu_add_user_properties() became quite repetitive after recent changes. Add a helper to hide the repetition away. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-ID: <20230912132423.268494-8-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 29 +++++++++++++---------------- 1 file changed, 13 insertions(+), 16 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 Property riscv_cpu_options[] = { DEFINE_PROP_END_OF_LIST(), }; +static void riscv_cpu_add_qdev_prop_array(DeviceState *dev, Property *array) +{ + g_assert(array); + + for (Property *prop = array; prop && prop->name; prop++) { + qdev_property_add_static(dev, prop); + } +} + #ifdef CONFIG_KVM static void cpu_set_cfg_unavailable(Object *obj, Visitor *v, const char *name, @@ -XXX,XX +XXX,XX @@ void kvm_riscv_cpu_add_kvm_properties(Object *obj) */ static void riscv_cpu_add_user_properties(Object *obj) { - Property *prop; DeviceState *dev = DEVICE(obj); #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_user_properties(Object *obj) riscv_cpu_add_misa_properties(obj); - for (prop = riscv_cpu_extensions; prop && prop->name; prop++) { - qdev_property_add_static(dev, prop); - } - - for (prop = riscv_cpu_options; prop && prop->name; prop++) { - qdev_property_add_static(dev, prop); - } - - for (prop = riscv_cpu_vendor_exts; prop && prop->name; prop++) { - qdev_property_add_static(dev, prop); - } - - for (prop = riscv_cpu_experimental_exts; prop && prop->name; prop++) { - qdev_property_add_static(dev, prop); - } + riscv_cpu_add_qdev_prop_array(dev, riscv_cpu_extensions); + riscv_cpu_add_qdev_prop_array(dev, riscv_cpu_options); + riscv_cpu_add_qdev_prop_array(dev, riscv_cpu_vendor_exts); + riscv_cpu_add_qdev_prop_array(dev, riscv_cpu_experimental_exts); } static Property riscv_cpu_properties[] = { -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Use a helper in riscv_cpu_add_kvm_properties() to eliminate some of its code repetition. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230912132423.268494-9-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 24 +++++++++++++----------- 1 file changed, 13 insertions(+), 11 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_add_kvm_unavail_prop(Object *obj, const char *prop_name) NULL, (void *)prop_name); } +static void riscv_cpu_add_kvm_unavail_prop_array(Object *obj, + Property *array) +{ + g_assert(array); + + for (Property *prop = array; prop && prop->name; prop++) { + riscv_cpu_add_kvm_unavail_prop(obj, prop->name); + } +} + void kvm_riscv_cpu_add_kvm_properties(Object *obj) { Property *prop; @@ -XXX,XX +XXX,XX @@ void kvm_riscv_cpu_add_kvm_properties(Object *obj) kvm_riscv_init_user_properties(obj); riscv_cpu_add_misa_properties(obj); - for (prop = riscv_cpu_extensions; prop && prop->name; prop++) { - riscv_cpu_add_kvm_unavail_prop(obj, prop->name); - } - - for (prop = riscv_cpu_vendor_exts; prop && prop->name; prop++) { - riscv_cpu_add_kvm_unavail_prop(obj, prop->name); - } - - for (prop = riscv_cpu_experimental_exts; prop && prop->name; prop++) { - riscv_cpu_add_kvm_unavail_prop(obj, prop->name); - } + riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_extensions); + riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_vendor_exts); + riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_experimental_exts); for (prop = riscv_cpu_options; prop && prop->name; prop++) { /* Check if KVM created the property already */ -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Inside riscv_cpu_validate_v() we're always throwing a log message if the user didn't set a vector version via 'vext_spec'. We're going to include one case with the 'max' CPU where env->vext_ver will be set in the cpu_init(). But that alone will not stop the "vector version is not specified" message from appearing. The usefulness of this log message is debatable for the generic CPUs, but for a 'max' CPU type, where we are supposed to deliver a CPU model with all features possible, it's strange to force users to set 'vext_spec' to get rid of this message. Change riscv_cpu_validate_v() to not throw this log message if env->vext_ver is already set. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Weiwei Li <liweiwei@iscas.ac.cn> Message-ID: <20230912132423.268494-10-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 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_disas_set_info(CPUState *s, disassemble_info *info) static void riscv_cpu_validate_v(CPURISCVState *env, RISCVCPUConfig *cfg, Error **errp) { - int vext_version = VEXT_VERSION_1_00_0; - if (!is_power_of_2(cfg->vlen)) { error_setg(errp, "Vector extension VLEN must be power of 2"); return; @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_validate_v(CPURISCVState *env, RISCVCPUConfig *cfg, } if (cfg->vext_spec) { if (!g_strcmp0(cfg->vext_spec, "v1.0")) { - vext_version = VEXT_VERSION_1_00_0; + env->vext_ver = VEXT_VERSION_1_00_0; } else { error_setg(errp, "Unsupported vector spec version '%s'", cfg->vext_spec); return; } - } else { + } else if (env->vext_ver == 0) { qemu_log("vector version is not specified, " "use the default value v1.0\n"); + + env->vext_ver = VEXT_VERSION_1_00_0; } - env->vext_ver = vext_version; } static void riscv_cpu_validate_priv_spec(RISCVCPU *cpu, Error **errp) -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> The 'max' CPU type is used by tooling to determine what's the most capable CPU a current QEMU version implements. Other archs such as ARM implements this type. Let's add it to RISC-V. What we consider "most capable CPU" in this context are related to ratified, non-vendor extensions. This means that we want the 'max' CPU to enable all (possible) ratified extensions by default. The reasoning behind this design is (1) vendor extensions can conflict with each other and we won't play favorities deciding which one is default or not and (2) non-ratified extensions are always prone to changes, not being stable enough to be enabled by default. All this said, we're still not able to enable all ratified extensions due to conflicts between them. Zfinx and all its dependencies aren't enabled because of a conflict with RVF. zce, zcmp and zcmt are also disabled due to RVD conflicts. When running with 64 bits we're also disabling zcf. MISA bits RVG, RVJ and RVV are also being set manually since they're default disabled. This is the resulting 'riscv,isa' DT for this new CPU: rv64imafdcvh_zicbom_zicboz_zicsr_zifencei_zihintpause_zawrs_zfa_ zfh_zfhmin_zca_zcb_zcd_zba_zbb_zbc_zbkb_zbkc_zbkx_zbs_zk_zkn_zknd_ zkne_zknh_zkr_zks_zksed_zksh_zkt_zve32f_zve64f_zve64d_ smstateen_sscofpmf_sstc_svadu_svinval_svnapot_svpbmt Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Weiwei Li <liweiwei@iscas.ac.cn> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-11-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu-qom.h | 1 + target/riscv/cpu.c | 56 ++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 57 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 CPU_RESOLVING_TYPE TYPE_RISCV_CPU #define TYPE_RISCV_CPU_ANY RISCV_CPU_TYPE_NAME("any") +#define TYPE_RISCV_CPU_MAX RISCV_CPU_TYPE_NAME("max") #define TYPE_RISCV_CPU_BASE32 RISCV_CPU_TYPE_NAME("rv32") #define TYPE_RISCV_CPU_BASE64 RISCV_CPU_TYPE_NAME("rv64") #define TYPE_RISCV_CPU_BASE128 RISCV_CPU_TYPE_NAME("x-rv128") 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 char * const riscv_intr_names[] = { }; static void riscv_cpu_add_user_properties(Object *obj); +static void riscv_init_max_cpu_extensions(Object *obj); const char *riscv_cpu_get_trap_name(target_ulong cause, bool async) { @@ -XXX,XX +XXX,XX @@ static void riscv_any_cpu_init(Object *obj) cpu->cfg.pmp = true; } +static void riscv_max_cpu_init(Object *obj) +{ + RISCVCPU *cpu = RISCV_CPU(obj); + CPURISCVState *env = &cpu->env; + RISCVMXL mlx = MXL_RV64; + +#ifdef TARGET_RISCV32 + mlx = MXL_RV32; +#endif + set_misa(env, mlx, 0); + riscv_cpu_add_user_properties(obj); + riscv_init_max_cpu_extensions(obj); + env->priv_ver = PRIV_VERSION_LATEST; +#ifndef CONFIG_USER_ONLY + set_satp_mode_max_supported(RISCV_CPU(obj), mlx == MXL_RV32 ? + VM_1_10_SV32 : VM_1_10_SV57); +#endif +} + #if defined(TARGET_RISCV64) static void rv64_base_cpu_init(Object *obj) { @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_user_properties(Object *obj) riscv_cpu_add_qdev_prop_array(dev, riscv_cpu_experimental_exts); } +/* + * The 'max' type CPU will have all possible ratified + * non-vendor extensions enabled. + */ +static void riscv_init_max_cpu_extensions(Object *obj) +{ + RISCVCPU *cpu = RISCV_CPU(obj); + CPURISCVState *env = &cpu->env; + Property *prop; + + /* Enable RVG, RVJ and RVV that are disabled by default */ + set_misa(env, env->misa_mxl, env->misa_ext | RVG | RVJ | RVV); + + for (prop = riscv_cpu_extensions; prop && prop->name; prop++) { + object_property_set_bool(obj, prop->name, true, NULL); + } + + /* set vector version */ + env->vext_ver = VEXT_VERSION_1_00_0; + + /* Zfinx is not compatible with F. Disable it */ + object_property_set_bool(obj, "zfinx", false, NULL); + object_property_set_bool(obj, "zdinx", false, NULL); + object_property_set_bool(obj, "zhinx", false, NULL); + object_property_set_bool(obj, "zhinxmin", false, NULL); + + object_property_set_bool(obj, "zce", false, NULL); + object_property_set_bool(obj, "zcmp", false, NULL); + object_property_set_bool(obj, "zcmt", false, NULL); + + if (env->misa_mxl != MXL_RV32) { + object_property_set_bool(obj, "zcf", false, NULL); + } +} + static Property riscv_cpu_properties[] = { DEFINE_PROP_BOOL("debug", RISCVCPU, cfg.debug, true), @@ -XXX,XX +XXX,XX @@ static const TypeInfo riscv_cpu_type_infos[] = { .abstract = true, }, DEFINE_DYNAMIC_CPU(TYPE_RISCV_CPU_ANY, riscv_any_cpu_init), + DEFINE_DYNAMIC_CPU(TYPE_RISCV_CPU_MAX, riscv_max_cpu_init), #if defined(CONFIG_KVM) DEFINE_CPU(TYPE_RISCV_CPU_HOST, riscv_host_cpu_init), #endif -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Add smoke tests to ensure that we'll not break the 'max' CPU type when adding new frozen/ratified RISC-V extensions. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Tested-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-ID: <20230912132423.268494-12-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- tests/avocado/tuxrun_baselines.py | 32 +++++++++++++++++++++++++++++++ 1 file changed, 32 insertions(+) diff --git a/tests/avocado/tuxrun_baselines.py b/tests/avocado/tuxrun_baselines.py index XXXXXXX..XXXXXXX 100644 --- a/tests/avocado/tuxrun_baselines.py +++ b/tests/avocado/tuxrun_baselines.py @@ -XXX,XX +XXX,XX @@ def test_riscv64(self): self.common_tuxrun(csums=sums) + def test_riscv32_maxcpu(self): + """ + :avocado: tags=arch:riscv32 + :avocado: tags=machine:virt + :avocado: tags=cpu:max + :avocado: tags=tuxboot:riscv32 + """ + sums = { "Image" : + "89599407d7334de629a40e7ad6503c73670359eb5f5ae9d686353a3d6deccbd5", + "fw_jump.elf" : + "f2ef28a0b77826f79d085d3e4aa686f1159b315eff9099a37046b18936676985", + "rootfs.ext4.zst" : + "7168d296d0283238ea73cd5a775b3dd608e55e04c7b92b76ecce31bb13108cba" } + + self.common_tuxrun(csums=sums) + + def test_riscv64_maxcpu(self): + """ + :avocado: tags=arch:riscv64 + :avocado: tags=machine:virt + :avocado: tags=cpu:max + :avocado: tags=tuxboot:riscv64 + """ + sums = { "Image" : + "cd634badc65e52fb63465ec99e309c0de0369f0841b7d9486f9729e119bac25e", + "fw_jump.elf" : + "6e3373abcab4305fe151b564a4c71110d833c21f2c0a1753b7935459e36aedcf", + "rootfs.ext4.zst" : + "b18e3a3bdf27be03da0b285e84cb71bf09eca071c3a087b42884b6982ed679eb" } + + self.common_tuxrun(csums=sums) + def test_s390(self): """ :avocado: tags=arch:s390x -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> The 'any' CPU type was introduced in commit dc5bd18fa5725 ("RISC-V CPU Core Definition"), being around since the beginning. It's not an easy CPU to use: it's undocumented and its name doesn't tell users much about what the CPU is supposed to bring. 'git log' doesn't help us either in knowing what was the original design of this CPU type. The closest we have is a comment from Alistair [1] where he recalls from memory that the 'any' CPU is supposed to behave like the newly added 'max' CPU. He also suggested that the 'any' CPU should be removed. The default CPUs are rv32 and rv64, so removing the 'any' CPU will have impact only on users that might have a script that uses '-cpu any'. And those users are better off using the default CPUs or the new 'max' CPU. We would love to just remove the code and be done with it, but one does not simply remove a feature in QEMU. We'll put the CPU in quarantine first, letting users know that we have the intent of removing it in the future. [1] https://lists.gnu.org/archive/html/qemu-devel/2023-07/msg02891.html Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Tested-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-ID: <20230912132423.268494-13-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- docs/about/deprecated.rst | 12 ++++++++++++ target/riscv/cpu.c | 5 +++++ 2 files changed, 17 insertions(+) diff --git a/docs/about/deprecated.rst b/docs/about/deprecated.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/about/deprecated.rst +++ b/docs/about/deprecated.rst @@ -XXX,XX +XXX,XX @@ QEMU's ``vhost`` feature, which would eliminate the high latency costs under which the 9p ``proxy`` backend currently suffers. However as of to date nobody has indicated plans for such kind of reimplementation unfortunately. +RISC-V 'any' CPU type ``-cpu any`` (since 8.2) +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +The 'any' CPU type was introduced back in 2018 and has been around since the +initial RISC-V QEMU port. Its usage has always been unclear: users don't know +what to expect from a CPU called 'any', and in fact the CPU does not do anything +special that isn't already done by the default CPUs rv32/rv64. + +After the introduction of the 'max' CPU type, RISC-V now has a good coverage +of generic CPUs: rv32 and rv64 as default CPUs and 'max' as a feature complete +CPU for both 32 and 64 bit builds. Users are then discouraged to use the 'any' +CPU type starting in 8.2. Block device options '''''''''''''''''''' 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) RISCVCPUClass *mcc = RISCV_CPU_GET_CLASS(dev); Error *local_err = NULL; + if (object_dynamic_cast(OBJECT(dev), TYPE_RISCV_CPU_ANY) != NULL) { + warn_report("The 'any' CPU is deprecated and will be " + "removed in the future."); + } + cpu_exec_realizefn(cs, &local_err); if (local_err != NULL) { error_propagate(errp, local_err); -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> We'll have future usage for a function where, given an offset of the struct RISCVCPUConfig, the flag is updated to a certain val. Change all existing callers to use edata->ext_enable_offset instead of 'edata'. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-14-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 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 const struct isa_ext_data isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(xventanacondops, PRIV_VERSION_1_12_0, ext_XVentanaCondOps), }; -static bool isa_ext_is_enabled(RISCVCPU *cpu, - const struct isa_ext_data *edata) +static bool isa_ext_is_enabled(RISCVCPU *cpu, uint32_t ext_offset) { - bool *ext_enabled = (void *)&cpu->cfg + edata->ext_enable_offset; + bool *ext_enabled = (void *)&cpu->cfg + ext_offset; return *ext_enabled; } -static void isa_ext_update_enabled(RISCVCPU *cpu, - const struct isa_ext_data *edata, bool en) +static void isa_ext_update_enabled(RISCVCPU *cpu, uint32_t ext_offset, + bool en) { - bool *ext_enabled = (void *)&cpu->cfg + edata->ext_enable_offset; + bool *ext_enabled = (void *)&cpu->cfg + ext_offset; *ext_enabled = en; } @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu) /* Force disable extensions if priv spec version does not match */ for (i = 0; i < ARRAY_SIZE(isa_edata_arr); i++) { - if (isa_ext_is_enabled(cpu, &isa_edata_arr[i]) && + if (isa_ext_is_enabled(cpu, isa_edata_arr[i].ext_enable_offset) && (env->priv_ver < isa_edata_arr[i].min_version)) { - isa_ext_update_enabled(cpu, &isa_edata_arr[i], false); + isa_ext_update_enabled(cpu, isa_edata_arr[i].ext_enable_offset, + false); #ifndef CONFIG_USER_ONLY warn_report("disabling %s extension for hart 0x" TARGET_FMT_lx " because privilege spec version does not match", @@ -XXX,XX +XXX,XX @@ static void riscv_isa_string_ext(RISCVCPU *cpu, char **isa_str, int i; for (i = 0; i < ARRAY_SIZE(isa_edata_arr); i++) { - if (isa_ext_is_enabled(cpu, &isa_edata_arr[i])) { + if (isa_ext_is_enabled(cpu, isa_edata_arr[i].ext_enable_offset)) { new = g_strconcat(old, "_", isa_edata_arr[i].name, NULL); g_free(old); old = new; -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> The RISC-V KVM driver uses a CPUCFG() macro that calculates the offset of a certain field in the struct RISCVCPUConfig. We're going to use this macro in target/riscv/cpu.c as well in the next patches. Make it public. Rename it to CPU_CFG_OFFSET() for more clarity while we're at it. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-ID: <20230912132423.268494-15-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 2 ++ target/riscv/cpu.c | 2 +- target/riscv/kvm.c | 8 +++----- 3 files changed, 6 insertions(+), 6 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 @@ const char *riscv_get_misa_ext_name(uint32_t bit); const char *riscv_get_misa_ext_description(uint32_t bit); +#define CPU_CFG_OFFSET(_prop) offsetof(struct RISCVCPUConfig, _prop) + /* Privileged specification version */ enum { PRIV_VERSION_1_10_0 = 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 @@ struct isa_ext_data { }; #define ISA_EXT_DATA_ENTRY(_name, _min_ver, _prop) \ - {#_name, _min_ver, offsetof(struct RISCVCPUConfig, _prop)} + {#_name, _min_ver, CPU_CFG_OFFSET(_prop)} /* * From vector_helper.c diff --git a/target/riscv/kvm.c b/target/riscv/kvm.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm.c +++ b/target/riscv/kvm.c @@ -XXX,XX +XXX,XX @@ static void kvm_riscv_update_cpu_misa_ext(RISCVCPU *cpu, CPUState *cs) } } -#define CPUCFG(_prop) offsetof(struct RISCVCPUConfig, _prop) - #define KVM_EXT_CFG(_name, _prop, _reg_id) \ - {.name = _name, .offset = CPUCFG(_prop), \ + {.name = _name, .offset = CPU_CFG_OFFSET(_prop), \ .kvm_reg_id = _reg_id} static KVMCPUConfig kvm_multi_ext_cfgs[] = { @@ -XXX,XX +XXX,XX @@ static void kvm_cpu_set_multi_ext_cfg(Object *obj, Visitor *v, static KVMCPUConfig kvm_cbom_blocksize = { .name = "cbom_blocksize", - .offset = CPUCFG(cbom_blocksize), + .offset = CPU_CFG_OFFSET(cbom_blocksize), .kvm_reg_id = KVM_REG_RISCV_CONFIG_REG(zicbom_block_size) }; static KVMCPUConfig kvm_cboz_blocksize = { .name = "cboz_blocksize", - .offset = CPUCFG(cboz_blocksize), + .offset = CPU_CFG_OFFSET(cboz_blocksize), .kvm_reg_id = KVM_REG_RISCV_CONFIG_REG(zicboz_block_size) }; -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> During realize() time we're activating a lot of extensions based on some criteria, e.g.: if (cpu->cfg.ext_zk) { cpu->cfg.ext_zkn = true; cpu->cfg.ext_zkr = true; cpu->cfg.ext_zkt = true; } This practice resulted in at least one case where we ended up enabling something we shouldn't: RVC enabling zca/zcd/zcf when using a CPU that has priv_spec older than 1.12.0. We're also not considering user choice. There's no way of doing it now but this is about to change in the next few patches. cpu_cfg_ext_auto_update() will check for priv version mismatches before enabling extensions. If we have a mismatch between the current priv version and the extension we want to enable, do not enable it. In the near future, this same function will also consider user choice when deciding if we're going to enable/disable an extension or not. For now let's use it to handle zca/zcd/zcf enablement if RVC is enabled. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-16-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 43 ++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 40 insertions(+), 3 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 isa_ext_update_enabled(RISCVCPU *cpu, uint32_t ext_offset, *ext_enabled = en; } +static int cpu_cfg_ext_get_min_version(uint32_t ext_offset) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(isa_edata_arr); i++) { + if (isa_edata_arr[i].ext_enable_offset != ext_offset) { + continue; + } + + return isa_edata_arr[i].min_version; + } + + g_assert_not_reached(); +} + +static void cpu_cfg_ext_auto_update(RISCVCPU *cpu, uint32_t ext_offset, + bool value) +{ + CPURISCVState *env = &cpu->env; + bool prev_val = isa_ext_is_enabled(cpu, ext_offset); + int min_version; + + if (prev_val == value) { + return; + } + + if (value && env->priv_ver != PRIV_VERSION_LATEST) { + /* Do not enable it if priv_ver is older than min_version */ + min_version = cpu_cfg_ext_get_min_version(ext_offset); + if (env->priv_ver < min_version) { + return; + } + } + + isa_ext_update_enabled(cpu, ext_offset, value); +} + const char * const riscv_int_regnames[] = { "x0/zero", "x1/ra", "x2/sp", "x3/gp", "x4/tp", "x5/t0", "x6/t1", "x7/t2", "x8/s0", "x9/s1", "x10/a0", "x11/a1", "x12/a2", "x13/a3", @@ -XXX,XX +XXX,XX @@ void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) /* zca, zcd and zcf has a PRIV 1.12.0 restriction */ if (riscv_has_ext(env, RVC) && env->priv_ver >= PRIV_VERSION_1_12_0) { - cpu->cfg.ext_zca = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zca), true); if (riscv_has_ext(env, RVF) && env->misa_mxl_max == MXL_RV32) { - cpu->cfg.ext_zcf = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcf), true); } if (riscv_has_ext(env, RVD)) { - cpu->cfg.ext_zcd = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcd), true); } } -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Let's change the other instances in realize() where we're enabling an extension based on a certain criteria (e.g. it's a dependency of another extension). We're leaving icsr and ifencei being enabled during RVG for later - we'll want to error out in that case. Every other extension enablement during realize is now done via cpu_cfg_ext_auto_update(). The end goal is that only cpu init() functions will handle extension flags directly via "cpu->cfg.ext_N = true|false". Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-17-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 50 +++++++++++++++++++++++----------------------- 1 file changed, 25 insertions(+), 25 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 @@ void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) } if (cpu->cfg.ext_zfh) { - cpu->cfg.ext_zfhmin = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zfhmin), true); } if (cpu->cfg.ext_zfhmin && !riscv_has_ext(env, RVF)) { @@ -XXX,XX +XXX,XX @@ void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) } /* The V vector extension depends on the Zve64d extension */ - cpu->cfg.ext_zve64d = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zve64d), true); } /* The Zve64d extension depends on the Zve64f extension */ if (cpu->cfg.ext_zve64d) { - cpu->cfg.ext_zve64f = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zve64f), true); } /* The Zve64f extension depends on the Zve32f extension */ if (cpu->cfg.ext_zve64f) { - cpu->cfg.ext_zve32f = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zve32f), true); } if (cpu->cfg.ext_zve64d && !riscv_has_ext(env, RVD)) { @@ -XXX,XX +XXX,XX @@ void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) } if (cpu->cfg.ext_zvfh) { - cpu->cfg.ext_zvfhmin = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zvfhmin), true); } if (cpu->cfg.ext_zvfhmin && !cpu->cfg.ext_zve32f) { @@ -XXX,XX +XXX,XX @@ void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) /* Set the ISA extensions, checks should have happened above */ if (cpu->cfg.ext_zhinx) { - cpu->cfg.ext_zhinxmin = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zca), true); } if ((cpu->cfg.ext_zdinx || cpu->cfg.ext_zhinxmin) && !cpu->cfg.ext_zfinx) { @@ -XXX,XX +XXX,XX @@ void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) } if (cpu->cfg.ext_zce) { - cpu->cfg.ext_zca = true; - cpu->cfg.ext_zcb = true; - cpu->cfg.ext_zcmp = true; - cpu->cfg.ext_zcmt = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zca), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcb), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcmp), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcmt), true); if (riscv_has_ext(env, RVF) && env->misa_mxl_max == MXL_RV32) { - cpu->cfg.ext_zcf = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcf), true); } } @@ -XXX,XX +XXX,XX @@ void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) } if (cpu->cfg.ext_zk) { - cpu->cfg.ext_zkn = true; - cpu->cfg.ext_zkr = true; - cpu->cfg.ext_zkt = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zkn), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zkr), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zkt), true); } if (cpu->cfg.ext_zkn) { - cpu->cfg.ext_zbkb = true; - cpu->cfg.ext_zbkc = true; - cpu->cfg.ext_zbkx = true; - cpu->cfg.ext_zkne = true; - cpu->cfg.ext_zknd = true; - cpu->cfg.ext_zknh = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkb), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkc), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkx), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zkne), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zknd), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zknh), true); } if (cpu->cfg.ext_zks) { - cpu->cfg.ext_zbkb = true; - cpu->cfg.ext_zbkc = true; - cpu->cfg.ext_zbkx = true; - cpu->cfg.ext_zksed = true; - cpu->cfg.ext_zksh = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkb), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkc), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkx), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zksed), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zksh), true); } /* -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> If we want to make better decisions when auto-enabling extensions during realize() we need a way to tell if an user set an extension manually. The RISC-V KVM driver has its own solution via a KVMCPUConfig struct that has an 'user_set' flag that is set during the Property set() callback. The set() callback also does init() time validations based on the current KVM driver capabilities. For TCG we would want a 'user_set' mechanic too, but we would look ad-hoc via cpu_cfg_ext_auto_update() if a certain extension was user set or not. If we copy what was made in the KVM side we would look for 'user_set' for one into 60+ extension structs spreaded in 3 arrays (riscv_cpu_extensions, riscv_cpu_experimental_exts, riscv_cpu_vendor_exts). We'll still need an extension struct but we won't be using the 'user_set' flag: - 'RISCVCPUMultiExtConfig' will be our specialized structure, similar to what we're already doing with the MISA extensions in 'RISCVCPUMisaExtConfig'. DEFINE_PROP_BOOL() for all 3 extensions arrays were replaced by MULTI_EXT_CFG_BOOL(), a macro that will init our specialized struct; - the 'multi_ext_user_opts' hash will be used to store the offset of each extension that the user set via the set() callback, cpu_set_multi_ext_cfg(). For now we're just initializing and populating it - next patch will use it to determine if a certain extension was user set; - cpu_add_multi_ext_prop() is a new helper that will replace the qdev_property_add_static() calls that our macros are doing to populate user properties. The macro was renamed to ADD_CPU_MULTIEXT_PROPS_ARRAY() for clarity. Note that the non-extension properties in riscv_cpu_options[] still need to be declared via qdev(). Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-18-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 258 ++++++++++++++++++++++++++++----------------- 1 file changed, 159 insertions(+), 99 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 const struct isa_ext_data isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(xventanacondops, PRIV_VERSION_1_12_0, ext_XVentanaCondOps), }; +/* Hash that stores user set extensions */ +static GHashTable *multi_ext_user_opts; + static bool isa_ext_is_enabled(RISCVCPU *cpu, uint32_t ext_offset) { bool *ext_enabled = (void *)&cpu->cfg + ext_offset; @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_init(Object *obj) qdev_init_gpio_in(DEVICE(obj), riscv_cpu_set_irq, IRQ_LOCAL_MAX + IRQ_LOCAL_GUEST_MAX); #endif /* CONFIG_USER_ONLY */ + + multi_ext_user_opts = g_hash_table_new(NULL, g_direct_equal); } typedef struct RISCVCPUMisaExtConfig { @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_misa_properties(Object *cpu_obj) } } -static Property riscv_cpu_extensions[] = { +typedef struct RISCVCPUMultiExtConfig { + const char *name; + uint32_t offset; + bool enabled; +} RISCVCPUMultiExtConfig; + +#define MULTI_EXT_CFG_BOOL(_name, _prop, _defval) \ + {.name = _name, .offset = CPU_CFG_OFFSET(_prop), \ + .enabled = _defval} + +static RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { /* Defaults for standard extensions */ - DEFINE_PROP_BOOL("sscofpmf", RISCVCPU, cfg.ext_sscofpmf, false), - DEFINE_PROP_BOOL("Zifencei", RISCVCPU, cfg.ext_ifencei, true), - DEFINE_PROP_BOOL("Zicsr", RISCVCPU, cfg.ext_icsr, true), - DEFINE_PROP_BOOL("Zihintntl", RISCVCPU, cfg.ext_zihintntl, true), - DEFINE_PROP_BOOL("Zihintpause", RISCVCPU, cfg.ext_zihintpause, true), - DEFINE_PROP_BOOL("Zawrs", RISCVCPU, cfg.ext_zawrs, true), - DEFINE_PROP_BOOL("Zfa", RISCVCPU, cfg.ext_zfa, true), - DEFINE_PROP_BOOL("Zfh", RISCVCPU, cfg.ext_zfh, false), - DEFINE_PROP_BOOL("Zfhmin", RISCVCPU, cfg.ext_zfhmin, false), - DEFINE_PROP_BOOL("Zve32f", RISCVCPU, cfg.ext_zve32f, false), - DEFINE_PROP_BOOL("Zve64f", RISCVCPU, cfg.ext_zve64f, false), - DEFINE_PROP_BOOL("Zve64d", RISCVCPU, cfg.ext_zve64d, false), - DEFINE_PROP_BOOL("sstc", RISCVCPU, cfg.ext_sstc, true), - - DEFINE_PROP_BOOL("smstateen", RISCVCPU, cfg.ext_smstateen, false), - DEFINE_PROP_BOOL("svadu", RISCVCPU, cfg.ext_svadu, true), - 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), - DEFINE_PROP_BOOL("zbc", RISCVCPU, cfg.ext_zbc, true), - DEFINE_PROP_BOOL("zbkb", RISCVCPU, cfg.ext_zbkb, false), - DEFINE_PROP_BOOL("zbkc", RISCVCPU, cfg.ext_zbkc, false), - DEFINE_PROP_BOOL("zbkx", RISCVCPU, cfg.ext_zbkx, false), - DEFINE_PROP_BOOL("zbs", RISCVCPU, cfg.ext_zbs, true), - DEFINE_PROP_BOOL("zk", RISCVCPU, cfg.ext_zk, false), - DEFINE_PROP_BOOL("zkn", RISCVCPU, cfg.ext_zkn, false), - DEFINE_PROP_BOOL("zknd", RISCVCPU, cfg.ext_zknd, false), - DEFINE_PROP_BOOL("zkne", RISCVCPU, cfg.ext_zkne, false), - DEFINE_PROP_BOOL("zknh", RISCVCPU, cfg.ext_zknh, false), - DEFINE_PROP_BOOL("zkr", RISCVCPU, cfg.ext_zkr, false), - DEFINE_PROP_BOOL("zks", RISCVCPU, cfg.ext_zks, false), - DEFINE_PROP_BOOL("zksed", RISCVCPU, cfg.ext_zksed, false), - DEFINE_PROP_BOOL("zksh", RISCVCPU, cfg.ext_zksh, false), - DEFINE_PROP_BOOL("zkt", RISCVCPU, cfg.ext_zkt, false), - - DEFINE_PROP_BOOL("zdinx", RISCVCPU, cfg.ext_zdinx, false), - DEFINE_PROP_BOOL("zfinx", RISCVCPU, cfg.ext_zfinx, false), - DEFINE_PROP_BOOL("zhinx", RISCVCPU, cfg.ext_zhinx, false), - DEFINE_PROP_BOOL("zhinxmin", RISCVCPU, cfg.ext_zhinxmin, false), - - DEFINE_PROP_BOOL("zicbom", RISCVCPU, cfg.ext_icbom, true), - DEFINE_PROP_BOOL("zicboz", RISCVCPU, cfg.ext_icboz, true), - - DEFINE_PROP_BOOL("zmmul", RISCVCPU, cfg.ext_zmmul, false), - - DEFINE_PROP_BOOL("zca", RISCVCPU, cfg.ext_zca, false), - DEFINE_PROP_BOOL("zcb", RISCVCPU, cfg.ext_zcb, false), - DEFINE_PROP_BOOL("zcd", RISCVCPU, cfg.ext_zcd, false), - DEFINE_PROP_BOOL("zce", RISCVCPU, cfg.ext_zce, false), - DEFINE_PROP_BOOL("zcf", RISCVCPU, cfg.ext_zcf, false), - DEFINE_PROP_BOOL("zcmp", RISCVCPU, cfg.ext_zcmp, false), - DEFINE_PROP_BOOL("zcmt", RISCVCPU, cfg.ext_zcmt, false), - DEFINE_PROP_BOOL("zicond", RISCVCPU, cfg.ext_zicond, false), + MULTI_EXT_CFG_BOOL("sscofpmf", ext_sscofpmf, false), + MULTI_EXT_CFG_BOOL("Zifencei", ext_ifencei, true), + MULTI_EXT_CFG_BOOL("Zicsr", ext_icsr, 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), + MULTI_EXT_CFG_BOOL("sstc", ext_sstc, true), + + MULTI_EXT_CFG_BOOL("smstateen", ext_smstateen, false), + MULTI_EXT_CFG_BOOL("svadu", ext_svadu, true), + 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("zba", ext_zba, true), + MULTI_EXT_CFG_BOOL("zbb", ext_zbb, true), + MULTI_EXT_CFG_BOOL("zbc", ext_zbc, true), + MULTI_EXT_CFG_BOOL("zbkb", ext_zbkb, false), + MULTI_EXT_CFG_BOOL("zbkc", ext_zbkc, false), + MULTI_EXT_CFG_BOOL("zbkx", ext_zbkx, false), + MULTI_EXT_CFG_BOOL("zbs", ext_zbs, true), + MULTI_EXT_CFG_BOOL("zk", ext_zk, false), + MULTI_EXT_CFG_BOOL("zkn", ext_zkn, false), + MULTI_EXT_CFG_BOOL("zknd", ext_zknd, false), + MULTI_EXT_CFG_BOOL("zkne", ext_zkne, false), + MULTI_EXT_CFG_BOOL("zknh", ext_zknh, false), + MULTI_EXT_CFG_BOOL("zkr", ext_zkr, false), + MULTI_EXT_CFG_BOOL("zks", ext_zks, false), + MULTI_EXT_CFG_BOOL("zksed", ext_zksed, false), + MULTI_EXT_CFG_BOOL("zksh", ext_zksh, false), + MULTI_EXT_CFG_BOOL("zkt", ext_zkt, false), + + MULTI_EXT_CFG_BOOL("zdinx", ext_zdinx, false), + MULTI_EXT_CFG_BOOL("zfinx", ext_zfinx, false), + MULTI_EXT_CFG_BOOL("zhinx", ext_zhinx, false), + MULTI_EXT_CFG_BOOL("zhinxmin", ext_zhinxmin, false), + + MULTI_EXT_CFG_BOOL("zicbom", ext_icbom, true), + MULTI_EXT_CFG_BOOL("zicboz", ext_icboz, true), + + MULTI_EXT_CFG_BOOL("zmmul", ext_zmmul, false), + + MULTI_EXT_CFG_BOOL("zca", ext_zca, false), + MULTI_EXT_CFG_BOOL("zcb", ext_zcb, false), + MULTI_EXT_CFG_BOOL("zcd", ext_zcd, false), + MULTI_EXT_CFG_BOOL("zce", ext_zce, false), + MULTI_EXT_CFG_BOOL("zcf", ext_zcf, false), + MULTI_EXT_CFG_BOOL("zcmp", ext_zcmp, false), + MULTI_EXT_CFG_BOOL("zcmt", ext_zcmt, false), + MULTI_EXT_CFG_BOOL("zicond", ext_zicond, false), DEFINE_PROP_END_OF_LIST(), }; -static Property riscv_cpu_vendor_exts[] = { - DEFINE_PROP_BOOL("xtheadba", RISCVCPU, cfg.ext_xtheadba, false), - DEFINE_PROP_BOOL("xtheadbb", RISCVCPU, cfg.ext_xtheadbb, false), - DEFINE_PROP_BOOL("xtheadbs", RISCVCPU, cfg.ext_xtheadbs, false), - DEFINE_PROP_BOOL("xtheadcmo", RISCVCPU, cfg.ext_xtheadcmo, false), - DEFINE_PROP_BOOL("xtheadcondmov", RISCVCPU, cfg.ext_xtheadcondmov, false), - DEFINE_PROP_BOOL("xtheadfmemidx", RISCVCPU, cfg.ext_xtheadfmemidx, false), - DEFINE_PROP_BOOL("xtheadfmv", RISCVCPU, cfg.ext_xtheadfmv, false), - DEFINE_PROP_BOOL("xtheadmac", RISCVCPU, cfg.ext_xtheadmac, false), - DEFINE_PROP_BOOL("xtheadmemidx", RISCVCPU, cfg.ext_xtheadmemidx, false), - DEFINE_PROP_BOOL("xtheadmempair", RISCVCPU, cfg.ext_xtheadmempair, false), - DEFINE_PROP_BOOL("xtheadsync", RISCVCPU, cfg.ext_xtheadsync, false), - DEFINE_PROP_BOOL("xventanacondops", RISCVCPU, cfg.ext_XVentanaCondOps, false), +static RISCVCPUMultiExtConfig riscv_cpu_vendor_exts[] = { + MULTI_EXT_CFG_BOOL("xtheadba", ext_xtheadba, false), + MULTI_EXT_CFG_BOOL("xtheadbb", ext_xtheadbb, false), + MULTI_EXT_CFG_BOOL("xtheadbs", ext_xtheadbs, false), + MULTI_EXT_CFG_BOOL("xtheadcmo", ext_xtheadcmo, false), + MULTI_EXT_CFG_BOOL("xtheadcondmov", ext_xtheadcondmov, false), + MULTI_EXT_CFG_BOOL("xtheadfmemidx", ext_xtheadfmemidx, false), + MULTI_EXT_CFG_BOOL("xtheadfmv", ext_xtheadfmv, false), + MULTI_EXT_CFG_BOOL("xtheadmac", ext_xtheadmac, false), + MULTI_EXT_CFG_BOOL("xtheadmemidx", ext_xtheadmemidx, false), + MULTI_EXT_CFG_BOOL("xtheadmempair", ext_xtheadmempair, false), + MULTI_EXT_CFG_BOOL("xtheadsync", ext_xtheadsync, false), + MULTI_EXT_CFG_BOOL("xventanacondops", ext_XVentanaCondOps, false), DEFINE_PROP_END_OF_LIST(), }; /* These are experimental so mark with 'x-' */ -static Property riscv_cpu_experimental_exts[] = { +static RISCVCPUMultiExtConfig riscv_cpu_experimental_exts[] = { /* ePMP 0.9.3 */ - DEFINE_PROP_BOOL("x-epmp", RISCVCPU, cfg.epmp, false), - DEFINE_PROP_BOOL("x-smaia", RISCVCPU, cfg.ext_smaia, false), - DEFINE_PROP_BOOL("x-ssaia", RISCVCPU, cfg.ext_ssaia, false), + MULTI_EXT_CFG_BOOL("x-epmp", epmp, false), + MULTI_EXT_CFG_BOOL("x-smaia", ext_smaia, false), + MULTI_EXT_CFG_BOOL("x-ssaia", ext_ssaia, false), - DEFINE_PROP_BOOL("x-zvfh", RISCVCPU, cfg.ext_zvfh, false), - DEFINE_PROP_BOOL("x-zvfhmin", RISCVCPU, cfg.ext_zvfhmin, false), + MULTI_EXT_CFG_BOOL("x-zvfh", ext_zvfh, false), + MULTI_EXT_CFG_BOOL("x-zvfhmin", ext_zvfhmin, false), - DEFINE_PROP_BOOL("x-zfbfmin", RISCVCPU, cfg.ext_zfbfmin, false), - DEFINE_PROP_BOOL("x-zvfbfmin", RISCVCPU, cfg.ext_zvfbfmin, false), - DEFINE_PROP_BOOL("x-zvfbfwma", RISCVCPU, cfg.ext_zvfbfwma, false), + MULTI_EXT_CFG_BOOL("x-zfbfmin", ext_zfbfmin, false), + MULTI_EXT_CFG_BOOL("x-zvfbfmin", ext_zvfbfmin, false), + MULTI_EXT_CFG_BOOL("x-zvfbfwma", ext_zvfbfwma, false), /* Vector cryptography extensions */ - DEFINE_PROP_BOOL("x-zvbb", RISCVCPU, cfg.ext_zvbb, false), - DEFINE_PROP_BOOL("x-zvbc", RISCVCPU, cfg.ext_zvbc, false), - DEFINE_PROP_BOOL("x-zvkg", RISCVCPU, cfg.ext_zvkg, false), - DEFINE_PROP_BOOL("x-zvkned", RISCVCPU, cfg.ext_zvkned, false), - DEFINE_PROP_BOOL("x-zvknha", RISCVCPU, cfg.ext_zvknha, false), - DEFINE_PROP_BOOL("x-zvknhb", RISCVCPU, cfg.ext_zvknhb, false), - DEFINE_PROP_BOOL("x-zvksed", RISCVCPU, cfg.ext_zvksed, false), - DEFINE_PROP_BOOL("x-zvksh", RISCVCPU, cfg.ext_zvksh, false), + MULTI_EXT_CFG_BOOL("x-zvbb", ext_zvbb, false), + MULTI_EXT_CFG_BOOL("x-zvbc", ext_zvbc, false), + MULTI_EXT_CFG_BOOL("x-zvkg", ext_zvkg, false), + MULTI_EXT_CFG_BOOL("x-zvkned", ext_zvkned, false), + MULTI_EXT_CFG_BOOL("x-zvknha", ext_zvknha, false), + MULTI_EXT_CFG_BOOL("x-zvknhb", ext_zvknhb, false), + MULTI_EXT_CFG_BOOL("x-zvksed", ext_zvksed, false), + MULTI_EXT_CFG_BOOL("x-zvksh", ext_zvksh, false), DEFINE_PROP_END_OF_LIST(), }; @@ -XXX,XX +XXX,XX @@ static Property riscv_cpu_options[] = { DEFINE_PROP_END_OF_LIST(), }; -static void riscv_cpu_add_qdev_prop_array(DeviceState *dev, Property *array) +static void cpu_set_multi_ext_cfg(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + const RISCVCPUMultiExtConfig *multi_ext_cfg = opaque; + bool value; + + if (!visit_type_bool(v, name, &value, errp)) { + return; + } + + isa_ext_update_enabled(RISCV_CPU(obj), multi_ext_cfg->offset, value); + + g_hash_table_insert(multi_ext_user_opts, + GUINT_TO_POINTER(multi_ext_cfg->offset), + (gpointer)value); +} + +static void cpu_get_multi_ext_cfg(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + const RISCVCPUMultiExtConfig *multi_ext_cfg = opaque; + bool value = isa_ext_is_enabled(RISCV_CPU(obj), multi_ext_cfg->offset); + + visit_type_bool(v, name, &value, errp); +} + +static void cpu_add_multi_ext_prop(Object *cpu_obj, + RISCVCPUMultiExtConfig *multi_cfg) +{ + object_property_add(cpu_obj, multi_cfg->name, "bool", + cpu_get_multi_ext_cfg, + cpu_set_multi_ext_cfg, + NULL, (void *)multi_cfg); + + /* + * Set def val directly instead of using + * object_property_set_bool() to save the set() + * callback hash for user inputs. + */ + isa_ext_update_enabled(RISCV_CPU(cpu_obj), multi_cfg->offset, + multi_cfg->enabled); +} + +static void riscv_cpu_add_multiext_prop_array(Object *obj, + RISCVCPUMultiExtConfig *array) { g_assert(array); - for (Property *prop = array; prop && prop->name; prop++) { - qdev_property_add_static(dev, prop); + for (RISCVCPUMultiExtConfig *prop = array; prop && prop->name; prop++) { + cpu_add_multi_ext_prop(obj, prop); } } @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_kvm_unavail_prop(Object *obj, const char *prop_name) } static void riscv_cpu_add_kvm_unavail_prop_array(Object *obj, - Property *array) + RISCVCPUMultiExtConfig *array) { g_assert(array); - for (Property *prop = array; prop && prop->name; prop++) { + for (RISCVCPUMultiExtConfig *prop = array; prop && prop->name; prop++) { riscv_cpu_add_kvm_unavail_prop(obj, prop->name); } } @@ -XXX,XX +XXX,XX @@ void kvm_riscv_cpu_add_kvm_properties(Object *obj) */ static void riscv_cpu_add_user_properties(Object *obj) { - DeviceState *dev = DEVICE(obj); - #ifndef CONFIG_USER_ONLY riscv_add_satp_mode_properties(obj); @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_user_properties(Object *obj) riscv_cpu_add_misa_properties(obj); - riscv_cpu_add_qdev_prop_array(dev, riscv_cpu_extensions); - riscv_cpu_add_qdev_prop_array(dev, riscv_cpu_options); - riscv_cpu_add_qdev_prop_array(dev, riscv_cpu_vendor_exts); - riscv_cpu_add_qdev_prop_array(dev, riscv_cpu_experimental_exts); + riscv_cpu_add_multiext_prop_array(obj, riscv_cpu_extensions); + riscv_cpu_add_multiext_prop_array(obj, riscv_cpu_vendor_exts); + riscv_cpu_add_multiext_prop_array(obj, riscv_cpu_experimental_exts); + + for (Property *prop = riscv_cpu_options; prop && prop->name; prop++) { + qdev_property_add_static(DEVICE(obj), prop); + } } /* @@ -XXX,XX +XXX,XX @@ static void riscv_init_max_cpu_extensions(Object *obj) { RISCVCPU *cpu = RISCV_CPU(obj); CPURISCVState *env = &cpu->env; - Property *prop; + RISCVCPUMultiExtConfig *prop; /* Enable RVG, RVJ and RVV that are disabled by default */ set_misa(env, env->misa_mxl, env->misa_ext | RVG | RVJ | RVV); -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Before adding support to detect if an extension was user set we need to handle how we're enabling extensions in riscv_init_max_cpu_extensions(). object_property_set_bool() calls the set() callback for the property, and we're going to use this callback to set the 'multi_ext_user_opts' hash. This means that, as is today, all extensions we're setting for the 'max' CPU will be seen as user set in the future. Let's change set_bool() to isa_ext_update_enabled() that will just enable/disable the flag on a certain offset. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-19-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 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_init_max_cpu_extensions(Object *obj) set_misa(env, env->misa_mxl, env->misa_ext | RVG | RVJ | RVV); for (prop = riscv_cpu_extensions; prop && prop->name; prop++) { - object_property_set_bool(obj, prop->name, true, NULL); + isa_ext_update_enabled(cpu, prop->offset, true); } /* set vector version */ env->vext_ver = VEXT_VERSION_1_00_0; /* Zfinx is not compatible with F. Disable it */ - object_property_set_bool(obj, "zfinx", false, NULL); - object_property_set_bool(obj, "zdinx", false, NULL); - object_property_set_bool(obj, "zhinx", false, NULL); - object_property_set_bool(obj, "zhinxmin", false, NULL); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zfinx), false); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zdinx), false); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zhinx), false); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zhinxmin), false); - object_property_set_bool(obj, "zce", false, NULL); - object_property_set_bool(obj, "zcmp", false, NULL); - object_property_set_bool(obj, "zcmt", false, NULL); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zce), false); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zcmp), false); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zcmt), false); if (env->misa_mxl != MXL_RV32) { - object_property_set_bool(obj, "zcf", false, NULL); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zcf), false); } } -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Add a new cpu_cfg_ext_is_user_set() helper to check if an extension was set by the user in the command line. Use it inside cpu_cfg_ext_auto_update() to verify if the user set a certain extension and, if that's the case, do not change its value. This will make us honor user choice instead of overwriting the values. Users will then be informed whether they're using an incompatible set of extensions instead of QEMU setting a magic value that works. The reason why we're not implementing user choice for MISA extensions right now is because, today, we do not silently change any MISA bit during realize() time (we do warn when enabling bits if RVG is enabled). We do that - a lot - with multi-letter extensions though, so we're handling the most immediate concern first. After this patch, we'll now error out if the user explicitly set 'zce' to true and 'zca' to false: $ ./build/qemu-system-riscv64 -M virt -cpu rv64,zce=true,zca=false -nographic qemu-system-riscv64: Zcf/Zcd/Zcb/Zcmp/Zcmt extensions require Zca extension This didn't happen before because we were enabling 'zca' if 'zce' was enabled regardless if the user set 'zca' to false. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-20-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 16 ++++++++++++++++ 1 file changed, 16 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 @@ static int cpu_cfg_ext_get_min_version(uint32_t ext_offset) g_assert_not_reached(); } +static bool cpu_cfg_ext_is_user_set(uint32_t ext_offset) +{ + return g_hash_table_contains(multi_ext_user_opts, + GUINT_TO_POINTER(ext_offset)); +} + static void cpu_cfg_ext_auto_update(RISCVCPU *cpu, uint32_t ext_offset, bool value) { @@ -XXX,XX +XXX,XX @@ static void cpu_cfg_ext_auto_update(RISCVCPU *cpu, uint32_t ext_offset, return; } + if (cpu_cfg_ext_is_user_set(ext_offset)) { + return; + } + if (value && env->priv_ver != PRIV_VERSION_LATEST) { /* Do not enable it if priv_ver is older than min_version */ min_version = cpu_cfg_ext_get_min_version(ext_offset); @@ -XXX,XX +XXX,XX @@ static RISCVCPUMisaExtConfig misa_ext_cfgs[] = { MISA_CFG(RVG, false), }; +/* + * We do not support user choice tracking for MISA + * extensions yet because, so far, we do not silently + * change MISA bits during realize() (RVG enables MISA + * bits but the user is warned about it). + */ static void riscv_cpu_add_misa_properties(Object *cpu_obj) { int i; -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Enabling RVG will enable a set of extensions that we're not checking if the user was okay enabling or not. And in this case we want to error out, instead of ignoring, otherwise we will be inconsistent enabling RVG without all its extensions. After this patch, disabling ifencei or icsr while enabling RVG will result in error: $ ./build/qemu-system-riscv64 -M virt -cpu rv64,g=true,Zifencei=false --nographic qemu-system-riscv64: RVG requires Zifencei but user set Zifencei to false Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20230912132423.268494-21-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 18 ++++++++++++++++-- 1 file changed, 16 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 @@ void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) riscv_has_ext(env, RVA) && riscv_has_ext(env, RVF) && riscv_has_ext(env, RVD) && cpu->cfg.ext_icsr && cpu->cfg.ext_ifencei)) { + + if (cpu_cfg_ext_is_user_set(CPU_CFG_OFFSET(ext_icsr)) && + !cpu->cfg.ext_icsr) { + error_setg(errp, "RVG requires Zicsr but user set Zicsr to false"); + return; + } + + if (cpu_cfg_ext_is_user_set(CPU_CFG_OFFSET(ext_ifencei)) && + !cpu->cfg.ext_ifencei) { + error_setg(errp, "RVG requires Zifencei but user set " + "Zifencei to false"); + return; + } + warn_report("Setting G will also set IMAFD_Zicsr_Zifencei"); - cpu->cfg.ext_icsr = true; - cpu->cfg.ext_ifencei = true; + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_icsr), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_ifencei), true); env->misa_ext |= RVI | RVM | RVA | RVF | RVD; env->misa_ext_mask |= RVI | RVM | RVA | RVF | RVD; -- 2.41.0 From: "liguang.zhang" <liguang.zhang@hexintek.com> This patch fixes guest reboot errors when using KVM. There are two issues when rebooting a guest using KVM 1. When the guest initiates a reboot the host is unable to stop the vcpu 2. When running a SMP guest the qemu monitor system_reset causes a vcpu crash This can be fixed by clearing the CSR values at reset and syncing the MPSTATE with the host. Signed-off-by: liguang.zhang <liguang.zhang@hexintek.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230913091332.17355-1-18622748025@163.com> [ Changes by AF - Fixup commit message - Fixup patch style ] Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/kvm_riscv.h | 1 + target/riscv/kvm.c | 44 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 45 insertions(+) diff --git a/target/riscv/kvm_riscv.h b/target/riscv/kvm_riscv.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm_riscv.h +++ b/target/riscv/kvm_riscv.h @@ -XXX,XX +XXX,XX @@ void kvm_riscv_aia_create(MachineState *machine, uint64_t group_shift, uint64_t aplic_base, uint64_t imsic_base, uint64_t guest_num); void riscv_kvm_aplic_request(void *opaque, int irq, int level); +int kvm_riscv_sync_mpstate_to_kvm(RISCVCPU *cpu, int state); #endif diff --git a/target/riscv/kvm.c b/target/riscv/kvm.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm.c +++ b/target/riscv/kvm.c @@ -XXX,XX +XXX,XX @@ void riscv_kvm_aplic_request(void *opaque, int irq, int level) kvm_set_irq(kvm_state, irq, !!level); } +static bool cap_has_mp_state; + static uint64_t kvm_riscv_reg_id(CPURISCVState *env, uint64_t type, uint64_t idx) { @@ -XXX,XX +XXX,XX @@ int kvm_arch_get_registers(CPUState *cs) return ret; } +int kvm_riscv_sync_mpstate_to_kvm(RISCVCPU *cpu, int state) +{ + if (cap_has_mp_state) { + struct kvm_mp_state mp_state = { + .mp_state = state + }; + + int ret = kvm_vcpu_ioctl(CPU(cpu), KVM_SET_MP_STATE, &mp_state); + if (ret) { + fprintf(stderr, "%s: failed to sync MP_STATE %d/%s\n", + __func__, ret, strerror(-ret)); + return -1; + } + } + + return 0; +} + int kvm_arch_put_registers(CPUState *cs, int level) { int ret = 0; @@ -XXX,XX +XXX,XX @@ int kvm_arch_put_registers(CPUState *cs, int level) return ret; } + if (KVM_PUT_RESET_STATE == level) { + RISCVCPU *cpu = RISCV_CPU(cs); + if (cs->cpu_index == 0) { + ret = kvm_riscv_sync_mpstate_to_kvm(cpu, KVM_MP_STATE_RUNNABLE); + } else { + ret = kvm_riscv_sync_mpstate_to_kvm(cpu, KVM_MP_STATE_STOPPED); + } + if (ret) { + return ret; + } + } + return ret; } @@ -XXX,XX +XXX,XX @@ int kvm_arch_get_default_type(MachineState *ms) int kvm_arch_init(MachineState *ms, KVMState *s) { + cap_has_mp_state = kvm_check_extension(s, KVM_CAP_MP_STATE); return 0; } @@ -XXX,XX +XXX,XX @@ int kvm_arch_handle_exit(CPUState *cs, struct kvm_run *run) void kvm_riscv_reset_vcpu(RISCVCPU *cpu) { CPURISCVState *env = &cpu->env; + int i; if (!kvm_enabled()) { return; } + for (i = 0; i < 32; i++) { + env->gpr[i] = 0; + } env->pc = cpu->env.kernel_addr; env->gpr[10] = kvm_arch_vcpu_id(CPU(cpu)); /* a0 */ env->gpr[11] = cpu->env.fdt_addr; /* a1 */ env->satp = 0; + env->mie = 0; + env->stvec = 0; + env->sscratch = 0; + env->sepc = 0; + env->scause = 0; + env->stval = 0; + env->mip = 0; } void kvm_riscv_set_irq(RISCVCPU *cpu, int irq, int level) -- 2.41.0 From: Alvin Chang <vivahavey@gmail.com> Fix the inverted order of pmpaddr13 and pmpaddr14 in csr_name(). Signed-off-by: Alvin Chang <alvinga@andestech.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230907084500.328-1-alvinga@andestech.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- disas/riscv.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/disas/riscv.c b/disas/riscv.c index XXXXXXX..XXXXXXX 100644 --- a/disas/riscv.c +++ b/disas/riscv.c @@ -XXX,XX +XXX,XX @@ static const char *csr_name(int csrno) case 0x03ba: return "pmpaddr10"; case 0x03bb: return "pmpaddr11"; case 0x03bc: return "pmpaddr12"; - case 0x03bd: return "pmpaddr14"; - case 0x03be: return "pmpaddr13"; + case 0x03bd: return "pmpaddr13"; + case 0x03be: return "pmpaddr14"; case 0x03bf: return "pmpaddr15"; case 0x0780: return "mtohost"; case 0x0781: return "mfromhost"; -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> target/riscv/cpu.c needs to handle all possible accelerators (TCG and KVM at this moment) during both init() and realize() time. This forces us to resort to a lot of "if tcg" and "if kvm" throughout the code, which isn't wrong, but can get cluttered over time. Splitting acceleration specific code from cpu.c to its own file will help to declutter the existing code and it will also make it easier to support KVM/TCG only builds in the future. We'll start by adding a new subdir called 'tcg' and a new file called 'tcg-cpu.c'. This file will be used to introduce a new accelerator class for TCG acceleration in RISC-V, allowing us to center all TCG exclusive code in its file instead of using 'cpu.c' for everything. This design is inpired by the work Claudio Fontana did in x86 a few years ago in commit f5cc5a5c1 ("i386: split cpu accelerators from cpu.c, using AccelCPUClass"). To avoid moving too much code at once we'll start by adding the new file and TCG AccelCPUClass declaration. The 'class_init' from the accel class will init 'tcg_ops', relieving the common riscv_cpu_class_init() from doing it. 'riscv_tcg_ops' is being exported from 'cpu.c' for now to avoid having to deal with moving code and files around right now. We'll focus on decoupling the realize() logic first. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-2-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 4 +++ target/riscv/cpu.c | 5 +--- target/riscv/tcg/tcg-cpu.c | 58 ++++++++++++++++++++++++++++++++++++ target/riscv/meson.build | 2 ++ target/riscv/tcg/meson.build | 2 ++ 5 files changed, 67 insertions(+), 4 deletions(-) create mode 100644 target/riscv/tcg/tcg-cpu.c create mode 100644 target/riscv/tcg/meson.build 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_pmu_event_idx { RISCV_PMU_EVENT_CACHE_ITLB_PREFETCH_MISS = 0x10021, }; +/* Export tcg_ops until we move everything to tcg/tcg-cpu.c */ +#include "hw/core/tcg-cpu-ops.h" +extern const struct TCGCPUOps riscv_tcg_ops; + /* CSR function table */ extern riscv_csr_operations csr_ops[CSR_TABLE_SIZE]; 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 struct SysemuCPUOps riscv_sysemu_ops = { }; #endif -#include "hw/core/tcg-cpu-ops.h" - -static const struct TCGCPUOps riscv_tcg_ops = { +const struct TCGCPUOps riscv_tcg_ops = { .initialize = riscv_translate_init, .synchronize_from_tb = riscv_cpu_synchronize_from_tb, .restore_state_to_opc = riscv_restore_state_to_opc, @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_class_init(ObjectClass *c, void *data) #endif cc->gdb_arch_name = riscv_gdb_arch_name; cc->gdb_get_dynamic_xml = riscv_gdb_get_dynamic_xml; - cc->tcg_ops = &riscv_tcg_ops; object_class_property_add(c, "mvendorid", "uint32", cpu_get_mvendorid, cpu_set_mvendorid, NULL, NULL); diff --git a/target/riscv/tcg/tcg-cpu.c b/target/riscv/tcg/tcg-cpu.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/riscv/tcg/tcg-cpu.c @@ -XXX,XX +XXX,XX @@ +/* + * riscv TCG cpu class initialization + * + * Copyright (c) 2016-2017 Sagar Karandikar, sagark@eecs.berkeley.edu + * Copyright (c) 2017-2018 SiFive, Inc. + * + * 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 "cpu.h" +#include "qemu/accel.h" +#include "hw/core/accel-cpu.h" + +static void tcg_cpu_init_ops(AccelCPUClass *accel_cpu, CPUClass *cc) +{ + /* + * All cpus use the same set of operations. + * riscv_tcg_ops is being imported from cpu.c for now. + */ + cc->tcg_ops = &riscv_tcg_ops; +} + +static void tcg_cpu_class_init(CPUClass *cc) +{ + cc->init_accel_cpu = tcg_cpu_init_ops; +} + +static void tcg_cpu_accel_class_init(ObjectClass *oc, void *data) +{ + AccelCPUClass *acc = ACCEL_CPU_CLASS(oc); + + acc->cpu_class_init = tcg_cpu_class_init; +} + +static const TypeInfo tcg_cpu_accel_type_info = { + .name = ACCEL_CPU_NAME("tcg"), + + .parent = TYPE_ACCEL_CPU, + .class_init = tcg_cpu_accel_class_init, + .abstract = true, +}; + +static void tcg_cpu_accel_register_types(void) +{ + type_register_static(&tcg_cpu_accel_type_info); +} +type_init(tcg_cpu_accel_register_types); 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 @@ riscv_system_ss.add(files( 'riscv-qmp-cmds.c', )) +subdir('tcg') + target_arch += {'riscv': riscv_ss} target_system_arch += {'riscv': riscv_system_ss} diff --git a/target/riscv/tcg/meson.build b/target/riscv/tcg/meson.build new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/riscv/tcg/meson.build @@ -XXX,XX +XXX,XX @@ +riscv_ss.add(when: 'CONFIG_TCG', if_true: files( + 'tcg-cpu.c')) -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> riscv_cpu_realize_tcg() was added to allow TCG cpus to have a different realize() path during the common riscv_cpu_realize(), making it a good choice to start moving TCG exclusive code to tcg-cpu.c. Rename it to tcg_cpu_realizefn() and assign it as a implementation of accel::cpu_realizefn(). tcg_cpu_realizefn() will then be called during riscv_cpu_realize() via cpu_exec_realizefn(). We'll use a similar approach with KVM in the near future. riscv_cpu_validate_set_extensions() is too big and with too many dependencies to be moved in this same patch. We'll do that next. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Message-ID: <20230925175709.35696-3-dbarboza@ventanamicro.com> [ Changes by AF: - Renames to fix build failures after rebase ] Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 128 ----------------------------------- target/riscv/tcg/tcg-cpu.c | 133 +++++++++++++++++++++++++++++++++++++ 2 files changed, 133 insertions(+), 128 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 @@ #include "qemu/log.h" #include "cpu.h" #include "cpu_vendorid.h" -#include "pmu.h" #include "internals.h" -#include "time_helper.h" #include "exec/exec-all.h" #include "qapi/error.h" #include "qapi/visitor.h" @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_validate_v(CPURISCVState *env, RISCVCPUConfig *cfg, } } -static void riscv_cpu_validate_priv_spec(RISCVCPU *cpu, Error **errp) -{ - CPURISCVState *env = &cpu->env; - int priv_version = -1; - - if (cpu->cfg.priv_spec) { - if (!g_strcmp0(cpu->cfg.priv_spec, "v1.12.0")) { - priv_version = PRIV_VERSION_1_12_0; - } else if (!g_strcmp0(cpu->cfg.priv_spec, "v1.11.0")) { - priv_version = PRIV_VERSION_1_11_0; - } else if (!g_strcmp0(cpu->cfg.priv_spec, "v1.10.0")) { - priv_version = PRIV_VERSION_1_10_0; - } else { - error_setg(errp, - "Unsupported privilege spec version '%s'", - cpu->cfg.priv_spec); - return; - } - - env->priv_ver = priv_version; - } -} - static void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu) { CPURISCVState *env = &cpu->env; @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu) } } -static void riscv_cpu_validate_misa_mxl(RISCVCPU *cpu, Error **errp) -{ - RISCVCPUClass *mcc = RISCV_CPU_GET_CLASS(cpu); - CPUClass *cc = CPU_CLASS(mcc); - CPURISCVState *env = &cpu->env; - - /* Validate that MISA_MXL is set properly. */ - switch (env->misa_mxl_max) { -#ifdef TARGET_RISCV64 - case MXL_RV64: - case MXL_RV128: - cc->gdb_core_xml_file = "riscv-64bit-cpu.xml"; - break; -#endif - case MXL_RV32: - cc->gdb_core_xml_file = "riscv-32bit-cpu.xml"; - break; - default: - g_assert_not_reached(); - } - - if (env->misa_mxl_max != env->misa_mxl) { - error_setg(errp, "misa_mxl_max must be equal to misa_mxl"); - return; - } -} - /* * Check consistency between chosen extensions while setting * cpu->cfg accordingly. @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_finalize_features(RISCVCPU *cpu, Error **errp) #endif } -static void riscv_cpu_validate_misa_priv(CPURISCVState *env, Error **errp) -{ - if (riscv_has_ext(env, RVH) && env->priv_ver < PRIV_VERSION_1_12_0) { - error_setg(errp, "H extension requires priv spec 1.12.0"); - return; - } -} - -static void riscv_cpu_realize_tcg(DeviceState *dev, Error **errp) -{ - RISCVCPU *cpu = RISCV_CPU(dev); - CPURISCVState *env = &cpu->env; - Error *local_err = NULL; - - if (object_dynamic_cast(OBJECT(dev), TYPE_RISCV_CPU_HOST)) { - error_setg(errp, "'host' CPU is not compatible with TCG acceleration"); - return; - } - - riscv_cpu_validate_misa_mxl(cpu, &local_err); - if (local_err != NULL) { - error_propagate(errp, local_err); - return; - } - - riscv_cpu_validate_priv_spec(cpu, &local_err); - if (local_err != NULL) { - error_propagate(errp, local_err); - return; - } - - riscv_cpu_validate_misa_priv(env, &local_err); - if (local_err != NULL) { - error_propagate(errp, local_err); - return; - } - - if (cpu->cfg.epmp && !cpu->cfg.pmp) { - /* - * Enhanced PMP should only be available - * on harts with PMP support - */ - error_setg(errp, "Invalid configuration: EPMP requires PMP support"); - return; - } - - riscv_cpu_validate_set_extensions(cpu, &local_err); - if (local_err != NULL) { - error_propagate(errp, local_err); - return; - } - -#ifndef CONFIG_USER_ONLY - CPU(dev)->tcg_cflags |= CF_PCREL; - - if (cpu->cfg.ext_sstc) { - riscv_timer_init(cpu); - } - - if (cpu->cfg.pmu_num) { - if (!riscv_pmu_init(cpu, cpu->cfg.pmu_num) && cpu->cfg.ext_sscofpmf) { - cpu->pmu_timer = timer_new_ns(QEMU_CLOCK_VIRTUAL, - riscv_pmu_timer_cb, cpu); - } - } -#endif -} - static void riscv_cpu_realize(DeviceState *dev, Error **errp) { CPUState *cs = CPU(dev); @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp) return; } - if (tcg_enabled()) { - riscv_cpu_realize_tcg(dev, &local_err); - if (local_err != NULL) { - error_propagate(errp, local_err); - return; - } - } - riscv_cpu_finalize_features(cpu, &local_err); if (local_err != NULL) { error_propagate(errp, local_err); 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 @@ */ #include "qemu/osdep.h" +#include "exec/exec-all.h" #include "cpu.h" +#include "pmu.h" +#include "time_helper.h" +#include "qapi/error.h" #include "qemu/accel.h" #include "hw/core/accel-cpu.h" + +static void riscv_cpu_validate_misa_priv(CPURISCVState *env, Error **errp) +{ + if (riscv_has_ext(env, RVH) && env->priv_ver < PRIV_VERSION_1_12_0) { + error_setg(errp, "H extension requires priv spec 1.12.0"); + return; + } +} + +static void riscv_cpu_validate_misa_mxl(RISCVCPU *cpu, Error **errp) +{ + RISCVCPUClass *mcc = RISCV_CPU_GET_CLASS(cpu); + CPUClass *cc = CPU_CLASS(mcc); + CPURISCVState *env = &cpu->env; + + /* Validate that MISA_MXL is set properly. */ + switch (env->misa_mxl_max) { +#ifdef TARGET_RISCV64 + case MXL_RV64: + case MXL_RV128: + cc->gdb_core_xml_file = "riscv-64bit-cpu.xml"; + break; +#endif + case MXL_RV32: + cc->gdb_core_xml_file = "riscv-32bit-cpu.xml"; + break; + default: + g_assert_not_reached(); + } + + if (env->misa_mxl_max != env->misa_mxl) { + error_setg(errp, "misa_mxl_max must be equal to misa_mxl"); + return; + } +} + +static void riscv_cpu_validate_priv_spec(RISCVCPU *cpu, Error **errp) +{ + CPURISCVState *env = &cpu->env; + int priv_version = -1; + + if (cpu->cfg.priv_spec) { + if (!g_strcmp0(cpu->cfg.priv_spec, "v1.12.0")) { + priv_version = PRIV_VERSION_1_12_0; + } else if (!g_strcmp0(cpu->cfg.priv_spec, "v1.11.0")) { + priv_version = PRIV_VERSION_1_11_0; + } else if (!g_strcmp0(cpu->cfg.priv_spec, "v1.10.0")) { + priv_version = PRIV_VERSION_1_10_0; + } else { + error_setg(errp, + "Unsupported privilege spec version '%s'", + cpu->cfg.priv_spec); + return; + } + + env->priv_ver = priv_version; + } +} + +/* + * We'll get here via the following path: + * + * riscv_cpu_realize() + * -> cpu_exec_realizefn() + * -> tcg_cpu_realize() (via accel_cpu_common_realize()) + */ +static bool tcg_cpu_realize(CPUState *cs, Error **errp) +{ + RISCVCPU *cpu = RISCV_CPU(cs); + CPURISCVState *env = &cpu->env; + Error *local_err = NULL; + + if (object_dynamic_cast(OBJECT(cpu), TYPE_RISCV_CPU_HOST)) { + error_setg(errp, "'host' CPU is not compatible with TCG acceleration"); + return false; + } + + riscv_cpu_validate_misa_mxl(cpu, &local_err); + if (local_err != NULL) { + error_propagate(errp, local_err); + return false; + } + + riscv_cpu_validate_priv_spec(cpu, &local_err); + if (local_err != NULL) { + error_propagate(errp, local_err); + return false; + } + + riscv_cpu_validate_misa_priv(env, &local_err); + if (local_err != NULL) { + error_propagate(errp, local_err); + return false; + } + + if (cpu->cfg.epmp && !cpu->cfg.pmp) { + /* + * Enhanced PMP should only be available + * on harts with PMP support + */ + error_setg(errp, "Invalid configuration: EPMP requires PMP support"); + return false; + } + + riscv_cpu_validate_set_extensions(cpu, &local_err); + if (local_err != NULL) { + error_propagate(errp, local_err); + return false; + } + +#ifndef CONFIG_USER_ONLY + CPU(cs)->tcg_cflags |= CF_PCREL; + + if (cpu->cfg.ext_sstc) { + riscv_timer_init(cpu); + } + + if (cpu->cfg.pmu_num) { + if (!riscv_pmu_init(cpu, cpu->cfg.pmu_num) && cpu->cfg.ext_sscofpmf) { + cpu->pmu_timer = timer_new_ns(QEMU_CLOCK_VIRTUAL, + riscv_pmu_timer_cb, cpu); + } + } +#endif + + return true; +} + static void tcg_cpu_init_ops(AccelCPUClass *accel_cpu, CPUClass *cc) { /* @@ -XXX,XX +XXX,XX @@ static void tcg_cpu_accel_class_init(ObjectClass *oc, void *data) AccelCPUClass *acc = ACCEL_CPU_CLASS(oc); acc->cpu_class_init = tcg_cpu_class_init; + acc->cpu_target_realize = tcg_cpu_realize; } static const TypeInfo tcg_cpu_accel_type_info = { -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> This function is the core of the RISC-V validations for TCG CPUs, and it has a lot going on. Functions in cpu.c were made public to allow them to be used by the KVM accelerator class later on. 'cpu_cfg_ext_get_min_version()' is notably hard to move it to another file due to its dependency with isa_edata_arr[] array, thus make it public and use it as is for now. riscv_cpu_validate_set_extensions() is kept public because it's used by csr.c in write_misa(). Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-4-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 8 +- target/riscv/tcg/tcg-cpu.h | 27 +++ target/riscv/cpu.c | 361 +------------------------------------ target/riscv/csr.c | 1 + target/riscv/tcg/tcg-cpu.c | 357 ++++++++++++++++++++++++++++++++++++ 5 files changed, 397 insertions(+), 357 deletions(-) create mode 100644 target/riscv/tcg/tcg-cpu.h 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 @@ bool riscv_cpu_tlb_fill(CPUState *cs, vaddr address, int size, bool probe, uintptr_t retaddr); char *riscv_isa_string(RISCVCPU *cpu); void riscv_cpu_list(void); -void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp); #define cpu_list riscv_cpu_list #define cpu_mmu_index riscv_cpu_mmu_index @@ -XXX,XX +XXX,XX @@ enum riscv_pmu_event_idx { #include "hw/core/tcg-cpu-ops.h" extern const struct TCGCPUOps riscv_tcg_ops; +/* used by tcg/tcg-cpu.c*/ +void isa_ext_update_enabled(RISCVCPU *cpu, uint32_t ext_offset, bool en); +bool cpu_cfg_ext_is_user_set(uint32_t ext_offset); +bool isa_ext_is_enabled(RISCVCPU *cpu, uint32_t ext_offset); +int cpu_cfg_ext_get_min_version(uint32_t ext_offset); +void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu); + /* CSR function table */ extern riscv_csr_operations csr_ops[CSR_TABLE_SIZE]; diff --git a/target/riscv/tcg/tcg-cpu.h b/target/riscv/tcg/tcg-cpu.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/riscv/tcg/tcg-cpu.h @@ -XXX,XX +XXX,XX @@ +/* + * riscv TCG cpu class initialization + * + * Copyright (c) 2023 Ventana Micro Systems Inc. + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef RISCV_TCG_CPU_H +#define RISCV_TCG_CPU_H + +#include "cpu.h" + +void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp); + +#endif 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 struct isa_ext_data isa_edata_arr[] = { /* Hash that stores user set extensions */ static GHashTable *multi_ext_user_opts; -static bool isa_ext_is_enabled(RISCVCPU *cpu, uint32_t ext_offset) +bool isa_ext_is_enabled(RISCVCPU *cpu, uint32_t ext_offset) { bool *ext_enabled = (void *)&cpu->cfg + ext_offset; return *ext_enabled; } -static void isa_ext_update_enabled(RISCVCPU *cpu, uint32_t ext_offset, - bool en) +void isa_ext_update_enabled(RISCVCPU *cpu, uint32_t ext_offset, bool en) { bool *ext_enabled = (void *)&cpu->cfg + ext_offset; *ext_enabled = en; } -static int cpu_cfg_ext_get_min_version(uint32_t ext_offset) +int cpu_cfg_ext_get_min_version(uint32_t ext_offset) { int i; @@ -XXX,XX +XXX,XX @@ static int cpu_cfg_ext_get_min_version(uint32_t ext_offset) g_assert_not_reached(); } -static bool cpu_cfg_ext_is_user_set(uint32_t ext_offset) +bool cpu_cfg_ext_is_user_set(uint32_t ext_offset) { return g_hash_table_contains(multi_ext_user_opts, GUINT_TO_POINTER(ext_offset)); } -static void cpu_cfg_ext_auto_update(RISCVCPU *cpu, uint32_t ext_offset, - bool value) -{ - CPURISCVState *env = &cpu->env; - bool prev_val = isa_ext_is_enabled(cpu, ext_offset); - int min_version; - - if (prev_val == value) { - return; - } - - if (cpu_cfg_ext_is_user_set(ext_offset)) { - return; - } - - if (value && env->priv_ver != PRIV_VERSION_LATEST) { - /* Do not enable it if priv_ver is older than min_version */ - min_version = cpu_cfg_ext_get_min_version(ext_offset); - if (env->priv_ver < min_version) { - return; - } - } - - isa_ext_update_enabled(cpu, ext_offset, value); -} - const char * const riscv_int_regnames[] = { "x0/zero", "x1/ra", "x2/sp", "x3/gp", "x4/tp", "x5/t0", "x6/t1", "x7/t2", "x8/s0", "x9/s1", "x10/a0", "x11/a1", "x12/a2", "x13/a3", @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_disas_set_info(CPUState *s, disassemble_info *info) } } -static void riscv_cpu_validate_v(CPURISCVState *env, RISCVCPUConfig *cfg, - Error **errp) -{ - if (!is_power_of_2(cfg->vlen)) { - error_setg(errp, "Vector extension VLEN must be power of 2"); - return; - } - if (cfg->vlen > RV_VLEN_MAX || cfg->vlen < 128) { - error_setg(errp, - "Vector extension implementation only supports VLEN " - "in the range [128, %d]", RV_VLEN_MAX); - return; - } - if (!is_power_of_2(cfg->elen)) { - error_setg(errp, "Vector extension ELEN must be power of 2"); - return; - } - if (cfg->elen > 64 || cfg->elen < 8) { - error_setg(errp, - "Vector extension implementation only supports ELEN " - "in the range [8, 64]"); - return; - } - if (cfg->vext_spec) { - if (!g_strcmp0(cfg->vext_spec, "v1.0")) { - env->vext_ver = VEXT_VERSION_1_00_0; - } else { - error_setg(errp, "Unsupported vector spec version '%s'", - cfg->vext_spec); - return; - } - } else if (env->vext_ver == 0) { - qemu_log("vector version is not specified, " - "use the default value v1.0\n"); - - env->vext_ver = VEXT_VERSION_1_00_0; - } -} - -static void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu) +void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu) { CPURISCVState *env = &cpu->env; int i; @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu) } } -/* - * Check consistency between chosen extensions while setting - * cpu->cfg accordingly. - */ -void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) -{ - CPURISCVState *env = &cpu->env; - Error *local_err = NULL; - - /* Do some ISA extension error checking */ - if (riscv_has_ext(env, RVG) && - !(riscv_has_ext(env, RVI) && riscv_has_ext(env, RVM) && - riscv_has_ext(env, RVA) && riscv_has_ext(env, RVF) && - riscv_has_ext(env, RVD) && - cpu->cfg.ext_icsr && cpu->cfg.ext_ifencei)) { - - if (cpu_cfg_ext_is_user_set(CPU_CFG_OFFSET(ext_icsr)) && - !cpu->cfg.ext_icsr) { - error_setg(errp, "RVG requires Zicsr but user set Zicsr to false"); - return; - } - - if (cpu_cfg_ext_is_user_set(CPU_CFG_OFFSET(ext_ifencei)) && - !cpu->cfg.ext_ifencei) { - error_setg(errp, "RVG requires Zifencei but user set " - "Zifencei to false"); - return; - } - - warn_report("Setting G will also set IMAFD_Zicsr_Zifencei"); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_icsr), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_ifencei), true); - - env->misa_ext |= RVI | RVM | RVA | RVF | RVD; - env->misa_ext_mask |= RVI | RVM | RVA | RVF | RVD; - } - - if (riscv_has_ext(env, RVI) && riscv_has_ext(env, RVE)) { - error_setg(errp, - "I and E extensions are incompatible"); - return; - } - - if (!riscv_has_ext(env, RVI) && !riscv_has_ext(env, RVE)) { - error_setg(errp, - "Either I or E extension must be set"); - return; - } - - if (riscv_has_ext(env, RVS) && !riscv_has_ext(env, RVU)) { - error_setg(errp, - "Setting S extension without U extension is illegal"); - return; - } - - if (riscv_has_ext(env, RVH) && !riscv_has_ext(env, RVI)) { - error_setg(errp, - "H depends on an I base integer ISA with 32 x registers"); - return; - } - - if (riscv_has_ext(env, RVH) && !riscv_has_ext(env, RVS)) { - error_setg(errp, "H extension implicitly requires S-mode"); - return; - } - - if (riscv_has_ext(env, RVF) && !cpu->cfg.ext_icsr) { - error_setg(errp, "F extension requires Zicsr"); - return; - } - - if ((cpu->cfg.ext_zawrs) && !riscv_has_ext(env, RVA)) { - error_setg(errp, "Zawrs extension requires A extension"); - return; - } - - if (cpu->cfg.ext_zfa && !riscv_has_ext(env, RVF)) { - error_setg(errp, "Zfa extension requires F extension"); - return; - } - - if (cpu->cfg.ext_zfh) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zfhmin), true); - } - - if (cpu->cfg.ext_zfhmin && !riscv_has_ext(env, RVF)) { - error_setg(errp, "Zfh/Zfhmin extensions require F extension"); - return; - } - - if (cpu->cfg.ext_zfbfmin && !riscv_has_ext(env, RVF)) { - error_setg(errp, "Zfbfmin extension depends on F extension"); - return; - } - - if (riscv_has_ext(env, RVD) && !riscv_has_ext(env, RVF)) { - error_setg(errp, "D extension requires F extension"); - return; - } - - if (riscv_has_ext(env, RVV)) { - riscv_cpu_validate_v(env, &cpu->cfg, &local_err); - if (local_err != NULL) { - error_propagate(errp, local_err); - return; - } - - /* The V vector extension depends on the Zve64d extension */ - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zve64d), true); - } - - /* The Zve64d extension depends on the Zve64f extension */ - if (cpu->cfg.ext_zve64d) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zve64f), true); - } - - /* The Zve64f extension depends on the Zve32f extension */ - if (cpu->cfg.ext_zve64f) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zve32f), true); - } - - if (cpu->cfg.ext_zve64d && !riscv_has_ext(env, RVD)) { - error_setg(errp, "Zve64d/V extensions require D extension"); - return; - } - - if (cpu->cfg.ext_zve32f && !riscv_has_ext(env, RVF)) { - error_setg(errp, "Zve32f/Zve64f extensions require F extension"); - return; - } - - if (cpu->cfg.ext_zvfh) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zvfhmin), true); - } - - if (cpu->cfg.ext_zvfhmin && !cpu->cfg.ext_zve32f) { - error_setg(errp, "Zvfh/Zvfhmin extensions require Zve32f extension"); - return; - } - - if (cpu->cfg.ext_zvfh && !cpu->cfg.ext_zfhmin) { - error_setg(errp, "Zvfh extensions requires Zfhmin extension"); - return; - } - - if (cpu->cfg.ext_zvfbfmin && !cpu->cfg.ext_zfbfmin) { - error_setg(errp, "Zvfbfmin extension depends on Zfbfmin extension"); - return; - } - - if (cpu->cfg.ext_zvfbfmin && !cpu->cfg.ext_zve32f) { - error_setg(errp, "Zvfbfmin extension depends on Zve32f extension"); - return; - } - - if (cpu->cfg.ext_zvfbfwma && !cpu->cfg.ext_zvfbfmin) { - error_setg(errp, "Zvfbfwma extension depends on Zvfbfmin extension"); - return; - } - - /* Set the ISA extensions, checks should have happened above */ - if (cpu->cfg.ext_zhinx) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zca), true); - } - - if ((cpu->cfg.ext_zdinx || cpu->cfg.ext_zhinxmin) && !cpu->cfg.ext_zfinx) { - error_setg(errp, "Zdinx/Zhinx/Zhinxmin extensions require Zfinx"); - return; - } - - if (cpu->cfg.ext_zfinx) { - if (!cpu->cfg.ext_icsr) { - error_setg(errp, "Zfinx extension requires Zicsr"); - return; - } - if (riscv_has_ext(env, RVF)) { - error_setg(errp, - "Zfinx cannot be supported together with F extension"); - return; - } - } - - if (cpu->cfg.ext_zce) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zca), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcb), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcmp), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcmt), true); - if (riscv_has_ext(env, RVF) && env->misa_mxl_max == MXL_RV32) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcf), true); - } - } - - /* zca, zcd and zcf has a PRIV 1.12.0 restriction */ - if (riscv_has_ext(env, RVC) && env->priv_ver >= PRIV_VERSION_1_12_0) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zca), true); - if (riscv_has_ext(env, RVF) && env->misa_mxl_max == MXL_RV32) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcf), true); - } - if (riscv_has_ext(env, RVD)) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcd), true); - } - } - - if (env->misa_mxl_max != MXL_RV32 && cpu->cfg.ext_zcf) { - error_setg(errp, "Zcf extension is only relevant to RV32"); - return; - } - - if (!riscv_has_ext(env, RVF) && cpu->cfg.ext_zcf) { - error_setg(errp, "Zcf extension requires F extension"); - return; - } - - if (!riscv_has_ext(env, RVD) && cpu->cfg.ext_zcd) { - error_setg(errp, "Zcd extension requires D extension"); - return; - } - - if ((cpu->cfg.ext_zcf || cpu->cfg.ext_zcd || cpu->cfg.ext_zcb || - cpu->cfg.ext_zcmp || cpu->cfg.ext_zcmt) && !cpu->cfg.ext_zca) { - error_setg(errp, "Zcf/Zcd/Zcb/Zcmp/Zcmt extensions require Zca " - "extension"); - return; - } - - if (cpu->cfg.ext_zcd && (cpu->cfg.ext_zcmp || cpu->cfg.ext_zcmt)) { - error_setg(errp, "Zcmp/Zcmt extensions are incompatible with " - "Zcd extension"); - return; - } - - if (cpu->cfg.ext_zcmt && !cpu->cfg.ext_icsr) { - error_setg(errp, "Zcmt extension requires Zicsr extension"); - return; - } - - /* - * In principle Zve*x would also suffice here, were they supported - * in qemu - */ - if ((cpu->cfg.ext_zvbb || cpu->cfg.ext_zvkg || cpu->cfg.ext_zvkned || - cpu->cfg.ext_zvknha || cpu->cfg.ext_zvksed || cpu->cfg.ext_zvksh) && - !cpu->cfg.ext_zve32f) { - error_setg(errp, - "Vector crypto extensions require V or Zve* extensions"); - return; - } - - if ((cpu->cfg.ext_zvbc || cpu->cfg.ext_zvknhb) && !cpu->cfg.ext_zve64f) { - error_setg( - errp, - "Zvbc and Zvknhb extensions require V or Zve64{f,d} extensions"); - return; - } - - if (cpu->cfg.ext_zk) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zkn), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zkr), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zkt), true); - } - - if (cpu->cfg.ext_zkn) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkb), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkc), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkx), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zkne), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zknd), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zknh), true); - } - - if (cpu->cfg.ext_zks) { - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkb), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkc), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkx), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zksed), true); - cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zksh), true); - } - - /* - * Disable isa extensions based on priv spec after we - * validated and set everything we need. - */ - riscv_cpu_disable_priv_spec_isa_exts(cpu); -} - #ifndef CONFIG_USER_ONLY static void riscv_cpu_satp_mode_finalize(RISCVCPU *cpu, Error **errp) { diff --git a/target/riscv/csr.c b/target/riscv/csr.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/csr.c +++ b/target/riscv/csr.c @@ -XXX,XX +XXX,XX @@ #include "qemu/log.h" #include "qemu/timer.h" #include "cpu.h" +#include "tcg/tcg-cpu.h" #include "pmu.h" #include "time_helper.h" #include "exec/exec-all.h" 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 @@ #include "qemu/osdep.h" #include "exec/exec-all.h" +#include "tcg-cpu.h" #include "cpu.h" #include "pmu.h" #include "time_helper.h" #include "qapi/error.h" #include "qemu/accel.h" +#include "qemu/error-report.h" +#include "qemu/log.h" #include "hw/core/accel-cpu.h" +static void cpu_cfg_ext_auto_update(RISCVCPU *cpu, uint32_t ext_offset, + bool value) +{ + CPURISCVState *env = &cpu->env; + bool prev_val = isa_ext_is_enabled(cpu, ext_offset); + int min_version; + + if (prev_val == value) { + return; + } + + if (cpu_cfg_ext_is_user_set(ext_offset)) { + return; + } + + if (value && env->priv_ver != PRIV_VERSION_LATEST) { + /* Do not enable it if priv_ver is older than min_version */ + min_version = cpu_cfg_ext_get_min_version(ext_offset); + if (env->priv_ver < min_version) { + return; + } + } + + isa_ext_update_enabled(cpu, ext_offset, value); +} + static void riscv_cpu_validate_misa_priv(CPURISCVState *env, Error **errp) { if (riscv_has_ext(env, RVH) && env->priv_ver < PRIV_VERSION_1_12_0) { @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_validate_priv_spec(RISCVCPU *cpu, Error **errp) } } +static void riscv_cpu_validate_v(CPURISCVState *env, RISCVCPUConfig *cfg, + Error **errp) +{ + if (!is_power_of_2(cfg->vlen)) { + error_setg(errp, "Vector extension VLEN must be power of 2"); + return; + } + + if (cfg->vlen > RV_VLEN_MAX || cfg->vlen < 128) { + error_setg(errp, + "Vector extension implementation only supports VLEN " + "in the range [128, %d]", RV_VLEN_MAX); + return; + } + + if (!is_power_of_2(cfg->elen)) { + error_setg(errp, "Vector extension ELEN must be power of 2"); + return; + } + + if (cfg->elen > 64 || cfg->elen < 8) { + error_setg(errp, + "Vector extension implementation only supports ELEN " + "in the range [8, 64]"); + return; + } + + if (cfg->vext_spec) { + if (!g_strcmp0(cfg->vext_spec, "v1.0")) { + env->vext_ver = VEXT_VERSION_1_00_0; + } else { + error_setg(errp, "Unsupported vector spec version '%s'", + cfg->vext_spec); + return; + } + } else if (env->vext_ver == 0) { + qemu_log("vector version is not specified, " + "use the default value v1.0\n"); + + env->vext_ver = VEXT_VERSION_1_00_0; + } +} + +/* + * Check consistency between chosen extensions while setting + * cpu->cfg accordingly. + */ +void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) +{ + CPURISCVState *env = &cpu->env; + Error *local_err = NULL; + + /* Do some ISA extension error checking */ + if (riscv_has_ext(env, RVG) && + !(riscv_has_ext(env, RVI) && riscv_has_ext(env, RVM) && + riscv_has_ext(env, RVA) && riscv_has_ext(env, RVF) && + riscv_has_ext(env, RVD) && + cpu->cfg.ext_icsr && cpu->cfg.ext_ifencei)) { + + if (cpu_cfg_ext_is_user_set(CPU_CFG_OFFSET(ext_icsr)) && + !cpu->cfg.ext_icsr) { + error_setg(errp, "RVG requires Zicsr but user set Zicsr to false"); + return; + } + + if (cpu_cfg_ext_is_user_set(CPU_CFG_OFFSET(ext_ifencei)) && + !cpu->cfg.ext_ifencei) { + error_setg(errp, "RVG requires Zifencei but user set " + "Zifencei to false"); + return; + } + + warn_report("Setting G will also set IMAFD_Zicsr_Zifencei"); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_icsr), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_ifencei), true); + + env->misa_ext |= RVI | RVM | RVA | RVF | RVD; + env->misa_ext_mask |= RVI | RVM | RVA | RVF | RVD; + } + + if (riscv_has_ext(env, RVI) && riscv_has_ext(env, RVE)) { + error_setg(errp, + "I and E extensions are incompatible"); + return; + } + + if (!riscv_has_ext(env, RVI) && !riscv_has_ext(env, RVE)) { + error_setg(errp, + "Either I or E extension must be set"); + return; + } + + if (riscv_has_ext(env, RVS) && !riscv_has_ext(env, RVU)) { + error_setg(errp, + "Setting S extension without U extension is illegal"); + return; + } + + if (riscv_has_ext(env, RVH) && !riscv_has_ext(env, RVI)) { + error_setg(errp, + "H depends on an I base integer ISA with 32 x registers"); + return; + } + + if (riscv_has_ext(env, RVH) && !riscv_has_ext(env, RVS)) { + error_setg(errp, "H extension implicitly requires S-mode"); + return; + } + + if (riscv_has_ext(env, RVF) && !cpu->cfg.ext_icsr) { + error_setg(errp, "F extension requires Zicsr"); + return; + } + + if ((cpu->cfg.ext_zawrs) && !riscv_has_ext(env, RVA)) { + error_setg(errp, "Zawrs extension requires A extension"); + return; + } + + if (cpu->cfg.ext_zfa && !riscv_has_ext(env, RVF)) { + error_setg(errp, "Zfa extension requires F extension"); + return; + } + + if (cpu->cfg.ext_zfh) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zfhmin), true); + } + + if (cpu->cfg.ext_zfhmin && !riscv_has_ext(env, RVF)) { + error_setg(errp, "Zfh/Zfhmin extensions require F extension"); + return; + } + + if (cpu->cfg.ext_zfbfmin && !riscv_has_ext(env, RVF)) { + error_setg(errp, "Zfbfmin extension depends on F extension"); + return; + } + + if (riscv_has_ext(env, RVD) && !riscv_has_ext(env, RVF)) { + error_setg(errp, "D extension requires F extension"); + return; + } + + if (riscv_has_ext(env, RVV)) { + riscv_cpu_validate_v(env, &cpu->cfg, &local_err); + if (local_err != NULL) { + error_propagate(errp, local_err); + return; + } + + /* The V vector extension depends on the Zve64d extension */ + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zve64d), true); + } + + /* The Zve64d extension depends on the Zve64f extension */ + if (cpu->cfg.ext_zve64d) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zve64f), true); + } + + /* The Zve64f extension depends on the Zve32f extension */ + if (cpu->cfg.ext_zve64f) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zve32f), true); + } + + if (cpu->cfg.ext_zve64d && !riscv_has_ext(env, RVD)) { + error_setg(errp, "Zve64d/V extensions require D extension"); + return; + } + + if (cpu->cfg.ext_zve32f && !riscv_has_ext(env, RVF)) { + error_setg(errp, "Zve32f/Zve64f extensions require F extension"); + return; + } + + if (cpu->cfg.ext_zvfh) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zvfhmin), true); + } + + if (cpu->cfg.ext_zvfhmin && !cpu->cfg.ext_zve32f) { + error_setg(errp, "Zvfh/Zvfhmin extensions require Zve32f extension"); + return; + } + + if (cpu->cfg.ext_zvfh && !cpu->cfg.ext_zfhmin) { + error_setg(errp, "Zvfh extensions requires Zfhmin extension"); + return; + } + + if (cpu->cfg.ext_zvfbfmin && !cpu->cfg.ext_zfbfmin) { + error_setg(errp, "Zvfbfmin extension depends on Zfbfmin extension"); + return; + } + + if (cpu->cfg.ext_zvfbfmin && !cpu->cfg.ext_zve32f) { + error_setg(errp, "Zvfbfmin extension depends on Zve32f extension"); + return; + } + + if (cpu->cfg.ext_zvfbfwma && !cpu->cfg.ext_zvfbfmin) { + error_setg(errp, "Zvfbfwma extension depends on Zvfbfmin extension"); + return; + } + + /* Set the ISA extensions, checks should have happened above */ + if (cpu->cfg.ext_zhinx) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zca), true); + } + + if ((cpu->cfg.ext_zdinx || cpu->cfg.ext_zhinxmin) && !cpu->cfg.ext_zfinx) { + error_setg(errp, "Zdinx/Zhinx/Zhinxmin extensions require Zfinx"); + return; + } + + if (cpu->cfg.ext_zfinx) { + if (!cpu->cfg.ext_icsr) { + error_setg(errp, "Zfinx extension requires Zicsr"); + return; + } + if (riscv_has_ext(env, RVF)) { + error_setg(errp, + "Zfinx cannot be supported together with F extension"); + return; + } + } + + if (cpu->cfg.ext_zce) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zca), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcb), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcmp), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcmt), true); + if (riscv_has_ext(env, RVF) && env->misa_mxl_max == MXL_RV32) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcf), true); + } + } + + /* zca, zcd and zcf has a PRIV 1.12.0 restriction */ + if (riscv_has_ext(env, RVC) && env->priv_ver >= PRIV_VERSION_1_12_0) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zca), true); + if (riscv_has_ext(env, RVF) && env->misa_mxl_max == MXL_RV32) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcf), true); + } + if (riscv_has_ext(env, RVD)) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zcd), true); + } + } + + if (env->misa_mxl_max != MXL_RV32 && cpu->cfg.ext_zcf) { + error_setg(errp, "Zcf extension is only relevant to RV32"); + return; + } + + if (!riscv_has_ext(env, RVF) && cpu->cfg.ext_zcf) { + error_setg(errp, "Zcf extension requires F extension"); + return; + } + + if (!riscv_has_ext(env, RVD) && cpu->cfg.ext_zcd) { + error_setg(errp, "Zcd extension requires D extension"); + return; + } + + if ((cpu->cfg.ext_zcf || cpu->cfg.ext_zcd || cpu->cfg.ext_zcb || + cpu->cfg.ext_zcmp || cpu->cfg.ext_zcmt) && !cpu->cfg.ext_zca) { + error_setg(errp, "Zcf/Zcd/Zcb/Zcmp/Zcmt extensions require Zca " + "extension"); + return; + } + + if (cpu->cfg.ext_zcd && (cpu->cfg.ext_zcmp || cpu->cfg.ext_zcmt)) { + error_setg(errp, "Zcmp/Zcmt extensions are incompatible with " + "Zcd extension"); + return; + } + + if (cpu->cfg.ext_zcmt && !cpu->cfg.ext_icsr) { + error_setg(errp, "Zcmt extension requires Zicsr extension"); + return; + } + + /* + * In principle Zve*x would also suffice here, were they supported + * in qemu + */ + if ((cpu->cfg.ext_zvbb || cpu->cfg.ext_zvkg || cpu->cfg.ext_zvkned || + cpu->cfg.ext_zvknha || cpu->cfg.ext_zvksed || cpu->cfg.ext_zvksh) && + !cpu->cfg.ext_zve32f) { + error_setg(errp, + "Vector crypto extensions require V or Zve* extensions"); + return; + } + + if ((cpu->cfg.ext_zvbc || cpu->cfg.ext_zvknhb) && !cpu->cfg.ext_zve64f) { + error_setg( + errp, + "Zvbc and Zvknhb extensions require V or Zve64{f,d} extensions"); + return; + } + + if (cpu->cfg.ext_zk) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zkn), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zkr), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zkt), true); + } + + if (cpu->cfg.ext_zkn) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkb), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkc), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkx), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zkne), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zknd), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zknh), true); + } + + if (cpu->cfg.ext_zks) { + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkb), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkc), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zbkx), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zksed), true); + cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_zksh), true); + } + + /* + * Disable isa extensions based on priv spec after we + * validated and set everything we need. + */ + riscv_cpu_disable_priv_spec_isa_exts(cpu); +} + /* * We'll get here via the following path: * -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Move the remaining of riscv_tcg_ops now that we have a working realize() implementation. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-5-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 4 --- target/riscv/cpu.c | 58 ------------------------------------ target/riscv/tcg/tcg-cpu.c | 60 +++++++++++++++++++++++++++++++++++++- 3 files changed, 59 insertions(+), 63 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 @@ enum riscv_pmu_event_idx { RISCV_PMU_EVENT_CACHE_ITLB_PREFETCH_MISS = 0x10021, }; -/* Export tcg_ops until we move everything to tcg/tcg-cpu.c */ -#include "hw/core/tcg-cpu-ops.h" -extern const struct TCGCPUOps riscv_tcg_ops; - /* used by tcg/tcg-cpu.c*/ void isa_ext_update_enabled(RISCVCPU *cpu, uint32_t ext_offset, bool en); bool cpu_cfg_ext_is_user_set(uint32_t ext_offset); 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 vaddr riscv_cpu_get_pc(CPUState *cs) return env->pc; } -static void riscv_cpu_synchronize_from_tb(CPUState *cs, - const TranslationBlock *tb) -{ - if (!(tb_cflags(tb) & CF_PCREL)) { - RISCVCPU *cpu = RISCV_CPU(cs); - CPURISCVState *env = &cpu->env; - RISCVMXL xl = FIELD_EX32(tb->flags, TB_FLAGS, XL); - - tcg_debug_assert(!(cs->tcg_cflags & CF_PCREL)); - - if (xl == MXL_RV32) { - env->pc = (int32_t) tb->pc; - } else { - env->pc = tb->pc; - } - } -} - static bool riscv_cpu_has_work(CPUState *cs) { #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ static bool riscv_cpu_has_work(CPUState *cs) #endif } -static void riscv_restore_state_to_opc(CPUState *cs, - const TranslationBlock *tb, - const uint64_t *data) -{ - RISCVCPU *cpu = RISCV_CPU(cs); - CPURISCVState *env = &cpu->env; - RISCVMXL xl = FIELD_EX32(tb->flags, TB_FLAGS, XL); - target_ulong pc; - - if (tb_cflags(tb) & CF_PCREL) { - pc = (env->pc & TARGET_PAGE_MASK) | data[0]; - } else { - pc = data[0]; - } - - if (xl == MXL_RV32) { - env->pc = (int32_t)pc; - } else { - env->pc = pc; - } - env->bins = data[1]; -} - static void riscv_cpu_reset_hold(Object *obj) { #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ static const struct SysemuCPUOps riscv_sysemu_ops = { }; #endif -const struct TCGCPUOps riscv_tcg_ops = { - .initialize = riscv_translate_init, - .synchronize_from_tb = riscv_cpu_synchronize_from_tb, - .restore_state_to_opc = riscv_restore_state_to_opc, - -#ifndef CONFIG_USER_ONLY - .tlb_fill = riscv_cpu_tlb_fill, - .cpu_exec_interrupt = riscv_cpu_exec_interrupt, - .do_interrupt = riscv_cpu_do_interrupt, - .do_transaction_failed = riscv_cpu_do_transaction_failed, - .do_unaligned_access = riscv_cpu_do_unaligned_access, - .debug_excp_handler = riscv_cpu_debug_excp_handler, - .debug_check_breakpoint = riscv_cpu_debug_check_breakpoint, - .debug_check_watchpoint = riscv_cpu_debug_check_watchpoint, -#endif /* !CONFIG_USER_ONLY */ -}; - static bool riscv_cpu_is_dynamic(Object *cpu_obj) { return object_dynamic_cast(cpu_obj, TYPE_RISCV_DYNAMIC_CPU) != NULL; 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 @@ #include "qemu/error-report.h" #include "qemu/log.h" #include "hw/core/accel-cpu.h" +#include "hw/core/tcg-cpu-ops.h" +#include "tcg/tcg.h" +static void riscv_cpu_synchronize_from_tb(CPUState *cs, + const TranslationBlock *tb) +{ + if (!(tb_cflags(tb) & CF_PCREL)) { + RISCVCPU *cpu = RISCV_CPU(cs); + CPURISCVState *env = &cpu->env; + RISCVMXL xl = FIELD_EX32(tb->flags, TB_FLAGS, XL); + + tcg_debug_assert(!(cs->tcg_cflags & CF_PCREL)); + + if (xl == MXL_RV32) { + env->pc = (int32_t) tb->pc; + } else { + env->pc = tb->pc; + } + } +} + +static void riscv_restore_state_to_opc(CPUState *cs, + const TranslationBlock *tb, + const uint64_t *data) +{ + RISCVCPU *cpu = RISCV_CPU(cs); + CPURISCVState *env = &cpu->env; + RISCVMXL xl = FIELD_EX32(tb->flags, TB_FLAGS, XL); + target_ulong pc; + + if (tb_cflags(tb) & CF_PCREL) { + pc = (env->pc & TARGET_PAGE_MASK) | data[0]; + } else { + pc = data[0]; + } + + if (xl == MXL_RV32) { + env->pc = (int32_t)pc; + } else { + env->pc = pc; + } + env->bins = data[1]; +} + +static const struct TCGCPUOps riscv_tcg_ops = { + .initialize = riscv_translate_init, + .synchronize_from_tb = riscv_cpu_synchronize_from_tb, + .restore_state_to_opc = riscv_restore_state_to_opc, + +#ifndef CONFIG_USER_ONLY + .tlb_fill = riscv_cpu_tlb_fill, + .cpu_exec_interrupt = riscv_cpu_exec_interrupt, + .do_interrupt = riscv_cpu_do_interrupt, + .do_transaction_failed = riscv_cpu_do_transaction_failed, + .do_unaligned_access = riscv_cpu_do_unaligned_access, + .debug_excp_handler = riscv_cpu_debug_excp_handler, + .debug_check_breakpoint = riscv_cpu_debug_check_breakpoint, + .debug_check_watchpoint = riscv_cpu_debug_check_watchpoint, +#endif /* !CONFIG_USER_ONLY */ +}; static void cpu_cfg_ext_auto_update(RISCVCPU *cpu, uint32_t ext_offset, bool value) @@ -XXX,XX +XXX,XX @@ static void tcg_cpu_init_ops(AccelCPUClass *accel_cpu, CPUClass *cc) { /* * All cpus use the same set of operations. - * riscv_tcg_ops is being imported from cpu.c for now. */ cc->tcg_ops = &riscv_tcg_ops; } -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> All generic CPUs call riscv_cpu_add_user_properties(). The 'max' CPU calls riscv_init_max_cpu_extensions(). Both can be moved to a common instance_post_init() callback, implemented in riscv_cpu_post_init(), called by all CPUs. The call order then becomes: riscv_cpu_init() -> cpu_init() of each CPU -> .instance_post_init() In the near future riscv_cpu_post_init() will call the init() function of the current accelerator, providing a hook for KVM and TCG accel classes to change the init() process of the CPU. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-6-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 43 ++++++++++++++++++++++++++++++++----------- 1 file changed, 32 insertions(+), 11 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_max_cpu_init(Object *obj) mlx = MXL_RV32; #endif set_misa(env, mlx, 0); - riscv_cpu_add_user_properties(obj); - riscv_init_max_cpu_extensions(obj); env->priv_ver = PRIV_VERSION_LATEST; #ifndef CONFIG_USER_ONLY set_satp_mode_max_supported(RISCV_CPU(obj), mlx == MXL_RV32 ? @@ -XXX,XX +XXX,XX @@ static void rv64_base_cpu_init(Object *obj) CPURISCVState *env = &RISCV_CPU(obj)->env; /* We set this in the realise function */ set_misa(env, MXL_RV64, 0); - riscv_cpu_add_user_properties(obj); /* Set latest version of privileged specification */ env->priv_ver = PRIV_VERSION_LATEST; #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ static void rv128_base_cpu_init(Object *obj) CPURISCVState *env = &RISCV_CPU(obj)->env; /* We set this in the realise function */ set_misa(env, MXL_RV128, 0); - riscv_cpu_add_user_properties(obj); /* Set latest version of privileged specification */ env->priv_ver = PRIV_VERSION_LATEST; #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ static void rv32_base_cpu_init(Object *obj) CPURISCVState *env = &RISCV_CPU(obj)->env; /* We set this in the realise function */ set_misa(env, MXL_RV32, 0); - riscv_cpu_add_user_properties(obj); /* Set latest version of privileged specification */ env->priv_ver = PRIV_VERSION_LATEST; #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ static void riscv_host_cpu_init(Object *obj) #elif defined(TARGET_RISCV64) set_misa(env, MXL_RV64, 0); #endif - riscv_cpu_add_user_properties(obj); } #endif /* CONFIG_KVM */ @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_set_irq(void *opaque, int irq, int level) } #endif /* CONFIG_USER_ONLY */ +static bool riscv_cpu_is_dynamic(Object *cpu_obj) +{ + return object_dynamic_cast(cpu_obj, TYPE_RISCV_DYNAMIC_CPU) != NULL; +} + +static bool riscv_cpu_has_max_extensions(Object *cpu_obj) +{ + return object_dynamic_cast(cpu_obj, TYPE_RISCV_CPU_MAX) != NULL; +} + +static bool riscv_cpu_has_user_properties(Object *cpu_obj) +{ + if (kvm_enabled() && + object_dynamic_cast(cpu_obj, TYPE_RISCV_CPU_HOST) != NULL) { + return true; + } + + return riscv_cpu_is_dynamic(cpu_obj); +} + +static void riscv_cpu_post_init(Object *obj) +{ + if (riscv_cpu_has_user_properties(obj)) { + riscv_cpu_add_user_properties(obj); + } + + if (riscv_cpu_has_max_extensions(obj)) { + riscv_init_max_cpu_extensions(obj); + } +} + static void riscv_cpu_init(Object *obj) { #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ static const struct SysemuCPUOps riscv_sysemu_ops = { }; #endif -static bool riscv_cpu_is_dynamic(Object *cpu_obj) -{ - return object_dynamic_cast(cpu_obj, TYPE_RISCV_DYNAMIC_CPU) != NULL; -} - static void cpu_set_mvendorid(Object *obj, Visitor *v, const char *name, void *opaque, Error **errp) { @@ -XXX,XX +XXX,XX @@ static const TypeInfo riscv_cpu_type_infos[] = { .instance_size = sizeof(RISCVCPU), .instance_align = __alignof(RISCVCPU), .instance_init = riscv_cpu_init, + .instance_post_init = riscv_cpu_post_init, .abstract = true, .class_size = sizeof(RISCVCPUClass), .class_init = riscv_cpu_class_init, -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> This CPU only exists if we're compiling with KVM so move it to the kvm specific file. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-7-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 15 --------------- target/riscv/kvm.c | 21 +++++++++++++++++++++ 2 files changed, 21 insertions(+), 15 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 rv32_imafcu_nommu_cpu_init(Object *obj) } #endif -#if defined(CONFIG_KVM) -static void riscv_host_cpu_init(Object *obj) -{ - CPURISCVState *env = &RISCV_CPU(obj)->env; -#if defined(TARGET_RISCV32) - set_misa(env, MXL_RV32, 0); -#elif defined(TARGET_RISCV64) - set_misa(env, MXL_RV64, 0); -#endif -} -#endif /* CONFIG_KVM */ - static ObjectClass *riscv_cpu_class_by_name(const char *cpu_model) { ObjectClass *oc; @@ -XXX,XX +XXX,XX @@ static const TypeInfo riscv_cpu_type_infos[] = { }, DEFINE_DYNAMIC_CPU(TYPE_RISCV_CPU_ANY, riscv_any_cpu_init), DEFINE_DYNAMIC_CPU(TYPE_RISCV_CPU_MAX, riscv_max_cpu_init), -#if defined(CONFIG_KVM) - DEFINE_CPU(TYPE_RISCV_CPU_HOST, riscv_host_cpu_init), -#endif #if defined(TARGET_RISCV32) DEFINE_DYNAMIC_CPU(TYPE_RISCV_CPU_BASE32, rv32_base_cpu_init), DEFINE_CPU(TYPE_RISCV_CPU_IBEX, rv32_ibex_cpu_init), diff --git a/target/riscv/kvm.c b/target/riscv/kvm.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm.c +++ b/target/riscv/kvm.c @@ -XXX,XX +XXX,XX @@ void kvm_riscv_aia_create(MachineState *machine, uint64_t group_shift, kvm_msi_via_irqfd_allowed = kvm_irqfds_enabled(); } + +static void riscv_host_cpu_init(Object *obj) +{ + CPURISCVState *env = &RISCV_CPU(obj)->env; + +#if defined(TARGET_RISCV32) + env->misa_mxl_max = env->misa_mxl = MXL_RV32; +#elif defined(TARGET_RISCV64) + env->misa_mxl_max = env->misa_mxl = MXL_RV64; +#endif +} + +static const TypeInfo riscv_kvm_cpu_type_infos[] = { + { + .name = TYPE_RISCV_CPU_HOST, + .parent = TYPE_RISCV_CPU, + .instance_init = riscv_host_cpu_init, + } +}; + +DEFINE_TYPES(riscv_kvm_cpu_type_infos) -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> We'll need to export these arrays to the accelerator classes in the next patches. Mark them as 'const' now because they should not be modified at runtime. Note that 'riscv_cpu_options' will also be exported, but can't be marked as 'const', because the properties are changed via qdev_property_add_static(). Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-8-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 22 +++++++++++++--------- 1 file changed, 13 insertions(+), 9 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 @@ typedef struct RISCVCPUMultiExtConfig { {.name = _name, .offset = CPU_CFG_OFFSET(_prop), \ .enabled = _defval} -static RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { +static const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { /* Defaults for standard extensions */ MULTI_EXT_CFG_BOOL("sscofpmf", ext_sscofpmf, false), MULTI_EXT_CFG_BOOL("Zifencei", ext_ifencei, true), @@ -XXX,XX +XXX,XX @@ static RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { DEFINE_PROP_END_OF_LIST(), }; -static RISCVCPUMultiExtConfig riscv_cpu_vendor_exts[] = { +static const RISCVCPUMultiExtConfig riscv_cpu_vendor_exts[] = { MULTI_EXT_CFG_BOOL("xtheadba", ext_xtheadba, false), MULTI_EXT_CFG_BOOL("xtheadbb", ext_xtheadbb, false), MULTI_EXT_CFG_BOOL("xtheadbs", ext_xtheadbs, false), @@ -XXX,XX +XXX,XX @@ static RISCVCPUMultiExtConfig riscv_cpu_vendor_exts[] = { }; /* These are experimental so mark with 'x-' */ -static RISCVCPUMultiExtConfig riscv_cpu_experimental_exts[] = { +static const RISCVCPUMultiExtConfig riscv_cpu_experimental_exts[] = { /* ePMP 0.9.3 */ MULTI_EXT_CFG_BOOL("x-epmp", epmp, false), MULTI_EXT_CFG_BOOL("x-smaia", ext_smaia, false), @@ -XXX,XX +XXX,XX @@ static void cpu_get_multi_ext_cfg(Object *obj, Visitor *v, const char *name, } static void cpu_add_multi_ext_prop(Object *cpu_obj, - RISCVCPUMultiExtConfig *multi_cfg) + const RISCVCPUMultiExtConfig *multi_cfg) { object_property_add(cpu_obj, multi_cfg->name, "bool", cpu_get_multi_ext_cfg, @@ -XXX,XX +XXX,XX @@ static void cpu_add_multi_ext_prop(Object *cpu_obj, } static void riscv_cpu_add_multiext_prop_array(Object *obj, - RISCVCPUMultiExtConfig *array) + const RISCVCPUMultiExtConfig *array) { + const RISCVCPUMultiExtConfig *prop; + g_assert(array); - for (RISCVCPUMultiExtConfig *prop = array; prop && prop->name; prop++) { + for (prop = array; prop && prop->name; prop++) { cpu_add_multi_ext_prop(obj, prop); } } @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_kvm_unavail_prop(Object *obj, const char *prop_name) } static void riscv_cpu_add_kvm_unavail_prop_array(Object *obj, - RISCVCPUMultiExtConfig *array) + const RISCVCPUMultiExtConfig *array) { + const RISCVCPUMultiExtConfig *prop; + g_assert(array); - for (RISCVCPUMultiExtConfig *prop = array; prop && prop->name; prop++) { + for (prop = array; prop && prop->name; prop++) { riscv_cpu_add_kvm_unavail_prop(obj, prop->name); } } @@ -XXX,XX +XXX,XX @@ static void riscv_init_max_cpu_extensions(Object *obj) { RISCVCPU *cpu = RISCV_CPU(obj); CPURISCVState *env = &cpu->env; - RISCVCPUMultiExtConfig *prop; + const RISCVCPUMultiExtConfig *prop; /* Enable RVG, RVJ and RVV that are disabled by default */ set_misa(env, env->misa_mxl, env->misa_ext | RVG | RVJ | RVV); -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> We'll introduce the KVM accelerator class with a 'cpu_instance_init' implementation that is going to be invoked during the common riscv_cpu_post_init() (via accel_cpu_instance_init()). This instance_init will execute KVM exclusive code that TCG doesn't care about, such as adding KVM specific properties, initing registers using a KVM scratch CPU and so on. The core of the forementioned cpu_instance_init impl is the current riscv_cpu_add_kvm_properties() that is being used by the common code via riscv_cpu_add_user_properties() in cpu.c. Move it to kvm.c, together will all the relevant artifacts, exporting and renaming it to kvm_riscv_cpu_add_kvm_properties() so cpu.c can keep using it for now. To make this work we'll need to export riscv_cpu_extensions, riscv_cpu_vendor_exts and riscv_cpu_experimental_exts from cpu.c as well. The TCG accelerator will also need to access those in the near future so this export will benefit us in the long run. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-9-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 14 +++++++ target/riscv/kvm_riscv.h | 3 -- target/riscv/cpu.c | 85 +++------------------------------------- target/riscv/kvm.c | 68 +++++++++++++++++++++++++++++++- 4 files changed, 86 insertions(+), 84 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 @@ #include "hw/core/cpu.h" #include "hw/registerfields.h" +#include "hw/qdev-properties.h" #include "exec/cpu-defs.h" #include "qemu/cpu-float.h" #include "qom/object.h" @@ -XXX,XX +XXX,XX @@ bool isa_ext_is_enabled(RISCVCPU *cpu, uint32_t ext_offset); int cpu_cfg_ext_get_min_version(uint32_t ext_offset); void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu); +typedef struct RISCVCPUMultiExtConfig { + const char *name; + uint32_t offset; + bool enabled; +} RISCVCPUMultiExtConfig; + +extern const RISCVCPUMultiExtConfig riscv_cpu_extensions[]; +extern const RISCVCPUMultiExtConfig riscv_cpu_vendor_exts[]; +extern const RISCVCPUMultiExtConfig riscv_cpu_experimental_exts[]; +extern Property riscv_cpu_options[]; + +void riscv_cpu_add_misa_properties(Object *cpu_obj); + /* CSR function table */ extern riscv_csr_operations csr_ops[CSR_TABLE_SIZE]; diff --git a/target/riscv/kvm_riscv.h b/target/riscv/kvm_riscv.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm_riscv.h +++ b/target/riscv/kvm_riscv.h @@ -XXX,XX +XXX,XX @@ #ifndef QEMU_KVM_RISCV_H #define QEMU_KVM_RISCV_H -/* Temporarily implemented in cpu.c */ void kvm_riscv_cpu_add_kvm_properties(Object *obj); - -void kvm_riscv_init_user_properties(Object *cpu_obj); void kvm_riscv_reset_vcpu(RISCVCPU *cpu); void kvm_riscv_set_irq(RISCVCPU *cpu, int irq, int level); void kvm_riscv_aia_create(MachineState *machine, uint64_t group_shift, 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 RISCVCPUMisaExtConfig misa_ext_cfgs[] = { * change MISA bits during realize() (RVG enables MISA * bits but the user is warned about it). */ -static void riscv_cpu_add_misa_properties(Object *cpu_obj) +void riscv_cpu_add_misa_properties(Object *cpu_obj) { int i; @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_misa_properties(Object *cpu_obj) } } -typedef struct RISCVCPUMultiExtConfig { - const char *name; - uint32_t offset; - bool enabled; -} RISCVCPUMultiExtConfig; - #define MULTI_EXT_CFG_BOOL(_name, _prop, _defval) \ {.name = _name, .offset = CPU_CFG_OFFSET(_prop), \ .enabled = _defval} -static const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { +const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { /* Defaults for standard extensions */ MULTI_EXT_CFG_BOOL("sscofpmf", ext_sscofpmf, false), MULTI_EXT_CFG_BOOL("Zifencei", ext_ifencei, true), @@ -XXX,XX +XXX,XX @@ static const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { DEFINE_PROP_END_OF_LIST(), }; -static const RISCVCPUMultiExtConfig riscv_cpu_vendor_exts[] = { +const RISCVCPUMultiExtConfig riscv_cpu_vendor_exts[] = { MULTI_EXT_CFG_BOOL("xtheadba", ext_xtheadba, false), MULTI_EXT_CFG_BOOL("xtheadbb", ext_xtheadbb, false), MULTI_EXT_CFG_BOOL("xtheadbs", ext_xtheadbs, false), @@ -XXX,XX +XXX,XX @@ static const RISCVCPUMultiExtConfig riscv_cpu_vendor_exts[] = { }; /* These are experimental so mark with 'x-' */ -static const RISCVCPUMultiExtConfig riscv_cpu_experimental_exts[] = { +const RISCVCPUMultiExtConfig riscv_cpu_experimental_exts[] = { /* ePMP 0.9.3 */ MULTI_EXT_CFG_BOOL("x-epmp", epmp, false), MULTI_EXT_CFG_BOOL("x-smaia", ext_smaia, false), @@ -XXX,XX +XXX,XX @@ static const RISCVCPUMultiExtConfig riscv_cpu_experimental_exts[] = { DEFINE_PROP_END_OF_LIST(), }; -static Property riscv_cpu_options[] = { +Property riscv_cpu_options[] = { DEFINE_PROP_UINT8("pmu-num", RISCVCPU, cfg.pmu_num, 16), DEFINE_PROP_BOOL("mmu", RISCVCPU, cfg.mmu, true), @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_multiext_prop_array(Object *obj, } } -#ifdef CONFIG_KVM -static void cpu_set_cfg_unavailable(Object *obj, Visitor *v, - const char *name, - void *opaque, Error **errp) -{ - const char *propname = opaque; - bool value; - - if (!visit_type_bool(v, name, &value, errp)) { - return; - } - - if (value) { - error_setg(errp, "extension %s is not available with KVM", - propname); - } -} - -static void riscv_cpu_add_kvm_unavail_prop(Object *obj, const char *prop_name) -{ - /* Check if KVM created the property already */ - if (object_property_find(obj, prop_name)) { - return; - } - - /* - * Set the default to disabled for every extension - * unknown to KVM and error out if the user attempts - * to enable any of them. - */ - object_property_add(obj, prop_name, "bool", - NULL, cpu_set_cfg_unavailable, - NULL, (void *)prop_name); -} - -static void riscv_cpu_add_kvm_unavail_prop_array(Object *obj, - const RISCVCPUMultiExtConfig *array) -{ - const RISCVCPUMultiExtConfig *prop; - - g_assert(array); - - for (prop = array; prop && prop->name; prop++) { - riscv_cpu_add_kvm_unavail_prop(obj, prop->name); - } -} - -void kvm_riscv_cpu_add_kvm_properties(Object *obj) -{ - Property *prop; - DeviceState *dev = DEVICE(obj); - - kvm_riscv_init_user_properties(obj); - riscv_cpu_add_misa_properties(obj); - - riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_extensions); - riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_vendor_exts); - riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_experimental_exts); - - for (prop = riscv_cpu_options; prop && prop->name; prop++) { - /* Check if KVM created the property already */ - if (object_property_find(obj, prop->name)) { - continue; - } - qdev_property_add_static(dev, prop); - } -} -#endif - /* * Add CPU properties with user-facing flags. * diff --git a/target/riscv/kvm.c b/target/riscv/kvm.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm.c +++ b/target/riscv/kvm.c @@ -XXX,XX +XXX,XX @@ static void kvm_riscv_update_cpu_cfg_isa_ext(RISCVCPU *cpu, CPUState *cs) } } +static void cpu_set_cfg_unavailable(Object *obj, Visitor *v, + const char *name, + void *opaque, Error **errp) +{ + const char *propname = opaque; + bool value; + + if (!visit_type_bool(v, name, &value, errp)) { + return; + } + + if (value) { + error_setg(errp, "extension %s is not available with KVM", + propname); + } +} + +static void riscv_cpu_add_kvm_unavail_prop(Object *obj, const char *prop_name) +{ + /* Check if KVM created the property already */ + if (object_property_find(obj, prop_name)) { + return; + } + + /* + * Set the default to disabled for every extension + * unknown to KVM and error out if the user attempts + * to enable any of them. + */ + object_property_add(obj, prop_name, "bool", + NULL, cpu_set_cfg_unavailable, + NULL, (void *)prop_name); +} + +static void riscv_cpu_add_kvm_unavail_prop_array(Object *obj, + const RISCVCPUMultiExtConfig *array) +{ + const RISCVCPUMultiExtConfig *prop; + + g_assert(array); + + for (prop = array; prop && prop->name; prop++) { + riscv_cpu_add_kvm_unavail_prop(obj, prop->name); + } +} + static void kvm_riscv_add_cpu_user_properties(Object *cpu_obj) { int i; @@ -XXX,XX +XXX,XX @@ static void kvm_riscv_init_multiext_cfg(RISCVCPU *cpu, KVMScratchCPU *kvmcpu) } } -void kvm_riscv_init_user_properties(Object *cpu_obj) +static void riscv_init_user_properties(Object *cpu_obj) { RISCVCPU *cpu = RISCV_CPU(cpu_obj); KVMScratchCPU kvmcpu; @@ -XXX,XX +XXX,XX @@ void kvm_riscv_aia_create(MachineState *machine, uint64_t group_shift, kvm_msi_via_irqfd_allowed = kvm_irqfds_enabled(); } +void kvm_riscv_cpu_add_kvm_properties(Object *obj) +{ + DeviceState *dev = DEVICE(obj); + + riscv_init_user_properties(obj); + riscv_cpu_add_misa_properties(obj); + + riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_extensions); + riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_vendor_exts); + riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_experimental_exts); + + for (Property *prop = riscv_cpu_options; prop && prop->name; prop++) { + /* Check if KVM created the property already */ + if (object_property_find(obj, prop->name)) { + continue; + } + qdev_property_add_static(dev, prop); + } +} + static void riscv_host_cpu_init(Object *obj) { CPURISCVState *env = &RISCV_CPU(obj)->env; -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> This function is used for both accelerators. Make it public, and call it from kvm_riscv_cpu_add_kvm_properties(). This will make it easier to split KVM specific code for the KVM accelerator class in the next patch. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-10-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 1 + target/riscv/cpu.c | 5 ++--- target/riscv/kvm.c | 1 + 3 files changed, 4 insertions(+), 3 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_experimental_exts[]; extern Property riscv_cpu_options[]; void riscv_cpu_add_misa_properties(Object *cpu_obj); +void riscv_add_satp_mode_properties(Object *obj); /* CSR function table */ extern riscv_csr_operations csr_ops[CSR_TABLE_SIZE]; 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 cpu_riscv_set_satp(Object *obj, Visitor *v, const char *name, satp_map->init |= 1 << satp; } -static void riscv_add_satp_mode_properties(Object *obj) +void riscv_add_satp_mode_properties(Object *obj) { RISCVCPU *cpu = RISCV_CPU(obj); @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_multiext_prop_array(Object *obj, static void riscv_cpu_add_user_properties(Object *obj) { #ifndef CONFIG_USER_ONLY - riscv_add_satp_mode_properties(obj); - if (kvm_enabled()) { kvm_riscv_cpu_add_kvm_properties(obj); return; } + riscv_add_satp_mode_properties(obj); #endif riscv_cpu_add_misa_properties(obj); diff --git a/target/riscv/kvm.c b/target/riscv/kvm.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm.c +++ b/target/riscv/kvm.c @@ -XXX,XX +XXX,XX @@ void kvm_riscv_cpu_add_kvm_properties(Object *obj) DeviceState *dev = DEVICE(obj); riscv_init_user_properties(obj); + riscv_add_satp_mode_properties(obj); riscv_cpu_add_misa_properties(obj); riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_extensions); -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> This file is not needed for some time now. Both kvm_riscv_reset_vcpu() and kvm_riscv_set_irq() have public declarations in kvm_riscv.h and are wrapped in 'if kvm_enabled()' blocks that the compiler will rip it out in non-KVM builds. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-11-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/kvm-stub.c | 30 ------------------------------ target/riscv/meson.build | 2 +- 2 files changed, 1 insertion(+), 31 deletions(-) delete mode 100644 target/riscv/kvm-stub.c diff --git a/target/riscv/kvm-stub.c b/target/riscv/kvm-stub.c deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/target/riscv/kvm-stub.c +++ /dev/null @@ -XXX,XX +XXX,XX @@ -/* - * QEMU KVM RISC-V specific function stubs - * - * Copyright (c) 2020 Huawei Technologies Co., Ltd - * - * 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 "cpu.h" -#include "kvm_riscv.h" - -void kvm_riscv_reset_vcpu(RISCVCPU *cpu) -{ - abort(); -} - -void kvm_riscv_set_irq(RISCVCPU *cpu, int irq, int level) -{ - abort(); -} 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 @@ riscv_ss.add(files( 'zce_helper.c', 'vcrypto_helper.c' )) -riscv_ss.add(when: 'CONFIG_KVM', if_true: files('kvm.c'), if_false: files('kvm-stub.c')) +riscv_ss.add(when: 'CONFIG_KVM', if_true: files('kvm.c')) riscv_system_ss = ss.source_set() riscv_system_ss.add(files( -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Add a KVM accelerator class like we did with TCG. The difference is that, at least for now, we won't be using a realize() implementation for this accelerator. We'll start by assiging kvm_riscv_cpu_add_kvm_properties(), renamed to kvm_cpu_instance_init(), as a 'cpu_instance_init' implementation. Change riscv_cpu_post_init() to invoke accel_cpu_instance_init(), which will go through the 'cpu_instance_init' impl of the current acceleration (if available) and execute it. The end result is that the KVM initial setup, i.e. starting registers and adding its specific properties, will be done via this hook. Add a 'tcg_enabled()' condition in riscv_cpu_post_init() to avoid calling riscv_cpu_add_user_properties() when running KVM. We'll remove this condition when the TCG accel class get its own 'cpu_instance_init' implementation. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-12-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/kvm_riscv.h | 1 - target/riscv/cpu.c | 8 +++----- target/riscv/kvm.c | 26 ++++++++++++++++++++++++-- 3 files changed, 27 insertions(+), 8 deletions(-) diff --git a/target/riscv/kvm_riscv.h b/target/riscv/kvm_riscv.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm_riscv.h +++ b/target/riscv/kvm_riscv.h @@ -XXX,XX +XXX,XX @@ #ifndef QEMU_KVM_RISCV_H #define QEMU_KVM_RISCV_H -void kvm_riscv_cpu_add_kvm_properties(Object *obj); void kvm_riscv_reset_vcpu(RISCVCPU *cpu); void kvm_riscv_set_irq(RISCVCPU *cpu, int irq, int level); void kvm_riscv_aia_create(MachineState *machine, uint64_t group_shift, 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 bool riscv_cpu_has_user_properties(Object *cpu_obj) static void riscv_cpu_post_init(Object *obj) { - if (riscv_cpu_has_user_properties(obj)) { + accel_cpu_instance_init(CPU(obj)); + + if (tcg_enabled() && riscv_cpu_has_user_properties(obj)) { riscv_cpu_add_user_properties(obj); } @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_multiext_prop_array(Object *obj, static void riscv_cpu_add_user_properties(Object *obj) { #ifndef CONFIG_USER_ONLY - if (kvm_enabled()) { - kvm_riscv_cpu_add_kvm_properties(obj); - return; - } riscv_add_satp_mode_properties(obj); #endif diff --git a/target/riscv/kvm.c b/target/riscv/kvm.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/kvm.c +++ b/target/riscv/kvm.c @@ -XXX,XX +XXX,XX @@ #include "sysemu/kvm_int.h" #include "cpu.h" #include "trace.h" +#include "hw/core/accel-cpu.h" #include "hw/pci/pci.h" #include "exec/memattrs.h" #include "exec/address-spaces.h" @@ -XXX,XX +XXX,XX @@ void kvm_riscv_aia_create(MachineState *machine, uint64_t group_shift, kvm_msi_via_irqfd_allowed = kvm_irqfds_enabled(); } -void kvm_riscv_cpu_add_kvm_properties(Object *obj) +static void kvm_cpu_instance_init(CPUState *cs) { + Object *obj = OBJECT(RISCV_CPU(cs)); DeviceState *dev = DEVICE(obj); riscv_init_user_properties(obj); @@ -XXX,XX +XXX,XX @@ void kvm_riscv_cpu_add_kvm_properties(Object *obj) riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_experimental_exts); for (Property *prop = riscv_cpu_options; prop && prop->name; prop++) { - /* Check if KVM created the property already */ + /* Check if we have a specific KVM handler for the option */ if (object_property_find(obj, prop->name)) { continue; } @@ -XXX,XX +XXX,XX @@ void kvm_riscv_cpu_add_kvm_properties(Object *obj) } } +static void kvm_cpu_accel_class_init(ObjectClass *oc, void *data) +{ + AccelCPUClass *acc = ACCEL_CPU_CLASS(oc); + + acc->cpu_instance_init = kvm_cpu_instance_init; +} + +static const TypeInfo kvm_cpu_accel_type_info = { + .name = ACCEL_CPU_NAME("kvm"), + + .parent = TYPE_ACCEL_CPU, + .class_init = kvm_cpu_accel_class_init, + .abstract = true, +}; +static void kvm_cpu_accel_register_types(void) +{ + type_register_static(&kvm_cpu_accel_type_info); +} +type_init(kvm_cpu_accel_register_types); + static void riscv_host_cpu_init(Object *obj) { CPURISCVState *env = &RISCV_CPU(obj)->env; -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Move the files to a 'kvm' dir to promote more code separation between accelerators and making our lives easier supporting build options such as --disable-tcg. Rename kvm.c to kvm-cpu.c to keep it in line with its TCG counterpart. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-13-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/{ => kvm}/kvm_riscv.h | 0 hw/intc/riscv_aplic.c | 2 +- hw/riscv/virt.c | 2 +- target/riscv/cpu.c | 2 +- target/riscv/{kvm.c => kvm/kvm-cpu.c} | 0 target/riscv/kvm/meson.build | 1 + target/riscv/meson.build | 2 +- 7 files changed, 5 insertions(+), 4 deletions(-) rename target/riscv/{ => kvm}/kvm_riscv.h (100%) rename target/riscv/{kvm.c => kvm/kvm-cpu.c} (100%) create mode 100644 target/riscv/kvm/meson.build diff --git a/target/riscv/kvm_riscv.h b/target/riscv/kvm/kvm_riscv.h similarity index 100% rename from target/riscv/kvm_riscv.h rename to target/riscv/kvm/kvm_riscv.h 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 @@ #include "target/riscv/cpu.h" #include "sysemu/sysemu.h" #include "sysemu/kvm.h" -#include "kvm_riscv.h" +#include "kvm/kvm_riscv.h" #include "migration/vmstate.h" #define APLIC_MAX_IDC (1UL << 14) 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 @@ #include "hw/riscv/virt.h" #include "hw/riscv/boot.h" #include "hw/riscv/numa.h" -#include "kvm_riscv.h" +#include "kvm/kvm_riscv.h" #include "hw/intc/riscv_aclint.h" #include "hw/intc/riscv_aplic.h" #include "hw/intc/riscv_imsic.h" 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 @@ #include "fpu/softfloat-helpers.h" #include "sysemu/kvm.h" #include "sysemu/tcg.h" -#include "kvm_riscv.h" +#include "kvm/kvm_riscv.h" #include "tcg/tcg.h" /* RISC-V CPU definitions */ diff --git a/target/riscv/kvm.c b/target/riscv/kvm/kvm-cpu.c similarity index 100% rename from target/riscv/kvm.c rename to target/riscv/kvm/kvm-cpu.c diff --git a/target/riscv/kvm/meson.build b/target/riscv/kvm/meson.build new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/riscv/kvm/meson.build @@ -0,0 +1 @@ +riscv_ss.add(when: 'CONFIG_KVM', if_true: files('kvm-cpu.c')) 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 @@ riscv_ss.add(files( 'zce_helper.c', 'vcrypto_helper.c' )) -riscv_ss.add(when: 'CONFIG_KVM', if_true: files('kvm.c')) riscv_system_ss = ss.source_set() riscv_system_ss.add(files( @@ -XXX,XX +XXX,XX @@ riscv_system_ss.add(files( )) subdir('tcg') +subdir('kvm') target_arch += {'riscv': riscv_ss} target_system_arch += {'riscv': riscv_system_ss} -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> riscv_cpu_add_misa_properties() is being used to fill the missing KVM MISA properties but it is a TCG helper that was adapted to do so. We'll move it to tcg-cpu.c in the next patches, meaning that KVM needs to fill the remaining MISA properties on its own. Do not use riscv_cpu_add_misa_properties(). Let's create a new array with all available MISA bits we support that can be read by KVM. The array is zero terminate to allow us to iterate through it without knowing its size. Then, inside kvm_riscv_add_cpu_user_properties(), we'll create all KVM MISA properties as usual and then use this array to add any missing MISA properties with the riscv_cpu_add_kvm_unavail_prop() helper. Note that we're creating misa_bits[], and not using the existing 'riscv_single_letter_exts[]', because the latter is tuned for riscv,isa related functions and it doesn't have all MISA bits we support. Commit 0e2c377023 ("target/riscv: misa to ISA string conversion fix") has the full context. While we're at it, move both satp and the multi-letter extension properties to kvm_riscv_add_cpu_user_properties() as well. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-14-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 3 ++- target/riscv/cpu.c | 2 ++ target/riscv/kvm/kvm-cpu.c | 22 ++++++++++++++-------- 3 files changed, 18 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 @@ #define RV(x) ((target_ulong)1 << (x - 'A')) /* - * Consider updating misa_ext_info_arr[] and misa_ext_cfgs[] + * Update misa_bits[], misa_ext_info_arr[] and misa_ext_cfgs[] * when adding new MISA bits here. */ #define RVI RV('I') @@ -XXX,XX +XXX,XX @@ #define RVJ RV('J') #define RVG RV('G') +extern const uint32_t misa_bits[]; const char *riscv_get_misa_ext_name(uint32_t bit); const char *riscv_get_misa_ext_description(uint32_t bit); 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 @@ /* RISC-V CPU definitions */ static const char riscv_single_letter_exts[] = "IEMAFDQCPVH"; +const uint32_t misa_bits[] = {RVI, RVE, RVM, RVA, RVF, RVD, RVV, + RVC, RVS, RVU, RVH, RVJ, RVG, 0}; struct isa_ext_data { const char *name; 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_add_cpu_user_properties(Object *cpu_obj) { int i; + riscv_add_satp_mode_properties(cpu_obj); + for (i = 0; i < ARRAY_SIZE(kvm_misa_ext_cfgs); i++) { KVMCPUConfig *misa_cfg = &kvm_misa_ext_cfgs[i]; int bit = misa_cfg->offset; @@ -XXX,XX +XXX,XX @@ static void kvm_riscv_add_cpu_user_properties(Object *cpu_obj) misa_cfg->description); } + for (i = 0; misa_bits[i] != 0; i++) { + const char *ext_name = riscv_get_misa_ext_name(misa_bits[i]); + riscv_cpu_add_kvm_unavail_prop(cpu_obj, ext_name); + } + for (i = 0; i < ARRAY_SIZE(kvm_multi_ext_cfgs); i++) { KVMCPUConfig *multi_cfg = &kvm_multi_ext_cfgs[i]; @@ -XXX,XX +XXX,XX @@ static void kvm_riscv_add_cpu_user_properties(Object *cpu_obj) object_property_add(cpu_obj, "cboz_blocksize", "uint16", NULL, kvm_cpu_set_cbomz_blksize, NULL, &kvm_cboz_blocksize); + + riscv_cpu_add_kvm_unavail_prop_array(cpu_obj, riscv_cpu_extensions); + riscv_cpu_add_kvm_unavail_prop_array(cpu_obj, riscv_cpu_vendor_exts); + riscv_cpu_add_kvm_unavail_prop_array(cpu_obj, riscv_cpu_experimental_exts); } static int kvm_riscv_get_regs_core(CPUState *cs) @@ -XXX,XX +XXX,XX @@ static void kvm_riscv_init_multiext_cfg(RISCVCPU *cpu, KVMScratchCPU *kvmcpu) } } -static void riscv_init_user_properties(Object *cpu_obj) +static void riscv_init_kvm_registers(Object *cpu_obj) { RISCVCPU *cpu = RISCV_CPU(cpu_obj); KVMScratchCPU kvmcpu; @@ -XXX,XX +XXX,XX @@ static void riscv_init_user_properties(Object *cpu_obj) return; } - kvm_riscv_add_cpu_user_properties(cpu_obj); kvm_riscv_init_machine_ids(cpu, &kvmcpu); kvm_riscv_init_misa_ext_mask(cpu, &kvmcpu); kvm_riscv_init_multiext_cfg(cpu, &kvmcpu); @@ -XXX,XX +XXX,XX @@ static void kvm_cpu_instance_init(CPUState *cs) Object *obj = OBJECT(RISCV_CPU(cs)); DeviceState *dev = DEVICE(obj); - riscv_init_user_properties(obj); - riscv_add_satp_mode_properties(obj); - riscv_cpu_add_misa_properties(obj); + riscv_init_kvm_registers(obj); - riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_extensions); - riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_vendor_exts); - riscv_cpu_add_kvm_unavail_prop_array(obj, riscv_cpu_experimental_exts); + kvm_riscv_add_cpu_user_properties(obj); for (Property *prop = riscv_cpu_options; prop && prop->name; prop++) { /* Check if we have a specific KVM handler for the option */ -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> We'll move riscv_init_max_cpu_extensions() to tcg-cpu.c in the next patch and set_misa() needs to be usable from there. Rename it to riscv_cpu_set_misa() and make it public. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-15-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 1 + target/riscv/cpu.c | 34 ++++++++++++++++++---------------- 2 files changed, 19 insertions(+), 16 deletions(-) diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.h +++ b/target/riscv/cpu.h @@ -XXX,XX +XXX,XX @@ void isa_ext_update_enabled(RISCVCPU *cpu, uint32_t ext_offset, bool en); bool cpu_cfg_ext_is_user_set(uint32_t ext_offset); bool isa_ext_is_enabled(RISCVCPU *cpu, uint32_t ext_offset); int cpu_cfg_ext_get_min_version(uint32_t ext_offset); +void riscv_cpu_set_misa(CPURISCVState *env, RISCVMXL mxl, uint32_t ext); void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu); typedef struct RISCVCPUMultiExtConfig { diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const char *riscv_cpu_get_trap_name(target_ulong cause, bool async) } } -static void set_misa(CPURISCVState *env, RISCVMXL mxl, uint32_t ext) +void riscv_cpu_set_misa(CPURISCVState *env, RISCVMXL mxl, uint32_t ext) { env->misa_mxl_max = env->misa_mxl = mxl; env->misa_ext_mask = env->misa_ext = ext; @@ -XXX,XX +XXX,XX @@ static void riscv_any_cpu_init(Object *obj) RISCVCPU *cpu = RISCV_CPU(obj); CPURISCVState *env = &cpu->env; #if defined(TARGET_RISCV32) - set_misa(env, MXL_RV32, RVI | RVM | RVA | RVF | RVD | RVC | RVU); + riscv_cpu_set_misa(env, MXL_RV32, RVI | RVM | RVA | RVF | RVD | RVC | RVU); #elif defined(TARGET_RISCV64) - set_misa(env, MXL_RV64, RVI | RVM | RVA | RVF | RVD | RVC | RVU); + riscv_cpu_set_misa(env, MXL_RV64, RVI | RVM | RVA | RVF | RVD | RVC | RVU); #endif #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ static void riscv_max_cpu_init(Object *obj) #ifdef TARGET_RISCV32 mlx = MXL_RV32; #endif - set_misa(env, mlx, 0); + riscv_cpu_set_misa(env, mlx, 0); env->priv_ver = PRIV_VERSION_LATEST; #ifndef CONFIG_USER_ONLY set_satp_mode_max_supported(RISCV_CPU(obj), mlx == MXL_RV32 ? @@ -XXX,XX +XXX,XX @@ static void rv64_base_cpu_init(Object *obj) { CPURISCVState *env = &RISCV_CPU(obj)->env; /* We set this in the realise function */ - set_misa(env, MXL_RV64, 0); + riscv_cpu_set_misa(env, MXL_RV64, 0); /* Set latest version of privileged specification */ env->priv_ver = PRIV_VERSION_LATEST; #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ static void rv64_sifive_u_cpu_init(Object *obj) { RISCVCPU *cpu = RISCV_CPU(obj); CPURISCVState *env = &cpu->env; - set_misa(env, MXL_RV64, RVI | RVM | RVA | RVF | RVD | RVC | RVS | RVU); + riscv_cpu_set_misa(env, MXL_RV64, + RVI | RVM | RVA | RVF | RVD | RVC | RVS | RVU); env->priv_ver = PRIV_VERSION_1_10_0; #ifndef CONFIG_USER_ONLY set_satp_mode_max_supported(RISCV_CPU(obj), VM_1_10_SV39); @@ -XXX,XX +XXX,XX @@ static void rv64_sifive_e_cpu_init(Object *obj) CPURISCVState *env = &RISCV_CPU(obj)->env; RISCVCPU *cpu = RISCV_CPU(obj); - set_misa(env, MXL_RV64, RVI | RVM | RVA | RVC | RVU); + riscv_cpu_set_misa(env, MXL_RV64, RVI | RVM | RVA | RVC | RVU); env->priv_ver = PRIV_VERSION_1_10_0; #ifndef CONFIG_USER_ONLY set_satp_mode_max_supported(cpu, VM_1_10_MBARE); @@ -XXX,XX +XXX,XX @@ static void rv64_thead_c906_cpu_init(Object *obj) CPURISCVState *env = &RISCV_CPU(obj)->env; RISCVCPU *cpu = RISCV_CPU(obj); - set_misa(env, MXL_RV64, RVG | RVC | RVS | RVU); + riscv_cpu_set_misa(env, MXL_RV64, RVG | RVC | RVS | RVU); env->priv_ver = PRIV_VERSION_1_11_0; cpu->cfg.ext_zfa = true; @@ -XXX,XX +XXX,XX @@ static void rv64_veyron_v1_cpu_init(Object *obj) CPURISCVState *env = &RISCV_CPU(obj)->env; RISCVCPU *cpu = RISCV_CPU(obj); - set_misa(env, MXL_RV64, RVG | RVC | RVS | RVU | RVH); + riscv_cpu_set_misa(env, MXL_RV64, RVG | RVC | RVS | RVU | RVH); env->priv_ver = PRIV_VERSION_1_12_0; /* Enable ISA extensions */ @@ -XXX,XX +XXX,XX @@ static void rv128_base_cpu_init(Object *obj) } CPURISCVState *env = &RISCV_CPU(obj)->env; /* We set this in the realise function */ - set_misa(env, MXL_RV128, 0); + riscv_cpu_set_misa(env, MXL_RV128, 0); /* Set latest version of privileged specification */ env->priv_ver = PRIV_VERSION_LATEST; #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ static void rv32_base_cpu_init(Object *obj) { CPURISCVState *env = &RISCV_CPU(obj)->env; /* We set this in the realise function */ - set_misa(env, MXL_RV32, 0); + riscv_cpu_set_misa(env, MXL_RV32, 0); /* Set latest version of privileged specification */ env->priv_ver = PRIV_VERSION_LATEST; #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ static void rv32_sifive_u_cpu_init(Object *obj) { RISCVCPU *cpu = RISCV_CPU(obj); CPURISCVState *env = &cpu->env; - set_misa(env, MXL_RV32, RVI | RVM | RVA | RVF | RVD | RVC | RVS | RVU); + riscv_cpu_set_misa(env, MXL_RV32, + RVI | RVM | RVA | RVF | RVD | RVC | RVS | RVU); env->priv_ver = PRIV_VERSION_1_10_0; #ifndef CONFIG_USER_ONLY set_satp_mode_max_supported(RISCV_CPU(obj), VM_1_10_SV32); @@ -XXX,XX +XXX,XX @@ static void rv32_sifive_e_cpu_init(Object *obj) CPURISCVState *env = &RISCV_CPU(obj)->env; RISCVCPU *cpu = RISCV_CPU(obj); - set_misa(env, MXL_RV32, RVI | RVM | RVA | RVC | RVU); + riscv_cpu_set_misa(env, MXL_RV32, RVI | RVM | RVA | RVC | RVU); env->priv_ver = PRIV_VERSION_1_10_0; #ifndef CONFIG_USER_ONLY set_satp_mode_max_supported(cpu, VM_1_10_MBARE); @@ -XXX,XX +XXX,XX @@ static void rv32_ibex_cpu_init(Object *obj) CPURISCVState *env = &RISCV_CPU(obj)->env; RISCVCPU *cpu = RISCV_CPU(obj); - set_misa(env, MXL_RV32, RVI | RVM | RVC | RVU); + riscv_cpu_set_misa(env, MXL_RV32, RVI | RVM | RVC | RVU); env->priv_ver = PRIV_VERSION_1_11_0; #ifndef CONFIG_USER_ONLY set_satp_mode_max_supported(cpu, VM_1_10_MBARE); @@ -XXX,XX +XXX,XX @@ static void rv32_imafcu_nommu_cpu_init(Object *obj) CPURISCVState *env = &RISCV_CPU(obj)->env; RISCVCPU *cpu = RISCV_CPU(obj); - set_misa(env, MXL_RV32, RVI | RVM | RVA | RVF | RVC | RVU); + riscv_cpu_set_misa(env, MXL_RV32, RVI | RVM | RVA | RVF | RVC | RVU); env->priv_ver = PRIV_VERSION_1_10_0; #ifndef CONFIG_USER_ONLY set_satp_mode_max_supported(cpu, VM_1_10_MBARE); @@ -XXX,XX +XXX,XX @@ static void riscv_init_max_cpu_extensions(Object *obj) const RISCVCPUMultiExtConfig *prop; /* Enable RVG, RVJ and RVV that are disabled by default */ - set_misa(env, env->misa_mxl, env->misa_ext | RVG | RVJ | RVV); + riscv_cpu_set_misa(env, env->misa_mxl, env->misa_ext | RVG | RVJ | RVV); for (prop = riscv_cpu_extensions; prop && prop->name; prop++) { isa_ext_update_enabled(cpu, prop->offset, true); -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> tcg_cpu_instance_init() will be the 'cpu_instance_init' impl for the TCG accelerator. It'll be called from within riscv_cpu_post_init(), via accel_cpu_instance_init(), similar to what happens with KVM. In fact, to preserve behavior, the implementation will be similar to what riscv_cpu_post_init() already does. In this patch we'll move riscv_cpu_add_user_properties() and riscv_init_max_cpu_extensions() and all their dependencies to tcg-cpu.c. All multi-extension properties code was moved. The 'multi_ext_user_opts' hash table was also moved to tcg-cpu.c since it's a TCG only structure, meaning that we won't have to worry about initializing a TCG hash table when running a KVM CPU anymore. riscv_cpu_add_user_properties() will remain in cpu.c for now due to how much code it requires to be moved at the same time. We'll do that in the next patch. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-16-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 1 - target/riscv/cpu.c | 150 ------------------------------------- target/riscv/tcg/tcg-cpu.c | 149 ++++++++++++++++++++++++++++++++++++ 3 files changed, 149 insertions(+), 151 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 @@ enum riscv_pmu_event_idx { /* used by tcg/tcg-cpu.c*/ void isa_ext_update_enabled(RISCVCPU *cpu, uint32_t ext_offset, bool en); -bool cpu_cfg_ext_is_user_set(uint32_t ext_offset); bool isa_ext_is_enabled(RISCVCPU *cpu, uint32_t ext_offset); int cpu_cfg_ext_get_min_version(uint32_t ext_offset); void riscv_cpu_set_misa(CPURISCVState *env, RISCVMXL mxl, uint32_t ext); 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 struct isa_ext_data isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(xventanacondops, PRIV_VERSION_1_12_0, ext_XVentanaCondOps), }; -/* Hash that stores user set extensions */ -static GHashTable *multi_ext_user_opts; - bool isa_ext_is_enabled(RISCVCPU *cpu, uint32_t ext_offset) { bool *ext_enabled = (void *)&cpu->cfg + ext_offset; @@ -XXX,XX +XXX,XX @@ int cpu_cfg_ext_get_min_version(uint32_t ext_offset) g_assert_not_reached(); } -bool cpu_cfg_ext_is_user_set(uint32_t ext_offset) -{ - return g_hash_table_contains(multi_ext_user_opts, - GUINT_TO_POINTER(ext_offset)); -} - const char * const riscv_int_regnames[] = { "x0/zero", "x1/ra", "x2/sp", "x3/gp", "x4/tp", "x5/t0", "x6/t1", "x7/t2", "x8/s0", "x9/s1", "x10/a0", "x11/a1", "x12/a2", "x13/a3", @@ -XXX,XX +XXX,XX @@ static const char * const riscv_intr_names[] = { "reserved" }; -static void riscv_cpu_add_user_properties(Object *obj); -static void riscv_init_max_cpu_extensions(Object *obj); - const char *riscv_cpu_get_trap_name(target_ulong cause, bool async) { if (async) { @@ -XXX,XX +XXX,XX @@ static bool riscv_cpu_is_dynamic(Object *cpu_obj) return object_dynamic_cast(cpu_obj, TYPE_RISCV_DYNAMIC_CPU) != NULL; } -static bool riscv_cpu_has_max_extensions(Object *cpu_obj) -{ - return object_dynamic_cast(cpu_obj, TYPE_RISCV_CPU_MAX) != NULL; -} - -static bool riscv_cpu_has_user_properties(Object *cpu_obj) -{ - if (kvm_enabled() && - object_dynamic_cast(cpu_obj, TYPE_RISCV_CPU_HOST) != NULL) { - return true; - } - - return riscv_cpu_is_dynamic(cpu_obj); -} - static void riscv_cpu_post_init(Object *obj) { accel_cpu_instance_init(CPU(obj)); - - if (tcg_enabled() && riscv_cpu_has_user_properties(obj)) { - riscv_cpu_add_user_properties(obj); - } - - if (riscv_cpu_has_max_extensions(obj)) { - riscv_init_max_cpu_extensions(obj); - } } static void riscv_cpu_init(Object *obj) @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_init(Object *obj) qdev_init_gpio_in(DEVICE(obj), riscv_cpu_set_irq, IRQ_LOCAL_MAX + IRQ_LOCAL_GUEST_MAX); #endif /* CONFIG_USER_ONLY */ - - multi_ext_user_opts = g_hash_table_new(NULL, g_direct_equal); } typedef struct RISCVCPUMisaExtConfig { @@ -XXX,XX +XXX,XX @@ Property riscv_cpu_options[] = { DEFINE_PROP_END_OF_LIST(), }; -static void cpu_set_multi_ext_cfg(Object *obj, Visitor *v, const char *name, - void *opaque, Error **errp) -{ - const RISCVCPUMultiExtConfig *multi_ext_cfg = opaque; - bool value; - - if (!visit_type_bool(v, name, &value, errp)) { - return; - } - - isa_ext_update_enabled(RISCV_CPU(obj), multi_ext_cfg->offset, value); - - g_hash_table_insert(multi_ext_user_opts, - GUINT_TO_POINTER(multi_ext_cfg->offset), - (gpointer)value); -} - -static void cpu_get_multi_ext_cfg(Object *obj, Visitor *v, const char *name, - void *opaque, Error **errp) -{ - const RISCVCPUMultiExtConfig *multi_ext_cfg = opaque; - bool value = isa_ext_is_enabled(RISCV_CPU(obj), multi_ext_cfg->offset); - - visit_type_bool(v, name, &value, errp); -} - -static void cpu_add_multi_ext_prop(Object *cpu_obj, - const RISCVCPUMultiExtConfig *multi_cfg) -{ - object_property_add(cpu_obj, multi_cfg->name, "bool", - cpu_get_multi_ext_cfg, - cpu_set_multi_ext_cfg, - NULL, (void *)multi_cfg); - - /* - * Set def val directly instead of using - * object_property_set_bool() to save the set() - * callback hash for user inputs. - */ - isa_ext_update_enabled(RISCV_CPU(cpu_obj), multi_cfg->offset, - multi_cfg->enabled); -} - -static void riscv_cpu_add_multiext_prop_array(Object *obj, - const RISCVCPUMultiExtConfig *array) -{ - const RISCVCPUMultiExtConfig *prop; - - g_assert(array); - - for (prop = array; prop && prop->name; prop++) { - cpu_add_multi_ext_prop(obj, prop); - } -} - -/* - * Add CPU properties with user-facing flags. - * - * This will overwrite existing env->misa_ext values with the - * defaults set via riscv_cpu_add_misa_properties(). - */ -static void riscv_cpu_add_user_properties(Object *obj) -{ -#ifndef CONFIG_USER_ONLY - riscv_add_satp_mode_properties(obj); -#endif - - riscv_cpu_add_misa_properties(obj); - - riscv_cpu_add_multiext_prop_array(obj, riscv_cpu_extensions); - riscv_cpu_add_multiext_prop_array(obj, riscv_cpu_vendor_exts); - riscv_cpu_add_multiext_prop_array(obj, riscv_cpu_experimental_exts); - - for (Property *prop = riscv_cpu_options; prop && prop->name; prop++) { - qdev_property_add_static(DEVICE(obj), prop); - } -} - -/* - * The 'max' type CPU will have all possible ratified - * non-vendor extensions enabled. - */ -static void riscv_init_max_cpu_extensions(Object *obj) -{ - RISCVCPU *cpu = RISCV_CPU(obj); - CPURISCVState *env = &cpu->env; - const RISCVCPUMultiExtConfig *prop; - - /* Enable RVG, RVJ and RVV that are disabled by default */ - riscv_cpu_set_misa(env, env->misa_mxl, env->misa_ext | RVG | RVJ | RVV); - - for (prop = riscv_cpu_extensions; prop && prop->name; prop++) { - isa_ext_update_enabled(cpu, prop->offset, true); - } - - /* set vector version */ - env->vext_ver = VEXT_VERSION_1_00_0; - - /* Zfinx is not compatible with F. Disable it */ - isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zfinx), false); - isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zdinx), false); - isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zhinx), false); - isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zhinxmin), false); - - isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zce), false); - isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zcmp), false); - isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zcmt), false); - - if (env->misa_mxl != MXL_RV32) { - isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zcf), false); - } -} - static Property riscv_cpu_properties[] = { DEFINE_PROP_BOOL("debug", RISCVCPU, cfg.debug, true), 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 @@ #include "pmu.h" #include "time_helper.h" #include "qapi/error.h" +#include "qapi/visitor.h" #include "qemu/accel.h" #include "qemu/error-report.h" #include "qemu/log.h" @@ -XXX,XX +XXX,XX @@ #include "hw/core/tcg-cpu-ops.h" #include "tcg/tcg.h" +/* Hash that stores user set extensions */ +static GHashTable *multi_ext_user_opts; + +static bool cpu_cfg_ext_is_user_set(uint32_t ext_offset) +{ + return g_hash_table_contains(multi_ext_user_opts, + GUINT_TO_POINTER(ext_offset)); +} + static void riscv_cpu_synchronize_from_tb(CPUState *cs, const TranslationBlock *tb) { @@ -XXX,XX +XXX,XX @@ static bool tcg_cpu_realize(CPUState *cs, Error **errp) return true; } +static void cpu_set_multi_ext_cfg(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + const RISCVCPUMultiExtConfig *multi_ext_cfg = opaque; + bool value; + + if (!visit_type_bool(v, name, &value, errp)) { + return; + } + + isa_ext_update_enabled(RISCV_CPU(obj), multi_ext_cfg->offset, value); + + g_hash_table_insert(multi_ext_user_opts, + GUINT_TO_POINTER(multi_ext_cfg->offset), + (gpointer)value); +} + +static void cpu_get_multi_ext_cfg(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + const RISCVCPUMultiExtConfig *multi_ext_cfg = opaque; + bool value = isa_ext_is_enabled(RISCV_CPU(obj), multi_ext_cfg->offset); + + visit_type_bool(v, name, &value, errp); +} + +static void cpu_add_multi_ext_prop(Object *cpu_obj, + const RISCVCPUMultiExtConfig *multi_cfg) +{ + object_property_add(cpu_obj, multi_cfg->name, "bool", + cpu_get_multi_ext_cfg, + cpu_set_multi_ext_cfg, + NULL, (void *)multi_cfg); + + /* + * Set def val directly instead of using + * object_property_set_bool() to save the set() + * callback hash for user inputs. + */ + isa_ext_update_enabled(RISCV_CPU(cpu_obj), multi_cfg->offset, + multi_cfg->enabled); +} + +static void riscv_cpu_add_multiext_prop_array(Object *obj, + const RISCVCPUMultiExtConfig *array) +{ + const RISCVCPUMultiExtConfig *prop; + + g_assert(array); + + for (prop = array; prop && prop->name; prop++) { + cpu_add_multi_ext_prop(obj, prop); + } +} + +/* + * Add CPU properties with user-facing flags. + * + * This will overwrite existing env->misa_ext values with the + * defaults set via riscv_cpu_add_misa_properties(). + */ +static void riscv_cpu_add_user_properties(Object *obj) +{ +#ifndef CONFIG_USER_ONLY + riscv_add_satp_mode_properties(obj); +#endif + + riscv_cpu_add_misa_properties(obj); + + riscv_cpu_add_multiext_prop_array(obj, riscv_cpu_extensions); + riscv_cpu_add_multiext_prop_array(obj, riscv_cpu_vendor_exts); + riscv_cpu_add_multiext_prop_array(obj, riscv_cpu_experimental_exts); + + for (Property *prop = riscv_cpu_options; prop && prop->name; prop++) { + qdev_property_add_static(DEVICE(obj), prop); + } +} + +/* + * The 'max' type CPU will have all possible ratified + * non-vendor extensions enabled. + */ +static void riscv_init_max_cpu_extensions(Object *obj) +{ + RISCVCPU *cpu = RISCV_CPU(obj); + CPURISCVState *env = &cpu->env; + const RISCVCPUMultiExtConfig *prop; + + /* Enable RVG, RVJ and RVV that are disabled by default */ + riscv_cpu_set_misa(env, env->misa_mxl, env->misa_ext | RVG | RVJ | RVV); + + for (prop = riscv_cpu_extensions; prop && prop->name; prop++) { + isa_ext_update_enabled(cpu, prop->offset, true); + } + + /* set vector version */ + env->vext_ver = VEXT_VERSION_1_00_0; + + /* Zfinx is not compatible with F. Disable it */ + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zfinx), false); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zdinx), false); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zhinx), false); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zhinxmin), false); + + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zce), false); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zcmp), false); + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zcmt), false); + + if (env->misa_mxl != MXL_RV32) { + isa_ext_update_enabled(cpu, CPU_CFG_OFFSET(ext_zcf), false); + } +} + +static bool riscv_cpu_has_max_extensions(Object *cpu_obj) +{ + return object_dynamic_cast(cpu_obj, TYPE_RISCV_CPU_MAX) != NULL; +} + +static bool riscv_cpu_has_user_properties(Object *cpu_obj) +{ + return object_dynamic_cast(cpu_obj, TYPE_RISCV_DYNAMIC_CPU) != NULL; +} + +static void tcg_cpu_instance_init(CPUState *cs) +{ + RISCVCPU *cpu = RISCV_CPU(cs); + Object *obj = OBJECT(cpu); + + if (riscv_cpu_has_user_properties(obj)) { + multi_ext_user_opts = g_hash_table_new(NULL, g_direct_equal); + riscv_cpu_add_user_properties(obj); + } + + if (riscv_cpu_has_max_extensions(obj)) { + riscv_init_max_cpu_extensions(obj); + } +} + static void tcg_cpu_init_ops(AccelCPUClass *accel_cpu, CPUClass *cc) { /* @@ -XXX,XX +XXX,XX @@ static void tcg_cpu_accel_class_init(ObjectClass *oc, void *data) AccelCPUClass *acc = ACCEL_CPU_CLASS(oc); acc->cpu_class_init = tcg_cpu_class_init; + acc->cpu_instance_init = tcg_cpu_instance_init; acc->cpu_target_realize = tcg_cpu_realize; } -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> The array isn't marked as 'const' because we're initializing their elements in riscv_cpu_add_misa_properties(), 'name' and 'description' fields. In a closer look we can see that we're not using these 2 fields after creating the MISA properties. And we can create the properties by using riscv_get_misa_ext_name() and riscv_get_misa_ext_description() directly. Remove the 'name' and 'description' fields from RISCVCPUMisaExtConfig and make misa_ext_cfgs[] a const array. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-17-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.c | 21 ++++++++------------- 1 file changed, 8 insertions(+), 13 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_init(Object *obj) } typedef struct RISCVCPUMisaExtConfig { - const char *name; - const char *description; target_ulong misa_bit; bool enabled; } RISCVCPUMisaExtConfig; @@ -XXX,XX +XXX,XX @@ const char *riscv_get_misa_ext_description(uint32_t bit) #define MISA_CFG(_bit, _enabled) \ {.misa_bit = _bit, .enabled = _enabled} -static RISCVCPUMisaExtConfig misa_ext_cfgs[] = { +static const RISCVCPUMisaExtConfig misa_ext_cfgs[] = { MISA_CFG(RVA, true), MISA_CFG(RVC, true), MISA_CFG(RVD, true), @@ -XXX,XX +XXX,XX @@ void riscv_cpu_add_misa_properties(Object *cpu_obj) int i; for (i = 0; i < ARRAY_SIZE(misa_ext_cfgs); i++) { - RISCVCPUMisaExtConfig *misa_cfg = &misa_ext_cfgs[i]; + const RISCVCPUMisaExtConfig *misa_cfg = &misa_ext_cfgs[i]; int bit = misa_cfg->misa_bit; - - misa_cfg->name = riscv_get_misa_ext_name(bit); - misa_cfg->description = riscv_get_misa_ext_description(bit); + const char *name = riscv_get_misa_ext_name(bit); + const char *desc = riscv_get_misa_ext_description(bit); /* Check if KVM already created the property */ - if (object_property_find(cpu_obj, misa_cfg->name)) { + if (object_property_find(cpu_obj, name)) { continue; } - object_property_add(cpu_obj, misa_cfg->name, "bool", + object_property_add(cpu_obj, name, "bool", cpu_get_misa_ext_cfg, cpu_set_misa_ext_cfg, NULL, (void *)misa_cfg); - object_property_set_description(cpu_obj, misa_cfg->name, - misa_cfg->description); - object_property_set_bool(cpu_obj, misa_cfg->name, - misa_cfg->enabled, NULL); + object_property_set_description(cpu_obj, name, desc); + object_property_set_bool(cpu_obj, name, misa_cfg->enabled, NULL); } } -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> All code related to MISA TCG properties is also moved. At this point, all TCG properties handling is done in tcg-cpu.c, all KVM properties handling is done in kvm-cpu.c. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-18-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 1 - target/riscv/cpu.c | 90 -------------------------------------- target/riscv/tcg/tcg-cpu.c | 90 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 90 insertions(+), 91 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_vendor_exts[]; extern const RISCVCPUMultiExtConfig riscv_cpu_experimental_exts[]; extern Property riscv_cpu_options[]; -void riscv_cpu_add_misa_properties(Object *cpu_obj); void riscv_add_satp_mode_properties(Object *obj); /* CSR function table */ 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) #endif /* CONFIG_USER_ONLY */ } -typedef struct RISCVCPUMisaExtConfig { - target_ulong misa_bit; - bool enabled; -} RISCVCPUMisaExtConfig; - -static void cpu_set_misa_ext_cfg(Object *obj, Visitor *v, const char *name, - void *opaque, Error **errp) -{ - const RISCVCPUMisaExtConfig *misa_ext_cfg = opaque; - target_ulong misa_bit = misa_ext_cfg->misa_bit; - RISCVCPU *cpu = RISCV_CPU(obj); - CPURISCVState *env = &cpu->env; - bool value; - - if (!visit_type_bool(v, name, &value, errp)) { - return; - } - - if (value) { - env->misa_ext |= misa_bit; - env->misa_ext_mask |= misa_bit; - } else { - env->misa_ext &= ~misa_bit; - env->misa_ext_mask &= ~misa_bit; - } -} - -static void cpu_get_misa_ext_cfg(Object *obj, Visitor *v, const char *name, - void *opaque, Error **errp) -{ - const RISCVCPUMisaExtConfig *misa_ext_cfg = opaque; - target_ulong misa_bit = misa_ext_cfg->misa_bit; - RISCVCPU *cpu = RISCV_CPU(obj); - CPURISCVState *env = &cpu->env; - bool value; - - value = env->misa_ext & misa_bit; - - visit_type_bool(v, name, &value, errp); -} - typedef struct misa_ext_info { const char *name; const char *description; @@ -XXX,XX +XXX,XX @@ const char *riscv_get_misa_ext_description(uint32_t bit) return val; } -#define MISA_CFG(_bit, _enabled) \ - {.misa_bit = _bit, .enabled = _enabled} - -static const RISCVCPUMisaExtConfig misa_ext_cfgs[] = { - MISA_CFG(RVA, true), - MISA_CFG(RVC, true), - MISA_CFG(RVD, true), - MISA_CFG(RVF, true), - MISA_CFG(RVI, true), - MISA_CFG(RVE, false), - MISA_CFG(RVM, true), - MISA_CFG(RVS, true), - MISA_CFG(RVU, true), - MISA_CFG(RVH, true), - MISA_CFG(RVJ, false), - MISA_CFG(RVV, false), - MISA_CFG(RVG, false), -}; - -/* - * We do not support user choice tracking for MISA - * extensions yet because, so far, we do not silently - * change MISA bits during realize() (RVG enables MISA - * bits but the user is warned about it). - */ -void riscv_cpu_add_misa_properties(Object *cpu_obj) -{ - int i; - - for (i = 0; i < ARRAY_SIZE(misa_ext_cfgs); i++) { - const RISCVCPUMisaExtConfig *misa_cfg = &misa_ext_cfgs[i]; - int bit = misa_cfg->misa_bit; - const char *name = riscv_get_misa_ext_name(bit); - const char *desc = riscv_get_misa_ext_description(bit); - - /* Check if KVM already created the property */ - if (object_property_find(cpu_obj, name)) { - continue; - } - - object_property_add(cpu_obj, name, "bool", - cpu_get_misa_ext_cfg, - cpu_set_misa_ext_cfg, - NULL, (void *)misa_cfg); - object_property_set_description(cpu_obj, name, desc); - object_property_set_bool(cpu_obj, name, misa_cfg->enabled, NULL); - } -} - #define MULTI_EXT_CFG_BOOL(_name, _prop, _defval) \ {.name = _name, .offset = CPU_CFG_OFFSET(_prop), \ .enabled = _defval} 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 tcg_cpu_realize(CPUState *cs, Error **errp) return true; } +typedef struct RISCVCPUMisaExtConfig { + target_ulong misa_bit; + bool enabled; +} RISCVCPUMisaExtConfig; + +static void cpu_set_misa_ext_cfg(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + const RISCVCPUMisaExtConfig *misa_ext_cfg = opaque; + target_ulong misa_bit = misa_ext_cfg->misa_bit; + RISCVCPU *cpu = RISCV_CPU(obj); + CPURISCVState *env = &cpu->env; + bool value; + + if (!visit_type_bool(v, name, &value, errp)) { + return; + } + + if (value) { + env->misa_ext |= misa_bit; + env->misa_ext_mask |= misa_bit; + } else { + env->misa_ext &= ~misa_bit; + env->misa_ext_mask &= ~misa_bit; + } +} + +static void cpu_get_misa_ext_cfg(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + const RISCVCPUMisaExtConfig *misa_ext_cfg = opaque; + target_ulong misa_bit = misa_ext_cfg->misa_bit; + RISCVCPU *cpu = RISCV_CPU(obj); + CPURISCVState *env = &cpu->env; + bool value; + + value = env->misa_ext & misa_bit; + + visit_type_bool(v, name, &value, errp); +} + +#define MISA_CFG(_bit, _enabled) \ + {.misa_bit = _bit, .enabled = _enabled} + +static const RISCVCPUMisaExtConfig misa_ext_cfgs[] = { + MISA_CFG(RVA, true), + MISA_CFG(RVC, true), + MISA_CFG(RVD, true), + MISA_CFG(RVF, true), + MISA_CFG(RVI, true), + MISA_CFG(RVE, false), + MISA_CFG(RVM, true), + MISA_CFG(RVS, true), + MISA_CFG(RVU, true), + MISA_CFG(RVH, true), + MISA_CFG(RVJ, false), + MISA_CFG(RVV, false), + MISA_CFG(RVG, false), +}; + +/* + * We do not support user choice tracking for MISA + * extensions yet because, so far, we do not silently + * change MISA bits during realize() (RVG enables MISA + * bits but the user is warned about it). + */ +static void riscv_cpu_add_misa_properties(Object *cpu_obj) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(misa_ext_cfgs); i++) { + const RISCVCPUMisaExtConfig *misa_cfg = &misa_ext_cfgs[i]; + int bit = misa_cfg->misa_bit; + const char *name = riscv_get_misa_ext_name(bit); + const char *desc = riscv_get_misa_ext_description(bit); + + /* Check if KVM already created the property */ + if (object_property_find(cpu_obj, name)) { + continue; + } + + object_property_add(cpu_obj, name, "bool", + cpu_get_misa_ext_cfg, + cpu_set_misa_ext_cfg, + NULL, (void *)misa_cfg); + object_property_set_description(cpu_obj, name, desc); + object_property_set_bool(cpu_obj, name, misa_cfg->enabled, NULL); + } +} + static void cpu_set_multi_ext_cfg(Object *obj, Visitor *v, const char *name, void *opaque, Error **errp) { -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> This array will be read by the TCG accel class, allowing it to handle priv spec verifications on its own. The array will remain here in cpu.c because it's also used by the riscv,isa string function. To export it we'll finish it with an empty element since ARRAY_SIZE() won't work outside of cpu.c. Get rid of its ARRAY_SIZE() usage now to alleviate the changes for the next patch. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230925175709.35696-19-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 7 +++++++ target/riscv/cpu.c | 47 +++++++++++++++++++++------------------------- 2 files changed, 28 insertions(+), 26 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_vendor_exts[]; extern const RISCVCPUMultiExtConfig riscv_cpu_experimental_exts[]; extern Property riscv_cpu_options[]; +typedef struct isa_ext_data { + const char *name; + int min_version; + int ext_enable_offset; +} RISCVIsaExtData; +extern const RISCVIsaExtData isa_edata_arr[]; + void riscv_add_satp_mode_properties(Object *obj); /* CSR function table */ 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 char riscv_single_letter_exts[] = "IEMAFDQCPVH"; const uint32_t misa_bits[] = {RVI, RVE, RVM, RVA, RVF, RVD, RVV, RVC, RVS, RVU, RVH, RVJ, RVG, 0}; -struct isa_ext_data { - const char *name; - int min_version; - int ext_enable_offset; -}; - -#define ISA_EXT_DATA_ENTRY(_name, _min_ver, _prop) \ - {#_name, _min_ver, CPU_CFG_OFFSET(_prop)} - /* * From vector_helper.c * Note that vector data is stored in host-endian 64-bit chunks, @@ -XXX,XX +XXX,XX @@ struct isa_ext_data { #define BYTE(x) (x) #endif +#define ISA_EXT_DATA_ENTRY(_name, _min_ver, _prop) \ + {#_name, _min_ver, CPU_CFG_OFFSET(_prop)} + /* * Here are the ordering rules of extension naming defined by RISC-V * specification : @@ -XXX,XX +XXX,XX @@ struct isa_ext_data { * Single letter extensions are checked in riscv_cpu_validate_misa_priv() * instead. */ -static const struct isa_ext_data isa_edata_arr[] = { +const RISCVIsaExtData isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(zicbom, PRIV_VERSION_1_12_0, ext_icbom), ISA_EXT_DATA_ENTRY(zicboz, PRIV_VERSION_1_12_0, ext_icboz), ISA_EXT_DATA_ENTRY(zicond, PRIV_VERSION_1_12_0, ext_zicond), @@ -XXX,XX +XXX,XX @@ static const struct isa_ext_data isa_edata_arr[] = { ISA_EXT_DATA_ENTRY(xtheadmempair, PRIV_VERSION_1_11_0, ext_xtheadmempair), ISA_EXT_DATA_ENTRY(xtheadsync, PRIV_VERSION_1_11_0, ext_xtheadsync), ISA_EXT_DATA_ENTRY(xventanacondops, PRIV_VERSION_1_12_0, ext_XVentanaCondOps), + + DEFINE_PROP_END_OF_LIST(), }; bool isa_ext_is_enabled(RISCVCPU *cpu, uint32_t ext_offset) @@ -XXX,XX +XXX,XX @@ void isa_ext_update_enabled(RISCVCPU *cpu, uint32_t ext_offset, bool en) int cpu_cfg_ext_get_min_version(uint32_t ext_offset) { - int i; + const RISCVIsaExtData *edata; - for (i = 0; i < ARRAY_SIZE(isa_edata_arr); i++) { - if (isa_edata_arr[i].ext_enable_offset != ext_offset) { + for (edata = isa_edata_arr; edata && edata->name; edata++) { + if (edata->ext_enable_offset != ext_offset) { continue; } - return isa_edata_arr[i].min_version; + return edata->min_version; } g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_disas_set_info(CPUState *s, disassemble_info *info) void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu) { CPURISCVState *env = &cpu->env; - int i; + const RISCVIsaExtData *edata; /* Force disable extensions if priv spec version does not match */ - for (i = 0; i < ARRAY_SIZE(isa_edata_arr); i++) { - if (isa_ext_is_enabled(cpu, isa_edata_arr[i].ext_enable_offset) && - (env->priv_ver < isa_edata_arr[i].min_version)) { - isa_ext_update_enabled(cpu, isa_edata_arr[i].ext_enable_offset, - false); + for (edata = isa_edata_arr; edata && edata->name; edata++) { + if (isa_ext_is_enabled(cpu, edata->ext_enable_offset) && + (env->priv_ver < edata->min_version)) { + isa_ext_update_enabled(cpu, edata->ext_enable_offset, false); #ifndef CONFIG_USER_ONLY warn_report("disabling %s extension for hart 0x" TARGET_FMT_lx " because privilege spec version does not match", - isa_edata_arr[i].name, env->mhartid); + edata->name, env->mhartid); #else warn_report("disabling %s extension because " "privilege spec version does not match", - isa_edata_arr[i].name); + edata->name); #endif } } @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_class_init(ObjectClass *c, void *data) static void riscv_isa_string_ext(RISCVCPU *cpu, char **isa_str, int max_str_len) { + const RISCVIsaExtData *edata; char *old = *isa_str; char *new = *isa_str; - int i; - for (i = 0; i < ARRAY_SIZE(isa_edata_arr); i++) { - if (isa_ext_is_enabled(cpu, isa_edata_arr[i].ext_enable_offset)) { - new = g_strconcat(old, "_", isa_edata_arr[i].name, NULL); + for (edata = isa_edata_arr; edata && edata->name; edata++) { + if (isa_ext_is_enabled(cpu, edata->ext_enable_offset)) { + new = g_strconcat(old, "_", edata->name, NULL); g_free(old); old = new; } -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Priv spec validation is TCG specific. Move it to the TCG accel class. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-ID: <20230925175709.35696-20-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 2 -- target/riscv/cpu.c | 38 -------------------------------------- target/riscv/tcg/tcg-cpu.c | 38 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 38 insertions(+), 40 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 @@ enum riscv_pmu_event_idx { /* used by tcg/tcg-cpu.c*/ void isa_ext_update_enabled(RISCVCPU *cpu, uint32_t ext_offset, bool en); bool isa_ext_is_enabled(RISCVCPU *cpu, uint32_t ext_offset); -int cpu_cfg_ext_get_min_version(uint32_t ext_offset); void riscv_cpu_set_misa(CPURISCVState *env, RISCVMXL mxl, uint32_t ext); -void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu); typedef struct RISCVCPUMultiExtConfig { 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 @@ void isa_ext_update_enabled(RISCVCPU *cpu, uint32_t ext_offset, bool en) *ext_enabled = en; } -int cpu_cfg_ext_get_min_version(uint32_t ext_offset) -{ - const RISCVIsaExtData *edata; - - for (edata = isa_edata_arr; edata && edata->name; edata++) { - if (edata->ext_enable_offset != ext_offset) { - continue; - } - - return edata->min_version; - } - - g_assert_not_reached(); -} - const char * const riscv_int_regnames[] = { "x0/zero", "x1/ra", "x2/sp", "x3/gp", "x4/tp", "x5/t0", "x6/t1", "x7/t2", "x8/s0", "x9/s1", "x10/a0", "x11/a1", "x12/a2", "x13/a3", @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_disas_set_info(CPUState *s, disassemble_info *info) } } -void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu) -{ - CPURISCVState *env = &cpu->env; - const RISCVIsaExtData *edata; - - /* Force disable extensions if priv spec version does not match */ - for (edata = isa_edata_arr; edata && edata->name; edata++) { - if (isa_ext_is_enabled(cpu, edata->ext_enable_offset) && - (env->priv_ver < edata->min_version)) { - isa_ext_update_enabled(cpu, edata->ext_enable_offset, false); -#ifndef CONFIG_USER_ONLY - warn_report("disabling %s extension for hart 0x" TARGET_FMT_lx - " because privilege spec version does not match", - edata->name, env->mhartid); -#else - warn_report("disabling %s extension because " - "privilege spec version does not match", - edata->name); -#endif - } - } -} - #ifndef CONFIG_USER_ONLY static void riscv_cpu_satp_mode_finalize(RISCVCPU *cpu, 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 const struct TCGCPUOps riscv_tcg_ops = { #endif /* !CONFIG_USER_ONLY */ }; +static int cpu_cfg_ext_get_min_version(uint32_t ext_offset) +{ + const RISCVIsaExtData *edata; + + for (edata = isa_edata_arr; edata && edata->name; edata++) { + if (edata->ext_enable_offset != ext_offset) { + continue; + } + + return edata->min_version; + } + + g_assert_not_reached(); +} + static void cpu_cfg_ext_auto_update(RISCVCPU *cpu, uint32_t ext_offset, bool value) { @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_validate_v(CPURISCVState *env, RISCVCPUConfig *cfg, } } +static void riscv_cpu_disable_priv_spec_isa_exts(RISCVCPU *cpu) +{ + CPURISCVState *env = &cpu->env; + const RISCVIsaExtData *edata; + + /* Force disable extensions if priv spec version does not match */ + for (edata = isa_edata_arr; edata && edata->name; edata++) { + if (isa_ext_is_enabled(cpu, edata->ext_enable_offset) && + (env->priv_ver < edata->min_version)) { + isa_ext_update_enabled(cpu, edata->ext_enable_offset, false); +#ifndef CONFIG_USER_ONLY + warn_report("disabling %s extension for hart 0x" TARGET_FMT_lx + " because privilege spec version does not match", + edata->name, env->mhartid); +#else + warn_report("disabling %s extension because " + "privilege spec version does not match", + edata->name); +#endif + } + } +} + /* * Check consistency between chosen extensions while setting * cpu->cfg accordingly. -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> We'll introduce generic errors that will output a CPU type name via its RISCVCPU pointer. Create a helper for that. Use the helper in tcg_cpu_realizefn() instead of hardcoding the 'host' CPU name. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230926183109.165878-2-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu.h | 1 + target/riscv/cpu.c | 11 +++++++++++ target/riscv/tcg/tcg-cpu.c | 4 +++- 3 files changed, 15 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 @@ typedef struct isa_ext_data { int ext_enable_offset; } RISCVIsaExtData; extern const RISCVIsaExtData isa_edata_arr[]; +char *riscv_cpu_get_name(RISCVCPU *cpu); void riscv_add_satp_mode_properties(Object *obj); 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 ObjectClass *riscv_cpu_class_by_name(const char *cpu_model) return oc; } +char *riscv_cpu_get_name(RISCVCPU *cpu) +{ + RISCVCPUClass *rcc = RISCV_CPU_GET_CLASS(cpu); + const char *typename = object_class_get_name(OBJECT_CLASS(rcc)); + + g_assert(g_str_has_suffix(typename, RISCV_CPU_TYPE_SUFFIX)); + + return g_strndup(typename, + strlen(typename) - strlen(RISCV_CPU_TYPE_SUFFIX)); +} + static void riscv_cpu_dump_state(CPUState *cs, FILE *f, int flags) { RISCVCPU *cpu = RISCV_CPU(cs); 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 tcg_cpu_realize(CPUState *cs, Error **errp) Error *local_err = NULL; if (object_dynamic_cast(OBJECT(cpu), TYPE_RISCV_CPU_HOST)) { - error_setg(errp, "'host' CPU is not compatible with TCG acceleration"); + g_autofree char *name = riscv_cpu_get_name(cpu); + error_setg(errp, "'%s' CPU is not compatible with TCG acceleration", + name); return false; } -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> At this moment we do not expose extension properties for vendor CPUs because that would allow users to change them via command line. The drawback is that if we were to add an API that shows all CPU properties, e.g. qmp-query-cpu-model-expansion, we won't be able to show extensions state of vendor CPUs. We have the required machinery to create extension properties for vendor CPUs while not allowing users to enable extensions. Disabling existing extensions is allowed since it can be useful for debugging. Change the set() callback cpu_set_multi_ext_cfg() to allow enabling extensions only for generic CPUs. In cpu_add_multi_ext_prop() let's not set the default values for the properties if we're not dealing with generic CPUs, otherwise the values set in cpu_init() of vendor CPUs will be overwritten. And finally, in tcg_cpu_instance_init(), add cpu user properties for all CPUs. For the veyron-v1 CPU, we're now able to disable existing extensions like smstateen: $ ./build/qemu-system-riscv64 --nographic -M virt \ -cpu veyron-v1,smstateen=false But setting extensions that the CPU didn't set during cpu_init(), like V, is not allowed: $ ./build/qemu-system-riscv64 --nographic -M virt \ -cpu veyron-v1,v=true qemu-system-riscv64: can't apply global veyron-v1-riscv-cpu.v=true: 'veyron-v1' CPU does not allow enabling extensions Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20230926183109.165878-3-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/tcg/tcg-cpu.c | 64 +++++++++++++++++++++++++++++--------- 1 file changed, 50 insertions(+), 14 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 @@ void riscv_cpu_validate_set_extensions(RISCVCPU *cpu, Error **errp) riscv_cpu_disable_priv_spec_isa_exts(cpu); } +static bool riscv_cpu_is_generic(Object *cpu_obj) +{ + return object_dynamic_cast(cpu_obj, TYPE_RISCV_DYNAMIC_CPU) != NULL; +} + /* * We'll get here via the following path: * @@ -XXX,XX +XXX,XX @@ static void cpu_set_misa_ext_cfg(Object *obj, Visitor *v, const char *name, target_ulong misa_bit = misa_ext_cfg->misa_bit; RISCVCPU *cpu = RISCV_CPU(obj); CPURISCVState *env = &cpu->env; - bool value; + bool generic_cpu = riscv_cpu_is_generic(obj); + bool prev_val, value; if (!visit_type_bool(v, name, &value, errp)) { return; } + prev_val = env->misa_ext & misa_bit; + + if (value == prev_val) { + return; + } + if (value) { + if (!generic_cpu) { + g_autofree char *cpuname = riscv_cpu_get_name(cpu); + error_setg(errp, "'%s' CPU does not allow enabling extensions", + cpuname); + return; + } + env->misa_ext |= misa_bit; env->misa_ext_mask |= misa_bit; } else { @@ -XXX,XX +XXX,XX @@ static const RISCVCPUMisaExtConfig misa_ext_cfgs[] = { */ static void riscv_cpu_add_misa_properties(Object *cpu_obj) { + bool use_def_vals = riscv_cpu_is_generic(cpu_obj); int i; for (i = 0; i < ARRAY_SIZE(misa_ext_cfgs); i++) { @@ -XXX,XX +XXX,XX @@ static void riscv_cpu_add_misa_properties(Object *cpu_obj) cpu_set_misa_ext_cfg, NULL, (void *)misa_cfg); object_property_set_description(cpu_obj, name, desc); - object_property_set_bool(cpu_obj, name, misa_cfg->enabled, NULL); + if (use_def_vals) { + object_property_set_bool(cpu_obj, name, misa_cfg->enabled, NULL); + } } } @@ -XXX,XX +XXX,XX @@ static void cpu_set_multi_ext_cfg(Object *obj, Visitor *v, const char *name, void *opaque, Error **errp) { const RISCVCPUMultiExtConfig *multi_ext_cfg = opaque; - bool value; + RISCVCPU *cpu = RISCV_CPU(obj); + bool generic_cpu = riscv_cpu_is_generic(obj); + bool prev_val, value; if (!visit_type_bool(v, name, &value, errp)) { return; } - isa_ext_update_enabled(RISCV_CPU(obj), multi_ext_cfg->offset, value); - g_hash_table_insert(multi_ext_user_opts, GUINT_TO_POINTER(multi_ext_cfg->offset), (gpointer)value); + + prev_val = isa_ext_is_enabled(cpu, multi_ext_cfg->offset); + + if (value == prev_val) { + return; + } + + if (value && !generic_cpu) { + g_autofree char *cpuname = riscv_cpu_get_name(cpu); + error_setg(errp, "'%s' CPU does not allow enabling extensions", + cpuname); + return; + } + + isa_ext_update_enabled(cpu, multi_ext_cfg->offset, value); } static void cpu_get_multi_ext_cfg(Object *obj, Visitor *v, const char *name, @@ -XXX,XX +XXX,XX @@ static void cpu_get_multi_ext_cfg(Object *obj, Visitor *v, const char *name, static void cpu_add_multi_ext_prop(Object *cpu_obj, const RISCVCPUMultiExtConfig *multi_cfg) { + bool generic_cpu = riscv_cpu_is_generic(cpu_obj); + 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) { + return; + } + /* * Set def val directly instead of using * object_property_set_bool() to save the set() @@ -XXX,XX +XXX,XX @@ static bool riscv_cpu_has_max_extensions(Object *cpu_obj) return object_dynamic_cast(cpu_obj, TYPE_RISCV_CPU_MAX) != NULL; } -static bool riscv_cpu_has_user_properties(Object *cpu_obj) -{ - return object_dynamic_cast(cpu_obj, TYPE_RISCV_DYNAMIC_CPU) != NULL; -} - static void tcg_cpu_instance_init(CPUState *cs) { RISCVCPU *cpu = RISCV_CPU(cs); Object *obj = OBJECT(cpu); - if (riscv_cpu_has_user_properties(obj)) { - multi_ext_user_opts = g_hash_table_new(NULL, g_direct_equal); - riscv_cpu_add_user_properties(obj); - } + multi_ext_user_opts = g_hash_table_new(NULL, g_direct_equal); + riscv_cpu_add_user_properties(obj); if (riscv_cpu_has_max_extensions(obj)) { riscv_init_max_cpu_extensions(obj); -- 2.41.0 From: Clément Chigot <chigot@adacore.com> As of now, the exit code was either EXIT_FAILURE when a panic shutdown was requested or EXIT_SUCCESS otherwise. However, some hardware could want to pass more complex exit codes. Thus, introduce a new shutdown request function allowing that. Signed-off-by: Clément Chigot <chigot@adacore.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20231003071427.188697-2-chigot@adacore.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- include/sysemu/runstate.h | 2 ++ system/runstate.c | 12 +++++++++++- 2 files changed, 13 insertions(+), 1 deletion(-) diff --git a/include/sysemu/runstate.h b/include/sysemu/runstate.h index XXXXXXX..XXXXXXX 100644 --- a/include/sysemu/runstate.h +++ b/include/sysemu/runstate.h @@ -XXX,XX +XXX,XX @@ void qemu_system_wakeup_request(WakeupReason reason, Error **errp); void qemu_system_wakeup_enable(WakeupReason reason, bool enabled); void qemu_register_wakeup_notifier(Notifier *notifier); void qemu_register_wakeup_support(void); +void qemu_system_shutdown_request_with_code(ShutdownCause reason, + int exit_code); void qemu_system_shutdown_request(ShutdownCause reason); void qemu_system_powerdown_request(void); void qemu_register_powerdown_notifier(Notifier *notifier); diff --git a/system/runstate.c b/system/runstate.c index XXXXXXX..XXXXXXX 100644 --- a/system/runstate.c +++ b/system/runstate.c @@ -XXX,XX +XXX,XX @@ void vm_state_notify(bool running, RunState state) static ShutdownCause reset_requested; static ShutdownCause shutdown_requested; +static int shutdown_exit_code = EXIT_SUCCESS; static int shutdown_signal; static pid_t shutdown_pid; static int powerdown_requested; @@ -XXX,XX +XXX,XX @@ void qemu_system_killed(int signal, pid_t pid) qemu_notify_event(); } +void qemu_system_shutdown_request_with_code(ShutdownCause reason, + int exit_code) +{ + shutdown_exit_code = exit_code; + qemu_system_shutdown_request(reason); +} + void qemu_system_shutdown_request(ShutdownCause reason) { trace_qemu_system_shutdown_request(reason); @@ -XXX,XX +XXX,XX @@ static bool main_loop_should_exit(int *status) if (shutdown_action == SHUTDOWN_ACTION_PAUSE) { vm_stop(RUN_STATE_SHUTDOWN); } else { - if (request == SHUTDOWN_CAUSE_GUEST_PANIC && + if (shutdown_exit_code != EXIT_SUCCESS) { + *status = shutdown_exit_code; + } else if (request == SHUTDOWN_CAUSE_GUEST_PANIC && panic_action == PANIC_ACTION_EXIT_FAILURE) { *status = EXIT_FAILURE; } -- 2.41.0 From: Clément Chigot <chigot@adacore.com> gdb_exit function aims to close gdb sessions and sends the exit code of the current execution. It's being called by qemu_cleanup once the main loop is over. Until now, the exit code sent was always 0. Now that hardware can shutdown this main loop with custom exit codes, these codes must be transfered to gdb as well. Signed-off-by: Clément Chigot <chigot@adacore.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20231003071427.188697-3-chigot@adacore.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- include/sysemu/sysemu.h | 2 +- system/main.c | 2 +- system/runstate.c | 4 ++-- 3 files changed, 4 insertions(+), 4 deletions(-) diff --git a/include/sysemu/sysemu.h b/include/sysemu/sysemu.h index XXXXXXX..XXXXXXX 100644 --- a/include/sysemu/sysemu.h +++ b/include/sysemu/sysemu.h @@ -XXX,XX +XXX,XX @@ bool defaults_enabled(void); void qemu_init(int argc, char **argv); int qemu_main_loop(void); -void qemu_cleanup(void); +void qemu_cleanup(int); extern QemuOptsList qemu_legacy_drive_opts; extern QemuOptsList qemu_common_drive_opts; diff --git a/system/main.c b/system/main.c index XXXXXXX..XXXXXXX 100644 --- a/system/main.c +++ b/system/main.c @@ -XXX,XX +XXX,XX @@ int qemu_default_main(void) int status; status = qemu_main_loop(); - qemu_cleanup(); + qemu_cleanup(status); return status; } diff --git a/system/runstate.c b/system/runstate.c index XXXXXXX..XXXXXXX 100644 --- a/system/runstate.c +++ b/system/runstate.c @@ -XXX,XX +XXX,XX @@ void qemu_init_subsystems(void) } -void qemu_cleanup(void) +void qemu_cleanup(int status) { - gdb_exit(0); + gdb_exit(status); /* * cleaning up the migration object cancels any existing migration -- 2.41.0 From: Clément Chigot <chigot@adacore.com> This replaces the exit calls by shutdown requests, ensuring a proper cleanup of Qemu. Otherwise, some connections like gdb could be broken before its final packet ("Wxx") is being sent. This part, being done inside qemu_cleanup function, can be reached only when the main loop exits after a shutdown request. Signed-off-by: Clément Chigot <chigot@adacore.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20231003071427.188697-4-chigot@adacore.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/misc/sifive_test.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/hw/misc/sifive_test.c b/hw/misc/sifive_test.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/sifive_test.c +++ b/hw/misc/sifive_test.c @@ -XXX,XX +XXX,XX @@ #include "qemu/module.h" #include "sysemu/runstate.h" #include "hw/misc/sifive_test.h" +#include "sysemu/sysemu.h" static uint64_t sifive_test_read(void *opaque, hwaddr addr, unsigned int size) { @@ -XXX,XX +XXX,XX @@ static void sifive_test_write(void *opaque, hwaddr addr, int code = (val64 >> 16) & 0xffff; switch (status) { case FINISHER_FAIL: - exit(code); + qemu_system_shutdown_request_with_code( + SHUTDOWN_CAUSE_GUEST_PANIC, code); + return; case FINISHER_PASS: - exit(0); + qemu_system_shutdown_request_with_code( + SHUTDOWN_CAUSE_GUEST_SHUTDOWN, code); + return; case FINISHER_RESET: qemu_system_reset_request(SHUTDOWN_CAUSE_GUEST_RESET); return; -- 2.41.0 From: Clément Chigot <chigot@adacore.com> This replaces the exit calls by shutdown requests, ensuring a proper cleanup of Qemu. Otherwise, some connections like gdb could be broken before its final packet ("Wxx") is being sent. This part, being done inside qemu_cleanup function, can be reached only when the main loop exits after a shutdown request. Signed-off-by: Clément Chigot <chigot@adacore.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20231003071427.188697-5-chigot@adacore.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- hw/char/riscv_htif.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/hw/char/riscv_htif.c b/hw/char/riscv_htif.c index XXXXXXX..XXXXXXX 100644 --- a/hw/char/riscv_htif.c +++ b/hw/char/riscv_htif.c @@ -XXX,XX +XXX,XX @@ #include "exec/address-spaces.h" #include "exec/tswap.h" #include "sysemu/dma.h" +#include "sysemu/runstate.h" #define RISCV_DEBUG_HTIF 0 #define HTIF_DEBUG(fmt, ...) \ @@ -XXX,XX +XXX,XX @@ static void htif_handle_tohost_write(HTIFState *s, uint64_t val_written) g_free(sig_data); } - exit(exit_code); + qemu_system_shutdown_request_with_code( + SHUTDOWN_CAUSE_GUEST_SHUTDOWN, exit_code); + return; } else { uint64_t syscall[8]; cpu_physical_memory_read(payload, syscall, sizeof(syscall)); -- 2.41.0 From: Clément Chigot <chigot@adacore.com> This replaces the exit calls by shutdown requests, ensuring a proper cleanup of Qemu. Features like net/vhost-vdpa.c are expecting qemu_cleanup to be called to remove their last residuals. Signed-off-by: Clément Chigot <chigot@adacore.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-ID: <20231003071427.188697-6-chigot@adacore.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- include/gdbstub/syscalls.h | 9 +++++++++ gdbstub/gdbstub.c | 5 +++-- gdbstub/system.c | 6 ++++++ gdbstub/user.c | 6 ++++++ 4 files changed, 24 insertions(+), 2 deletions(-) diff --git a/include/gdbstub/syscalls.h b/include/gdbstub/syscalls.h index XXXXXXX..XXXXXXX 100644 --- a/include/gdbstub/syscalls.h +++ b/include/gdbstub/syscalls.h @@ -XXX,XX +XXX,XX @@ int use_gdb_syscalls(void); */ void gdb_exit(int code); +/** + * gdb_qemu_exit: ask qemu to exit + * @code: exit code reported + * + * This requests qemu to exit. This function is allowed to return as + * the exit request might be processed asynchronously by qemu backend. + */ +void gdb_qemu_exit(int code); + #endif /* _SYSCALLS_H_ */ diff --git a/gdbstub/gdbstub.c b/gdbstub/gdbstub.c index XXXXXXX..XXXXXXX 100644 --- a/gdbstub/gdbstub.c +++ b/gdbstub/gdbstub.c @@ -XXX,XX +XXX,XX @@ static void handle_v_kill(GArray *params, void *user_ctx) gdb_put_packet("OK"); error_report("QEMU: Terminated via GDBstub"); gdb_exit(0); - exit(0); + gdb_qemu_exit(0); } static const GdbCmdParseEntry gdb_v_commands_table[] = { @@ -XXX,XX +XXX,XX @@ static int gdb_handle_packet(const char *line_buf) /* Kill the target */ error_report("QEMU: Terminated via GDBstub"); gdb_exit(0); - exit(0); + gdb_qemu_exit(0); + break; case 'D': { static const GdbCmdParseEntry detach_cmd_desc = { diff --git a/gdbstub/system.c b/gdbstub/system.c index XXXXXXX..XXXXXXX 100644 --- a/gdbstub/system.c +++ b/gdbstub/system.c @@ -XXX,XX +XXX,XX @@ void gdb_exit(int code) qemu_chr_fe_deinit(&gdbserver_system_state.chr, true); } +void gdb_qemu_exit(int code) +{ + qemu_system_shutdown_request_with_code(SHUTDOWN_CAUSE_GUEST_SHUTDOWN, + code); +} + /* * Memory access */ diff --git a/gdbstub/user.c b/gdbstub/user.c index XXXXXXX..XXXXXXX 100644 --- a/gdbstub/user.c +++ b/gdbstub/user.c @@ -XXX,XX +XXX,XX @@ void gdb_exit(int code) gdb_put_packet(buf); gdbserver_state.allow_stop_reply = false; } + +} + +void gdb_qemu_exit(int code) +{ + exit(code); } int gdb_handlesig(CPUState *cpu, int sig) -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Our error message is returning the value of 'ret', which will be always -1 in case of error, and will not be that useful: qemu-system-riscv64: Unable to read ISA_EXT KVM register ssaia, error -1 Improve the error message by outputting 'errno' instead of 'ret'. Use strerrorname_np() to output the error name instead of the error code. This will give us what we need to know right away: qemu-system-riscv64: Unable to read ISA_EXT KVM register ssaia, error code: ENOENT Given that we're going to exit(1) in this condition instead of attempting to recover, remove the 'kvm_riscv_destroy_scratch_vcpu()' call. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-ID: <20231003132148.797921-2-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/kvm/kvm-cpu.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) 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_init_multiext_cfg(RISCVCPU *cpu, KVMScratchCPU *kvmcpu) val = false; } else { error_report("Unable to read ISA_EXT KVM register %s, " - "error %d", multi_ext_cfg->name, ret); - kvm_riscv_destroy_scratch_vcpu(kvmcpu); + "error code: %s", multi_ext_cfg->name, + strerrorname_np(errno)); exit(EXIT_FAILURE); } } else { -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> KVM for RISC-V started supporting KVM_GET_REG_LIST in Linux 6.6. It consists of a KVM ioctl() that retrieves a list of all available regs for get_one_reg/set_one_reg. Regs that aren't present in the list aren't supported in the host. This simplifies our lives when initing the KVM regs since we don't have to always attempt a KVM_GET_ONE_REG for all regs QEMU knows. We'll only attempt a get_one_reg() if we're sure the reg is supported, i.e. it was retrieved by KVM_GET_REG_LIST. Any error in get_one_reg() will then always considered fatal, instead of having to handle special error codes that might indicate a non-fatal failure. Start by moving the current kvm_riscv_init_multiext_cfg() logic into a new kvm_riscv_read_multiext_legacy() helper. We'll prioritize using KVM_GET_REG_LIST, so check if we have it available and, in case we don't, use the legacy() logic. Otherwise, retrieve the available reg list and use it to check if the host supports our known KVM regs, doing the usual get_one_reg() for the supported regs and setting cpu->cfg accordingly. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20231003132148.797921-3-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/kvm/kvm-cpu.c | 96 +++++++++++++++++++++++++++++++++++++- 1 file changed, 95 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_read_cbomz_blksize(RISCVCPU *cpu, KVMScratchCPU *kvmcpu, } } -static void kvm_riscv_init_multiext_cfg(RISCVCPU *cpu, KVMScratchCPU *kvmcpu) +static void kvm_riscv_read_multiext_legacy(RISCVCPU *cpu, + KVMScratchCPU *kvmcpu) { CPURISCVState *env = &cpu->env; uint64_t val; @@ -XXX,XX +XXX,XX @@ static void kvm_riscv_init_multiext_cfg(RISCVCPU *cpu, KVMScratchCPU *kvmcpu) } } +static int uint64_cmp(const void *a, const void *b) +{ + uint64_t val1 = *(const uint64_t *)a; + uint64_t val2 = *(const uint64_t *)b; + + if (val1 < val2) { + return -1; + } + + if (val1 > val2) { + return 1; + } + + return 0; +} + +static void kvm_riscv_init_multiext_cfg(RISCVCPU *cpu, KVMScratchCPU *kvmcpu) +{ + KVMCPUConfig *multi_ext_cfg; + struct kvm_one_reg reg; + struct kvm_reg_list rl_struct; + struct kvm_reg_list *reglist; + uint64_t val, reg_id, *reg_search; + int i, ret; + + rl_struct.n = 0; + ret = ioctl(kvmcpu->cpufd, KVM_GET_REG_LIST, &rl_struct); + + /* + * If KVM_GET_REG_LIST isn't supported we'll get errno 22 + * (EINVAL). Use read_legacy() in this case. + */ + if (errno == EINVAL) { + return kvm_riscv_read_multiext_legacy(cpu, kvmcpu); + } else if (errno != E2BIG) { + /* + * E2BIG is an expected error message for the API since we + * don't know the number of registers. The right amount will + * be written in rl_struct.n. + * + * Error out if we get any other errno. + */ + error_report("Error when accessing get-reg-list, code: %s", + strerrorname_np(errno)); + exit(EXIT_FAILURE); + } + + reglist = g_malloc(sizeof(struct kvm_reg_list) + + rl_struct.n * sizeof(uint64_t)); + reglist->n = rl_struct.n; + ret = ioctl(kvmcpu->cpufd, KVM_GET_REG_LIST, reglist); + if (ret) { + error_report("Error when reading KVM_GET_REG_LIST, code %s ", + strerrorname_np(errno)); + exit(EXIT_FAILURE); + } + + /* sort reglist to use bsearch() */ + qsort(&reglist->reg, reglist->n, sizeof(uint64_t), uint64_cmp); + + for (i = 0; i < ARRAY_SIZE(kvm_multi_ext_cfgs); i++) { + multi_ext_cfg = &kvm_multi_ext_cfgs[i]; + reg_id = kvm_riscv_reg_id(&cpu->env, KVM_REG_RISCV_ISA_EXT, + multi_ext_cfg->kvm_reg_id); + reg_search = bsearch(&reg_id, reglist->reg, reglist->n, + sizeof(uint64_t), uint64_cmp); + if (!reg_search) { + continue; + } + + reg.id = reg_id; + reg.addr = (uint64_t)&val; + ret = ioctl(kvmcpu->cpufd, KVM_GET_ONE_REG, &reg); + if (ret != 0) { + error_report("Unable to read ISA_EXT KVM register %s, " + "error code: %s", multi_ext_cfg->name, + strerrorname_np(errno)); + exit(EXIT_FAILURE); + } + + multi_ext_cfg->supported = true; + kvm_cpu_cfg_set(cpu, multi_ext_cfg, val); + } + + if (cpu->cfg.ext_icbom) { + kvm_riscv_read_cbomz_blksize(cpu, kvmcpu, &kvm_cbom_blocksize); + } + + if (cpu->cfg.ext_icboz) { + kvm_riscv_read_cbomz_blksize(cpu, kvmcpu, &kvm_cboz_blocksize); + } +} + static void riscv_init_kvm_registers(Object *cpu_obj) { RISCVCPU *cpu = RISCV_CPU(cpu_obj); -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Vendor CPUs that set RVG are displaying user warnings about other extensions that RVG must enable, one warning per CPU. E.g.: $ ./build/qemu-system-riscv64 -smp 8 -M virt -cpu veyron-v1 -nographic qemu-system-riscv64: warning: Setting G will also set IMAFD_Zicsr_Zifencei qemu-system-riscv64: warning: Setting G will also set IMAFD_Zicsr_Zifencei qemu-system-riscv64: warning: Setting G will also set IMAFD_Zicsr_Zifencei qemu-system-riscv64: warning: Setting G will also set IMAFD_Zicsr_Zifencei qemu-system-riscv64: warning: Setting G will also set IMAFD_Zicsr_Zifencei qemu-system-riscv64: warning: Setting G will also set IMAFD_Zicsr_Zifencei qemu-system-riscv64: warning: Setting G will also set IMAFD_Zicsr_Zifencei qemu-system-riscv64: warning: Setting G will also set IMAFD_Zicsr_Zifencei This happens because we decided a while ago that, for simplicity, vendor CPUs could set RVG instead of setting each G extension individually in their cpu_init(). Our warning isn't taking that into account, and we're bugging users with a warning that we're causing ourselves. In a closer look we conclude that this warning is not warranted in any other circumstance since we're just following the ISA [1], which states in chapter 24: "One goal of the RISC-V project is that it be used as a stable software development target. For this purpose, we define a combination of a base ISA (RV32I or RV64I) plus selected standard extensions (IMAFD, Zicsr, Zifencei) as a 'general-purpose' ISA, and we use the abbreviation G for the IMAFDZicsr Zifencei combination of instruction-set extensions." With this in mind, enabling IMAFD_Zicsr_Zifencei if the user explicitly enables 'G' is an expected behavior and the warning is unneeded. Any user caught by surprise should refer to the ISA. Remove the warning when handling RVG. [1] https://github.com/riscv/riscv-isa-manual/releases/download/Ratified-IMAFDQC/riscv-spec-20191213.pdf Reported-by: Paul A. Clarke <pclarke@ventanamicro.com> Suggested-by: Andrew Jones <ajones@ventanamicro.com> Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20231003122539.775932-1-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/tcg/tcg-cpu.c | 1 - 1 file changed, 1 deletion(-) 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) return; } - warn_report("Setting G will also set IMAFD_Zicsr_Zifencei"); cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_icsr), true); cpu_cfg_ext_auto_update(cpu, CPU_CFG_OFFSET(ext_ifencei), true); -- 2.41.0 From: "Richard W.M. Jones" <rjones@redhat.com> RISCV_CPU(cs) uses a checked cast. When QOM cast debugging is enabled this adds about 5% total overhead when emulating RV64 on x86-64 host. Using a RISC-V guest with 16 vCPUs, 16 GB of guest RAM, virtio-blk disk. The guest has a copy of the qemu source tree. The test involves compiling the qemu source tree with 'make clean; time make -j16'. Before making this change the compile step took 449 & 447 seconds over two consecutive runs. After making this change: 428 & 421 seconds. The saving is over 5%. Thanks: Paolo Bonzini Thanks: Philippe Mathieu-Daudé Signed-off-by: Richard W.M. Jones <rjones@redhat.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-ID: <20231009124859.3373696-2-rjones@redhat.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/cpu_helper.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) 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 @@ int riscv_cpu_mmu_index(CPURISCVState *env, bool ifetch) void cpu_get_tb_cpu_state(CPURISCVState *env, vaddr *pc, uint64_t *cs_base, uint32_t *pflags) { - CPUState *cs = env_cpu(env); - RISCVCPU *cpu = RISCV_CPU(cs); + RISCVCPU *cpu = env_archcpu(env); RISCVExtStatus fs, vs; uint32_t flags = 0; -- 2.41.0 From: Daniel Henrique Barboza <dbarboza@ventanamicro.com> At this moment there are eleven CPU extension properties that starts with capital 'Z': Zifencei, Zicsr, Zihintntl, Zihintpause, Zawrs, Zfa, Zfh, Zfhmin, Zve32f, Zve64f and Zve64d. All other extensions are named with lower-case letters. We want all properties to be named with lower-case letters since it's consistent with the riscv-isa string that we create in the FDT. Having these 11 properties to be exceptions can be confusing. Deprecate all of them. Create their lower-case counterpart to be used as maintained CPU properties. When trying to use any deprecated property a warning message will be displayed, recommending users to switch to the lower-case variant: ./build/qemu-system-riscv64 -M virt -cpu rv64,Zifencei=true --nographic qemu-system-riscv64: warning: CPU property 'Zifencei' is deprecated. Please use 'zifencei' instead This will give users some time to change their scripts before we remove the capital 'Z' properties entirely. Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Andrew Jones <ajones@ventanamicro.com> Message-ID: <20231009112817.8896-2-dbarboza@ventanamicro.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- docs/about/deprecated.rst | 23 ++++++++++++++++++++++ target/riscv/cpu.h | 1 + target/riscv/cpu.c | 39 +++++++++++++++++++++++++++----------- target/riscv/tcg/tcg-cpu.c | 31 +++++++++++++++++++++++++++++- 4 files changed, 82 insertions(+), 12 deletions(-) diff --git a/docs/about/deprecated.rst b/docs/about/deprecated.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/about/deprecated.rst +++ b/docs/about/deprecated.rst @@ -XXX,XX +XXX,XX @@ of generic CPUs: rv32 and rv64 as default CPUs and 'max' as a feature complete CPU for both 32 and 64 bit builds. Users are then discouraged to use the 'any' CPU type starting in 8.2. +RISC-V CPU properties which start with capital 'Z' (since 8.2) +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +All RISC-V CPU properties which start with capital 'Z' are being deprecated +starting in 8.2. The reason is that they were wrongly added with capital 'Z' +in the past. CPU properties were later added with lower-case names, which +is the format we want to use from now on. + +Users which try to use these deprecated properties will receive a warning +recommending to switch to their stable counterparts: + +- "Zifencei" should be replaced with "zifencei" +- "Zicsr" should be replaced with "zicsr" +- "Zihintntl" should be replaced with "zihintntl" +- "Zihintpause" should be replaced with "zihintpause" +- "Zawrs" should be replaced with "zawrs" +- "Zfa" should be replaced with "zfa" +- "Zfh" should be replaced with "zfh" +- "Zfhmin" should be replaced with "zfhmin" +- "Zve32f" should be replaced with "zve32f" +- "Zve64f" should be replaced with "zve64f" +- "Zve64d" should be replaced with "zve64d" + Block device options '''''''''''''''''''' 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 RISCVCPUMultiExtConfig { 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_deprecated_exts[]; extern Property riscv_cpu_options[]; typedef struct isa_ext_data { diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu.c +++ b/target/riscv/cpu.c @@ -XXX,XX +XXX,XX @@ const char *riscv_get_misa_ext_description(uint32_t bit) const RISCVCPUMultiExtConfig riscv_cpu_extensions[] = { /* Defaults for standard extensions */ MULTI_EXT_CFG_BOOL("sscofpmf", ext_sscofpmf, false), - MULTI_EXT_CFG_BOOL("Zifencei", ext_ifencei, true), - MULTI_EXT_CFG_BOOL("Zicsr", ext_icsr, 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), + MULTI_EXT_CFG_BOOL("zifencei", ext_ifencei, true), + MULTI_EXT_CFG_BOOL("zicsr", ext_icsr, 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), MULTI_EXT_CFG_BOOL("sstc", ext_sstc, true), MULTI_EXT_CFG_BOOL("smstateen", ext_smstateen, false), @@ -XXX,XX +XXX,XX @@ const RISCVCPUMultiExtConfig riscv_cpu_experimental_exts[] = { DEFINE_PROP_END_OF_LIST(), }; +/* Deprecated entries marked for future removal */ +const RISCVCPUMultiExtConfig riscv_cpu_deprecated_exts[] = { + MULTI_EXT_CFG_BOOL("Zifencei", ext_ifencei, true), + MULTI_EXT_CFG_BOOL("Zicsr", ext_icsr, 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), + + DEFINE_PROP_END_OF_LIST(), +}; + Property riscv_cpu_options[] = { DEFINE_PROP_UINT8("pmu-num", RISCVCPU, cfg.pmu_num, 16), 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_misa_properties(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); + } + g_hash_table_insert(multi_ext_user_opts, GUINT_TO_POINTER(multi_ext_cfg->offset), (gpointer)value); @@ -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) { + if (!generic_cpu || deprecated_ext) { 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); + for (Property *prop = riscv_cpu_options; prop && prop->name; prop++) { qdev_property_add_static(DEVICE(obj), prop); } -- 2.41.0 From: Max Chou <max.chou@sifive.com> The operator (fwmacc16) of vfwmaccbf16.vf helper function should be replaced by fwmaccbf16. Fixes: adf772b0f7 ("target/riscv: Add support for Zvfbfwma extension") Signed-off-by: Max Chou <max.chou@sifive.com> Reviewed-by: LIU Zhiwei <zhiwei_liu@linux.alibaba.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-ID: <20231005095734.567575-1-max.chou@sifive.com> Signed-off-by: Alistair Francis <alistair.francis@wdc.com> --- target/riscv/vector_helper.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) 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 @@ static uint32_t fwmaccbf16(uint16_t a, uint16_t b, uint32_t d, float_status *s) RVVCALL(OPFVV3, vfwmaccbf16_vv, WOP_UUU_H, H4, H2, H2, fwmaccbf16) GEN_VEXT_VV_ENV(vfwmaccbf16_vv, 4) -RVVCALL(OPFVF3, vfwmaccbf16_vf, WOP_UUU_H, H4, H2, fwmacc16) +RVVCALL(OPFVF3, vfwmaccbf16_vf, WOP_UUU_H, H4, H2, fwmaccbf16) GEN_VEXT_VF(vfwmaccbf16_vf, 4) static uint32_t fwnmacc16(uint16_t a, uint16_t b, uint32_t d, float_status *s) -- 2.41.0 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