arm64/fp: Generate definitons for floating point system registers
Subsystems (3)
cover arm64/fp: Generate definitons for floating point system registers
This series builds on top of the recently applied series for system register generation, converting the floating point registers to automatic generation. There's nothing remarkable in here, just a bunch of straightforward cleanups and the addition of support for RAZ fields followed by the actual conversions.
It does skip over the SVCRSM, SVCRZA and SVCRSMZA aliases for now since I need to think a little about what to do with those.
Since the series includes conversion of the SME registers it is based on for-next/core rather than for-next/sysreg-gen to pick up for-next/sme as well.
Mark Brown (12):
arm64/fp: Make SVE and SME length register definition match
architecture
arm64/fp: Rename SVE and SME LEN field name to _WIDTH arm64/sme: Drop SYS_ from SMIDR_EL1 defines arm64/sme: Standardise bitfield names for SVCR arm64/sme: Remove _EL0 from name of SVCR - FIXME sysreg.h arm64/sysreg: Support generation of RAZ fields arm64/sme: Automatically generate defines for SMCR arm64/sme: Automatically generate SMIDR_EL1 defines arm64/sme: Automatically generate SMPRIMAP_EL2 definitions arm64/sme: Generate SMPRI_EL1 definitions arm64/sme: Generate defintions for SVCR arm64/sve: Generate ZCR definitions
arch/arm64/include/asm/el2_setup.h | 2 +- arch/arm64/include/asm/fpsimd.h | 4 +- arch/arm64/include/asm/processor.h | 2 +- arch/arm64/include/asm/sysreg.h | 42 ++-------------- arch/arm64/kernel/cpufeature.c | 4 +- arch/arm64/kernel/fpsimd.c | 26 +++++----- arch/arm64/kernel/ptrace.c | 8 ++-- arch/arm64/kernel/signal.c | 14 +++--- arch/arm64/kernel/syscall.c | 4 +- arch/arm64/kvm/fpsimd.c | 4 +- arch/arm64/kvm/sys_regs.c | 2 +- arch/arm64/tools/gen-sysreg.awk | 7 +++ arch/arm64/tools/sysreg | 77 ++++++++++++++++++++++++++++++ 13 files changed, 122 insertions(+), 74 deletions(-)
base-commit: 01ce2f5af527043a591208fe26788f55101b90bd
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH 1/12 arm64/fp: Make SVE and SME length register definition match architecture
Currently (as of DDI0487H.a) the architecture defines the vector length control field in ZCR and SMCR as being 4 bits wide with an additional 5 bits reserved above it marked as RAZ/WI for future expansion. The kernel currently attempts to anticipate such expansion by treating these extra bits as part of the LEN field but this will be inconvenient when we start generating the defines and would cause problems in the event that the architecture goes a different direction with these fields. Let's instead change the defines to reflect the currently defined architecture, we can update in future as needed.
No change in behaviour should be seen in any system, even emulated systems using the maximum allowed vector length for the current architecture.
--- arch/arm64/include/asm/sysreg.h | 18 ++++-------------- 1 file changed, 4 insertions(+), 14 deletions(-)
arch/arm64/include/asm/sysreg.h +4 -14
@@ -1113,26 +1113,16 @@#define DCZID_DZP_SHIFT 4#define DCZID_BS_SHIFT 0 -/* -* The ZCR_ELx_LEN_* definitions intentionally include bits [8:4] which -* are reserved by the SVE architecture for future expansion of the LEN -* field, with compatible semantics. -*/ #define ZCR_ELx_LEN_SHIFT 0 -#define ZCR_ELx_LEN_SIZE 9 -#define ZCR_ELx_LEN_MASK 0x1ff +#define ZCR_ELx_LEN_SIZE 4 +#define ZCR_ELx_LEN_MASK 0xf#define SMCR_ELx_FA64_SHIFT 31#define SMCR_ELx_FA64_MASK (1 << SMCR_ELx_FA64_SHIFT) -/* -* The SMCR_ELx_LEN_* definitions intentionally include bits [8:4] which -* are reserved by the SME architecture for future expansion of the LEN -* field, with compatible semantics. -*/ #define SMCR_ELx_LEN_SHIFT 0 -#define SMCR_ELx_LEN_SIZE 9 -#define SMCR_ELx_LEN_MASK 0x1ff +#define SMCR_ELx_LEN_SIZE 4 +#define SMCR_ELx_LEN_MASK 0xf#define CPACR_EL1_FPEN_EL1EN (BIT(20))/* enable EL1 access */ #define CPACR_EL1_FPEN_EL0EN (BIT(21))/* enable EL0 access, if EL1EN set */
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH 2/12 arm64/fp: Rename SVE and SME LEN field name to _WIDTH
In reply to Mark BrownThe SVE and SVE length configuration field LEN have constants specifying their width called _SIZE rather than the more normal _WIDTH, in preparation for automatic generation rename to _WIDTH. No functional change.
--- arch/arm64/include/asm/sysreg.h | 4 ++-- arch/arm64/kernel/cpufeature.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-)
arch/arm64/include/asm/sysreg.h +2 -2
@@ -1114,14 +1114,14 @@#define DCZID_BS_SHIFT 0#define ZCR_ELx_LEN_SHIFT 0 -#define ZCR_ELx_LEN_SIZE 4 +#define ZCR_ELx_LEN_WIDTH 4#define ZCR_ELx_LEN_MASK 0xf#define SMCR_ELx_FA64_SHIFT 31#define SMCR_ELx_FA64_MASK (1 << SMCR_ELx_FA64_SHIFT)#define SMCR_ELx_LEN_SHIFT 0 -#define SMCR_ELx_LEN_SIZE 4 +#define SMCR_ELx_LEN_WIDTH 4#define SMCR_ELx_LEN_MASK 0xf#define CPACR_EL1_FPEN_EL1EN (BIT(20))/* enable EL1 access */
arch/arm64/kernel/cpufeature.c +2 -2
@@ -577,13 +577,13 @@ static const struct arm64_ftr_bits ftr_id_dfr1[] = {static const struct arm64_ftr_bits ftr_zcr[] = {ARM64_FTR_BITS (FTR_HIDDEN ,FTR_NONSTRICT ,FTR_LOWER_SAFE , - ZCR_ELx_LEN_SHIFT, ZCR_ELx_LEN_SIZE,0 ),/* LEN */ + ZCR_ELx_LEN_SHIFT, ZCR_ELx_LEN_WIDTH,0 ),/* LEN */ ARM64_FTR_END , };static const struct arm64_ftr_bits ftr_smcr[] = {ARM64_FTR_BITS (FTR_HIDDEN ,FTR_NONSTRICT ,FTR_LOWER_SAFE , - SMCR_ELx_LEN_SHIFT, SMCR_ELx_LEN_SIZE,0 ),/* LEN */ + SMCR_ELx_LEN_SHIFT, SMCR_ELx_LEN_WIDTH,0 ),/* LEN */ ARM64_FTR_END , };
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH 3/12 arm64/sme: Drop SYS_ from SMIDR_EL1 defines
In reply to Mark BrownWe currently have a non-standard SYS_ prefix in the constants generated for SMIDR_EL1 bitfields. Drop this in preparation for automatic register definition generation, no functional change.
--- arch/arm64/include/asm/el2_setup.h | 2 +- arch/arm64/include/asm/sysreg.h | 6 +++--- 2 files changed, 4 insertions(+), 4 deletions(-)
arch/arm64/include/asm/el2_setup.h +1 -1
@@ -171,7 +171,7 @@msr_s SYS_SMCR_EL2 , x1// length for EL1. mrs_s x1, SYS_SMIDR_EL1 // Priority mapping supported? - ubfx x1, x1, #SYS_SMIDR_EL1_SMPS_SHIFT, #1 + ubfx x1, x1, #SMIDR_EL1_SMPS_SHIFT, #1 cbz x1, .Lskip_sme_\@ msr_s SYS_SMPRIMAP_EL2, xzr // Make all priorities equal
arch/arm64/include/asm/sysreg.h +3 -3
@@ -467,9 +467,9 @@#define SYS_SMIDR_EL1 sys_reg(3, 1, 0, 0, 6)#define SYS_AIDR_EL1 sys_reg(3, 1, 0, 0, 7) -#define SYS_SMIDR_EL1_IMPLEMENTER_SHIFT 24 -#define SYS_SMIDR_EL1_SMPS_SHIFT 15 -#define SYS_SMIDR_EL1_AFFINITY_SHIFT 0 +#define SMIDR_EL1_IMPLEMENTER_SHIFT 24 +#define SMIDR_EL1_SMPS_SHIFT 15 +#define SMIDR_EL1_AFFINITY_SHIFT 0#define SYS_CSSELR_EL1 sys_reg(3, 2, 0, 0, 0)
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH 4/12 arm64/sme: Standardise bitfield names for SVCR
In reply to Mark BrownThe bitfield definitions for SVCR have a SYS_ added to the names of the constant which will be a problem for automatic generation. Remove the prefixes, no functional change.
--- arch/arm64/include/asm/fpsimd.h | 4 ++-- arch/arm64/include/asm/processor.h | 2 +- arch/arm64/include/asm/sysreg.h | 4 ++-- arch/arm64/kernel/fpsimd.c | 6 +++--- 4 files changed, 8 insertions(+), 8 deletions(-)
arch/arm64/include/asm/fpsimd.h +2 -2
@@ -67,12 +67,12 @@ extern void fpsimd_save_and_flush_cpu_state(void);static inline bool thread_sm_enabled (struct thread_struct *thread) { -return system_supports_sme () && (thread->svcr &SYS_SVCR_EL0_SM_MASK ); +return system_supports_sme () && (thread->svcr &SVCR_EL0_SM_MASK ); }static inline bool thread_za_enabled (struct thread_struct *thread) { -return system_supports_sme () && (thread->svcr &SYS_SVCR_EL0_ZA_MASK ); +return system_supports_sme () && (thread->svcr &SVCR_EL0_ZA_MASK ); }/* Maximum VL that SVE/SME VL-agnostic software can transparently support */
arch/arm64/include/asm/processor.h +1 -1
@@ -192,7 +192,7 @@ static inline unsigned int thread_get_sme_vl(struct thread_struct *thread)static inline unsigned int thread_get_cur_vl (struct thread_struct *thread) { -if (system_supports_sme () && (thread->svcr &SYS_SVCR_EL0_SM_MASK )) +if (system_supports_sme () && (thread->svcr &SVCR_EL0_SM_MASK ))return thread_get_sme_vl (thread);else return thread_get_sve_vl (thread);
arch/arm64/include/asm/sysreg.h +2 -2
@@ -480,8 +480,8 @@#define SYS_RNDRRS_EL0 sys_reg(3, 3, 2, 4, 1)#define SYS_SVCR_EL0 sys_reg(3, 3, 4, 2, 2) -#define SYS_SVCR_EL0_ZA_MASK 2 -#define SYS_SVCR_EL0_SM_MASK 1 +#define SVCR_EL0_ZA_MASK 2 +#define SVCR_EL0_SM_MASK 1#define SYS_PMCR_EL0 sys_reg(3, 3, 9, 12, 0)#define SYS_PMCNTENSET_EL0 sys_reg(3, 3, 9, 12, 1)
arch/arm64/kernel/fpsimd.c +3 -3
@@ -1893,7 +1893,7 @@ void __efi_fpsimd_begin(void) svcr =read_sysreg_s (SYS_SVCR_EL0 );if (!system_supports_fa64 ()) - ffr = svcr &SYS_SVCR_EL0_SM_MASK ; + ffr = svcr &SVCR_EL0_SM_MASK ;__this_cpu_write (efi_sm_state, ffr); }
@@ -1904,7 +1904,7 @@ void __efi_fpsimd_begin(void)if (system_supports_sme ())sysreg_clear_set_s (SYS_SVCR_EL0 , -SYS_SVCR_EL0_SM_MASK ,0 ); +SVCR_EL0_SM_MASK ,0 ); }else {fpsimd_save_state (this_cpu_ptr (&efi_fpsimd_state));
@@ -1939,7 +1939,7 @@ void __efi_fpsimd_end(void)if (__this_cpu_read (efi_sm_state)) {sysreg_clear_set_s (SYS_SVCR_EL0 ,0 , -SYS_SVCR_EL0_SM_MASK ); +SVCR_EL0_SM_MASK );if (!system_supports_fa64 ()) ffr = efi_sm_state; }
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH 5/12 arm64/sme: Remove _EL0 from name of SVCR - FIXME sysreg.h
In reply to Mark BrownThe defines for SVCR call it SVCR_EL0 however the architecture calls the register SVCR with no _EL0 suffix. In preparation for generating the sysreg definitions rename to match the architecture, no functional change.
--- arch/arm64/include/asm/fpsimd.h | 4 ++-- arch/arm64/include/asm/processor.h | 2 +- arch/arm64/include/asm/sysreg.h | 6 +++--- arch/arm64/kernel/fpsimd.c | 26 +++++++++++++------------- arch/arm64/kernel/ptrace.c | 8 ++++---- arch/arm64/kernel/signal.c | 14 +++++++------- arch/arm64/kernel/syscall.c | 4 ++-- arch/arm64/kvm/fpsimd.c | 4 ++-- arch/arm64/kvm/sys_regs.c | 2 +- 9 files changed, 35 insertions(+), 35 deletions(-)
arch/arm64/include/asm/fpsimd.h +2 -2
@@ -67,12 +67,12 @@ extern void fpsimd_save_and_flush_cpu_state(void);static inline bool thread_sm_enabled (struct thread_struct *thread) { -return system_supports_sme () && (thread->svcr &SVCR_EL0_SM_MASK ); +return system_supports_sme () && (thread->svcr &SVCR_SM_MASK ); }static inline bool thread_za_enabled (struct thread_struct *thread) { -return system_supports_sme () && (thread->svcr &SVCR_EL0_ZA_MASK ); +return system_supports_sme () && (thread->svcr &SVCR_ZA_MASK ); }/* Maximum VL that SVE/SME VL-agnostic software can transparently support */
arch/arm64/include/asm/processor.h +1 -1
@@ -192,7 +192,7 @@ static inline unsigned int thread_get_sme_vl(struct thread_struct *thread)static inline unsigned int thread_get_cur_vl (struct thread_struct *thread) { -if (system_supports_sme () && (thread->svcr &SVCR_EL0_SM_MASK )) +if (system_supports_sme () && (thread->svcr &SVCR_SM_MASK ))return thread_get_sme_vl (thread);else return thread_get_sve_vl (thread);
arch/arm64/include/asm/sysreg.h +3 -3
@@ -479,9 +479,9 @@#define SYS_RNDR_EL0 sys_reg(3, 3, 2, 4, 0)#define SYS_RNDRRS_EL0 sys_reg(3, 3, 2, 4, 1) -#define SYS_SVCR_EL0 sys_reg(3, 3, 4, 2, 2) -#define SVCR_EL0_ZA_MASK 2 -#define SVCR_EL0_SM_MASK 1 +#define SYS_SVCR sys_reg(3, 3, 4, 2, 2) +#define SVCR_ZA_MASK 2 +#define SVCR_SM_MASK 1#define SYS_PMCR_EL0 sys_reg(3, 3, 9, 12, 0)#define SYS_PMCNTENSET_EL0 sys_reg(3, 3, 9, 12, 1)
arch/arm64/kernel/fpsimd.c +13 -13
@@ -398,7 +398,7 @@ static void task_fpsimd_load(void)if (test_thread_flag (TIF_SME ))sme_set_vq (sve_vq_from_vl (sme_vl) -1 ); -write_sysreg_s (current->thread.svcr,SYS_SVCR_EL0 ); +write_sysreg_s (current->thread.svcr,SYS_SVCR );if (thread_za_enabled (¤t->thread))za_load_state (current->thread.za_state);
@@ -450,15 +450,15 @@ static void fpsimd_save(void)if (system_supports_sme ()) {u64 *svcr = last->svcr; - *svcr =read_sysreg_s (SYS_SVCR_EL0 ); + *svcr =read_sysreg_s (SYS_SVCR ); - *svcr =read_sysreg_s (SYS_SVCR_EL0 ); + *svcr =read_sysreg_s (SYS_SVCR ); -if (*svcr &SYS_SVCR_EL0_ZA_MASK ) +if (*svcr &SVCR_ZA_MASK )za_save_state (last->za_state);/* If we are in streaming mode override regular SVE. */ -if (*svcr &SYS_SVCR_EL0_SM_MASK ) { +if (*svcr &SVCR_SM_MASK ) { save_sve_regs = true; save_ffr =system_supports_fa64 (); vl = last->sme_vl;
@@ -840,8 +840,8 @@ int vec_set_vector_length(struct task_struct *task, enum vec_type type,sve_to_fpsimd (task);if (system_supports_sme () && type ==ARM64_VEC_SME ) { - task->thread.svcr &= ~(SYS_SVCR_EL0_SM_MASK | -SYS_SVCR_EL0_ZA_MASK ); + task->thread.svcr &= ~(SVCR_SM_MASK | +SVCR_ZA_MASK );clear_thread_flag (TIF_SME ); }
@@ -1890,10 +1890,10 @@ void __efi_fpsimd_begin(void)__this_cpu_write (efi_sve_state_used, true);if (system_supports_sme ()) { - svcr =read_sysreg_s (SYS_SVCR_EL0 ); + svcr =read_sysreg_s (SYS_SVCR );if (!system_supports_fa64 ()) - ffr = svcr &SVCR_EL0_SM_MASK ; + ffr = svcr &SVCR_SM_MASK ;__this_cpu_write (efi_sm_state, ffr); }
@@ -1903,8 +1903,8 @@ void __efi_fpsimd_begin(void) ffr);if (system_supports_sme ()) -sysreg_clear_set_s (SYS_SVCR_EL0 , -SVCR_EL0_SM_MASK ,0 ); +sysreg_clear_set_s (SYS_SVCR , +SVCR_SM_MASK ,0 ); }else {fpsimd_save_state (this_cpu_ptr (&efi_fpsimd_state));
@@ -1937,9 +1937,9 @@ void __efi_fpsimd_end(void)*/ if (system_supports_sme ()) {if (__this_cpu_read (efi_sm_state)) { -sysreg_clear_set_s (SYS_SVCR_EL0 , +sysreg_clear_set_s (SYS_SVCR ,0 , -SVCR_EL0_SM_MASK ); +SVCR_SM_MASK );if (!system_supports_fa64 ()) ffr = efi_sm_state; }
arch/arm64/kernel/ptrace.c +4 -4
@@ -867,10 +867,10 @@ static int sve_set_common(struct task_struct *target,switch (type) {case ARM64_VEC_SVE : - target->thread.svcr &= ~SYS_SVCR_EL0_SM_MASK ; + target->thread.svcr &= ~SVCR_SM_MASK ;break ;case ARM64_VEC_SME : - target->thread.svcr |=SYS_SVCR_EL0_SM_MASK ; + target->thread.svcr |=SVCR_SM_MASK ;break ;default :WARN_ON_ONCE (1 );
@@ -1100,7 +1100,7 @@ static int za_set(struct task_struct *target,/* If there is no data then disable ZA */ if (!count) { - target->thread.svcr &= ~SYS_SVCR_EL0_ZA_MASK ; + target->thread.svcr &= ~SVCR_ZA_MASK ; goto out; }
@@ -1125,7 +1125,7 @@ static int za_set(struct task_struct *target,/* Mark ZA as active and let userspace use it */ set_tsk_thread_flag (target,TIF_SME ); - target->thread.svcr |=SYS_SVCR_EL0_ZA_MASK ; + target->thread.svcr |=SVCR_ZA_MASK ; out:fpsimd_flush_task_state (target);
arch/arm64/kernel/signal.c +7 -7
@@ -288,7 +288,7 @@ static int restore_sve_fpsimd_context(struct user_ctxs *user)if (sve.head.size <=sizeof (*user->sve)) {clear_thread_flag (TIF_SVE ); - current->thread.svcr &= ~SYS_SVCR_EL0_SM_MASK ; + current->thread.svcr &= ~SVCR_SM_MASK ; goto fpsimd_only; }
@@ -321,7 +321,7 @@ static int restore_sve_fpsimd_context(struct user_ctxs *user)return -EFAULT ;if (sve.flags &SVE_SIG_FLAG_SM ) - current->thread.svcr |=SYS_SVCR_EL0_SM_MASK ; + current->thread.svcr |=SVCR_SM_MASK ;else set_thread_flag (TIF_SVE );
@@ -398,7 +398,7 @@ static int restore_za_context(struct user_ctxs __user *user)return -EINVAL ;if (za.head.size <=sizeof (*user->za)) { - current->thread.svcr &= ~SYS_SVCR_EL0_ZA_MASK ; + current->thread.svcr &= ~SVCR_ZA_MASK ;return 0 ; }
@@ -419,7 +419,7 @@ static int restore_za_context(struct user_ctxs __user *user)sme_alloc (current);if (!current->thread.za_state) { - current->thread.svcr &= ~SYS_SVCR_EL0_ZA_MASK ; + current->thread.svcr &= ~SVCR_ZA_MASK ;clear_thread_flag (TIF_SME );return -ENOMEM ; }
@@ -432,7 +432,7 @@ static int restore_za_context(struct user_ctxs __user *user)return -EFAULT ;set_thread_flag (TIF_SME ); - current->thread.svcr |=SYS_SVCR_EL0_ZA_MASK ; + current->thread.svcr |=SVCR_ZA_MASK ;return 0 ; }
@@ -922,8 +922,8 @@ static void setup_return(struct pt_regs *regs, struct k_sigaction *ka,/* Signal handlers are invoked with ZA and streaming mode disabled */ if (system_supports_sme ()) { - current->thread.svcr &= ~(SYS_SVCR_EL0_ZA_MASK | -SYS_SVCR_EL0_SM_MASK ); + current->thread.svcr &= ~(SVCR_ZA_MASK | +SVCR_SM_MASK );sme_smstop (); }
arch/arm64/kernel/syscall.c +2 -2
@@ -174,9 +174,9 @@ static inline void fp_user_discard(void)* need updating. */ if (system_supports_sme () &&test_thread_flag (TIF_SME )) { -u64 svcr =read_sysreg_s (SYS_SVCR_EL0 ); +u64 svcr =read_sysreg_s (SYS_SVCR ); -if (svcr &SYS_SVCR_EL0_SM_MASK ) +if (svcr &SVCR_SM_MASK )sme_smstop_sm (); }
arch/arm64/kvm/fpsimd.c +2 -2
@@ -96,8 +96,8 @@ void kvm_arch_vcpu_load_fp(struct kvm_vcpu *vcpu)if (read_sysreg (cpacr_el1) &CPACR_EL1_SMEN_EL0EN ) vcpu->arch.flags |=KVM_ARM64_HOST_SME_ENABLED ; -if (read_sysreg_s (SYS_SVCR_EL0 ) & - (SYS_SVCR_EL0_SM_MASK |SYS_SVCR_EL0_ZA_MASK )) { +if (read_sysreg_s (SYS_SVCR ) & + (SVCR_SM_MASK |SVCR_ZA_MASK )) { vcpu->arch.flags &= ~KVM_ARM64_FP_HOST ;fpsimd_save_and_flush_cpu_state (); }
arch/arm64/kvm/sys_regs.c +1 -1
@@ -1685,7 +1685,7 @@ static const struct sys_reg_desc sys_reg_descs[] = { { SYS_DESC(SYS_SMIDR_EL1), undef_access }, { SYS_DESC(SYS_CSSELR_EL1), access_csselr, reset_unknown, CSSELR_EL1 }, { SYS_DESC(SYS_CTR_EL0), access_ctr }, - { SYS_DESC(SYS_SVCR_EL0), undef_access }, + { SYS_DESC(SYS_SVCR), undef_access }, { PMU_SYS_REG(SYS_PMCR_EL0), .access = access_pmcr, .reset = reset_pmcr, .reg = PMCR_EL0 },
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Mark Rutland
reviewerRe: [PATCH v1 05/12] arm64/sme: Remove _EL0 from name of SVCR - FIXME sysreg.h
Quoted 5 lines · Mark Brown
The defines for SVCR call it
SVCR_EL0however the architecture calls the register SVCR with no_EL0suffix. In preparation for generating the sysreg definitions rename to match the architecture, no functional change.Signed-off-by Mark Brown <[email protected]>
What's the "FIXME sysreg.h" in the title for? Is that an accidental leftover, or is there something to do there?
Other than that this looks fine to me.
Thanks,
Mark.
Quoted 283 lines
--- arch/arm64/include/asm/fpsimd.h | 4 ++-- arch/arm64/include/asm/processor.h | 2 +- arch/arm64/include/asm/sysreg.h | 6 +++--- arch/arm64/kernel/fpsimd.c | 26 +++++++++++++------------- arch/arm64/kernel/ptrace.c | 8 ++++---- arch/arm64/kernel/signal.c | 14 +++++++------- arch/arm64/kernel/syscall.c | 4 ++-- arch/arm64/kvm/fpsimd.c | 4 ++-- arch/arm64/kvm/sys_regs.c | 2 +- 9 files changed, 35 insertions(+), 35 deletions(-)arch/arm64/include/asm/fpsimd.h +2
-2@@ -67,12 +67,12 @@ extern void fpsimd_save_and_flush_cpu_state(void);static inline bool thread_sm_enabled (struct thread_struct *thread) { -return system_supports_sme () && (thread->svcr &SVCR_EL0_SM_MASK ); +return system_supports_sme () && (thread->svcr &SVCR_SM_MASK ); }static inline bool thread_za_enabled (struct thread_struct *thread) { -return system_supports_sme () && (thread->svcr &SVCR_EL0_ZA_MASK ); +return system_supports_sme () && (thread->svcr &SVCR_ZA_MASK ); }/* Maximum VL that SVE/SME VL-agnostic software can transparently support */ arch/arm64/include/asm/processor.h +1
-1@@ -192,7 +192,7 @@ static inline unsigned int thread_get_sme_vl(struct thread_struct *thread)static inline unsigned int thread_get_cur_vl (struct thread_struct *thread) { -if (system_supports_sme () && (thread->svcr &SVCR_EL0_SM_MASK )) +if (system_supports_sme () && (thread->svcr &SVCR_SM_MASK ))return thread_get_sme_vl (thread);else return thread_get_sve_vl (thread);arch/arm64/include/asm/sysreg.h +3
-3@@ -479,9 +479,9 @@#define SYS_RNDR_EL0 sys_reg(3, 3, 2, 4, 0)#define SYS_RNDRRS_EL0 sys_reg(3, 3, 2, 4, 1) -#define SYS_SVCR_EL0 sys_reg(3, 3, 4, 2, 2) -#define SVCR_EL0_ZA_MASK 2 -#define SVCR_EL0_SM_MASK 1 +#define SYS_SVCR sys_reg(3, 3, 4, 2, 2) +#define SVCR_ZA_MASK 2 +#define SVCR_SM_MASK 1#define SYS_PMCR_EL0 sys_reg(3, 3, 9, 12, 0)#define SYS_PMCNTENSET_EL0 sys_reg(3, 3, 9, 12, 1)arch/arm64/kernel/fpsimd.c +13
-13@@ -398,7 +398,7 @@ static void task_fpsimd_load(void)if (test_thread_flag (TIF_SME ))sme_set_vq (sve_vq_from_vl (sme_vl) -1 ); -write_sysreg_s (current->thread.svcr,SYS_SVCR_EL0 ); +write_sysreg_s (current->thread.svcr,SYS_SVCR );if (thread_za_enabled (¤t->thread))za_load_state (current->thread.za_state);@@ -450,15 +450,15 @@ static void fpsimd_save(void)if (system_supports_sme ()) {u64 *svcr = last->svcr; - *svcr =read_sysreg_s (SYS_SVCR_EL0 ); + *svcr =read_sysreg_s (SYS_SVCR ); - *svcr =read_sysreg_s (SYS_SVCR_EL0 ); + *svcr =read_sysreg_s (SYS_SVCR ); -if (*svcr &SYS_SVCR_EL0_ZA_MASK ) +if (*svcr &SVCR_ZA_MASK )za_save_state (last->za_state);/* If we are in streaming mode override regular SVE. */ -if (*svcr &SYS_SVCR_EL0_SM_MASK ) { +if (*svcr &SVCR_SM_MASK ) { save_sve_regs = true; save_ffr =system_supports_fa64 (); vl = last->sme_vl;@@ -840,8 +840,8 @@ int vec_set_vector_length(struct task_struct *task, enum vec_type type,sve_to_fpsimd (task);if (system_supports_sme () && type ==ARM64_VEC_SME ) { - task->thread.svcr &= ~(SYS_SVCR_EL0_SM_MASK | -SYS_SVCR_EL0_ZA_MASK ); + task->thread.svcr &= ~(SVCR_SM_MASK | +SVCR_ZA_MASK );clear_thread_flag (TIF_SME ); }@@ -1890,10 +1890,10 @@ void __efi_fpsimd_begin(void)__this_cpu_write (efi_sve_state_used, true);if (system_supports_sme ()) { - svcr =read_sysreg_s (SYS_SVCR_EL0 ); + svcr =read_sysreg_s (SYS_SVCR );if (!system_supports_fa64 ()) - ffr = svcr &SVCR_EL0_SM_MASK ; + ffr = svcr &SVCR_SM_MASK ;__this_cpu_write (efi_sm_state, ffr); }@@ -1903,8 +1903,8 @@ void __efi_fpsimd_begin(void) ffr);if (system_supports_sme ()) -sysreg_clear_set_s (SYS_SVCR_EL0 , -SVCR_EL0_SM_MASK ,0 ); +sysreg_clear_set_s (SYS_SVCR , +SVCR_SM_MASK ,0 ); }else {fpsimd_save_state (this_cpu_ptr (&efi_fpsimd_state));@@ -1937,9 +1937,9 @@ void __efi_fpsimd_end(void)*/ if (system_supports_sme ()) {if (__this_cpu_read (efi_sm_state)) { -sysreg_clear_set_s (SYS_SVCR_EL0 , +sysreg_clear_set_s (SYS_SVCR ,0 , -SVCR_EL0_SM_MASK ); +SVCR_SM_MASK );if (!system_supports_fa64 ()) ffr = efi_sm_state; }arch/arm64/kernel/ptrace.c +4
-4@@ -867,10 +867,10 @@ static int sve_set_common(struct task_struct *target,switch (type) {case ARM64_VEC_SVE : - target->thread.svcr &= ~SYS_SVCR_EL0_SM_MASK ; + target->thread.svcr &= ~SVCR_SM_MASK ;break ;case ARM64_VEC_SME : - target->thread.svcr |=SYS_SVCR_EL0_SM_MASK ; + target->thread.svcr |=SVCR_SM_MASK ;break ;default :WARN_ON_ONCE (1 );@@ -1100,7 +1100,7 @@ static int za_set(struct task_struct *target,/* If there is no data then disable ZA */ if (!count) { - target->thread.svcr &= ~SYS_SVCR_EL0_ZA_MASK ; + target->thread.svcr &= ~SVCR_ZA_MASK ; goto out; }@@ -1125,7 +1125,7 @@ static int za_set(struct task_struct *target,/* Mark ZA as active and let userspace use it */ set_tsk_thread_flag (target,TIF_SME ); - target->thread.svcr |=SYS_SVCR_EL0_ZA_MASK ; + target->thread.svcr |=SVCR_ZA_MASK ; out:fpsimd_flush_task_state (target);arch/arm64/kernel/signal.c +7
-7@@ -288,7 +288,7 @@ static int restore_sve_fpsimd_context(struct user_ctxs *user)if (sve.head.size <=sizeof (*user->sve)) {clear_thread_flag (TIF_SVE ); - current->thread.svcr &= ~SYS_SVCR_EL0_SM_MASK ; + current->thread.svcr &= ~SVCR_SM_MASK ; goto fpsimd_only; }@@ -321,7 +321,7 @@ static int restore_sve_fpsimd_context(struct user_ctxs *user)return -EFAULT ;if (sve.flags &SVE_SIG_FLAG_SM ) - current->thread.svcr |=SYS_SVCR_EL0_SM_MASK ; + current->thread.svcr |=SVCR_SM_MASK ;else set_thread_flag (TIF_SVE );@@ -398,7 +398,7 @@ static int restore_za_context(struct user_ctxs __user *user)return -EINVAL ;if (za.head.size <=sizeof (*user->za)) { - current->thread.svcr &= ~SYS_SVCR_EL0_ZA_MASK ; + current->thread.svcr &= ~SVCR_ZA_MASK ;return 0 ; }@@ -419,7 +419,7 @@ static int restore_za_context(struct user_ctxs __user *user)sme_alloc (current);if (!current->thread.za_state) { - current->thread.svcr &= ~SYS_SVCR_EL0_ZA_MASK ; + current->thread.svcr &= ~SVCR_ZA_MASK ;clear_thread_flag (TIF_SME );return -ENOMEM ; }@@ -432,7 +432,7 @@ static int restore_za_context(struct user_ctxs __user *user)return -EFAULT ;set_thread_flag (TIF_SME ); - current->thread.svcr |=SYS_SVCR_EL0_ZA_MASK ; + current->thread.svcr |=SVCR_ZA_MASK ;return 0 ; }@@ -922,8 +922,8 @@ static void setup_return(struct pt_regs *regs, struct k_sigaction *ka,/* Signal handlers are invoked with ZA and streaming mode disabled */ if (system_supports_sme ()) { - current->thread.svcr &= ~(SYS_SVCR_EL0_ZA_MASK | -SYS_SVCR_EL0_SM_MASK ); + current->thread.svcr &= ~(SVCR_ZA_MASK | +SVCR_SM_MASK );sme_smstop (); }arch/arm64/kernel/syscall.c +2
-2@@ -174,9 +174,9 @@ static inline void fp_user_discard(void)* need updating. */ if (system_supports_sme () &&test_thread_flag (TIF_SME )) { -u64 svcr =read_sysreg_s (SYS_SVCR_EL0 ); +u64 svcr =read_sysreg_s (SYS_SVCR ); -if (svcr &SYS_SVCR_EL0_SM_MASK ) +if (svcr &SVCR_SM_MASK )sme_smstop_sm (); }arch/arm64/kvm/fpsimd.c +2
-2@@ -96,8 +96,8 @@ void kvm_arch_vcpu_load_fp(struct kvm_vcpu *vcpu)if (read_sysreg (cpacr_el1) &CPACR_EL1_SMEN_EL0EN ) vcpu->arch.flags |=KVM_ARM64_HOST_SME_ENABLED ; -if (read_sysreg_s (SYS_SVCR_EL0 ) & - (SYS_SVCR_EL0_SM_MASK |SYS_SVCR_EL0_ZA_MASK )) { +if (read_sysreg_s (SYS_SVCR ) & + (SVCR_SM_MASK |SVCR_ZA_MASK )) { vcpu->arch.flags &= ~KVM_ARM64_FP_HOST ;fpsimd_save_and_flush_cpu_state (); }arch/arm64/kvm/sys_regs.c +1
-1@@ -1685,7 +1685,7 @@ static const struct sys_reg_desc sys_reg_descs[] = { { SYS_DESC(SYS_SMIDR_EL1), undef_access }, { SYS_DESC(SYS_CSSELR_EL1), access_csselr, reset_unknown, CSSELR_EL1 }, { SYS_DESC(SYS_CTR_EL0), access_ctr }, - { SYS_DESC(SYS_SVCR_EL0), undef_access }, + { SYS_DESC(SYS_SVCR), undef_access }, { PMU_SYS_REG(SYS_PMCR_EL0), .access = access_pmcr, .reset = reset_pmcr, .reg = PMCR_EL0 },Signature · 1 line
2.30.2
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH v1 05/12] arm64/sme: Remove _EL0 from name of SVCR - FIXME sysreg.h
Quoted 7 lines · Mark Rutland
Quoted 3 lines · Mark Brown
The defines for SVCR call it
SVCR_EL0however the architecture calls the register SVCR with no_EL0suffix. In preparation for generating the sysreg definitions rename to match the architecture, no functional change.What's the "FIXME
sysreg.h" in the title for? Is that an accidental leftover, or is there something to do there?
That's an accidental leftover, sorry.
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH 6/12 arm64/sysreg: Support generation of RAZ fields
In reply to Mark BrownAdd a statement for RAZ bitfields to the automatic register generation script. Nothing is emitted to the header for these fields.
--- arch/arm64/tools/gen-sysreg.awk | 7 +++++++ 1 file changed, 7 insertions(+)
arch/arm64/tools/gen-sysreg.awk +7 -0
@@ -226,6 +226,13 @@ END { next } +/^Raz/ && (block == "Sysreg" || block == "SysregFields") { + expect_fields(2) + parse_bitdef(reg, field, $2) + + next +} + /^Enum/ { change_block("Enum", "Sysreg", "Enum") expect_fields(3)
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH v1 06/12] arm64/sysreg: Support generation of RAZ fields
Quoted 4 lines · Mark Brown
Add a statement for RAZ bitfields to the automatic register generation script. Nothing is emitted to the header for these fields.
Signed-off-by Mark Brown <[email protected]>
Makes sense to me:
Mark.
Quoted 25 lines
--- arch/arm64/tools/gen-sysreg.awk | 7 +++++++ 1 file changed, 7 insertions(+)arch/arm64/tools/gen-sysreg.awk +7
-0@@ -226,6 +226,13 @@ END { next } +/^Raz/ && (block == "Sysreg" || block == "SysregFields") { + expect_fields(2) + parse_bitdef(reg, field, $2) + + next +} + /^Enum/ { change_block("Enum", "Sysreg", "Enum") expect_fields(3)Signature · 1 line
2.30.2
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH 7/12 arm64/sme: Automatically generate defines for SMCR
In reply to Mark BrownConvert SMCR to use the register definition code, no functional change.
--- arch/arm64/include/asm/sysreg.h | 10 ---------- arch/arm64/tools/sysreg | 20 ++++++++++++++++++++ 2 files changed, 20 insertions(+), 10 deletions(-)
arch/arm64/include/asm/sysreg.h +0 -10
@@ -216,7 +216,6 @@#define SYS_ZCR_EL1 sys_reg(3, 0, 1, 2, 0)#define SYS_TRFCR_EL1 sys_reg(3, 0, 1, 2, 1)#define SYS_SMPRI_EL1 sys_reg(3, 0, 1, 2, 4) -#define SYS_SMCR_EL1 sys_reg(3, 0, 1, 2, 6)#define SYS_TCR_EL1 sys_reg(3, 0, 2, 0, 2)
@@ -571,7 +570,6 @@#define SYS_TRFCR_EL2 sys_reg(3, 4, 1, 2, 1)#define SYS_HCRX_EL2 sys_reg(3, 4, 1, 2, 2)#define SYS_SMPRIMAP_EL2 sys_reg(3, 4, 1, 2, 5) -#define SYS_SMCR_EL2 sys_reg(3, 4, 1, 2, 6)#define SYS_DACR32_EL2 sys_reg(3, 4, 3, 0, 0)#define SYS_HDFGRTR_EL2 sys_reg(3, 4, 3, 1, 4)#define SYS_HDFGWTR_EL2 sys_reg(3, 4, 3, 1, 5)
@@ -631,7 +629,6 @@#define SYS_SCTLR_EL12 sys_reg(3, 5, 1, 0, 0)#define SYS_CPACR_EL12 sys_reg(3, 5, 1, 0, 2)#define SYS_ZCR_EL12 sys_reg(3, 5, 1, 2, 0) -#define SYS_SMCR_EL12 sys_reg(3, 5, 1, 2, 6)#define SYS_TTBR0_EL12 sys_reg(3, 5, 2, 0, 0)#define SYS_TTBR1_EL12 sys_reg(3, 5, 2, 0, 1)#define SYS_TCR_EL12 sys_reg(3, 5, 2, 0, 2)
@@ -1117,13 +1114,6 @@#define ZCR_ELx_LEN_WIDTH 4#define ZCR_ELx_LEN_MASK 0xf -#define SMCR_ELx_FA64_SHIFT 31 -#define SMCR_ELx_FA64_MASK (1 << SMCR_ELx_FA64_SHIFT) - -#define SMCR_ELx_LEN_SHIFT 0 -#define SMCR_ELx_LEN_WIDTH 4 -#define SMCR_ELx_LEN_MASK 0xf -#define CPACR_EL1_FPEN_EL1EN (BIT(20))/* enable EL1 access */ #define CPACR_EL1_FPEN_EL0EN (BIT(21))/* enable EL0 access, if EL1EN set */
arch/arm64/tools/sysreg +20 -0
@@ -185,6 +185,26 @@ Field 1 A Field 0 M EndSysreg +SysregFields SMCR_ELx +Res0 63:32 +Field 31 FA64 +Res0 30:9 +Raz 8:4 +Field 3:0 LEN +EndSysregFields + +Sysreg SMCR_EL1 3 0 1 2 6 +Fields SMCR_ELx +EndSysreg + +Sysreg SMCR_EL2 3 4 1 2 6 +Fields SMCR_ELx +EndSysreg + +Sysreg SMCR_EL12 3 5 1 2 6 +Fields SMCR_ELx +EndSysreg + SysregFields TTBRx_EL1 Field 63:48 ASID Field 47:1 BADDR
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH v1 07/12] arm64/sme: Automatically generate defines for SMCR
Quoted 67 lines · Mark Brown
Convert SMCR to use the register definition code, no functional change.
Signed-off-by Mark Brown <[email protected]>--- arch/arm64/include/asm/sysreg.h | 10 ---------- arch/arm64/tools/sysreg | 20 ++++++++++++++++++++ 2 files changed, 20 insertions(+), 10 deletions(-)arch/arm64/include/asm/sysreg.h +0
-10@@ -216,7 +216,6 @@#define SYS_ZCR_EL1 sys_reg(3, 0, 1, 2, 0)#define SYS_TRFCR_EL1 sys_reg(3, 0, 1, 2, 1)#define SYS_SMPRI_EL1 sys_reg(3, 0, 1, 2, 4) -#define SYS_SMCR_EL1 sys_reg(3, 0, 1, 2, 6)#define SYS_TCR_EL1 sys_reg(3, 0, 2, 0, 2)@@ -571,7 +570,6 @@#define SYS_TRFCR_EL2 sys_reg(3, 4, 1, 2, 1)#define SYS_HCRX_EL2 sys_reg(3, 4, 1, 2, 2)#define SYS_SMPRIMAP_EL2 sys_reg(3, 4, 1, 2, 5) -#define SYS_SMCR_EL2 sys_reg(3, 4, 1, 2, 6)#define SYS_DACR32_EL2 sys_reg(3, 4, 3, 0, 0)#define SYS_HDFGRTR_EL2 sys_reg(3, 4, 3, 1, 4)#define SYS_HDFGWTR_EL2 sys_reg(3, 4, 3, 1, 5)@@ -631,7 +629,6 @@#define SYS_SCTLR_EL12 sys_reg(3, 5, 1, 0, 0)#define SYS_CPACR_EL12 sys_reg(3, 5, 1, 0, 2)#define SYS_ZCR_EL12 sys_reg(3, 5, 1, 2, 0) -#define SYS_SMCR_EL12 sys_reg(3, 5, 1, 2, 6)#define SYS_TTBR0_EL12 sys_reg(3, 5, 2, 0, 0)#define SYS_TTBR1_EL12 sys_reg(3, 5, 2, 0, 1)#define SYS_TCR_EL12 sys_reg(3, 5, 2, 0, 2)@@ -1117,13 +1114,6 @@#define ZCR_ELx_LEN_WIDTH 4#define ZCR_ELx_LEN_MASK 0xf -#define SMCR_ELx_FA64_SHIFT 31 -#define SMCR_ELx_FA64_MASK (1 << SMCR_ELx_FA64_SHIFT) - -#define SMCR_ELx_LEN_SHIFT 0 -#define SMCR_ELx_LEN_WIDTH 4 -#define SMCR_ELx_LEN_MASK 0xf -#define CPACR_EL1_FPEN_EL1EN (BIT(20))/* enable EL1 access */ #define CPACR_EL1_FPEN_EL0EN (BIT(21))/* enable EL0 access, if EL1EN set */ arch/arm64/tools/sysreg +9
-0@@ -185,6 +185,26 @@ Field 1 A Field 0 M EndSysreg +SysregFields SMCR_ELx +Res0 63:32 +Field 31 FA64 +Res0 30:9 +Raz 8:4 +Field 3:0 LEN +EndSysregFields + +Sysreg SMCR_EL1 3 0 1 2 6
arch/arm64/tools/sysreg Field 1 A · line 206
Quoted 10 lines
arch/arm64/tools/sysreg +10
-0+Fields SMCR_ELx +EndSysreg + +Sysreg SMCR_EL2 3 4 1 2 6 +Fields SMCR_ELx +EndSysreg + +Sysreg SMCR_EL12 3 5 1 2 6 +Fields SMCR_ELx +EndSysreg
These all look right to me, per ARM DDI 0487H.a, so:
One minor thing for the benefit of other reviewers: the ARM ARM describes SMCR_ELx[8:4] as RAZ/WI. I think it's fine to capture that as RAZ (since the script won't output anything, and just needs some field type to check we've described all bits), but if other folk want that captured as RAZ/WI specifically that's also fine by me.
Thanks,
Mark.
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH 8/12 arm64/sme: Automatically generate SMIDR_EL1 defines
In reply to Mark BrownAutomatically generate the defines for SMIDR_EL1, no functional change.
--- arch/arm64/include/asm/sysreg.h | 1 - arch/arm64/tools/sysreg | 9 +++++++++ 2 files changed, 9 insertions(+), 1 deletion(-)
arch/arm64/include/asm/sysreg.h +0 -1
@@ -463,7 +463,6 @@#define SYS_CCSIDR_EL1 sys_reg(3, 1, 0, 0, 0)#define SYS_CLIDR_EL1 sys_reg(3, 1, 0, 0, 1)#define SYS_GMID_EL1 sys_reg(3, 1, 0, 0, 4) -#define SYS_SMIDR_EL1 sys_reg(3, 1, 0, 0, 6)#define SYS_AIDR_EL1 sys_reg(3, 1, 0, 0, 7)#define SMIDR_EL1_IMPLEMENTER_SHIFT 24
arch/arm64/tools/sysreg +9 -0
@@ -197,6 +197,15 @@ Sysreg SMCR_EL1 3 0 1 2 6 Fields SMCR_ELx EndSysreg +Sysreg SMIDR_EL1 3 1 0 0 6 +Res0 63:32 +Field 31:24 IMPLEMENTER +Field 23:16 REVISION +Field 15 SMPS +Res0 14:12 +Field 11:0 AFFINITY +EndSysreg + Sysreg SMCR_EL2 3 4 1 2 6 Fields SMCR_ELx EndSysreg
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH v1 08/12] arm64/sme: Automatically generate SMIDR_EL1 defines
Quoted 26 lines · Mark Brown
Automatically generate the defines for
SMIDR_EL1, no functional change.Signed-off-by Mark Brown <[email protected]>--- arch/arm64/include/asm/sysreg.h | 1 - arch/arm64/tools/sysreg | 9 +++++++++ 2 files changed, 9 insertions(+), 1 deletion(-)arch/arm64/include/asm/sysreg.h +0
-1@@ -463,7 +463,6 @@#define SYS_CCSIDR_EL1 sys_reg(3, 1, 0, 0, 0)#define SYS_CLIDR_EL1 sys_reg(3, 1, 0, 0, 1)#define SYS_GMID_EL1 sys_reg(3, 1, 0, 0, 4) -#define SYS_SMIDR_EL1 sys_reg(3, 1, 0, 0, 6)#define SYS_AIDR_EL1 sys_reg(3, 1, 0, 0, 7)#define SMIDR_EL1_IMPLEMENTER_SHIFT 24arch/arm64/tools/sysreg +0
-0@@ -197,6 +197,15 @@ Sysreg SMCR_EL1 3 0 1 2 6 Fields SMCR_ELx
arch/arm64/tools/sysreg Sysreg SMCR_EL1 3 0 1 2 6 · line 207
Quoted 10 lines
arch/arm64/tools/sysreg +8
-0EndSysreg +Sysreg SMIDR_EL1 3 1 0 0 6 +Res0 63:32 +Field 31:24 IMPLEMENTER +Field 23:16 REVISION +Field 15 SMPS +Res0 14:12 +Field 11:0 AFFINITY +EndSysreg
These all look right to me per ARM DDI 0487H.a, section D13.2.128, pages D13-5937 to D13-5938.
Mark.
Quoted 7 lines
arch/arm64/tools/sysreg +1
-0+ Sysreg SMCR_EL2 3 4 1 2 6 Fields SMCR_ELx EndSysregSignature · 1 line
2.30.2
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH 9/12 arm64/sme: Automatically generate SMPRIMAP_EL2 definitions
In reply to Mark BrownNo functional change should be seen from converting SMPRIMAP_EL2 to be generated.
--- arch/arm64/include/asm/sysreg.h | 1 - arch/arm64/tools/sysreg | 19 +++++++++++++++++++ 2 files changed, 19 insertions(+), 1 deletion(-)
arch/arm64/include/asm/sysreg.h +0 -1
@@ -568,7 +568,6 @@#define SYS_ZCR_EL2 sys_reg(3, 4, 1, 2, 0)#define SYS_TRFCR_EL2 sys_reg(3, 4, 1, 2, 1)#define SYS_HCRX_EL2 sys_reg(3, 4, 1, 2, 2) -#define SYS_SMPRIMAP_EL2 sys_reg(3, 4, 1, 2, 5)#define SYS_DACR32_EL2 sys_reg(3, 4, 3, 0, 0)#define SYS_HDFGRTR_EL2 sys_reg(3, 4, 3, 1, 4)#define SYS_HDFGWTR_EL2 sys_reg(3, 4, 3, 1, 5)
arch/arm64/tools/sysreg +19 -0
@@ -206,6 +206,25 @@ Res0 14:12 Field 11:0 AFFINITY EndSysreg +Sysreg SMPRIMAP_EL2 3 4 1 2 5 +Field 63:60 P15 +Field 59:56 P14 +Field 55:52 P13 +Field 51:48 P12 +Field 47:44 P11 +Field 43:40 P10 +Field 39:36 F9 +Field 35:32 P8 +Field 31:28 P7 +Field 27:24 P6 +Field 23:20 P5 +Field 19:16 P4 +Field 15:12 P3 +Field 11:8 P2 +Field 7:4 P1 +Field 3:0 P0 +EndSysreg + Sysreg SMCR_EL2 3 4 1 2 6 Fields SMCR_ELx EndSysreg
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH v1 09/12] arm64/sme: Automatically generate SMPRIMAP_EL2 definitions
Quoted 37 lines · Mark Brown
No functional change should be seen from converting
SMPRIMAP_EL2to be generated.Signed-off-by Mark Brown <[email protected]>--- arch/arm64/include/asm/sysreg.h | 1 - arch/arm64/tools/sysreg | 19 +++++++++++++++++++ 2 files changed, 19 insertions(+), 1 deletion(-)arch/arm64/include/asm/sysreg.h +0
-1@@ -568,7 +568,6 @@#define SYS_ZCR_EL2 sys_reg(3, 4, 1, 2, 0)#define SYS_TRFCR_EL2 sys_reg(3, 4, 1, 2, 1)#define SYS_HCRX_EL2 sys_reg(3, 4, 1, 2, 2) -#define SYS_SMPRIMAP_EL2 sys_reg(3, 4, 1, 2, 5)#define SYS_DACR32_EL2 sys_reg(3, 4, 3, 0, 0)#define SYS_HDFGRTR_EL2 sys_reg(3, 4, 3, 1, 4)#define SYS_HDFGWTR_EL2 sys_reg(3, 4, 3, 1, 5)arch/arm64/tools/sysreg +8
-0@@ -206,6 +206,25 @@ Res0 14:12 Field 11:0 AFFINITY EndSysreg +Sysreg SMPRIMAP_EL2 3 4 1 2 5 +Field 63:60 P15 +Field 59:56 P14 +Field 55:52 P13 +Field 51:48 P12 +Field 47:44 P11 +Field 43:40 P10 +Field 39:36 F9
arch/arm64/tools/sysreg Res0 14:12 · line 226
Quoted 10 lines
arch/arm64/tools/sysreg +10
-0+Field 35:32 P8 +Field 31:28 P7 +Field 27:24 P6 +Field 23:20 P5 +Field 19:16 P4 +Field 15:12 P3 +Field 11:8 P2 +Field 7:4 P1 +Field 3:0 P0 +EndSysreg
These all look right to me per ARM DDI 0487H.a, section D13.2.129, pages D13-5939 to D13-5942.
Mark.
Quoted 7 lines
arch/arm64/tools/sysreg +1
-0+ Sysreg SMCR_EL2 3 4 1 2 6 Fields SMCR_ELx EndSysregSignature · 1 line
2.30.2
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH 10/12 arm64/sme: Generate SMPRI_EL1 definitions
In reply to Mark BrownConvert SMPRI_EL1 to be generated. No functional change.
--- arch/arm64/include/asm/sysreg.h | 3 --- arch/arm64/tools/sysreg | 5 +++++ 2 files changed, 5 insertions(+), 3 deletions(-)
arch/arm64/include/asm/sysreg.h +0 -3
@@ -215,7 +215,6 @@#define SYS_ZCR_EL1 sys_reg(3, 0, 1, 2, 0)#define SYS_TRFCR_EL1 sys_reg(3, 0, 1, 2, 1) -#define SYS_SMPRI_EL1 sys_reg(3, 0, 1, 2, 4)#define SYS_TCR_EL1 sys_reg(3, 0, 2, 0, 2)
@@ -406,8 +405,6 @@#define TRBIDR_ALIGN_MASK GENMASK(3, 0)#define TRBIDR_ALIGN_SHIFT 0 -#define SMPRI_EL1_PRIORITY_MASK 0xf -#define SYS_PMINTENSET_EL1 sys_reg(3, 0, 9, 14, 1)#define SYS_PMINTENCLR_EL1 sys_reg(3, 0, 9, 14, 2)
arch/arm64/tools/sysreg +5 -0
@@ -185,6 +185,11 @@ Field 1 A Field 0 M EndSysreg +Sysreg SMPRI_EL1 3 0 1 2 4 +Res0 63:4 +Field 3:0 PRIORITY +EndSysreg + SysregFields SMCR_ELx Res0 63:32 Field 31 FA64
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH v1 10/12] arm64/sme: Generate SMPRI_EL1 definitions
Quoted 34 lines · Mark Brown
Convert
SMPRI_EL1to be generated. No functional change.Signed-off-by Mark Brown <[email protected]>--- arch/arm64/include/asm/sysreg.h | 3 --- arch/arm64/tools/sysreg | 5 +++++ 2 files changed, 5 insertions(+), 3 deletions(-)arch/arm64/include/asm/sysreg.h +0
-3@@ -215,7 +215,6 @@#define SYS_ZCR_EL1 sys_reg(3, 0, 1, 2, 0)#define SYS_TRFCR_EL1 sys_reg(3, 0, 1, 2, 1) -#define SYS_SMPRI_EL1 sys_reg(3, 0, 1, 2, 4)#define SYS_TCR_EL1 sys_reg(3, 0, 2, 0, 2)@@ -406,8 +405,6 @@#define TRBIDR_ALIGN_MASK GENMASK(3, 0)#define TRBIDR_ALIGN_SHIFT 0 -#define SMPRI_EL1_PRIORITY_MASK 0xf -#define SYS_PMINTENSET_EL1 sys_reg(3, 0, 9, 14, 1)#define SYS_PMINTENCLR_EL1 sys_reg(3, 0, 9, 14, 2)arch/arm64/tools/sysreg +0
-0
arch/arm64/tools/sysreg Field 1 A · line 191
Quoted 7 lines
arch/arm64/tools/sysreg +4
-0@@ -185,6 +185,11 @@ Field 1 A Field 0 M EndSysreg +Sysreg SMPRI_EL1 3 0 1 2 4 +Res0 63:4 +Field 3:0 PRIORITY +EndSysreg
These all look right to me per ARM DDI 0487H.a, section D13.2.130, pages D13-5943 to D13-5945.
Mark.
Quoted 7 lines
arch/arm64/tools/sysreg +1
-0+ SysregFields SMCR_ELx Res0 63:32 Field 31 FA64Signature · 1 line
2.30.2
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH 11/12 arm64/sme: Generate defintions for SVCR
In reply to Mark BrownConvert SVCR to automatic generation, no functional change.
--- arch/arm64/include/asm/sysreg.h | 4 ---- arch/arm64/tools/sysreg | 6 ++++++ 2 files changed, 6 insertions(+), 4 deletions(-)
arch/arm64/include/asm/sysreg.h +0 -4
@@ -474,10 +474,6 @@#define SYS_RNDR_EL0 sys_reg(3, 3, 2, 4, 0)#define SYS_RNDRRS_EL0 sys_reg(3, 3, 2, 4, 1) -#define SYS_SVCR sys_reg(3, 3, 4, 2, 2) -#define SVCR_ZA_MASK 2 -#define SVCR_SM_MASK 1 -#define SYS_PMCR_EL0 sys_reg(3, 3, 9, 12, 0)#define SYS_PMCNTENSET_EL0 sys_reg(3, 3, 9, 12, 1)#define SYS_PMCNTENCLR_EL0 sys_reg(3, 3, 9, 12, 2)
arch/arm64/tools/sysreg +6 -0
@@ -211,6 +211,12 @@ Res0 14:12 Field 11:0 AFFINITY EndSysreg +Sysreg SVCR 3 3 4 2 2 +Res0 63:2 +Field 1 ZA +Field 0 SM +EndSysreg + Sysreg SMPRIMAP_EL2 3 4 1 2 5 Field 63:60 P15 Field 59:56 P14
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH v1 11/12] arm64/sme: Generate defintions for SVCR
Quoted 28 lines · Mark Brown
Convert SVCR to automatic generation, no functional change.
Signed-off-by Mark Brown <[email protected]>--- arch/arm64/include/asm/sysreg.h | 4 ---- arch/arm64/tools/sysreg | 6 ++++++ 2 files changed, 6 insertions(+), 4 deletions(-)arch/arm64/include/asm/sysreg.h +0
-4@@ -474,10 +474,6 @@#define SYS_RNDR_EL0 sys_reg(3, 3, 2, 4, 0)#define SYS_RNDRRS_EL0 sys_reg(3, 3, 2, 4, 1) -#define SYS_SVCR sys_reg(3, 3, 4, 2, 2) -#define SVCR_ZA_MASK 2 -#define SVCR_SM_MASK 1 -#define SYS_PMCR_EL0 sys_reg(3, 3, 9, 12, 0)#define SYS_PMCNTENSET_EL0 sys_reg(3, 3, 9, 12, 1)#define SYS_PMCNTENCLR_EL0 sys_reg(3, 3, 9, 12, 2)arch/arm64/tools/sysreg +0
-0
arch/arm64/tools/sysreg Res0 14:12 · line 218
Quoted 8 lines
arch/arm64/tools/sysreg +5
-0@@ -211,6 +211,12 @@ Res0 14:12 Field 11:0 AFFINITY EndSysreg +Sysreg SVCR 3 3 4 2 2 +Res0 63:2 +Field 1 ZA +Field 0 SM +EndSysreg
These all look right to me per ARM DDI 0487H.a, section C5.2.25, pages C5-747 to C5-749.
Mark.
Quoted 7 lines
arch/arm64/tools/sysreg +1
-0+ Sysreg SMPRIMAP_EL2 3 4 1 2 5 Field 63:60 P15 Field 59:56 P14Signature · 1 line
2.30.2
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
PATCH 12/12 arm64/sve: Generate ZCR definitions
In reply to Mark BrownConvert the various ZCR instances to automatic generation, no functional changes expected.
--- arch/arm64/include/asm/sysreg.h | 7 ------- arch/arm64/tools/sysreg | 18 ++++++++++++++++++ 2 files changed, 18 insertions(+), 7 deletions(-)
arch/arm64/include/asm/sysreg.h +0 -7
@@ -213,7 +213,6 @@#define SYS_RGSR_EL1 sys_reg(3, 0, 1, 0, 5)#define SYS_GCR_EL1 sys_reg(3, 0, 1, 0, 6) -#define SYS_ZCR_EL1 sys_reg(3, 0, 1, 2, 0)#define SYS_TRFCR_EL1 sys_reg(3, 0, 1, 2, 1)#define SYS_TCR_EL1 sys_reg(3, 0, 2, 0, 2)
@@ -558,7 +557,6 @@#define SYS_HFGRTR_EL2 sys_reg(3, 4, 1, 1, 4)#define SYS_HFGWTR_EL2 sys_reg(3, 4, 1, 1, 5)#define SYS_HFGITR_EL2 sys_reg(3, 4, 1, 1, 6) -#define SYS_ZCR_EL2 sys_reg(3, 4, 1, 2, 0)#define SYS_TRFCR_EL2 sys_reg(3, 4, 1, 2, 1)#define SYS_HCRX_EL2 sys_reg(3, 4, 1, 2, 2)#define SYS_DACR32_EL2 sys_reg(3, 4, 3, 0, 0)
@@ -619,7 +617,6 @@/* VHE encodings for architectural EL0/1 system registers */ #define SYS_SCTLR_EL12 sys_reg(3, 5, 1, 0, 0)#define SYS_CPACR_EL12 sys_reg(3, 5, 1, 0, 2) -#define SYS_ZCR_EL12 sys_reg(3, 5, 1, 2, 0)#define SYS_TTBR0_EL12 sys_reg(3, 5, 2, 0, 0)#define SYS_TTBR1_EL12 sys_reg(3, 5, 2, 0, 1)#define SYS_TCR_EL12 sys_reg(3, 5, 2, 0, 2)
@@ -1101,10 +1098,6 @@#define DCZID_DZP_SHIFT 4#define DCZID_BS_SHIFT 0 -#define ZCR_ELx_LEN_SHIFT 0 -#define ZCR_ELx_LEN_WIDTH 4 -#define ZCR_ELx_LEN_MASK 0xf -#define CPACR_EL1_FPEN_EL1EN (BIT(20))/* enable EL1 access */ #define CPACR_EL1_FPEN_EL0EN (BIT(21))/* enable EL0 access, if EL1EN set */
arch/arm64/tools/sysreg +18 -0
@@ -190,6 +190,16 @@ Res0 63:4 Field 3:0 PRIORITY EndSysreg +SysregFields ZCR_ELx +Res0 63:9 +Raz 8:4 +Field 3:0 LEN +EndSysregFields + +Sysreg ZCR_EL1 3 0 1 2 0 +Fields ZCR_ELx +EndSysreg + SysregFields SMCR_ELx Res0 63:32 Field 31 FA64
@@ -217,6 +227,10 @@ Field 1 ZA Field 0 SM EndSysreg +Sysreg ZCR_EL2 3 4 1 2 0 +Fields ZCR_ELx +EndSysreg + Sysreg SMPRIMAP_EL2 3 4 1 2 5 Field 63:60 P15 Field 59:56 P14
@@ -240,6 +254,10 @@ Sysreg SMCR_EL2 3 4 1 2 6 Fields SMCR_ELx EndSysreg +Sysreg ZCR_EL12 3 5 1 2 0 +Fields ZCR_ELx +EndSysreg + Sysreg SMCR_EL12 3 5 1 2 6 Fields SMCR_ELx EndSysreg
Signature · 7 lines
2.30.2 _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Re: [PATCH v1 12/12] arm64/sve: Generate ZCR definitions
Quoted 81 lines · Mark Brown
Convert the various ZCR instances to automatic generation, no functional changes expected.
Signed-off-by Mark Brown <[email protected]>--- arch/arm64/include/asm/sysreg.h | 7 ------- arch/arm64/tools/sysreg | 18 ++++++++++++++++++ 2 files changed, 18 insertions(+), 7 deletions(-)arch/arm64/include/asm/sysreg.h +0
-7@@ -213,7 +213,6 @@#define SYS_RGSR_EL1 sys_reg(3, 0, 1, 0, 5)#define SYS_GCR_EL1 sys_reg(3, 0, 1, 0, 6) -#define SYS_ZCR_EL1 sys_reg(3, 0, 1, 2, 0)#define SYS_TRFCR_EL1 sys_reg(3, 0, 1, 2, 1)#define SYS_TCR_EL1 sys_reg(3, 0, 2, 0, 2)@@ -558,7 +557,6 @@#define SYS_HFGRTR_EL2 sys_reg(3, 4, 1, 1, 4)#define SYS_HFGWTR_EL2 sys_reg(3, 4, 1, 1, 5)#define SYS_HFGITR_EL2 sys_reg(3, 4, 1, 1, 6) -#define SYS_ZCR_EL2 sys_reg(3, 4, 1, 2, 0)#define SYS_TRFCR_EL2 sys_reg(3, 4, 1, 2, 1)#define SYS_HCRX_EL2 sys_reg(3, 4, 1, 2, 2)#define SYS_DACR32_EL2 sys_reg(3, 4, 3, 0, 0)@@ -619,7 +617,6 @@/* VHE encodings for architectural EL0/1 system registers */ #define SYS_SCTLR_EL12 sys_reg(3, 5, 1, 0, 0)#define SYS_CPACR_EL12 sys_reg(3, 5, 1, 0, 2) -#define SYS_ZCR_EL12 sys_reg(3, 5, 1, 2, 0)#define SYS_TTBR0_EL12 sys_reg(3, 5, 2, 0, 0)#define SYS_TTBR1_EL12 sys_reg(3, 5, 2, 0, 1)#define SYS_TCR_EL12 sys_reg(3, 5, 2, 0, 2)@@ -1101,10 +1098,6 @@#define DCZID_DZP_SHIFT 4#define DCZID_BS_SHIFT 0 -#define ZCR_ELx_LEN_SHIFT 0 -#define ZCR_ELx_LEN_WIDTH 4 -#define ZCR_ELx_LEN_MASK 0xf -#define CPACR_EL1_FPEN_EL1EN (BIT(20))/* enable EL1 access */ #define CPACR_EL1_FPEN_EL0EN (BIT(21))/* enable EL0 access, if EL1EN set */ arch/arm64/tools/sysreg +14
-0@@ -190,6 +190,16 @@ Res0 63:4 Field 3:0 PRIORITY EndSysreg +SysregFields ZCR_ELx +Res0 63:9 +Raz 8:4 +Field 3:0 LEN +EndSysregFields + +Sysreg ZCR_EL1 3 0 1 2 0 +Fields ZCR_ELx +EndSysreg + SysregFields SMCR_ELx Res0 63:32 Field 31 FA64@@ -217,6 +227,10 @@ Field 1 ZA Field 0 SM EndSysreg +Sysreg ZCR_EL2 3 4 1 2 0 +Fields ZCR_ELx +EndSysreg + Sysreg SMPRIMAP_EL2 3 4 1 2 5 Field 63:60 P15 Field 59:56 P14
arch/arm64/tools/sysreg Sysreg SMCR_EL2 3 4 1 2 6 · line 263
Quoted 10 lines
arch/arm64/tools/sysreg +4
-0@@ -240,6 +254,10 @@ Sysreg SMCR_EL2 3 4 1 2 6 Fields SMCR_ELx EndSysreg +Sysreg ZCR_EL12 3 5 1 2 0 +Fields ZCR_ELx +EndSysreg + Sysreg SMCR_EL12 3 5 1 2 6 Fields SMCR_ELx EndSysreg
These all look right to me per ARM DDI 0487H.a:
* ZCR_EL1: pages D13-6071 to D13-6072
* ZCR_EL12: pages D13-6073 to D13-6074
* ZCR_EL2: pages D13-6076 to D13-6077
As on another patch, I'm fine either way with capturing RAZ/WI as RAZ.
Mark.
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Quoted 10 lines · Mark Brown
This series builds on top of the recently applied series for system register generation, converting the floating point registers to automatic generation. There's nothing remarkable in here, just a bunch of straightforward cleanups and the addition of support for RAZ fields followed by the actual conversions.
It does skip over the SVCRSM, SVCRZA and SVCRSMZA aliases for now since I need to think a little about what to do with those.
[...]
Applied to arm64 (for-next/sysreg-gen), thanks!
(I merged for-next/sme into this branch for dependencies)
[01/12] arm64/fp: Make SVE and SME length register definition match architecture
https://git.kernel.org/arm64/c/f171f9e4097d [02/12] arm64/fp: Rename SVE and SME LEN field name to _WIDTH https://git.kernel.org/arm64/c/5b06dcfd9e0a [03/12] arm64/sme: Drop SYS_ from SMIDR_EL1 defines https://git.kernel.org/arm64/c/a6dab6cc0f4c [04/12] arm64/sme: Standardise bitfield names for SVCR https://git.kernel.org/arm64/c/e65fc01bf271 [05/12] arm64/sme: Remove _EL0 from name of SVCR - FIXME sysreg.h https://git.kernel.org/arm64/c/ec0067a63e5a [06/12] arm64/sysreg: Support generation of RAZ fields https://git.kernel.org/arm64/c/9e2c0819ac85 [07/12] arm64/sme: Automatically generate defines for SMCR https://git.kernel.org/arm64/c/0d1322e7ea75 [08/12] arm64/sme: Automatically generate SMIDR_EL1 defines https://git.kernel.org/arm64/c/c37b8700b723 [09/12] arm64/sme: Automatically generate SMPRIMAP_EL2 definitions https://git.kernel.org/arm64/c/8e053810e6ce [10/12] arm64/sme: Generate SMPRI_EL1 definitions https://git.kernel.org/arm64/c/9321f0492b89 [11/12] arm64/sme: Generate defintions for SVCR https://git.kernel.org/arm64/c/11e12a91c118 [12/12] arm64/sve: Generate ZCR definitions https://git.kernel.org/arm64/c/89e9fb327421
Signature · 7 lines
Catalin _______________________________________________ linux-arm-kernel mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-arm-kernel