ARM queue: mostly patches from me, but also the Smartfusion2 board. thanks -- PMM The following changes since commit 9ee660e7c138595224b65ddc1c5712549f0a278c: Merge remote-tracking branch 'remotes/yongbok/tags/mips-20170921' into staging (2017-09-21 14:40:32 +0100) are available in the git repository at: git://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20170921 for you to fetch changes up to 6d262dcb7d108eda93813574c2061398084dc795: msf2: Add Emcraft's Smartfusion2 SOM kit (2017-09-21 16:36:56 +0100) ---------------------------------------------------------------- target-arm queue: * more preparatory work for v8M support * convert some omap devices away from old_mmio * remove out of date ARM ARM section references in comments * add the Smartfusion2 board ---------------------------------------------------------------- Peter Maydell (26): target/arm: Implement MSR/MRS access to NS banked registers nvic: Add banked exception states nvic: Add cached vectpending_is_s_banked state nvic: Add cached vectpending_prio state nvic: Implement AIRCR changes for v8M nvic: Make ICSR.RETTOBASE handle banked exceptions nvic: Implement NVIC_ITNS<n> registers nvic: Handle banked exceptions in nvic_recompute_state() nvic: Make set_pending and clear_pending take a secure parameter nvic: Make SHPR registers banked nvic: Compare group priority for escalation to HF nvic: In escalation to HardFault, support HF not being priority -1 nvic: Implement v8M changes to fixed priority exceptions nvic: Disable the non-secure HardFault if AIRCR.BFHFNMINS is clear nvic: Handle v8M changes in nvic_exec_prio() target/arm: Handle banking in negative-execution-priority check in cpu_mmu_index() nvic: Make ICSR banked for v8M nvic: Make SHCSR banked for v8M nvic: Support banked exceptions in acknowledge and complete target/arm: Remove out of date ARM ARM section references in A64 decoder hw/arm/palm.c: Don't use old_mmio for static_ops hw/gpio/omap_gpio.c: Don't use old_mmio hw/timer/omap_synctimer.c: Don't use old_mmio hw/timer/omap_gptimer: Don't use old_mmio hw/i2c/omap_i2c.c: Don't use old_mmio hw/arm/omap2.c: Don't use old_mmio Subbaraya Sundeep (5): msf2: Add Smartfusion2 System timer msf2: Microsemi Smartfusion2 System Register block msf2: Add Smartfusion2 SPI controller msf2: Add Smartfusion2 SoC msf2: Add Emcraft's Smartfusion2 SOM kit hw/arm/Makefile.objs | 1 + hw/misc/Makefile.objs | 1 + hw/ssi/Makefile.objs | 1 + hw/timer/Makefile.objs | 1 + include/hw/arm/msf2-soc.h | 67 +++ include/hw/intc/armv7m_nvic.h | 33 +- include/hw/misc/msf2-sysreg.h | 77 ++++ include/hw/ssi/mss-spi.h | 58 +++ include/hw/timer/mss-timer.h | 64 +++ target/arm/cpu.h | 62 ++- hw/arm/msf2-soc.c | 238 +++++++++++ hw/arm/msf2-som.c | 105 +++++ hw/arm/omap2.c | 49 ++- hw/arm/palm.c | 30 +- hw/gpio/omap_gpio.c | 26 +- hw/i2c/omap_i2c.c | 44 +- hw/intc/armv7m_nvic.c | 913 ++++++++++++++++++++++++++++++++++------ hw/misc/msf2-sysreg.c | 160 +++++++ hw/ssi/mss-spi.c | 404 ++++++++++++++++++ hw/timer/mss-timer.c | 289 +++++++++++++ hw/timer/omap_gptimer.c | 49 ++- hw/timer/omap_synctimer.c | 35 +- target/arm/cpu.c | 7 + target/arm/helper.c | 142 ++++++- target/arm/translate-a64.c | 227 +++++----- default-configs/arm-softmmu.mak | 1 + hw/intc/trace-events | 13 +- hw/misc/trace-events | 5 + 28 files changed, 2735 insertions(+), 367 deletions(-) create mode 100644 include/hw/arm/msf2-soc.h create mode 100644 include/hw/misc/msf2-sysreg.h create mode 100644 include/hw/ssi/mss-spi.h create mode 100644 include/hw/timer/mss-timer.h create mode 100644 hw/arm/msf2-soc.c create mode 100644 hw/arm/msf2-som.c create mode 100644 hw/misc/msf2-sysreg.c create mode 100644 hw/ssi/mss-spi.c create mode 100644 hw/timer/mss-timer.c In v8M the MSR and MRS instructions have extra register value encodings to allow secure code to access the non-secure banked version of various special registers. (We don't implement the MSPLIM_NS or PSPLIM_NS aliases, because we don't currently implement the stack limit registers at all.) Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-2-git-send-email-peter.maydell@linaro.org --- target/arm/helper.c | 110 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 110 insertions(+) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ uint32_t HELPER(v7m_mrs)(CPUARMState *env, uint32_t reg) break; case 20: /* CONTROL */ return env->v7m.control[env->v7m.secure]; + case 0x94: /* CONTROL_NS */ + /* We have to handle this here because unprivileged Secure code + * can read the NS CONTROL register. + */ + if (!env->v7m.secure) { + return 0; + } + return env->v7m.control[M_REG_NS]; } if (el == 0) { return 0; /* unprivileged reads others as zero */ } + if (arm_feature(env, ARM_FEATURE_M_SECURITY)) { + switch (reg) { + case 0x88: /* MSP_NS */ + if (!env->v7m.secure) { + return 0; + } + return env->v7m.other_ss_msp; + case 0x89: /* PSP_NS */ + if (!env->v7m.secure) { + return 0; + } + return env->v7m.other_ss_psp; + case 0x90: /* PRIMASK_NS */ + if (!env->v7m.secure) { + return 0; + } + return env->v7m.primask[M_REG_NS]; + case 0x91: /* BASEPRI_NS */ + if (!env->v7m.secure) { + return 0; + } + return env->v7m.basepri[M_REG_NS]; + case 0x93: /* FAULTMASK_NS */ + if (!env->v7m.secure) { + return 0; + } + return env->v7m.faultmask[M_REG_NS]; + case 0x98: /* SP_NS */ + { + /* This gives the non-secure SP selected based on whether we're + * currently in handler mode or not, using the NS CONTROL.SPSEL. + */ + bool spsel = env->v7m.control[M_REG_NS] & R_V7M_CONTROL_SPSEL_MASK; + + if (!env->v7m.secure) { + return 0; + } + if (!arm_v7m_is_handler_mode(env) && spsel) { + return env->v7m.other_ss_psp; + } else { + return env->v7m.other_ss_msp; + } + } + default: + break; + } + } + switch (reg) { case 8: /* MSP */ return (env->v7m.control[env->v7m.secure] & R_V7M_CONTROL_SPSEL_MASK) ? @@ -XXX,XX +XXX,XX @@ void HELPER(v7m_msr)(CPUARMState *env, uint32_t maskreg, uint32_t val) return; } + if (arm_feature(env, ARM_FEATURE_M_SECURITY)) { + switch (reg) { + case 0x88: /* MSP_NS */ + if (!env->v7m.secure) { + return; + } + env->v7m.other_ss_msp = val; + return; + case 0x89: /* PSP_NS */ + if (!env->v7m.secure) { + return; + } + env->v7m.other_ss_psp = val; + return; + case 0x90: /* PRIMASK_NS */ + if (!env->v7m.secure) { + return; + } + env->v7m.primask[M_REG_NS] = val & 1; + return; + case 0x91: /* BASEPRI_NS */ + if (!env->v7m.secure) { + return; + } + env->v7m.basepri[M_REG_NS] = val & 0xff; + return; + case 0x93: /* FAULTMASK_NS */ + if (!env->v7m.secure) { + return; + } + env->v7m.faultmask[M_REG_NS] = val & 1; + return; + case 0x98: /* SP_NS */ + { + /* This gives the non-secure SP selected based on whether we're + * currently in handler mode or not, using the NS CONTROL.SPSEL. + */ + bool spsel = env->v7m.control[M_REG_NS] & R_V7M_CONTROL_SPSEL_MASK; + + if (!env->v7m.secure) { + return; + } + if (!arm_v7m_is_handler_mode(env) && spsel) { + env->v7m.other_ss_psp = val; + } else { + env->v7m.other_ss_msp = val; + } + return; + } + default: + break; + } + } + switch (reg) { case 0 ... 7: /* xPSR sub-fields */ /* only APSR is actually writable */ -- 2.7.4 For the v8M security extension, some exceptions must be banked between security states. Add the new vecinfo array which holds the state for the banked exceptions and migrate it if the CPU the NVIC is attached to implements the security extension. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> --- include/hw/intc/armv7m_nvic.h | 14 ++++++++++++ hw/intc/armv7m_nvic.c | 53 ++++++++++++++++++++++++++++++++++++++++++- 2 files changed, 66 insertions(+), 1 deletion(-) diff --git a/include/hw/intc/armv7m_nvic.h b/include/hw/intc/armv7m_nvic.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/intc/armv7m_nvic.h +++ b/include/hw/intc/armv7m_nvic.h @@ -XXX,XX +XXX,XX @@ /* Highest permitted number of exceptions (architectural limit) */ #define NVIC_MAX_VECTORS 512 +/* Number of internal exceptions */ +#define NVIC_INTERNAL_VECTORS 16 typedef struct VecInfo { /* Exception priorities can range from -3 to 255; only the unmodifiable @@ -XXX,XX +XXX,XX @@ typedef struct NVICState { ARMCPU *cpu; VecInfo vectors[NVIC_MAX_VECTORS]; + /* If the v8M security extension is implemented, some of the internal + * exceptions are banked between security states (ie there exists both + * a Secure and a NonSecure version of the exception and its state): + * HardFault, MemManage, UsageFault, SVCall, PendSV, SysTick (R_PJHV) + * The rest (including all the external exceptions) are not banked, though + * they may be configurable to target either Secure or NonSecure state. + * We store the secure exception state in sec_vectors[] for the banked + * exceptions, and otherwise use only vectors[] (including for exceptions + * like SecureFault that unconditionally target Secure state). + * Entries in sec_vectors[] for non-banked exception numbers are unused. + */ + VecInfo sec_vectors[NVIC_INTERNAL_VECTORS]; uint32_t prigroup; /* vectpending and exception_prio are both cached state that can diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ * For historical reasons QEMU tends to use "interrupt" and * "exception" more or less interchangeably. */ -#define NVIC_FIRST_IRQ 16 +#define NVIC_FIRST_IRQ NVIC_INTERNAL_VECTORS #define NVIC_MAX_IRQ (NVIC_MAX_VECTORS - NVIC_FIRST_IRQ) /* Effective running priority of the CPU when no exception is active @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_VecInfo = { } }; +static bool nvic_security_needed(void *opaque) +{ + NVICState *s = opaque; + + return arm_feature(&s->cpu->env, ARM_FEATURE_M_SECURITY); +} + +static int nvic_security_post_load(void *opaque, int version_id) +{ + NVICState *s = opaque; + int i; + + /* Check for out of range priority settings */ + if (s->sec_vectors[ARMV7M_EXCP_HARD].prio != -1) { + return 1; + } + for (i = ARMV7M_EXCP_MEM; i < ARRAY_SIZE(s->sec_vectors); i++) { + if (s->sec_vectors[i].prio & ~0xff) { + return 1; + } + } + return 0; +} + +static const VMStateDescription vmstate_nvic_security = { + .name = "nvic/m-security", + .version_id = 1, + .minimum_version_id = 1, + .needed = nvic_security_needed, + .post_load = &nvic_security_post_load, + .fields = (VMStateField[]) { + VMSTATE_STRUCT_ARRAY(sec_vectors, NVICState, NVIC_INTERNAL_VECTORS, 1, + vmstate_VecInfo, VecInfo), + VMSTATE_END_OF_LIST() + } +}; + static const VMStateDescription vmstate_nvic = { .name = "armv7m_nvic", .version_id = 4, @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_nvic = { vmstate_VecInfo, VecInfo), VMSTATE_UINT32(prigroup, NVICState), VMSTATE_END_OF_LIST() + }, + .subsections = (const VMStateDescription*[]) { + &vmstate_nvic_security, + NULL } }; @@ -XXX,XX +XXX,XX @@ static void armv7m_nvic_reset(DeviceState *dev) s->vectors[ARMV7M_EXCP_NMI].prio = -2; s->vectors[ARMV7M_EXCP_HARD].prio = -1; + if (arm_feature(&s->cpu->env, ARM_FEATURE_M_SECURITY)) { + s->sec_vectors[ARMV7M_EXCP_HARD].enabled = 1; + s->sec_vectors[ARMV7M_EXCP_SVC].enabled = 1; + s->sec_vectors[ARMV7M_EXCP_PENDSV].enabled = 1; + s->sec_vectors[ARMV7M_EXCP_SYSTICK].enabled = 1; + + /* AIRCR.BFHFNMINS resets to 0 so Secure HF is priority -1 (R_CMTC) */ + s->sec_vectors[ARMV7M_EXCP_HARD].prio = -1; + } + /* Strictly speaking the reset handler should be enabled. * However, we don't simulate soft resets through the NVIC, * and the reset vector should never be pended. -- 2.7.4 With banked exceptions, just the exception number in s->vectpending is no longer sufficient to uniquely identify the pending exception. Add a vectpending_is_s_banked bool which is true if the exception is using the sec_vectors[] array. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Message-id: 1505240046-11454-4-git-send-email-peter.maydell@linaro.org --- include/hw/intc/armv7m_nvic.h | 11 +++++++++-- hw/intc/armv7m_nvic.c | 1 + 2 files changed, 10 insertions(+), 2 deletions(-) diff --git a/include/hw/intc/armv7m_nvic.h b/include/hw/intc/armv7m_nvic.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/intc/armv7m_nvic.h +++ b/include/hw/intc/armv7m_nvic.h @@ -XXX,XX +XXX,XX @@ typedef struct NVICState { VecInfo sec_vectors[NVIC_INTERNAL_VECTORS]; uint32_t prigroup; - /* vectpending and exception_prio are both cached state that can - * be recalculated from the vectors[] array and the prigroup field. + /* The following fields are all cached state that can be recalculated + * from the vectors[] and sec_vectors[] arrays and the prigroup field: + * - vectpending + * - vectpending_is_secure + * - exception_prio */ unsigned int vectpending; /* highest prio pending enabled exception */ + /* true if vectpending is a banked secure exception, ie it is in + * sec_vectors[] rather than vectors[] + */ + bool vectpending_is_s_banked; int exception_prio; /* group prio of the highest prio active exception */ MemoryRegion sysregmem; diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ static void armv7m_nvic_reset(DeviceState *dev) s->exception_prio = NVIC_NOEXC_PRIO; s->vectpending = 0; + s->vectpending_is_s_banked = false; } static void nvic_systick_trigger(void *opaque, int n, int level) -- 2.7.4 Instead of looking up the pending priority in nvic_pending_prio(), cache it in a new state struct field. The calculation of the pending priority given the interrupt number is more complicated in v8M with the security extension, so the caching will be worthwhile. This changes nvic_pending_prio() from returning a full (group + subpriority) priority value to returning a group priority. This doesn't require changes to its callsites because we use it only in comparisons of the form execution_prio > nvic_pending_prio() and execution priority is always a group priority, so a test (exec prio > full prio) is true if and only if (execprio > group_prio). (Architecturally the expected comparison is with the group priority for this sort of "would we preempt" test; we were only doing a test with a full priority as an optimisation to avoid the mask, which is possible precisely because the two comparisons always give the same answer.) Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-5-git-send-email-peter.maydell@linaro.org --- include/hw/intc/armv7m_nvic.h | 2 ++ hw/intc/armv7m_nvic.c | 23 +++++++++++++---------- hw/intc/trace-events | 2 +- 3 files changed, 16 insertions(+), 11 deletions(-) diff --git a/include/hw/intc/armv7m_nvic.h b/include/hw/intc/armv7m_nvic.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/intc/armv7m_nvic.h +++ b/include/hw/intc/armv7m_nvic.h @@ -XXX,XX +XXX,XX @@ typedef struct NVICState { * - vectpending * - vectpending_is_secure * - exception_prio + * - vectpending_prio */ unsigned int vectpending; /* highest prio pending enabled exception */ /* true if vectpending is a banked secure exception, ie it is in @@ -XXX,XX +XXX,XX @@ typedef struct NVICState { */ bool vectpending_is_s_banked; int exception_prio; /* group prio of the highest prio active exception */ + int vectpending_prio; /* group prio of the exeception in vectpending */ MemoryRegion sysregmem; MemoryRegion sysreg_ns_mem; diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ static const uint8_t nvic_id[] = { static int nvic_pending_prio(NVICState *s) { - /* return the priority of the current pending interrupt, + /* return the group priority of the current pending interrupt, * or NVIC_NOEXC_PRIO if no interrupt is pending */ - return s->vectpending ? s->vectors[s->vectpending].prio : NVIC_NOEXC_PRIO; + return s->vectpending_prio; } /* Return the value of the ISCR RETTOBASE bit: @@ -XXX,XX +XXX,XX @@ static void nvic_recompute_state(NVICState *s) active_prio &= nvic_gprio_mask(s); } + if (pend_prio > 0) { + pend_prio &= nvic_gprio_mask(s); + } + s->vectpending = pend_irq; + s->vectpending_prio = pend_prio; s->exception_prio = active_prio; - trace_nvic_recompute_state(s->vectpending, s->exception_prio); + trace_nvic_recompute_state(s->vectpending, + s->vectpending_prio, + s->exception_prio); } /* Return the current execution priority of the CPU @@ -XXX,XX +XXX,XX @@ void armv7m_nvic_acknowledge_irq(void *opaque) CPUARMState *env = &s->cpu->env; const int pending = s->vectpending; const int running = nvic_exec_prio(s); - int pendgroupprio; VecInfo *vec; assert(pending > ARMV7M_EXCP_RESET && pending < s->num_irq); @@ -XXX,XX +XXX,XX @@ void armv7m_nvic_acknowledge_irq(void *opaque) assert(vec->enabled); assert(vec->pending); - pendgroupprio = vec->prio; - if (pendgroupprio > 0) { - pendgroupprio &= nvic_gprio_mask(s); - } - assert(pendgroupprio < running); + assert(s->vectpending_prio < running); - trace_nvic_acknowledge_irq(pending, vec->prio); + trace_nvic_acknowledge_irq(pending, s->vectpending_prio); vec->active = 1; vec->pending = 0; @@ -XXX,XX +XXX,XX @@ static void armv7m_nvic_reset(DeviceState *dev) s->exception_prio = NVIC_NOEXC_PRIO; s->vectpending = 0; s->vectpending_is_s_banked = false; + s->vectpending_prio = NVIC_NOEXC_PRIO; } static void nvic_systick_trigger(void *opaque, int n, int level) diff --git a/hw/intc/trace-events b/hw/intc/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/trace-events +++ b/hw/intc/trace-events @@ -XXX,XX +XXX,XX @@ gicv3_redist_set_irq(uint32_t cpu, int irq, int level) "GICv3 redistributor 0x%x gicv3_redist_send_sgi(uint32_t cpu, int irq) "GICv3 redistributor 0x%x pending SGI %d" # hw/intc/armv7m_nvic.c -nvic_recompute_state(int vectpending, int exception_prio) "NVIC state recomputed: vectpending %d exception_prio %d" +nvic_recompute_state(int vectpending, int vectpending_prio, int exception_prio) "NVIC state recomputed: vectpending %d vectpending_prio %d exception_prio %d" nvic_set_prio(int irq, uint8_t prio) "NVIC set irq %d priority %d" nvic_irq_update(int vectpending, int pendprio, int exception_prio, int level) "NVIC vectpending %d pending prio %d exception_prio %d: setting irq line to %d" nvic_escalate_prio(int irq, int irqprio, int runprio) "NVIC escalating irq %d to HardFault: insufficient priority %d >= %d" -- 2.7.4 The Application Interrupt and Reset Control Register has some changes for v8M: * new bits SYSRESETREQS, BFHFNMINS and PRIS: these all have real state if the security extension is implemented and otherwise are constant * the PRIGROUP field is banked between security states * non-secure code can be blocked from using the SYSRESET bit to reset the system if SYSRESETREQS is set Implement the new state and the changes to register read and write. For the moment we ignore the effects of the secure PRIGROUP. We will implement the effects of PRIS and BFHFNMIS later. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-6-git-send-email-peter.maydell@linaro.org --- include/hw/intc/armv7m_nvic.h | 3 ++- target/arm/cpu.h | 12 +++++++++++ hw/intc/armv7m_nvic.c | 49 +++++++++++++++++++++++++++++++++---------- target/arm/cpu.c | 7 +++++++ 4 files changed, 59 insertions(+), 12 deletions(-) diff --git a/include/hw/intc/armv7m_nvic.h b/include/hw/intc/armv7m_nvic.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/intc/armv7m_nvic.h +++ b/include/hw/intc/armv7m_nvic.h @@ -XXX,XX +XXX,XX @@ typedef struct NVICState { * Entries in sec_vectors[] for non-banked exception numbers are unused. */ VecInfo sec_vectors[NVIC_INTERNAL_VECTORS]; - uint32_t prigroup; + /* The PRIGROUP field in AIRCR is banked */ + uint32_t prigroup[M_REG_NUM_BANKS]; /* The following fields are all cached state that can be recalculated * from the vectors[] and sec_vectors[] arrays and the prigroup field: diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ typedef struct CPUARMState { int exception; uint32_t primask[M_REG_NUM_BANKS]; uint32_t faultmask[M_REG_NUM_BANKS]; + uint32_t aircr; /* only holds r/w state if security extn implemented */ uint32_t secure; /* Is CPU in Secure state? (not guest visible) */ } v7m; @@ -XXX,XX +XXX,XX @@ FIELD(V7M_CCR, STKALIGN, 9, 1) FIELD(V7M_CCR, DC, 16, 1) FIELD(V7M_CCR, IC, 17, 1) +/* V7M AIRCR bits */ +FIELD(V7M_AIRCR, VECTRESET, 0, 1) +FIELD(V7M_AIRCR, VECTCLRACTIVE, 1, 1) +FIELD(V7M_AIRCR, SYSRESETREQ, 2, 1) +FIELD(V7M_AIRCR, SYSRESETREQS, 3, 1) +FIELD(V7M_AIRCR, PRIGROUP, 8, 3) +FIELD(V7M_AIRCR, BFHFNMINS, 13, 1) +FIELD(V7M_AIRCR, PRIS, 14, 1) +FIELD(V7M_AIRCR, ENDIANNESS, 15, 1) +FIELD(V7M_AIRCR, VECTKEY, 16, 16) + /* V7M CFSR bits for MMFSR */ FIELD(V7M_CFSR, IACCVIOL, 0, 1) FIELD(V7M_CFSR, DACCVIOL, 1, 1) diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ static bool nvic_isrpending(NVICState *s) */ static inline uint32_t nvic_gprio_mask(NVICState *s) { - return ~0U << (s->prigroup + 1); + return ~0U << (s->prigroup[M_REG_NS] + 1); } /* Recompute vectpending and exception_prio */ @@ -XXX,XX +XXX,XX @@ static uint32_t nvic_readl(NVICState *s, uint32_t offset, MemTxAttrs attrs) return val; case 0xd08: /* Vector Table Offset. */ return cpu->env.v7m.vecbase[attrs.secure]; - case 0xd0c: /* Application Interrupt/Reset Control. */ - return 0xfa050000 | (s->prigroup << 8); + case 0xd0c: /* Application Interrupt/Reset Control (AIRCR) */ + val = 0xfa050000 | (s->prigroup[attrs.secure] << 8); + if (attrs.secure) { + /* s->aircr stores PRIS, BFHFNMINS, SYSRESETREQS */ + val |= cpu->env.v7m.aircr; + } else { + if (arm_feature(&cpu->env, ARM_FEATURE_V8)) { + /* BFHFNMINS is R/O from NS; other bits are RAZ/WI. If + * security isn't supported then BFHFNMINS is RAO (and + * the bit in env.v7m.aircr is always set). + */ + val |= cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK; + } + } + return val; case 0xd10: /* System Control. */ /* TODO: Implement SLEEPONEXIT. */ return 0; @@ -XXX,XX +XXX,XX @@ static void nvic_writel(NVICState *s, uint32_t offset, uint32_t value, case 0xd08: /* Vector Table Offset. */ cpu->env.v7m.vecbase[attrs.secure] = value & 0xffffff80; break; - case 0xd0c: /* Application Interrupt/Reset Control. */ - if ((value >> 16) == 0x05fa) { - if (value & 4) { - qemu_irq_pulse(s->sysresetreq); + case 0xd0c: /* Application Interrupt/Reset Control (AIRCR) */ + if ((value >> R_V7M_AIRCR_VECTKEY_SHIFT) == 0x05fa) { + if (value & R_V7M_AIRCR_SYSRESETREQ_MASK) { + if (attrs.secure || + !(cpu->env.v7m.aircr & R_V7M_AIRCR_SYSRESETREQS_MASK)) { + qemu_irq_pulse(s->sysresetreq); + } } - if (value & 2) { + if (value & R_V7M_AIRCR_VECTCLRACTIVE_MASK) { qemu_log_mask(LOG_GUEST_ERROR, "Setting VECTCLRACTIVE when not in DEBUG mode " "is UNPREDICTABLE\n"); } - if (value & 1) { + if (value & R_V7M_AIRCR_VECTRESET_MASK) { + /* NB: this bit is RES0 in v8M */ qemu_log_mask(LOG_GUEST_ERROR, "Setting VECTRESET when not in DEBUG mode " "is UNPREDICTABLE\n"); } - s->prigroup = extract32(value, 8, 3); + s->prigroup[attrs.secure] = extract32(value, + R_V7M_AIRCR_PRIGROUP_SHIFT, + R_V7M_AIRCR_PRIGROUP_LENGTH); + if (attrs.secure) { + /* These bits are only writable by secure */ + cpu->env.v7m.aircr = value & + (R_V7M_AIRCR_SYSRESETREQS_MASK | + R_V7M_AIRCR_BFHFNMINS_MASK | + R_V7M_AIRCR_PRIS_MASK); + } nvic_irq_update(s); } break; @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_nvic_security = { .fields = (VMStateField[]) { VMSTATE_STRUCT_ARRAY(sec_vectors, NVICState, NVIC_INTERNAL_VECTORS, 1, vmstate_VecInfo, VecInfo), + VMSTATE_UINT32(prigroup[M_REG_S], NVICState), VMSTATE_END_OF_LIST() } }; @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_nvic = { .fields = (VMStateField[]) { VMSTATE_STRUCT_ARRAY(vectors, NVICState, NVIC_MAX_VECTORS, 1, vmstate_VecInfo, VecInfo), - VMSTATE_UINT32(prigroup, NVICState), + VMSTATE_UINT32(prigroup[M_REG_NS], NVICState), VMSTATE_END_OF_LIST() }, .subsections = (const VMStateDescription*[]) { diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void arm_cpu_reset(CPUState *s) if (arm_feature(env, ARM_FEATURE_M_SECURITY)) { env->v7m.secure = true; + } else { + /* This bit resets to 0 if security is supported, but 1 if + * it is not. The bit is not present in v7M, but we set it + * here so we can avoid having to make checks on it conditional + * on ARM_FEATURE_V8 (we don't let the guest see the bit). + */ + env->v7m.aircr = R_V7M_AIRCR_BFHFNMINS_MASK; } /* In v7M the reset value of this bit is IMPDEF, but ARM recommends -- 2.7.4 Update the code in nvic_rettobase() so that it checks the sec_vectors[] array as well as the vectors[] array if needed. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-7-git-send-email-peter.maydell@linaro.org --- hw/intc/armv7m_nvic.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ static int nvic_pending_prio(NVICState *s) static bool nvic_rettobase(NVICState *s) { int irq, nhand = 0; + bool check_sec = arm_feature(&s->cpu->env, ARM_FEATURE_M_SECURITY); for (irq = ARMV7M_EXCP_RESET; irq < s->num_irq; irq++) { - if (s->vectors[irq].active) { + if (s->vectors[irq].active || + (check_sec && irq < NVIC_INTERNAL_VECTORS && + s->sec_vectors[irq].active)) { nhand++; if (nhand == 2) { return 0; -- 2.7.4 For v8M, the NVIC has a new set of registers per interrupt, NVIC_ITNS<n>. These determine whether the interrupt targets Secure or Non-secure state. Implement the register read/write code for these, and make them cause NVIC_IABR, NVIC_ICER, NVIC_ISER, NVIC_ICPR, NVIC_IPR and NVIC_ISPR to RAZ/WI for non-secure accesses to fields corresponding to interrupts which are configured to target secure state. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-8-git-send-email-peter.maydell@linaro.org --- include/hw/intc/armv7m_nvic.h | 3 ++ hw/intc/armv7m_nvic.c | 74 +++++++++++++++++++++++++++++++++++++++---- 2 files changed, 70 insertions(+), 7 deletions(-) diff --git a/include/hw/intc/armv7m_nvic.h b/include/hw/intc/armv7m_nvic.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/intc/armv7m_nvic.h +++ b/include/hw/intc/armv7m_nvic.h @@ -XXX,XX +XXX,XX @@ typedef struct NVICState { /* The PRIGROUP field in AIRCR is banked */ uint32_t prigroup[M_REG_NUM_BANKS]; + /* v8M NVIC_ITNS state (stored as a bool per bit) */ + bool itns[NVIC_MAX_VECTORS]; + /* The following fields are all cached state that can be recalculated * from the vectors[] and sec_vectors[] arrays and the prigroup field: * - vectpending diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ static uint32_t nvic_readl(NVICState *s, uint32_t offset, MemTxAttrs attrs) switch (offset) { case 4: /* Interrupt Control Type. */ return ((s->num_irq - NVIC_FIRST_IRQ) / 32) - 1; + case 0x380 ... 0x3bf: /* NVIC_ITNS<n> */ + { + int startvec = 32 * (offset - 0x380) + NVIC_FIRST_IRQ; + int i; + + if (!arm_feature(&cpu->env, ARM_FEATURE_V8)) { + goto bad_offset; + } + if (!attrs.secure) { + return 0; + } + val = 0; + for (i = 0; i < 32 && startvec + i < s->num_irq; i++) { + if (s->itns[startvec + i]) { + val |= (1 << i); + } + } + return val; + } case 0xd00: /* CPUID Base. */ return cpu->midr; case 0xd04: /* Interrupt Control State. */ @@ -XXX,XX +XXX,XX @@ static void nvic_writel(NVICState *s, uint32_t offset, uint32_t value, ARMCPU *cpu = s->cpu; switch (offset) { + case 0x380 ... 0x3bf: /* NVIC_ITNS<n> */ + { + int startvec = 32 * (offset - 0x380) + NVIC_FIRST_IRQ; + int i; + + if (!arm_feature(&cpu->env, ARM_FEATURE_V8)) { + goto bad_offset; + } + if (!attrs.secure) { + break; + } + for (i = 0; i < 32 && startvec + i < s->num_irq; i++) { + s->itns[startvec + i] = (value >> i) & 1; + } + nvic_irq_update(s); + break; + } case 0xd04: /* Interrupt Control State. */ if (value & (1 << 31)) { armv7m_nvic_set_pending(s, ARMV7M_EXCP_NMI); @@ -XXX,XX +XXX,XX @@ static MemTxResult nvic_sysreg_read(void *opaque, hwaddr addr, startvec = offset - 0x180 + NVIC_FIRST_IRQ; /* vector # */ for (i = 0, end = size * 8; i < end && startvec + i < s->num_irq; i++) { - if (s->vectors[startvec + i].enabled) { + if (s->vectors[startvec + i].enabled && + (attrs.secure || s->itns[startvec + i])) { val |= (1 << i); } } @@ -XXX,XX +XXX,XX @@ static MemTxResult nvic_sysreg_read(void *opaque, hwaddr addr, val = 0; startvec = offset - 0x280 + NVIC_FIRST_IRQ; /* vector # */ for (i = 0, end = size * 8; i < end && startvec + i < s->num_irq; i++) { - if (s->vectors[startvec + i].pending) { + if (s->vectors[startvec + i].pending && + (attrs.secure || s->itns[startvec + i])) { val |= (1 << i); } } @@ -XXX,XX +XXX,XX @@ static MemTxResult nvic_sysreg_read(void *opaque, hwaddr addr, startvec = offset - 0x300 + NVIC_FIRST_IRQ; /* vector # */ for (i = 0, end = size * 8; i < end && startvec + i < s->num_irq; i++) { - if (s->vectors[startvec + i].active) { + if (s->vectors[startvec + i].active && + (attrs.secure || s->itns[startvec + i])) { val |= (1 << i); } } @@ -XXX,XX +XXX,XX @@ static MemTxResult nvic_sysreg_read(void *opaque, hwaddr addr, startvec = offset - 0x400 + NVIC_FIRST_IRQ; /* vector # */ for (i = 0; i < size && startvec + i < s->num_irq; i++) { - val |= s->vectors[startvec + i].prio << (8 * i); + if (attrs.secure || s->itns[startvec + i]) { + val |= s->vectors[startvec + i].prio << (8 * i); + } } break; case 0xd18 ... 0xd23: /* System Handler Priority. */ @@ -XXX,XX +XXX,XX @@ static MemTxResult nvic_sysreg_write(void *opaque, hwaddr addr, startvec = 8 * (offset - 0x180) + NVIC_FIRST_IRQ; for (i = 0, end = size * 8; i < end && startvec + i < s->num_irq; i++) { - if (value & (1 << i)) { + if (value & (1 << i) && + (attrs.secure || s->itns[startvec + i])) { s->vectors[startvec + i].enabled = setval; } } @@ -XXX,XX +XXX,XX @@ static MemTxResult nvic_sysreg_write(void *opaque, hwaddr addr, startvec = 8 * (offset - 0x280) + NVIC_FIRST_IRQ; /* vector # */ for (i = 0, end = size * 8; i < end && startvec + i < s->num_irq; i++) { - if (value & (1 << i)) { + if (value & (1 << i) && + (attrs.secure || s->itns[startvec + i])) { s->vectors[startvec + i].pending = setval; } } @@ -XXX,XX +XXX,XX @@ static MemTxResult nvic_sysreg_write(void *opaque, hwaddr addr, startvec = 8 * (offset - 0x400) + NVIC_FIRST_IRQ; /* vector # */ for (i = 0; i < size && startvec + i < s->num_irq; i++) { - set_prio(s, startvec + i, (value >> (i * 8)) & 0xff); + if (attrs.secure || s->itns[startvec + i]) { + set_prio(s, startvec + i, (value >> (i * 8)) & 0xff); + } } nvic_irq_update(s); return MEMTX_OK; @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_nvic_security = { VMSTATE_STRUCT_ARRAY(sec_vectors, NVICState, NVIC_INTERNAL_VECTORS, 1, vmstate_VecInfo, VecInfo), VMSTATE_UINT32(prigroup[M_REG_S], NVICState), + VMSTATE_BOOL_ARRAY(itns, NVICState, NVIC_MAX_VECTORS), VMSTATE_END_OF_LIST() } }; @@ -XXX,XX +XXX,XX @@ static void armv7m_nvic_reset(DeviceState *dev) s->vectpending = 0; s->vectpending_is_s_banked = false; s->vectpending_prio = NVIC_NOEXC_PRIO; + + if (arm_feature(&s->cpu->env, ARM_FEATURE_M_SECURITY)) { + memset(s->itns, 0, sizeof(s->itns)); + } else { + /* This state is constant and not guest accessible in a non-security + * NVIC; we set the bits to true to avoid having to do a feature + * bit check in the NVIC enable/pend/etc register accessors. + */ + int i; + + for (i = NVIC_FIRST_IRQ; i < ARRAY_SIZE(s->itns); i++) { + s->itns[i] = true; + } + } } static void nvic_systick_trigger(void *opaque, int n, int level) -- 2.7.4 Update the nvic_recompute_state() code to handle the security extension and its associated banked registers. Code that uses the resulting cached state (ie the irq acknowledge and complete code) will be updated in a later commit. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-9-git-send-email-peter.maydell@linaro.org --- hw/intc/armv7m_nvic.c | 151 ++++++++++++++++++++++++++++++++++++++++++++++++-- hw/intc/trace-events | 1 + 2 files changed, 147 insertions(+), 5 deletions(-) diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ * (higher than the highest possible priority value) */ #define NVIC_NOEXC_PRIO 0x100 +/* Maximum priority of non-secure exceptions when AIRCR.PRIS is set */ +#define NVIC_NS_PRIO_LIMIT 0x80 static const uint8_t nvic_id[] = { 0x00, 0xb0, 0x1b, 0x00, 0x0d, 0xe0, 0x05, 0xb1 @@ -XXX,XX +XXX,XX @@ static bool nvic_isrpending(NVICState *s) return false; } +static bool exc_is_banked(int exc) +{ + /* Return true if this is one of the limited set of exceptions which + * are banked (and thus have state in sec_vectors[]) + */ + return exc == ARMV7M_EXCP_HARD || + exc == ARMV7M_EXCP_MEM || + exc == ARMV7M_EXCP_USAGE || + exc == ARMV7M_EXCP_SVC || + exc == ARMV7M_EXCP_PENDSV || + exc == ARMV7M_EXCP_SYSTICK; +} + /* Return a mask word which clears the subpriority bits from * a priority value for an M-profile exception, leaving only * the group priority. */ -static inline uint32_t nvic_gprio_mask(NVICState *s) +static inline uint32_t nvic_gprio_mask(NVICState *s, bool secure) +{ + return ~0U << (s->prigroup[secure] + 1); +} + +static bool exc_targets_secure(NVICState *s, int exc) +{ + /* Return true if this non-banked exception targets Secure state. */ + if (!arm_feature(&s->cpu->env, ARM_FEATURE_M_SECURITY)) { + return false; + } + + if (exc >= NVIC_FIRST_IRQ) { + return !s->itns[exc]; + } + + /* Function shouldn't be called for banked exceptions. */ + assert(!exc_is_banked(exc)); + + switch (exc) { + case ARMV7M_EXCP_NMI: + case ARMV7M_EXCP_BUS: + return !(s->cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK); + case ARMV7M_EXCP_SECURE: + return true; + case ARMV7M_EXCP_DEBUG: + /* TODO: controlled by DEMCR.SDME, which we don't yet implement */ + return false; + default: + /* reset, and reserved (unused) low exception numbers. + * We'll get called by code that loops through all the exception + * numbers, but it doesn't matter what we return here as these + * non-existent exceptions will never be pended or active. + */ + return true; + } +} + +static int exc_group_prio(NVICState *s, int rawprio, bool targets_secure) +{ + /* Return the group priority for this exception, given its raw + * (group-and-subgroup) priority value and whether it is targeting + * secure state or not. + */ + if (rawprio < 0) { + return rawprio; + } + rawprio &= nvic_gprio_mask(s, targets_secure); + /* AIRCR.PRIS causes us to squash all NS priorities into the + * lower half of the total range + */ + if (!targets_secure && + (s->cpu->env.v7m.aircr & R_V7M_AIRCR_PRIS_MASK)) { + rawprio = (rawprio >> 1) + NVIC_NS_PRIO_LIMIT; + } + return rawprio; +} + +/* Recompute vectpending and exception_prio for a CPU which implements + * the Security extension + */ +static void nvic_recompute_state_secure(NVICState *s) { - return ~0U << (s->prigroup[M_REG_NS] + 1); + int i, bank; + int pend_prio = NVIC_NOEXC_PRIO; + int active_prio = NVIC_NOEXC_PRIO; + int pend_irq = 0; + bool pending_is_s_banked = false; + + /* R_CQRV: precedence is by: + * - lowest group priority; if both the same then + * - lowest subpriority; if both the same then + * - lowest exception number; if both the same (ie banked) then + * - secure exception takes precedence + * Compare pseudocode RawExecutionPriority. + * Annoyingly, now we have two prigroup values (for S and NS) + * we can't do the loop comparison on raw priority values. + */ + for (i = 1; i < s->num_irq; i++) { + for (bank = M_REG_S; bank >= M_REG_NS; bank--) { + VecInfo *vec; + int prio; + bool targets_secure; + + if (bank == M_REG_S) { + if (!exc_is_banked(i)) { + continue; + } + vec = &s->sec_vectors[i]; + targets_secure = true; + } else { + vec = &s->vectors[i]; + targets_secure = !exc_is_banked(i) && exc_targets_secure(s, i); + } + + prio = exc_group_prio(s, vec->prio, targets_secure); + if (vec->enabled && vec->pending && prio < pend_prio) { + pend_prio = prio; + pend_irq = i; + pending_is_s_banked = (bank == M_REG_S); + } + if (vec->active && prio < active_prio) { + active_prio = prio; + } + } + } + + s->vectpending_is_s_banked = pending_is_s_banked; + s->vectpending = pend_irq; + s->vectpending_prio = pend_prio; + s->exception_prio = active_prio; + + trace_nvic_recompute_state_secure(s->vectpending, + s->vectpending_is_s_banked, + s->vectpending_prio, + s->exception_prio); } /* Recompute vectpending and exception_prio */ @@ -XXX,XX +XXX,XX @@ static void nvic_recompute_state(NVICState *s) int active_prio = NVIC_NOEXC_PRIO; int pend_irq = 0; + /* In theory we could write one function that handled both + * the "security extension present" and "not present"; however + * the security related changes significantly complicate the + * recomputation just by themselves and mixing both cases together + * would be even worse, so we retain a separate non-secure-only + * version for CPUs which don't implement the security extension. + */ + if (arm_feature(&s->cpu->env, ARM_FEATURE_M_SECURITY)) { + nvic_recompute_state_secure(s); + return; + } + for (i = 1; i < s->num_irq; i++) { VecInfo *vec = &s->vectors[i]; @@ -XXX,XX +XXX,XX @@ static void nvic_recompute_state(NVICState *s) } if (active_prio > 0) { - active_prio &= nvic_gprio_mask(s); + active_prio &= nvic_gprio_mask(s, false); } if (pend_prio > 0) { - pend_prio &= nvic_gprio_mask(s); + pend_prio &= nvic_gprio_mask(s, false); } s->vectpending = pend_irq; @@ -XXX,XX +XXX,XX @@ static inline int nvic_exec_prio(NVICState *s) } else if (env->v7m.primask[env->v7m.secure]) { running = 0; } else if (env->v7m.basepri[env->v7m.secure] > 0) { - running = env->v7m.basepri[env->v7m.secure] & nvic_gprio_mask(s); + running = env->v7m.basepri[env->v7m.secure] & + nvic_gprio_mask(s, env->v7m.secure); } else { running = NVIC_NOEXC_PRIO; /* lower than any possible priority */ } diff --git a/hw/intc/trace-events b/hw/intc/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/trace-events +++ b/hw/intc/trace-events @@ -XXX,XX +XXX,XX @@ gicv3_redist_send_sgi(uint32_t cpu, int irq) "GICv3 redistributor 0x%x pending S # hw/intc/armv7m_nvic.c nvic_recompute_state(int vectpending, int vectpending_prio, int exception_prio) "NVIC state recomputed: vectpending %d vectpending_prio %d exception_prio %d" +nvic_recompute_state_secure(int vectpending, bool vectpending_is_s_banked, int vectpending_prio, int exception_prio) "NVIC state recomputed: vectpending %d is_s_banked %d vectpending_prio %d exception_prio %d" nvic_set_prio(int irq, uint8_t prio) "NVIC set irq %d priority %d" nvic_irq_update(int vectpending, int pendprio, int exception_prio, int level) "NVIC vectpending %d pending prio %d exception_prio %d: setting irq line to %d" nvic_escalate_prio(int irq, int irqprio, int runprio) "NVIC escalating irq %d to HardFault: insufficient priority %d >= %d" -- 2.7.4 Make the armv7m_nvic_set_pending() and armv7m_nvic_clear_pending() functions take a bool indicating whether to pend the secure or non-secure version of a banked interrupt, and update the callsites accordingly. In most callsites we can simply pass the correct security state in; in a couple of cases we use TODO comments to indicate that we will return the code in a subsequent commit. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-10-git-send-email-peter.maydell@linaro.org --- target/arm/cpu.h | 14 ++++++++++- hw/intc/armv7m_nvic.c | 64 ++++++++++++++++++++++++++++++++++++++------------- target/arm/helper.c | 24 +++++++++++-------- hw/intc/trace-events | 4 ++-- 4 files changed, 77 insertions(+), 29 deletions(-) diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ static inline bool armv7m_nvic_can_take_pending_exception(void *opaque) return true; } #endif -void armv7m_nvic_set_pending(void *opaque, int irq); +/** + * armv7m_nvic_set_pending: mark the specified exception as pending + * @opaque: the NVIC + * @irq: the exception number to mark pending + * @secure: false for non-banked exceptions or for the nonsecure + * version of a banked exception, true for the secure version of a banked + * exception. + * + * Marks the specified exception as pending. Note that we will assert() + * if @secure is true and @irq does not specify one of the fixed set + * of architecturally banked exceptions. + */ +void armv7m_nvic_set_pending(void *opaque, int irq, bool secure); void armv7m_nvic_acknowledge_irq(void *opaque); /** * armv7m_nvic_complete_irq: complete specified interrupt or exception diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ static void nvic_irq_update(NVICState *s) qemu_set_irq(s->excpout, lvl); } -static void armv7m_nvic_clear_pending(void *opaque, int irq) +/** + * armv7m_nvic_clear_pending: mark the specified exception as not pending + * @opaque: the NVIC + * @irq: the exception number to mark as not pending + * @secure: false for non-banked exceptions or for the nonsecure + * version of a banked exception, true for the secure version of a banked + * exception. + * + * Marks the specified exception as not pending. Note that we will assert() + * if @secure is true and @irq does not specify one of the fixed set + * of architecturally banked exceptions. + */ +static void armv7m_nvic_clear_pending(void *opaque, int irq, bool secure) { NVICState *s = (NVICState *)opaque; VecInfo *vec; assert(irq > ARMV7M_EXCP_RESET && irq < s->num_irq); - vec = &s->vectors[irq]; - trace_nvic_clear_pending(irq, vec->enabled, vec->prio); + if (secure) { + assert(exc_is_banked(irq)); + vec = &s->sec_vectors[irq]; + } else { + vec = &s->vectors[irq]; + } + trace_nvic_clear_pending(irq, secure, vec->enabled, vec->prio); if (vec->pending) { vec->pending = 0; nvic_irq_update(s); } } -void armv7m_nvic_set_pending(void *opaque, int irq) +void armv7m_nvic_set_pending(void *opaque, int irq, bool secure) { NVICState *s = (NVICState *)opaque; + bool banked = exc_is_banked(irq); VecInfo *vec; assert(irq > ARMV7M_EXCP_RESET && irq < s->num_irq); + assert(!secure || banked); - vec = &s->vectors[irq]; - trace_nvic_set_pending(irq, vec->enabled, vec->prio); + vec = (banked && secure) ? &s->sec_vectors[irq] : &s->vectors[irq]; + trace_nvic_set_pending(irq, secure, vec->enabled, vec->prio); if (irq >= ARMV7M_EXCP_HARD && irq < ARMV7M_EXCP_PENDSV) { /* If a synchronous exception is pending then it may be @@ -XXX,XX +XXX,XX @@ void armv7m_nvic_set_pending(void *opaque, int irq) "(current priority %d)\n", irq, running); } - /* We can do the escalation, so we take HardFault instead */ + /* We can do the escalation, so we take HardFault instead. + * If BFHFNMINS is set then we escalate to the banked HF for + * the target security state of the original exception; otherwise + * we take a Secure HardFault. + */ irq = ARMV7M_EXCP_HARD; - vec = &s->vectors[irq]; + if (arm_feature(&s->cpu->env, ARM_FEATURE_M_SECURITY) && + (secure || + !(s->cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK))) { + vec = &s->sec_vectors[irq]; + } else { + vec = &s->vectors[irq]; + } + /* HF may be banked but there is only one shared HFSR */ s->cpu->env.v7m.hfsr |= R_V7M_HFSR_FORCED_MASK; } } @@ -XXX,XX +XXX,XX @@ static void set_irq_level(void *opaque, int n, int level) if (level != vec->level) { vec->level = level; if (level) { - armv7m_nvic_set_pending(s, n); + armv7m_nvic_set_pending(s, n, false); } } } @@ -XXX,XX +XXX,XX @@ static void nvic_writel(NVICState *s, uint32_t offset, uint32_t value, } case 0xd04: /* Interrupt Control State. */ if (value & (1 << 31)) { - armv7m_nvic_set_pending(s, ARMV7M_EXCP_NMI); + armv7m_nvic_set_pending(s, ARMV7M_EXCP_NMI, false); } if (value & (1 << 28)) { - armv7m_nvic_set_pending(s, ARMV7M_EXCP_PENDSV); + armv7m_nvic_set_pending(s, ARMV7M_EXCP_PENDSV, attrs.secure); } else if (value & (1 << 27)) { - armv7m_nvic_clear_pending(s, ARMV7M_EXCP_PENDSV); + armv7m_nvic_clear_pending(s, ARMV7M_EXCP_PENDSV, attrs.secure); } if (value & (1 << 26)) { - armv7m_nvic_set_pending(s, ARMV7M_EXCP_SYSTICK); + armv7m_nvic_set_pending(s, ARMV7M_EXCP_SYSTICK, attrs.secure); } else if (value & (1 << 25)) { - armv7m_nvic_clear_pending(s, ARMV7M_EXCP_SYSTICK); + armv7m_nvic_clear_pending(s, ARMV7M_EXCP_SYSTICK, attrs.secure); } break; case 0xd08: /* Vector Table Offset. */ @@ -XXX,XX +XXX,XX @@ static void nvic_writel(NVICState *s, uint32_t offset, uint32_t value, { int excnum = (value & 0x1ff) + NVIC_FIRST_IRQ; if (excnum < s->num_irq) { - armv7m_nvic_set_pending(s, excnum); + armv7m_nvic_set_pending(s, excnum, false); } break; } @@ -XXX,XX +XXX,XX @@ static void nvic_systick_trigger(void *opaque, int n, int level) /* SysTick just asked us to pend its exception. * (This is different from an external interrupt line's * behaviour.) + * TODO: when we implement the banked systicks we must make + * this pend the correct banked exception. */ - armv7m_nvic_set_pending(s, ARMV7M_EXCP_SYSTICK); + armv7m_nvic_set_pending(s, ARMV7M_EXCP_SYSTICK, false); } } diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) * stack, directly take a usage fault on the current stack. */ env->v7m.cfsr[env->v7m.secure] |= R_V7M_CFSR_INVPC_MASK; - armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_USAGE); + armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_USAGE, env->v7m.secure); v7m_exception_taken(cpu, excret); qemu_log_mask(CPU_LOG_INT, "...taking UsageFault on existing " "stackframe: failed exception return integrity check\n"); @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) * exception return excret specified then this is a UsageFault. */ if (return_to_handler != arm_v7m_is_handler_mode(env)) { - /* Take an INVPC UsageFault by pushing the stack again. */ - armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_USAGE); + /* Take an INVPC UsageFault by pushing the stack again. + * TODO: the v8M version of this code should target the + * background state for this exception. + */ + armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_USAGE, false); env->v7m.cfsr[env->v7m.secure] |= R_V7M_CFSR_INVPC_MASK; v7m_push_stack(cpu); v7m_exception_taken(cpu, excret); @@ -XXX,XX +XXX,XX @@ void arm_v7m_cpu_do_interrupt(CPUState *cs) handle it. */ switch (cs->exception_index) { case EXCP_UDEF: - armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_USAGE); + armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_USAGE, env->v7m.secure); env->v7m.cfsr[env->v7m.secure] |= R_V7M_CFSR_UNDEFINSTR_MASK; break; case EXCP_NOCP: - armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_USAGE); + armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_USAGE, env->v7m.secure); env->v7m.cfsr[env->v7m.secure] |= R_V7M_CFSR_NOCP_MASK; break; case EXCP_INVSTATE: - armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_USAGE); + armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_USAGE, env->v7m.secure); env->v7m.cfsr[env->v7m.secure] |= R_V7M_CFSR_INVSTATE_MASK; break; case EXCP_SWI: /* The PC already points to the next instruction. */ - armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_SVC); + armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_SVC, env->v7m.secure); break; case EXCP_PREFETCH_ABORT: case EXCP_DATA_ABORT: @@ -XXX,XX +XXX,XX @@ void arm_v7m_cpu_do_interrupt(CPUState *cs) env->v7m.bfar); break; } - armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_BUS); + armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_BUS, false); break; default: /* All other FSR values are either MPU faults or "can't happen @@ -XXX,XX +XXX,XX @@ void arm_v7m_cpu_do_interrupt(CPUState *cs) env->v7m.mmfar[env->v7m.secure]); break; } - armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_MEM); + armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_MEM, + env->v7m.secure); break; } break; @@ -XXX,XX +XXX,XX @@ void arm_v7m_cpu_do_interrupt(CPUState *cs) return; } } - armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_DEBUG); + armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_DEBUG, false); break; case EXCP_IRQ: break; diff --git a/hw/intc/trace-events b/hw/intc/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/trace-events +++ b/hw/intc/trace-events @@ -XXX,XX +XXX,XX @@ nvic_set_prio(int irq, uint8_t prio) "NVIC set irq %d priority %d" nvic_irq_update(int vectpending, int pendprio, int exception_prio, int level) "NVIC vectpending %d pending prio %d exception_prio %d: setting irq line to %d" nvic_escalate_prio(int irq, int irqprio, int runprio) "NVIC escalating irq %d to HardFault: insufficient priority %d >= %d" nvic_escalate_disabled(int irq) "NVIC escalating irq %d to HardFault: disabled" -nvic_set_pending(int irq, int en, int prio) "NVIC set pending irq %d (enabled: %d priority %d)" -nvic_clear_pending(int irq, int en, int prio) "NVIC clear pending irq %d (enabled: %d priority %d)" +nvic_set_pending(int irq, bool secure, int en, int prio) "NVIC set pending irq %d secure-bank %d (enabled: %d priority %d)" +nvic_clear_pending(int irq, bool secure, int en, int prio) "NVIC clear pending irq %d secure-bank %d (enabled: %d priority %d)" nvic_set_pending_level(int irq) "NVIC set pending: irq %d higher prio than vectpending: setting irq line to 1" nvic_acknowledge_irq(int irq, int prio) "NVIC acknowledge IRQ: %d now active (prio %d)" nvic_complete_irq(int irq) "NVIC complete IRQ %d" -- 2.7.4 Make the set_prio() function take a bool indicating whether to pend the secure or non-secure version of a banked interrupt, and use this to implement the correct banking semantics for the SHPR registers. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-11-git-send-email-peter.maydell@linaro.org --- hw/intc/armv7m_nvic.c | 96 ++++++++++++++++++++++++++++++++++++++++++++++----- hw/intc/trace-events | 2 +- 2 files changed, 88 insertions(+), 10 deletions(-) diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ int armv7m_nvic_raw_execution_priority(void *opaque) return s->exception_prio; } -/* caller must call nvic_irq_update() after this */ -static void set_prio(NVICState *s, unsigned irq, uint8_t prio) +/* caller must call nvic_irq_update() after this. + * secure indicates the bank to use for banked exceptions (we assert if + * we are passed secure=true for a non-banked exception). + */ +static void set_prio(NVICState *s, unsigned irq, bool secure, uint8_t prio) { assert(irq > ARMV7M_EXCP_NMI); /* only use for configurable prios */ assert(irq < s->num_irq); - s->vectors[irq].prio = prio; + if (secure) { + assert(exc_is_banked(irq)); + s->sec_vectors[irq].prio = prio; + } else { + s->vectors[irq].prio = prio; + } + + trace_nvic_set_prio(irq, secure, prio); +} + +/* Return the current raw priority register value. + * secure indicates the bank to use for banked exceptions (we assert if + * we are passed secure=true for a non-banked exception). + */ +static int get_prio(NVICState *s, unsigned irq, bool secure) +{ + assert(irq > ARMV7M_EXCP_NMI); /* only use for configurable prios */ + assert(irq < s->num_irq); - trace_nvic_set_prio(irq, prio); + if (secure) { + assert(exc_is_banked(irq)); + return s->sec_vectors[irq].prio; + } else { + return s->vectors[irq].prio; + } } /* Recompute state and assert irq line accordingly. @@ -XXX,XX +XXX,XX @@ static bool nvic_user_access_ok(NVICState *s, hwaddr offset, MemTxAttrs attrs) } } +static int shpr_bank(NVICState *s, int exc, MemTxAttrs attrs) +{ + /* Behaviour for the SHPR register field for this exception: + * return M_REG_NS to use the nonsecure vector (including for + * non-banked exceptions), M_REG_S for the secure version of + * a banked exception, and -1 if this field should RAZ/WI. + */ + switch (exc) { + case ARMV7M_EXCP_MEM: + case ARMV7M_EXCP_USAGE: + case ARMV7M_EXCP_SVC: + case ARMV7M_EXCP_PENDSV: + case ARMV7M_EXCP_SYSTICK: + /* Banked exceptions */ + return attrs.secure; + case ARMV7M_EXCP_BUS: + /* Not banked, RAZ/WI from nonsecure if BFHFNMINS is zero */ + if (!attrs.secure && + !(s->cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK)) { + return -1; + } + return M_REG_NS; + case ARMV7M_EXCP_SECURE: + /* Not banked, RAZ/WI from nonsecure */ + if (!attrs.secure) { + return -1; + } + return M_REG_NS; + case ARMV7M_EXCP_DEBUG: + /* Not banked. TODO should RAZ/WI if DEMCR.SDME is set */ + return M_REG_NS; + case 8 ... 10: + case 13: + /* RES0 */ + return -1; + default: + /* Not reachable due to decode of SHPR register addresses */ + g_assert_not_reached(); + } +} + static MemTxResult nvic_sysreg_read(void *opaque, hwaddr addr, uint64_t *data, unsigned size, MemTxAttrs attrs) @@ -XXX,XX +XXX,XX @@ static MemTxResult nvic_sysreg_read(void *opaque, hwaddr addr, } } break; - case 0xd18 ... 0xd23: /* System Handler Priority. */ + case 0xd18 ... 0xd23: /* System Handler Priority (SHPR1, SHPR2, SHPR3) */ val = 0; for (i = 0; i < size; i++) { - val |= s->vectors[(offset - 0xd14) + i].prio << (i * 8); + unsigned hdlidx = (offset - 0xd14) + i; + int sbank = shpr_bank(s, hdlidx, attrs); + + if (sbank < 0) { + continue; + } + val = deposit32(val, i * 8, 8, get_prio(s, hdlidx, sbank)); } break; case 0xfe0 ... 0xfff: /* ID. */ @@ -XXX,XX +XXX,XX @@ static MemTxResult nvic_sysreg_write(void *opaque, hwaddr addr, for (i = 0; i < size && startvec + i < s->num_irq; i++) { if (attrs.secure || s->itns[startvec + i]) { - set_prio(s, startvec + i, (value >> (i * 8)) & 0xff); + set_prio(s, startvec + i, false, (value >> (i * 8)) & 0xff); } } nvic_irq_update(s); return MEMTX_OK; - case 0xd18 ... 0xd23: /* System Handler Priority. */ + case 0xd18 ... 0xd23: /* System Handler Priority (SHPR1, SHPR2, SHPR3) */ for (i = 0; i < size; i++) { unsigned hdlidx = (offset - 0xd14) + i; - set_prio(s, hdlidx, (value >> (i * 8)) & 0xff); + int newprio = extract32(value, i * 8, 8); + int sbank = shpr_bank(s, hdlidx, attrs); + + if (sbank < 0) { + continue; + } + set_prio(s, hdlidx, sbank, newprio); } nvic_irq_update(s); return MEMTX_OK; diff --git a/hw/intc/trace-events b/hw/intc/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/trace-events +++ b/hw/intc/trace-events @@ -XXX,XX +XXX,XX @@ gicv3_redist_send_sgi(uint32_t cpu, int irq) "GICv3 redistributor 0x%x pending S # hw/intc/armv7m_nvic.c nvic_recompute_state(int vectpending, int vectpending_prio, int exception_prio) "NVIC state recomputed: vectpending %d vectpending_prio %d exception_prio %d" nvic_recompute_state_secure(int vectpending, bool vectpending_is_s_banked, int vectpending_prio, int exception_prio) "NVIC state recomputed: vectpending %d is_s_banked %d vectpending_prio %d exception_prio %d" -nvic_set_prio(int irq, uint8_t prio) "NVIC set irq %d priority %d" +nvic_set_prio(int irq, bool secure, uint8_t prio) "NVIC set irq %d secure-bank %d priority %d" nvic_irq_update(int vectpending, int pendprio, int exception_prio, int level) "NVIC vectpending %d pending prio %d exception_prio %d: setting irq line to %d" nvic_escalate_prio(int irq, int irqprio, int runprio) "NVIC escalating irq %d to HardFault: insufficient priority %d >= %d" nvic_escalate_disabled(int irq) "NVIC escalating irq %d to HardFault: disabled" -- 2.7.4 In armv7m_nvic_set_pending() we have to compare the priority of an exception against the execution priority to decide whether it needs to be escalated to HardFault. In the specification this is a comparison against the exception's group priority; for v7M we implemented it as a comparison against the raw exception priority because the two comparisons will always give the same answer. For v8M the existence of AIRCR.PRIS and the possibility of different PRIGROUP values for secure and nonsecure exceptions means we need to explicitly calculate the vector's group priority for this check. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-12-git-send-email-peter.maydell@linaro.org --- hw/intc/armv7m_nvic.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ void armv7m_nvic_set_pending(void *opaque, int irq, bool secure) int running = nvic_exec_prio(s); bool escalate = false; - if (vec->prio >= running) { + if (exc_group_prio(s, vec->prio, secure) >= running) { trace_nvic_escalate_prio(irq, vec->prio, running); escalate = true; } else if (!vec->enabled) { -- 2.7.4 When escalating to HardFault, we must go into Lockup if we can't take the synchronous HardFault because the current execution priority is already at or below the priority of HardFault. In v7M HF is always priority -1 so a simple < 0 comparison sufficed; in v8M the priority of HardFault can vary depending on whether it is a Secure or NonSecure HardFault, so we must check against the priority of the HardFault exception vector we're about to use. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-13-git-send-email-peter.maydell@linaro.org --- hw/intc/armv7m_nvic.c | 23 ++++++++++++----------- 1 file changed, 12 insertions(+), 11 deletions(-) diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ void armv7m_nvic_set_pending(void *opaque, int irq, bool secure) } if (escalate) { - if (running < 0) { - /* We want to escalate to HardFault but we can't take a - * synchronous HardFault at this point either. This is a - * Lockup condition due to a guest bug. We don't model - * Lockup, so report via cpu_abort() instead. - */ - cpu_abort(&s->cpu->parent_obj, - "Lockup: can't escalate %d to HardFault " - "(current priority %d)\n", irq, running); - } - /* We can do the escalation, so we take HardFault instead. + /* We need to escalate this exception to a synchronous HardFault. * If BFHFNMINS is set then we escalate to the banked HF for * the target security state of the original exception; otherwise * we take a Secure HardFault. @@ -XXX,XX +XXX,XX @@ void armv7m_nvic_set_pending(void *opaque, int irq, bool secure) } else { vec = &s->vectors[irq]; } + if (running <= vec->prio) { + /* We want to escalate to HardFault but we can't take the + * synchronous HardFault at this point either. This is a + * Lockup condition due to a guest bug. We don't model + * Lockup, so report via cpu_abort() instead. + */ + cpu_abort(&s->cpu->parent_obj, + "Lockup: can't escalate %d to HardFault " + "(current priority %d)\n", irq, running); + } + /* HF may be banked but there is only one shared HFSR */ s->cpu->env.v7m.hfsr |= R_V7M_HFSR_FORCED_MASK; } -- 2.7.4 In v7M, the fixed-priority exceptions are: Reset: -3 NMI: -2 HardFault: -1 In v8M, this changes because Secure HardFault may need to be prioritised above NMI: Reset: -4 Secure HardFault if AIRCR.BFHFNMINS == 1: -3 NMI: -2 Secure HardFault if AIRCR.BFHFNMINS == 0: -1 NonSecure HardFault: -1 Make these changes, including support for changing the priority of Secure HardFault as AIRCR.BFHFNMINS changes. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-14-git-send-email-peter.maydell@linaro.org --- hw/intc/armv7m_nvic.c | 22 +++++++++++++++++++--- 1 file changed, 19 insertions(+), 3 deletions(-) diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ static void nvic_writel(NVICState *s, uint32_t offset, uint32_t value, (R_V7M_AIRCR_SYSRESETREQS_MASK | R_V7M_AIRCR_BFHFNMINS_MASK | R_V7M_AIRCR_PRIS_MASK); + /* BFHFNMINS changes the priority of Secure HardFault */ + if (cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK) { + s->sec_vectors[ARMV7M_EXCP_HARD].prio = -3; + } else { + s->sec_vectors[ARMV7M_EXCP_HARD].prio = -1; + } } nvic_irq_update(s); } @@ -XXX,XX +XXX,XX @@ static int nvic_post_load(void *opaque, int version_id) { NVICState *s = opaque; unsigned i; + int resetprio; /* Check for out of range priority settings */ - if (s->vectors[ARMV7M_EXCP_RESET].prio != -3 || + resetprio = arm_feature(&s->cpu->env, ARM_FEATURE_V8) ? -4 : -3; + + if (s->vectors[ARMV7M_EXCP_RESET].prio != resetprio || s->vectors[ARMV7M_EXCP_NMI].prio != -2 || s->vectors[ARMV7M_EXCP_HARD].prio != -1) { return 1; @@ -XXX,XX +XXX,XX @@ static int nvic_security_post_load(void *opaque, int version_id) int i; /* Check for out of range priority settings */ - if (s->sec_vectors[ARMV7M_EXCP_HARD].prio != -1) { + if (s->sec_vectors[ARMV7M_EXCP_HARD].prio != -1 + && s->sec_vectors[ARMV7M_EXCP_HARD].prio != -3) { + /* We can't cross-check against AIRCR.BFHFNMINS as we don't know + * if the CPU state has been migrated yet; a mismatch won't + * cause the emulation to blow up, though. + */ return 1; } for (i = ARMV7M_EXCP_MEM; i < ARRAY_SIZE(s->sec_vectors); i++) { @@ -XXX,XX +XXX,XX @@ static Property props_nvic[] = { static void armv7m_nvic_reset(DeviceState *dev) { + int resetprio; NVICState *s = NVIC(dev); s->vectors[ARMV7M_EXCP_NMI].enabled = 1; @@ -XXX,XX +XXX,XX @@ static void armv7m_nvic_reset(DeviceState *dev) s->vectors[ARMV7M_EXCP_PENDSV].enabled = 1; s->vectors[ARMV7M_EXCP_SYSTICK].enabled = 1; - s->vectors[ARMV7M_EXCP_RESET].prio = -3; + resetprio = arm_feature(&s->cpu->env, ARM_FEATURE_V8) ? -4 : -3; + s->vectors[ARMV7M_EXCP_RESET].prio = resetprio; s->vectors[ARMV7M_EXCP_NMI].prio = -2; s->vectors[ARMV7M_EXCP_HARD].prio = -1; -- 2.7.4 If AIRCR.BFHFNMINS is clear, then although NonSecure HardFault can still be pended via SHCSR.HARDFAULTPENDED it mustn't actually preempt execution. The simple way to achieve this is to clear the enable bit for it, since the enable bit isn't guest visible. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-15-git-send-email-peter.maydell@linaro.org --- hw/intc/armv7m_nvic.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ static void nvic_writel(NVICState *s, uint32_t offset, uint32_t value, (R_V7M_AIRCR_SYSRESETREQS_MASK | R_V7M_AIRCR_BFHFNMINS_MASK | R_V7M_AIRCR_PRIS_MASK); - /* BFHFNMINS changes the priority of Secure HardFault */ + /* BFHFNMINS changes the priority of Secure HardFault, and + * allows a pending Non-secure HardFault to preempt (which + * we implement by marking it enabled). + */ if (cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK) { s->sec_vectors[ARMV7M_EXCP_HARD].prio = -3; + s->vectors[ARMV7M_EXCP_HARD].enabled = 1; } else { s->sec_vectors[ARMV7M_EXCP_HARD].prio = -1; + s->vectors[ARMV7M_EXCP_HARD].enabled = 0; } } nvic_irq_update(s); @@ -XXX,XX +XXX,XX @@ static void armv7m_nvic_reset(DeviceState *dev) NVICState *s = NVIC(dev); s->vectors[ARMV7M_EXCP_NMI].enabled = 1; - s->vectors[ARMV7M_EXCP_HARD].enabled = 1; /* MEM, BUS, and USAGE are enabled through * the System Handler Control register */ @@ -XXX,XX +XXX,XX @@ static void armv7m_nvic_reset(DeviceState *dev) /* AIRCR.BFHFNMINS resets to 0 so Secure HF is priority -1 (R_CMTC) */ s->sec_vectors[ARMV7M_EXCP_HARD].prio = -1; + /* If AIRCR.BFHFNMINS is 0 then NS HF is (effectively) disabled */ + s->vectors[ARMV7M_EXCP_HARD].enabled = 0; + } else { + s->vectors[ARMV7M_EXCP_HARD].enabled = 1; } /* Strictly speaking the reset handler should be enabled. -- 2.7.4 Update nvic_exec_prio() to support the v8M changes: * BASEPRI, FAULTMASK and PRIMASK are all banked * AIRCR.PRIS can affect NS priorities * AIRCR.BFHFNMINS affects FAULTMASK behaviour These changes mean that it's no longer possible to definitely say that if FAULTMASK is set it overrides PRIMASK, and if PRIMASK is set it overrides BASEPRI (since if PRIMASK_NS is set and AIRCR.PRIS is set then whether that 0x80 priority should take effect or the priority in BASEPRI_S depends on the value of BASEPRI_S, for instance). So we switch to the same approach used by the pseudocode of working through BASEPRI, PRIMASK and FAULTMASK and overriding the previous values if needed. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-16-git-send-email-peter.maydell@linaro.org --- hw/intc/armv7m_nvic.c | 51 ++++++++++++++++++++++++++++++++++++++++++--------- 1 file changed, 42 insertions(+), 9 deletions(-) diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ static void nvic_recompute_state(NVICState *s) static inline int nvic_exec_prio(NVICState *s) { CPUARMState *env = &s->cpu->env; - int running; + int running = NVIC_NOEXC_PRIO; - if (env->v7m.faultmask[env->v7m.secure]) { - running = -1; - } else if (env->v7m.primask[env->v7m.secure]) { + if (env->v7m.basepri[M_REG_NS] > 0) { + running = exc_group_prio(s, env->v7m.basepri[M_REG_NS], M_REG_NS); + } + + if (env->v7m.basepri[M_REG_S] > 0) { + int basepri = exc_group_prio(s, env->v7m.basepri[M_REG_S], M_REG_S); + if (running > basepri) { + running = basepri; + } + } + + if (env->v7m.primask[M_REG_NS]) { + if (env->v7m.aircr & R_V7M_AIRCR_PRIS_MASK) { + if (running > NVIC_NS_PRIO_LIMIT) { + running = NVIC_NS_PRIO_LIMIT; + } + } else { + running = 0; + } + } + + if (env->v7m.primask[M_REG_S]) { running = 0; - } else if (env->v7m.basepri[env->v7m.secure] > 0) { - running = env->v7m.basepri[env->v7m.secure] & - nvic_gprio_mask(s, env->v7m.secure); - } else { - running = NVIC_NOEXC_PRIO; /* lower than any possible priority */ } + + if (env->v7m.faultmask[M_REG_NS]) { + if (env->v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK) { + running = -1; + } else { + if (env->v7m.aircr & R_V7M_AIRCR_PRIS_MASK) { + if (running > NVIC_NS_PRIO_LIMIT) { + running = NVIC_NS_PRIO_LIMIT; + } + } else { + running = 0; + } + } + } + + if (env->v7m.faultmask[M_REG_S]) { + running = (env->v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK) ? -3 : -1; + } + /* consider priority of active handler */ return MIN(running, s->exception_prio); } -- 2.7.4 Now that we have a banked FAULTMASK register and banked exceptions, we can implement the correct check in cpu_mmu_index() for whether the MPU_CTRL.HFNMIENA bit's effect should apply. This bit causes handlers which have requested a negative execution priority to run with the MPU disabled. In v8M the test has to check this for the current security state and so takes account of banking. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-17-git-send-email-peter.maydell@linaro.org --- target/arm/cpu.h | 21 ++++++++++++++++----- hw/intc/armv7m_nvic.c | 29 +++++++++++++++++++++++++++++ 2 files changed, 45 insertions(+), 5 deletions(-) diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ int armv7m_nvic_complete_irq(void *opaque, int irq); * (v8M ARM ARM I_PKLD.) */ int armv7m_nvic_raw_execution_priority(void *opaque); +/** + * armv7m_nvic_neg_prio_requested: return true if the requested execution + * priority is negative for the specified security state. + * @opaque: the NVIC + * @secure: the security state to test + * This corresponds to the pseudocode IsReqExecPriNeg(). + */ +#ifndef CONFIG_USER_ONLY +bool armv7m_nvic_neg_prio_requested(void *opaque, bool secure); +#else +static inline bool armv7m_nvic_neg_prio_requested(void *opaque, bool secure) +{ + return false; +} +#endif /* Interface for defining coprocessor registers. * Registers are defined in tables of arm_cp_reginfo structs @@ -XXX,XX +XXX,XX @@ static inline int cpu_mmu_index(CPUARMState *env, bool ifetch) if (arm_feature(env, ARM_FEATURE_M)) { ARMMMUIdx mmu_idx = el == 0 ? ARMMMUIdx_MUser : ARMMMUIdx_MPriv; - /* Execution priority is negative if FAULTMASK is set or - * we're in a HardFault or NMI handler. - */ - if ((env->v7m.exception > 0 && env->v7m.exception <= 3) - || env->v7m.faultmask[env->v7m.secure]) { + if (armv7m_nvic_neg_prio_requested(env->nvic, env->v7m.secure)) { mmu_idx = ARMMMUIdx_MNegPri; } diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ static inline int nvic_exec_prio(NVICState *s) return MIN(running, s->exception_prio); } +bool armv7m_nvic_neg_prio_requested(void *opaque, bool secure) +{ + /* Return true if the requested execution priority is negative + * for the specified security state, ie that security state + * has an active NMI or HardFault or has set its FAULTMASK. + * Note that this is not the same as whether the execution + * priority is actually negative (for instance AIRCR.PRIS may + * mean we don't allow FAULTMASK_NS to actually make the execution + * priority negative). Compare pseudocode IsReqExcPriNeg(). + */ + NVICState *s = opaque; + + if (s->cpu->env.v7m.faultmask[secure]) { + return true; + } + + if (secure ? s->sec_vectors[ARMV7M_EXCP_HARD].active : + s->vectors[ARMV7M_EXCP_HARD].active) { + return true; + } + + if (s->vectors[ARMV7M_EXCP_NMI].active && + exc_targets_secure(s, ARMV7M_EXCP_NMI) == secure) { + return true; + } + + return false; +} + bool armv7m_nvic_can_take_pending_exception(void *opaque) { NVICState *s = opaque; -- 2.7.4 The ICSR NVIC register is banked for v8M. This doesn't require any new state, but it does mean that some bits are controlled by BFHNFNMINS and some bits must work with the correct banked exception. There is also a new in v8M PENDNMICLR bit. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-18-git-send-email-peter.maydell@linaro.org --- hw/intc/armv7m_nvic.c | 45 ++++++++++++++++++++++++++++++++------------- 1 file changed, 32 insertions(+), 13 deletions(-) diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ static uint32_t nvic_readl(NVICState *s, uint32_t offset, MemTxAttrs attrs) } case 0xd00: /* CPUID Base. */ return cpu->midr; - case 0xd04: /* Interrupt Control State. */ + case 0xd04: /* Interrupt Control State (ICSR) */ /* VECTACTIVE */ val = cpu->env.v7m.exception; /* VECTPENDING */ @@ -XXX,XX +XXX,XX @@ static uint32_t nvic_readl(NVICState *s, uint32_t offset, MemTxAttrs attrs) if (nvic_rettobase(s)) { val |= (1 << 11); } - /* PENDSTSET */ - if (s->vectors[ARMV7M_EXCP_SYSTICK].pending) { - val |= (1 << 26); - } - /* PENDSVSET */ - if (s->vectors[ARMV7M_EXCP_PENDSV].pending) { - val |= (1 << 28); + if (attrs.secure) { + /* PENDSTSET */ + if (s->sec_vectors[ARMV7M_EXCP_SYSTICK].pending) { + val |= (1 << 26); + } + /* PENDSVSET */ + if (s->sec_vectors[ARMV7M_EXCP_PENDSV].pending) { + val |= (1 << 28); + } + } else { + /* PENDSTSET */ + if (s->vectors[ARMV7M_EXCP_SYSTICK].pending) { + val |= (1 << 26); + } + /* PENDSVSET */ + if (s->vectors[ARMV7M_EXCP_PENDSV].pending) { + val |= (1 << 28); + } } /* NMIPENDSET */ - if (s->vectors[ARMV7M_EXCP_NMI].pending) { + if ((cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK) && + s->vectors[ARMV7M_EXCP_NMI].pending) { val |= (1 << 31); } - /* ISRPREEMPT not implemented */ + /* ISRPREEMPT: RES0 when halting debug not implemented */ + /* STTNS: RES0 for the Main Extension */ return val; case 0xd08: /* Vector Table Offset. */ return cpu->env.v7m.vecbase[attrs.secure]; @@ -XXX,XX +XXX,XX @@ static void nvic_writel(NVICState *s, uint32_t offset, uint32_t value, nvic_irq_update(s); break; } - case 0xd04: /* Interrupt Control State. */ - if (value & (1 << 31)) { - armv7m_nvic_set_pending(s, ARMV7M_EXCP_NMI, false); + case 0xd04: /* Interrupt Control State (ICSR) */ + if (cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK) { + if (value & (1 << 31)) { + armv7m_nvic_set_pending(s, ARMV7M_EXCP_NMI, false); + } else if (value & (1 << 30) && + arm_feature(&cpu->env, ARM_FEATURE_V8)) { + /* PENDNMICLR didn't exist in v7M */ + armv7m_nvic_clear_pending(s, ARMV7M_EXCP_NMI, false); + } } if (value & (1 << 28)) { armv7m_nvic_set_pending(s, ARMV7M_EXCP_PENDSV, attrs.secure); -- 2.7.4 Handle banking of SHCSR: some register bits are banked between Secure and Non-Secure, and some are only accessible to Secure. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-19-git-send-email-peter.maydell@linaro.org --- hw/intc/armv7m_nvic.c | 221 ++++++++++++++++++++++++++++++++++++++------------ 1 file changed, 169 insertions(+), 52 deletions(-) diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ static uint32_t nvic_readl(NVICState *s, uint32_t offset, MemTxAttrs attrs) val = cpu->env.v7m.ccr[attrs.secure]; val |= cpu->env.v7m.ccr[M_REG_NS] & R_V7M_CCR_BFHFNMIGN_MASK; return val; - case 0xd24: /* System Handler Status. */ + case 0xd24: /* System Handler Control and State (SHCSR) */ val = 0; - if (s->vectors[ARMV7M_EXCP_MEM].active) { - val |= (1 << 0); - } - if (s->vectors[ARMV7M_EXCP_BUS].active) { - val |= (1 << 1); - } - if (s->vectors[ARMV7M_EXCP_USAGE].active) { - val |= (1 << 3); + if (attrs.secure) { + if (s->sec_vectors[ARMV7M_EXCP_MEM].active) { + val |= (1 << 0); + } + if (s->sec_vectors[ARMV7M_EXCP_HARD].active) { + val |= (1 << 2); + } + if (s->sec_vectors[ARMV7M_EXCP_USAGE].active) { + val |= (1 << 3); + } + if (s->sec_vectors[ARMV7M_EXCP_SVC].active) { + val |= (1 << 7); + } + if (s->sec_vectors[ARMV7M_EXCP_PENDSV].active) { + val |= (1 << 10); + } + if (s->sec_vectors[ARMV7M_EXCP_SYSTICK].active) { + val |= (1 << 11); + } + if (s->sec_vectors[ARMV7M_EXCP_USAGE].pending) { + val |= (1 << 12); + } + if (s->sec_vectors[ARMV7M_EXCP_MEM].pending) { + val |= (1 << 13); + } + if (s->sec_vectors[ARMV7M_EXCP_SVC].pending) { + val |= (1 << 15); + } + if (s->sec_vectors[ARMV7M_EXCP_MEM].enabled) { + val |= (1 << 16); + } + if (s->sec_vectors[ARMV7M_EXCP_USAGE].enabled) { + val |= (1 << 18); + } + if (s->sec_vectors[ARMV7M_EXCP_HARD].pending) { + val |= (1 << 21); + } + /* SecureFault is not banked but is always RAZ/WI to NS */ + if (s->vectors[ARMV7M_EXCP_SECURE].active) { + val |= (1 << 4); + } + if (s->vectors[ARMV7M_EXCP_SECURE].enabled) { + val |= (1 << 19); + } + if (s->vectors[ARMV7M_EXCP_SECURE].pending) { + val |= (1 << 20); + } + } else { + if (s->vectors[ARMV7M_EXCP_MEM].active) { + val |= (1 << 0); + } + if (arm_feature(&cpu->env, ARM_FEATURE_V8)) { + /* HARDFAULTACT, HARDFAULTPENDED not present in v7M */ + if (s->vectors[ARMV7M_EXCP_HARD].active) { + val |= (1 << 2); + } + if (s->vectors[ARMV7M_EXCP_HARD].pending) { + val |= (1 << 21); + } + } + if (s->vectors[ARMV7M_EXCP_USAGE].active) { + val |= (1 << 3); + } + if (s->vectors[ARMV7M_EXCP_SVC].active) { + val |= (1 << 7); + } + if (s->vectors[ARMV7M_EXCP_PENDSV].active) { + val |= (1 << 10); + } + if (s->vectors[ARMV7M_EXCP_SYSTICK].active) { + val |= (1 << 11); + } + if (s->vectors[ARMV7M_EXCP_USAGE].pending) { + val |= (1 << 12); + } + if (s->vectors[ARMV7M_EXCP_MEM].pending) { + val |= (1 << 13); + } + if (s->vectors[ARMV7M_EXCP_SVC].pending) { + val |= (1 << 15); + } + if (s->vectors[ARMV7M_EXCP_MEM].enabled) { + val |= (1 << 16); + } + if (s->vectors[ARMV7M_EXCP_USAGE].enabled) { + val |= (1 << 18); + } } - if (s->vectors[ARMV7M_EXCP_SVC].active) { - val |= (1 << 7); + if (attrs.secure || (cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK)) { + if (s->vectors[ARMV7M_EXCP_BUS].active) { + val |= (1 << 1); + } + if (s->vectors[ARMV7M_EXCP_BUS].pending) { + val |= (1 << 14); + } + if (s->vectors[ARMV7M_EXCP_BUS].enabled) { + val |= (1 << 17); + } + if (arm_feature(&cpu->env, ARM_FEATURE_V8) && + s->vectors[ARMV7M_EXCP_NMI].active) { + /* NMIACT is not present in v7M */ + val |= (1 << 5); + } } + + /* TODO: this is RAZ/WI from NS if DEMCR.SDME is set */ if (s->vectors[ARMV7M_EXCP_DEBUG].active) { val |= (1 << 8); } - if (s->vectors[ARMV7M_EXCP_PENDSV].active) { - val |= (1 << 10); - } - if (s->vectors[ARMV7M_EXCP_SYSTICK].active) { - val |= (1 << 11); - } - if (s->vectors[ARMV7M_EXCP_USAGE].pending) { - val |= (1 << 12); - } - if (s->vectors[ARMV7M_EXCP_MEM].pending) { - val |= (1 << 13); - } - if (s->vectors[ARMV7M_EXCP_BUS].pending) { - val |= (1 << 14); - } - if (s->vectors[ARMV7M_EXCP_SVC].pending) { - val |= (1 << 15); - } - if (s->vectors[ARMV7M_EXCP_MEM].enabled) { - val |= (1 << 16); - } - if (s->vectors[ARMV7M_EXCP_BUS].enabled) { - val |= (1 << 17); - } - if (s->vectors[ARMV7M_EXCP_USAGE].enabled) { - val |= (1 << 18); - } return val; case 0xd28: /* Configurable Fault Status. */ /* The BFSR bits [15:8] are shared between security states @@ -XXX,XX +XXX,XX @@ static void nvic_writel(NVICState *s, uint32_t offset, uint32_t value, cpu->env.v7m.ccr[attrs.secure] = value; break; - case 0xd24: /* System Handler Control. */ - s->vectors[ARMV7M_EXCP_MEM].active = (value & (1 << 0)) != 0; - s->vectors[ARMV7M_EXCP_BUS].active = (value & (1 << 1)) != 0; - s->vectors[ARMV7M_EXCP_USAGE].active = (value & (1 << 3)) != 0; - s->vectors[ARMV7M_EXCP_SVC].active = (value & (1 << 7)) != 0; + case 0xd24: /* System Handler Control and State (SHCSR) */ + if (attrs.secure) { + s->sec_vectors[ARMV7M_EXCP_MEM].active = (value & (1 << 0)) != 0; + /* Secure HardFault active bit cannot be written */ + s->sec_vectors[ARMV7M_EXCP_USAGE].active = (value & (1 << 3)) != 0; + s->sec_vectors[ARMV7M_EXCP_SVC].active = (value & (1 << 7)) != 0; + s->sec_vectors[ARMV7M_EXCP_PENDSV].active = + (value & (1 << 10)) != 0; + s->sec_vectors[ARMV7M_EXCP_SYSTICK].active = + (value & (1 << 11)) != 0; + s->sec_vectors[ARMV7M_EXCP_USAGE].pending = + (value & (1 << 12)) != 0; + s->sec_vectors[ARMV7M_EXCP_MEM].pending = (value & (1 << 13)) != 0; + s->sec_vectors[ARMV7M_EXCP_SVC].pending = (value & (1 << 15)) != 0; + s->sec_vectors[ARMV7M_EXCP_MEM].enabled = (value & (1 << 16)) != 0; + s->sec_vectors[ARMV7M_EXCP_BUS].enabled = (value & (1 << 17)) != 0; + s->sec_vectors[ARMV7M_EXCP_USAGE].enabled = + (value & (1 << 18)) != 0; + /* SecureFault not banked, but RAZ/WI to NS */ + s->vectors[ARMV7M_EXCP_SECURE].active = (value & (1 << 4)) != 0; + s->vectors[ARMV7M_EXCP_SECURE].enabled = (value & (1 << 19)) != 0; + s->vectors[ARMV7M_EXCP_SECURE].pending = (value & (1 << 20)) != 0; + } else { + s->vectors[ARMV7M_EXCP_MEM].active = (value & (1 << 0)) != 0; + if (arm_feature(&cpu->env, ARM_FEATURE_V8)) { + /* HARDFAULTPENDED is not present in v7M */ + s->vectors[ARMV7M_EXCP_HARD].pending = (value & (1 << 21)) != 0; + } + s->vectors[ARMV7M_EXCP_USAGE].active = (value & (1 << 3)) != 0; + s->vectors[ARMV7M_EXCP_SVC].active = (value & (1 << 7)) != 0; + s->vectors[ARMV7M_EXCP_PENDSV].active = (value & (1 << 10)) != 0; + s->vectors[ARMV7M_EXCP_SYSTICK].active = (value & (1 << 11)) != 0; + s->vectors[ARMV7M_EXCP_USAGE].pending = (value & (1 << 12)) != 0; + s->vectors[ARMV7M_EXCP_MEM].pending = (value & (1 << 13)) != 0; + s->vectors[ARMV7M_EXCP_SVC].pending = (value & (1 << 15)) != 0; + s->vectors[ARMV7M_EXCP_MEM].enabled = (value & (1 << 16)) != 0; + s->vectors[ARMV7M_EXCP_USAGE].enabled = (value & (1 << 18)) != 0; + } + if (attrs.secure || (cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK)) { + s->vectors[ARMV7M_EXCP_BUS].active = (value & (1 << 1)) != 0; + s->vectors[ARMV7M_EXCP_BUS].pending = (value & (1 << 14)) != 0; + s->vectors[ARMV7M_EXCP_BUS].enabled = (value & (1 << 17)) != 0; + } + /* NMIACT can only be written if the write is of a zero, with + * BFHFNMINS 1, and by the CPU in secure state via the NS alias. + */ + if (!attrs.secure && cpu->env.v7m.secure && + (cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK) && + (value & (1 << 5)) == 0) { + s->vectors[ARMV7M_EXCP_NMI].active = 0; + } + /* HARDFAULTACT can only be written if the write is of a zero + * to the non-secure HardFault state by the CPU in secure state. + * The only case where we can be targeting the non-secure HF state + * when in secure state is if this is a write via the NS alias + * and BFHFNMINS is 1. + */ + if (!attrs.secure && cpu->env.v7m.secure && + (cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK) && + (value & (1 << 2)) == 0) { + s->vectors[ARMV7M_EXCP_HARD].active = 0; + } + + /* TODO: this is RAZ/WI from NS if DEMCR.SDME is set */ s->vectors[ARMV7M_EXCP_DEBUG].active = (value & (1 << 8)) != 0; - s->vectors[ARMV7M_EXCP_PENDSV].active = (value & (1 << 10)) != 0; - s->vectors[ARMV7M_EXCP_SYSTICK].active = (value & (1 << 11)) != 0; - s->vectors[ARMV7M_EXCP_USAGE].pending = (value & (1 << 12)) != 0; - s->vectors[ARMV7M_EXCP_MEM].pending = (value & (1 << 13)) != 0; - s->vectors[ARMV7M_EXCP_BUS].pending = (value & (1 << 14)) != 0; - s->vectors[ARMV7M_EXCP_SVC].pending = (value & (1 << 15)) != 0; - s->vectors[ARMV7M_EXCP_MEM].enabled = (value & (1 << 16)) != 0; - s->vectors[ARMV7M_EXCP_BUS].enabled = (value & (1 << 17)) != 0; - s->vectors[ARMV7M_EXCP_USAGE].enabled = (value & (1 << 18)) != 0; nvic_irq_update(s); break; case 0xd28: /* Configurable Fault Status. */ -- 2.7.4 Update armv7m_nvic_acknowledge_irq() and armv7m_nvic_complete_irq() to handle banked exceptions: * acknowledge needs to use the correct vector, which may be in sec_vectors[] * acknowledge needs to return to its caller whether the exception should be taken to secure or non-secure state * complete needs its caller to tell it whether the exception being completed is a secure one or not Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505240046-11454-20-git-send-email-peter.maydell@linaro.org --- target/arm/cpu.h | 15 +++++++++++++-- hw/intc/armv7m_nvic.c | 26 ++++++++++++++++++++------ target/arm/helper.c | 8 +++++--- hw/intc/trace-events | 4 ++-- 4 files changed, 40 insertions(+), 13 deletions(-) diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ static inline bool armv7m_nvic_can_take_pending_exception(void *opaque) * of architecturally banked exceptions. */ void armv7m_nvic_set_pending(void *opaque, int irq, bool secure); -void armv7m_nvic_acknowledge_irq(void *opaque); +/** + * armv7m_nvic_acknowledge_irq: make highest priority pending exception active + * @opaque: the NVIC + * + * Move the current highest priority pending exception from the pending + * state to the active state, and update v7m.exception to indicate that + * it is the exception currently being handled. + * + * Returns: true if exception should be taken to Secure state, false for NS + */ +bool armv7m_nvic_acknowledge_irq(void *opaque); /** * armv7m_nvic_complete_irq: complete specified interrupt or exception * @opaque: the NVIC * @irq: the exception number to complete + * @secure: true if this exception was secure * * Returns: -1 if the irq was not active * 1 if completing this irq brought us back to base (no active irqs) * 0 if there is still an irq active after this one was completed * (Ignoring -1, this is the same as the RETTOBASE value before completion.) */ -int armv7m_nvic_complete_irq(void *opaque, int irq); +int armv7m_nvic_complete_irq(void *opaque, int irq, bool secure); /** * armv7m_nvic_raw_execution_priority: return the raw execution priority * @opaque: the NVIC diff --git a/hw/intc/armv7m_nvic.c b/hw/intc/armv7m_nvic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/armv7m_nvic.c +++ b/hw/intc/armv7m_nvic.c @@ -XXX,XX +XXX,XX @@ void armv7m_nvic_set_pending(void *opaque, int irq, bool secure) } /* Make pending IRQ active. */ -void armv7m_nvic_acknowledge_irq(void *opaque) +bool armv7m_nvic_acknowledge_irq(void *opaque) { NVICState *s = (NVICState *)opaque; CPUARMState *env = &s->cpu->env; const int pending = s->vectpending; const int running = nvic_exec_prio(s); VecInfo *vec; + bool targets_secure; assert(pending > ARMV7M_EXCP_RESET && pending < s->num_irq); - vec = &s->vectors[pending]; + if (s->vectpending_is_s_banked) { + vec = &s->sec_vectors[pending]; + targets_secure = true; + } else { + vec = &s->vectors[pending]; + targets_secure = !exc_is_banked(s->vectpending) && + exc_targets_secure(s, s->vectpending); + } assert(vec->enabled); assert(vec->pending); assert(s->vectpending_prio < running); - trace_nvic_acknowledge_irq(pending, s->vectpending_prio); + trace_nvic_acknowledge_irq(pending, s->vectpending_prio, targets_secure); vec->active = 1; vec->pending = 0; @@ -XXX,XX +XXX,XX @@ void armv7m_nvic_acknowledge_irq(void *opaque) env->v7m.exception = s->vectpending; nvic_irq_update(s); + + return targets_secure; } -int armv7m_nvic_complete_irq(void *opaque, int irq) +int armv7m_nvic_complete_irq(void *opaque, int irq, bool secure) { NVICState *s = (NVICState *)opaque; VecInfo *vec; @@ -XXX,XX +XXX,XX @@ int armv7m_nvic_complete_irq(void *opaque, int irq) assert(irq > ARMV7M_EXCP_RESET && irq < s->num_irq); - vec = &s->vectors[irq]; + if (secure && exc_is_banked(irq)) { + vec = &s->sec_vectors[irq]; + } else { + vec = &s->vectors[irq]; + } - trace_nvic_complete_irq(irq); + trace_nvic_complete_irq(irq, secure); if (!vec->active) { /* Tell the caller this was an illegal exception return */ diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) bool return_to_sp_process = false; bool return_to_handler = false; bool rettobase = false; + bool exc_secure = false; /* We can only get here from an EXCP_EXCEPTION_EXIT, and * gen_bx_excret() enforces the architectural rule @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) * which security state's faultmask to clear. (v8M ARM ARM R_KBNF.) */ if (arm_feature(env, ARM_FEATURE_M_SECURITY)) { - int es = excret & R_V7M_EXCRET_ES_MASK; + exc_secure = excret & R_V7M_EXCRET_ES_MASK; if (armv7m_nvic_raw_execution_priority(env->nvic) >= 0) { - env->v7m.faultmask[es] = 0; + env->v7m.faultmask[exc_secure] = 0; } } else { env->v7m.faultmask[M_REG_NS] = 0; } } - switch (armv7m_nvic_complete_irq(env->nvic, env->v7m.exception)) { + switch (armv7m_nvic_complete_irq(env->nvic, env->v7m.exception, + exc_secure)) { case -1: /* attempt to exit an exception that isn't active */ ufault = true; diff --git a/hw/intc/trace-events b/hw/intc/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/trace-events +++ b/hw/intc/trace-events @@ -XXX,XX +XXX,XX @@ nvic_escalate_disabled(int irq) "NVIC escalating irq %d to HardFault: disabled" nvic_set_pending(int irq, bool secure, int en, int prio) "NVIC set pending irq %d secure-bank %d (enabled: %d priority %d)" nvic_clear_pending(int irq, bool secure, int en, int prio) "NVIC clear pending irq %d secure-bank %d (enabled: %d priority %d)" nvic_set_pending_level(int irq) "NVIC set pending: irq %d higher prio than vectpending: setting irq line to 1" -nvic_acknowledge_irq(int irq, int prio) "NVIC acknowledge IRQ: %d now active (prio %d)" -nvic_complete_irq(int irq) "NVIC complete IRQ %d" +nvic_acknowledge_irq(int irq, int prio, bool targets_secure) "NVIC acknowledge IRQ: %d now active (prio %d targets_secure %d)" +nvic_complete_irq(int irq, bool secure) "NVIC complete IRQ %d (secure %d)" nvic_set_irq_level(int irq, int level) "NVIC external irq %d level set to %d" nvic_sysreg_read(uint64_t addr, uint32_t value, unsigned size) "NVIC sysreg read addr 0x%" PRIx64 " data 0x%" PRIx32 " size %u" nvic_sysreg_write(uint64_t addr, uint32_t value, unsigned size) "NVIC sysreg write addr 0x%" PRIx64 " data 0x%" PRIx32 " size %u" -- 2.7.4 In the A64 decoder, we have a lot of references to section numbers from version A.a of the v8A ARM ARM (DDI0487). This version of the document is now long obsolete (we are currently on revision B.a), and various intervening versions renumbered all the sections. The most recent B.a version of the document doesn't assign section numbers at all to the individual instruction classes in the way that the various A.x versions did. The simplest thing to do is just to delete all the out of date C.x.x references. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Message-id: 20170915150849.23557-1-peter.maydell@linaro.org --- target/arm/translate-a64.c | 227 +++++++++++++++++++++++---------------------- 1 file changed, 114 insertions(+), 113 deletions(-) diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static inline AArch64DecodeFn *lookup_disas_fn(const AArch64DecodeTable *table, } /* - * the instruction disassembly implemented here matches - * the instruction encoding classifications in chapter 3 (C3) - * of the ARM Architecture Reference Manual (DDI0487A_a) + * The instruction disassembly implemented here matches + * the instruction encoding classifications in chapter C4 + * of the ARM Architecture Reference Manual (DDI0487B_a); + * classification names and decode diagrams here should generally + * match up with those in the manual. */ -/* C3.2.7 Unconditional branch (immediate) +/* Unconditional branch (immediate) * 31 30 26 25 0 * +----+-----------+-------------------------------------+ * | op | 0 0 1 0 1 | imm26 | @@ -XXX,XX +XXX,XX @@ static void disas_uncond_b_imm(DisasContext *s, uint32_t insn) uint64_t addr = s->pc + sextract32(insn, 0, 26) * 4 - 4; if (insn & (1U << 31)) { - /* C5.6.26 BL Branch with link */ + /* BL Branch with link */ tcg_gen_movi_i64(cpu_reg(s, 30), s->pc); } - /* C5.6.20 B Branch / C5.6.26 BL Branch with link */ + /* B Branch / BL Branch with link */ gen_goto_tb(s, 0, addr); } -/* C3.2.1 Compare & branch (immediate) +/* Compare and branch (immediate) * 31 30 25 24 23 5 4 0 * +----+-------------+----+---------------------+--------+ * | sf | 0 1 1 0 1 0 | op | imm19 | Rt | @@ -XXX,XX +XXX,XX @@ static void disas_comp_b_imm(DisasContext *s, uint32_t insn) gen_goto_tb(s, 1, addr); } -/* C3.2.5 Test & branch (immediate) +/* Test and branch (immediate) * 31 30 25 24 23 19 18 5 4 0 * +----+-------------+----+-------+-------------+------+ * | b5 | 0 1 1 0 1 1 | op | b40 | imm14 | Rt | @@ -XXX,XX +XXX,XX @@ static void disas_test_b_imm(DisasContext *s, uint32_t insn) gen_goto_tb(s, 1, addr); } -/* C3.2.2 / C5.6.19 Conditional branch (immediate) +/* Conditional branch (immediate) * 31 25 24 23 5 4 3 0 * +---------------+----+---------------------+----+------+ * | 0 1 0 1 0 1 0 | o1 | imm19 | o0 | cond | @@ -XXX,XX +XXX,XX @@ static void disas_cond_b_imm(DisasContext *s, uint32_t insn) } } -/* C5.6.68 HINT */ +/* HINT instruction group, including various allocated HINTs */ static void handle_hint(DisasContext *s, uint32_t insn, unsigned int op1, unsigned int op2, unsigned int crm) { @@ -XXX,XX +XXX,XX @@ static void handle_sync(DisasContext *s, uint32_t insn, } } -/* C5.6.130 MSR (immediate) - move immediate to processor state field */ +/* MSR (immediate) - move immediate to processor state field */ static void handle_msr_i(DisasContext *s, uint32_t insn, unsigned int op1, unsigned int op2, unsigned int crm) { @@ -XXX,XX +XXX,XX @@ static void gen_set_nzcv(TCGv_i64 tcg_rt) tcg_temp_free_i32(nzcv); } -/* C5.6.129 MRS - move from system register - * C5.6.131 MSR (register) - move to system register - * C5.6.204 SYS - * C5.6.205 SYSL +/* MRS - move from system register + * MSR (register) - move to system register + * SYS + * SYSL * These are all essentially the same insn in 'read' and 'write' * versions, with varying op0 fields. */ @@ -XXX,XX +XXX,XX @@ static void handle_sys(DisasContext *s, uint32_t insn, bool isread, } } -/* C3.2.4 System +/* System * 31 22 21 20 19 18 16 15 12 11 8 7 5 4 0 * +---------------------+---+-----+-----+-------+-------+-----+------+ * | 1 1 0 1 0 1 0 1 0 0 | L | op0 | op1 | CRn | CRm | op2 | Rt | @@ -XXX,XX +XXX,XX @@ static void disas_system(DisasContext *s, uint32_t insn) return; } switch (crn) { - case 2: /* C5.6.68 HINT */ + case 2: /* HINT (including allocated hints like NOP, YIELD, etc) */ handle_hint(s, insn, op1, op2, crm); break; case 3: /* CLREX, DSB, DMB, ISB */ handle_sync(s, insn, op1, op2, crm); break; - case 4: /* C5.6.130 MSR (immediate) */ + case 4: /* MSR (immediate) */ handle_msr_i(s, insn, op1, op2, crm); break; default: @@ -XXX,XX +XXX,XX @@ static void disas_system(DisasContext *s, uint32_t insn) handle_sys(s, insn, l, op0, op1, op2, crn, crm, rt); } -/* C3.2.3 Exception generation +/* Exception generation * * 31 24 23 21 20 5 4 2 1 0 * +-----------------+-----+------------------------+-----+----+ @@ -XXX,XX +XXX,XX @@ static void disas_exc(DisasContext *s, uint32_t insn) } } -/* C3.2.7 Unconditional branch (register) +/* Unconditional branch (register) * 31 25 24 21 20 16 15 10 9 5 4 0 * +---------------+-------+-------+-------+------+-------+ * | 1 1 0 1 0 1 1 | opc | op2 | op3 | Rn | op4 | @@ -XXX,XX +XXX,XX @@ static void disas_uncond_b_reg(DisasContext *s, uint32_t insn) s->base.is_jmp = DISAS_JUMP; } -/* C3.2 Branches, exception generating and system instructions */ +/* Branches, exception generating and system instructions */ static void disas_b_exc_sys(DisasContext *s, uint32_t insn) { switch (extract32(insn, 25, 7)) { @@ -XXX,XX +XXX,XX @@ static bool disas_ldst_compute_iss_sf(int size, bool is_signed, int opc) return regsize == 64; } -/* C3.3.6 Load/store exclusive +/* Load/store exclusive * * 31 30 29 24 23 22 21 20 16 15 14 10 9 5 4 0 * +-----+-------------+----+---+----+------+----+-------+------+------+ @@ -XXX,XX +XXX,XX @@ static void disas_ldst_excl(DisasContext *s, uint32_t insn) } /* - * C3.3.5 Load register (literal) + * Load register (literal) * * 31 30 29 27 26 25 24 23 5 4 0 * +-----+-------+---+-----+-------------------+-------+ @@ -XXX,XX +XXX,XX @@ static void disas_ld_lit(DisasContext *s, uint32_t insn) } /* - * C5.6.80 LDNP (Load Pair - non-temporal hint) - * C5.6.81 LDP (Load Pair - non vector) - * C5.6.82 LDPSW (Load Pair Signed Word - non vector) - * C5.6.176 STNP (Store Pair - non-temporal hint) - * C5.6.177 STP (Store Pair - non vector) - * C6.3.165 LDNP (Load Pair of SIMD&FP - non-temporal hint) - * C6.3.165 LDP (Load Pair of SIMD&FP) - * C6.3.284 STNP (Store Pair of SIMD&FP - non-temporal hint) - * C6.3.284 STP (Store Pair of SIMD&FP) + * LDNP (Load Pair - non-temporal hint) + * LDP (Load Pair - non vector) + * LDPSW (Load Pair Signed Word - non vector) + * STNP (Store Pair - non-temporal hint) + * STP (Store Pair - non vector) + * LDNP (Load Pair of SIMD&FP - non-temporal hint) + * LDP (Load Pair of SIMD&FP) + * STNP (Store Pair of SIMD&FP - non-temporal hint) + * STP (Store Pair of SIMD&FP) * * 31 30 29 27 26 25 24 23 22 21 15 14 10 9 5 4 0 * +-----+-------+---+---+-------+---+-----------------------------+ @@ -XXX,XX +XXX,XX @@ static void disas_ldst_pair(DisasContext *s, uint32_t insn) } /* - * C3.3.8 Load/store (immediate post-indexed) - * C3.3.9 Load/store (immediate pre-indexed) - * C3.3.12 Load/store (unscaled immediate) + * Load/store (immediate post-indexed) + * Load/store (immediate pre-indexed) + * Load/store (unscaled immediate) * * 31 30 29 27 26 25 24 23 22 21 20 12 11 10 9 5 4 0 * +----+-------+---+-----+-----+---+--------+-----+------+------+ @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_imm9(DisasContext *s, uint32_t insn, } /* - * C3.3.10 Load/store (register offset) + * Load/store (register offset) * * 31 30 29 27 26 25 24 23 22 21 20 16 15 13 12 11 10 9 5 4 0 * +----+-------+---+-----+-----+---+------+-----+--+-----+----+----+ @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg_roffset(DisasContext *s, uint32_t insn, } /* - * C3.3.13 Load/store (unsigned immediate) + * Load/store (unsigned immediate) * * 31 30 29 27 26 25 24 23 22 21 10 9 5 * +----+-------+---+-----+-----+------------+-------+------+ @@ -XXX,XX +XXX,XX @@ static void disas_ldst_reg(DisasContext *s, uint32_t insn) } } -/* C3.3.1 AdvSIMD load/store multiple structures +/* AdvSIMD load/store multiple structures * * 31 30 29 23 22 21 16 15 12 11 10 9 5 4 0 * +---+---+---------------+---+-------------+--------+------+------+------+ * | 0 | Q | 0 0 1 1 0 0 0 | L | 0 0 0 0 0 0 | opcode | size | Rn | Rt | * +---+---+---------------+---+-------------+--------+------+------+------+ * - * C3.3.2 AdvSIMD load/store multiple structures (post-indexed) + * AdvSIMD load/store multiple structures (post-indexed) * * 31 30 29 23 22 21 20 16 15 12 11 10 9 5 4 0 * +---+---+---------------+---+---+---------+--------+------+------+------+ @@ -XXX,XX +XXX,XX @@ static void disas_ldst_multiple_struct(DisasContext *s, uint32_t insn) tcg_temp_free_i64(tcg_addr); } -/* C3.3.3 AdvSIMD load/store single structure +/* AdvSIMD load/store single structure * * 31 30 29 23 22 21 20 16 15 13 12 11 10 9 5 4 0 * +---+---+---------------+-----+-----------+-----+---+------+------+------+ * | 0 | Q | 0 0 1 1 0 1 0 | L R | 0 0 0 0 0 | opc | S | size | Rn | Rt | * +---+---+---------------+-----+-----------+-----+---+------+------+------+ * - * C3.3.4 AdvSIMD load/store single structure (post-indexed) + * AdvSIMD load/store single structure (post-indexed) * * 31 30 29 23 22 21 20 16 15 13 12 11 10 9 5 4 0 * +---+---+---------------+-----+-----------+-----+---+------+------+------+ @@ -XXX,XX +XXX,XX @@ static void disas_ldst_single_struct(DisasContext *s, uint32_t insn) tcg_temp_free_i64(tcg_addr); } -/* C3.3 Loads and stores */ +/* Loads and stores */ static void disas_ldst(DisasContext *s, uint32_t insn) { switch (extract32(insn, 24, 6)) { @@ -XXX,XX +XXX,XX @@ static void disas_ldst(DisasContext *s, uint32_t insn) } } -/* C3.4.6 PC-rel. addressing +/* PC-rel. addressing * 31 30 29 28 24 23 5 4 0 * +----+-------+-----------+-------------------+------+ * | op | immlo | 1 0 0 0 0 | immhi | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_pc_rel_adr(DisasContext *s, uint32_t insn) } /* - * C3.4.1 Add/subtract (immediate) + * Add/subtract (immediate) * * 31 30 29 28 24 23 22 21 10 9 5 4 0 * +--+--+--+-----------+-----+-------------+-----+-----+ @@ -XXX,XX +XXX,XX @@ static bool logic_imm_decode_wmask(uint64_t *result, unsigned int immn, return true; } -/* C3.4.4 Logical (immediate) +/* Logical (immediate) * 31 30 29 28 23 22 21 16 15 10 9 5 4 0 * +----+-----+-------------+---+------+------+------+------+ * | sf | opc | 1 0 0 1 0 0 | N | immr | imms | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_logic_imm(DisasContext *s, uint32_t insn) } /* - * C3.4.5 Move wide (immediate) + * Move wide (immediate) * * 31 30 29 28 23 22 21 20 5 4 0 * +--+-----+-------------+-----+----------------+------+ @@ -XXX,XX +XXX,XX @@ static void disas_movw_imm(DisasContext *s, uint32_t insn) } } -/* C3.4.2 Bitfield +/* Bitfield * 31 30 29 28 23 22 21 16 15 10 9 5 4 0 * +----+-----+-------------+---+------+------+------+------+ * | sf | opc | 1 0 0 1 1 0 | N | immr | imms | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_bitfield(DisasContext *s, uint32_t insn) } } -/* C3.4.3 Extract +/* Extract * 31 30 29 28 23 22 21 20 16 15 10 9 5 4 0 * +----+------+-------------+---+----+------+--------+------+------+ * | sf | op21 | 1 0 0 1 1 1 | N | o0 | Rm | imms | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_extract(DisasContext *s, uint32_t insn) } } -/* C3.4 Data processing - immediate */ +/* Data processing - immediate */ static void disas_data_proc_imm(DisasContext *s, uint32_t insn) { switch (extract32(insn, 23, 6)) { @@ -XXX,XX +XXX,XX @@ static void shift_reg_imm(TCGv_i64 dst, TCGv_i64 src, int sf, } } -/* C3.5.10 Logical (shifted register) +/* Logical (shifted register) * 31 30 29 28 24 23 22 21 20 16 15 10 9 5 4 0 * +----+-----+-----------+-------+---+------+--------+------+------+ * | sf | opc | 0 1 0 1 0 | shift | N | Rm | imm6 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_logic_reg(DisasContext *s, uint32_t insn) } /* - * C3.5.1 Add/subtract (extended register) + * Add/subtract (extended register) * * 31|30|29|28 24|23 22|21|20 16|15 13|12 10|9 5|4 0| * +--+--+--+-----------+-----+--+-------+------+------+----+----+ @@ -XXX,XX +XXX,XX @@ static void disas_add_sub_ext_reg(DisasContext *s, uint32_t insn) } /* - * C3.5.2 Add/subtract (shifted register) + * Add/subtract (shifted register) * * 31 30 29 28 24 23 22 21 20 16 15 10 9 5 4 0 * +--+--+--+-----------+-----+--+-------+---------+------+------+ @@ -XXX,XX +XXX,XX @@ static void disas_add_sub_reg(DisasContext *s, uint32_t insn) tcg_temp_free_i64(tcg_result); } -/* C3.5.9 Data-processing (3 source) - - 31 30 29 28 24 23 21 20 16 15 14 10 9 5 4 0 - +--+------+-----------+------+------+----+------+------+------+ - |sf| op54 | 1 1 0 1 1 | op31 | Rm | o0 | Ra | Rn | Rd | - +--+------+-----------+------+------+----+------+------+------+ - +/* Data-processing (3 source) + * + * 31 30 29 28 24 23 21 20 16 15 14 10 9 5 4 0 + * +--+------+-----------+------+------+----+------+------+------+ + * |sf| op54 | 1 1 0 1 1 | op31 | Rm | o0 | Ra | Rn | Rd | + * +--+------+-----------+------+------+----+------+------+------+ */ static void disas_data_proc_3src(DisasContext *s, uint32_t insn) { @@ -XXX,XX +XXX,XX @@ static void disas_data_proc_3src(DisasContext *s, uint32_t insn) tcg_temp_free_i64(tcg_tmp); } -/* C3.5.3 - Add/subtract (with carry) +/* Add/subtract (with carry) * 31 30 29 28 27 26 25 24 23 22 21 20 16 15 10 9 5 4 0 * +--+--+--+------------------------+------+---------+------+-----+ * |sf|op| S| 1 1 0 1 0 0 0 0 | rm | opcode2 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_adc_sbc(DisasContext *s, uint32_t insn) } } -/* C3.5.4 - C3.5.5 Conditional compare (immediate / register) +/* Conditional compare (immediate / register) * 31 30 29 28 27 26 25 24 23 22 21 20 16 15 12 11 10 9 5 4 3 0 * +--+--+--+------------------------+--------+------+----+--+------+--+-----+ * |sf|op| S| 1 1 0 1 0 0 1 0 |imm5/rm | cond |i/r |o2| Rn |o3|nzcv | @@ -XXX,XX +XXX,XX @@ static void disas_cc(DisasContext *s, uint32_t insn) tcg_temp_free_i32(tcg_t2); } -/* C3.5.6 Conditional select +/* Conditional select * 31 30 29 28 21 20 16 15 12 11 10 9 5 4 0 * +----+----+---+-----------------+------+------+-----+------+------+ * | sf | op | S | 1 1 0 1 0 1 0 0 | Rm | cond | op2 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void handle_rbit(DisasContext *s, unsigned int sf, } } -/* C5.6.149 REV with sf==1, opcode==3 ("REV64") */ +/* REV with sf==1, opcode==3 ("REV64") */ static void handle_rev64(DisasContext *s, unsigned int sf, unsigned int rn, unsigned int rd) { @@ -XXX,XX +XXX,XX @@ static void handle_rev64(DisasContext *s, unsigned int sf, tcg_gen_bswap64_i64(cpu_reg(s, rd), cpu_reg(s, rn)); } -/* C5.6.149 REV with sf==0, opcode==2 - * C5.6.151 REV32 (sf==1, opcode==2) +/* REV with sf==0, opcode==2 + * REV32 (sf==1, opcode==2) */ static void handle_rev32(DisasContext *s, unsigned int sf, unsigned int rn, unsigned int rd) @@ -XXX,XX +XXX,XX @@ static void handle_rev32(DisasContext *s, unsigned int sf, } } -/* C5.6.150 REV16 (opcode==1) */ +/* REV16 (opcode==1) */ static void handle_rev16(DisasContext *s, unsigned int sf, unsigned int rn, unsigned int rd) { @@ -XXX,XX +XXX,XX @@ static void handle_rev16(DisasContext *s, unsigned int sf, tcg_temp_free_i64(tcg_tmp); } -/* C3.5.7 Data-processing (1 source) +/* Data-processing (1 source) * 31 30 29 28 21 20 16 15 10 9 5 4 0 * +----+---+---+-----------------+---------+--------+------+------+ * | sf | 1 | S | 1 1 0 1 0 1 1 0 | opcode2 | opcode | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void handle_div(DisasContext *s, bool is_signed, unsigned int sf, } } -/* C5.6.115 LSLV, C5.6.118 LSRV, C5.6.17 ASRV, C5.6.154 RORV */ +/* LSLV, LSRV, ASRV, RORV */ static void handle_shift_reg(DisasContext *s, enum a64_shift_type shift_type, unsigned int sf, unsigned int rm, unsigned int rn, unsigned int rd) @@ -XXX,XX +XXX,XX @@ static void handle_crc32(DisasContext *s, tcg_temp_free_i32(tcg_bytes); } -/* C3.5.8 Data-processing (2 source) +/* Data-processing (2 source) * 31 30 29 28 21 20 16 15 10 9 5 4 0 * +----+---+---+-----------------+------+--------+------+------+ * | sf | 0 | S | 1 1 0 1 0 1 1 0 | Rm | opcode | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_data_proc_2src(DisasContext *s, uint32_t insn) } } -/* C3.5 Data processing - register */ +/* Data processing - register */ static void disas_data_proc_reg(DisasContext *s, uint32_t insn) { switch (extract32(insn, 24, 5)) { @@ -XXX,XX +XXX,XX @@ static void handle_fp_compare(DisasContext *s, bool is_double, tcg_temp_free_i64(tcg_flags); } -/* C3.6.22 Floating point compare +/* Floating point compare * 31 30 29 28 24 23 22 21 20 16 15 14 13 10 9 5 4 0 * +---+---+---+-----------+------+---+------+-----+---------+------+-------+ * | M | 0 | S | 1 1 1 1 0 | type | 1 | Rm | op | 1 0 0 0 | Rn | op2 | @@ -XXX,XX +XXX,XX @@ static void disas_fp_compare(DisasContext *s, uint32_t insn) handle_fp_compare(s, type, rn, rm, opc & 1, opc & 2); } -/* C3.6.23 Floating point conditional compare +/* Floating point conditional compare * 31 30 29 28 24 23 22 21 20 16 15 12 11 10 9 5 4 3 0 * +---+---+---+-----------+------+---+------+------+-----+------+----+------+ * | M | 0 | S | 1 1 1 1 0 | type | 1 | Rm | cond | 0 1 | Rn | op | nzcv | @@ -XXX,XX +XXX,XX @@ static void disas_fp_ccomp(DisasContext *s, uint32_t insn) } } -/* C3.6.24 Floating point conditional select +/* Floating point conditional select * 31 30 29 28 24 23 22 21 20 16 15 12 11 10 9 5 4 0 * +---+---+---+-----------+------+---+------+------+-----+------+------+ * | M | 0 | S | 1 1 1 1 0 | type | 1 | Rm | cond | 1 1 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_fp_csel(DisasContext *s, uint32_t insn) tcg_temp_free_i64(t_true); } -/* C3.6.25 Floating-point data-processing (1 source) - single precision */ +/* Floating-point data-processing (1 source) - single precision */ static void handle_fp_1src_single(DisasContext *s, int opcode, int rd, int rn) { TCGv_ptr fpst; @@ -XXX,XX +XXX,XX @@ static void handle_fp_1src_single(DisasContext *s, int opcode, int rd, int rn) tcg_temp_free_i32(tcg_res); } -/* C3.6.25 Floating-point data-processing (1 source) - double precision */ +/* Floating-point data-processing (1 source) - double precision */ static void handle_fp_1src_double(DisasContext *s, int opcode, int rd, int rn) { TCGv_ptr fpst; @@ -XXX,XX +XXX,XX @@ static void handle_fp_fcvt(DisasContext *s, int opcode, } } -/* C3.6.25 Floating point data-processing (1 source) +/* Floating point data-processing (1 source) * 31 30 29 28 24 23 22 21 20 15 14 10 9 5 4 0 * +---+---+---+-----------+------+---+--------+-----------+------+------+ * | M | 0 | S | 1 1 1 1 0 | type | 1 | opcode | 1 0 0 0 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_fp_1src(DisasContext *s, uint32_t insn) } } -/* C3.6.26 Floating-point data-processing (2 source) - single precision */ +/* Floating-point data-processing (2 source) - single precision */ static void handle_fp_2src_single(DisasContext *s, int opcode, int rd, int rn, int rm) { @@ -XXX,XX +XXX,XX @@ static void handle_fp_2src_single(DisasContext *s, int opcode, tcg_temp_free_i32(tcg_res); } -/* C3.6.26 Floating-point data-processing (2 source) - double precision */ +/* Floating-point data-processing (2 source) - double precision */ static void handle_fp_2src_double(DisasContext *s, int opcode, int rd, int rn, int rm) { @@ -XXX,XX +XXX,XX @@ static void handle_fp_2src_double(DisasContext *s, int opcode, tcg_temp_free_i64(tcg_res); } -/* C3.6.26 Floating point data-processing (2 source) +/* Floating point data-processing (2 source) * 31 30 29 28 24 23 22 21 20 16 15 12 11 10 9 5 4 0 * +---+---+---+-----------+------+---+------+--------+-----+------+------+ * | M | 0 | S | 1 1 1 1 0 | type | 1 | Rm | opcode | 1 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_fp_2src(DisasContext *s, uint32_t insn) } } -/* C3.6.27 Floating-point data-processing (3 source) - single precision */ +/* Floating-point data-processing (3 source) - single precision */ static void handle_fp_3src_single(DisasContext *s, bool o0, bool o1, int rd, int rn, int rm, int ra) { @@ -XXX,XX +XXX,XX @@ static void handle_fp_3src_single(DisasContext *s, bool o0, bool o1, tcg_temp_free_i32(tcg_res); } -/* C3.6.27 Floating-point data-processing (3 source) - double precision */ +/* Floating-point data-processing (3 source) - double precision */ static void handle_fp_3src_double(DisasContext *s, bool o0, bool o1, int rd, int rn, int rm, int ra) { @@ -XXX,XX +XXX,XX @@ static void handle_fp_3src_double(DisasContext *s, bool o0, bool o1, tcg_temp_free_i64(tcg_res); } -/* C3.6.27 Floating point data-processing (3 source) +/* Floating point data-processing (3 source) * 31 30 29 28 24 23 22 21 20 16 15 14 10 9 5 4 0 * +---+---+---+-----------+------+----+------+----+------+------+------+ * | M | 0 | S | 1 1 1 1 1 | type | o1 | Rm | o0 | Ra | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_fp_3src(DisasContext *s, uint32_t insn) } } -/* C3.6.28 Floating point immediate +/* Floating point immediate * 31 30 29 28 24 23 22 21 20 13 12 10 9 5 4 0 * +---+---+---+-----------+------+---+------------+-------+------+------+ * | M | 0 | S | 1 1 1 1 0 | type | 1 | imm8 | 1 0 0 | imm5 | Rd | @@ -XXX,XX +XXX,XX @@ static void handle_fpfpcvt(DisasContext *s, int rd, int rn, int opcode, tcg_temp_free_i32(tcg_shift); } -/* C3.6.29 Floating point <-> fixed point conversions +/* Floating point <-> fixed point conversions * 31 30 29 28 24 23 22 21 20 19 18 16 15 10 9 5 4 0 * +----+---+---+-----------+------+---+-------+--------+-------+------+------+ * | sf | 0 | S | 1 1 1 1 0 | type | 0 | rmode | opcode | scale | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void handle_fmov(DisasContext *s, int rd, int rn, int type, bool itof) } } -/* C3.6.30 Floating point <-> integer conversions +/* Floating point <-> integer conversions * 31 30 29 28 24 23 22 21 20 19 18 16 15 10 9 5 4 0 * +----+---+---+-----------+------+---+-------+-----+-------------+----+----+ * | sf | 0 | S | 1 1 1 1 0 | type | 1 | rmode | opc | 0 0 0 0 0 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void do_ext64(DisasContext *s, TCGv_i64 tcg_left, TCGv_i64 tcg_right, tcg_temp_free_i64(tcg_tmp); } -/* C3.6.1 EXT +/* EXT * 31 30 29 24 23 22 21 20 16 15 14 11 10 9 5 4 0 * +---+---+-------------+-----+---+------+---+------+---+------+------+ * | 0 | Q | 1 0 1 1 1 0 | op2 | 0 | Rm | 0 | imm4 | 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_simd_ext(DisasContext *s, uint32_t insn) tcg_temp_free_i64(tcg_resh); } -/* C3.6.2 TBL/TBX +/* TBL/TBX * 31 30 29 24 23 22 21 20 16 15 14 13 12 11 10 9 5 4 0 * +---+---+-------------+-----+---+------+---+-----+----+-----+------+------+ * | 0 | Q | 0 0 1 1 1 0 | op2 | 0 | Rm | 0 | len | op | 0 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_simd_tb(DisasContext *s, uint32_t insn) tcg_temp_free_i64(tcg_resh); } -/* C3.6.3 ZIP/UZP/TRN +/* ZIP/UZP/TRN * 31 30 29 24 23 22 21 20 16 15 14 12 11 10 9 5 4 0 * +---+---+-------------+------+---+------+---+------------------+------+ * | 0 | Q | 0 0 1 1 1 0 | size | 0 | Rm | 0 | opc | 1 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void do_minmaxop(DisasContext *s, TCGv_i32 tcg_elt1, TCGv_i32 tcg_elt2, } } -/* C3.6.4 AdvSIMD across lanes +/* AdvSIMD across lanes * 31 30 29 28 24 23 22 21 17 16 12 11 10 9 5 4 0 * +---+---+---+-----------+------+-----------+--------+-----+------+------+ * | 0 | Q | U | 0 1 1 1 0 | size | 1 1 0 0 0 | opcode | 1 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_simd_across_lanes(DisasContext *s, uint32_t insn) tcg_temp_free_i64(tcg_res); } -/* C6.3.31 DUP (Element, Vector) +/* DUP (Element, Vector) * * 31 30 29 21 20 16 15 10 9 5 4 0 * +---+---+-------------------+--------+-------------+------+------+ @@ -XXX,XX +XXX,XX @@ static void handle_simd_dupe(DisasContext *s, int is_q, int rd, int rn, tcg_temp_free_i64(tmp); } -/* C6.3.31 DUP (element, scalar) +/* DUP (element, scalar) * 31 21 20 16 15 10 9 5 4 0 * +-----------------------+--------+-------------+------+------+ * | 0 1 0 1 1 1 1 0 0 0 0 | imm5 | 0 0 0 0 0 1 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void handle_simd_dupes(DisasContext *s, int rd, int rn, tcg_temp_free_i64(tmp); } -/* C6.3.32 DUP (General) +/* DUP (General) * * 31 30 29 21 20 16 15 10 9 5 4 0 * +---+---+-------------------+--------+-------------+------+------+ @@ -XXX,XX +XXX,XX @@ static void handle_simd_dupg(DisasContext *s, int is_q, int rd, int rn, } } -/* C6.3.150 INS (Element) +/* INS (Element) * * 31 21 20 16 15 14 11 10 9 5 4 0 * +-----------------------+--------+------------+---+------+------+ @@ -XXX,XX +XXX,XX @@ static void handle_simd_inse(DisasContext *s, int rd, int rn, } -/* C6.3.151 INS (General) +/* INS (General) * * 31 21 20 16 15 10 9 5 4 0 * +-----------------------+--------+-------------+------+------+ @@ -XXX,XX +XXX,XX @@ static void handle_simd_insg(DisasContext *s, int rd, int rn, int imm5) } /* - * C6.3.321 UMOV (General) - * C6.3.237 SMOV (General) + * UMOV (General) + * SMOV (General) * * 31 30 29 21 20 16 15 12 10 9 5 4 0 * +---+---+-------------------+--------+-------------+------+------+ @@ -XXX,XX +XXX,XX @@ static void handle_simd_umov_smov(DisasContext *s, int is_q, int is_signed, } } -/* C3.6.5 AdvSIMD copy +/* AdvSIMD copy * 31 30 29 28 21 20 16 15 14 11 10 9 5 4 0 * +---+---+----+-----------------+------+---+------+---+------+------+ * | 0 | Q | op | 0 1 1 1 0 0 0 0 | imm5 | 0 | imm4 | 1 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_simd_copy(DisasContext *s, uint32_t insn) } } -/* C3.6.6 AdvSIMD modified immediate +/* AdvSIMD modified immediate * 31 30 29 28 19 18 16 15 12 11 10 9 5 4 0 * +---+---+----+---------------------+-----+-------+----+---+-------+------+ * | 0 | Q | op | 0 1 1 1 1 0 0 0 0 0 | abc | cmode | o2 | 1 | defgh | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_simd_mod_imm(DisasContext *s, uint32_t insn) tcg_temp_free_i64(tcg_imm); } -/* C3.6.7 AdvSIMD scalar copy +/* AdvSIMD scalar copy * 31 30 29 28 21 20 16 15 14 11 10 9 5 4 0 * +-----+----+-----------------+------+---+------+---+------+------+ * | 0 1 | op | 1 1 1 1 0 0 0 0 | imm5 | 0 | imm4 | 1 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_simd_scalar_copy(DisasContext *s, uint32_t insn) handle_simd_dupes(s, rd, rn, imm5); } -/* C3.6.8 AdvSIMD scalar pairwise +/* AdvSIMD scalar pairwise * 31 30 29 28 24 23 22 21 17 16 12 11 10 9 5 4 0 * +-----+---+-----------+------+-----------+--------+-----+------+------+ * | 0 1 | U | 1 1 1 1 0 | size | 1 1 0 0 0 | opcode | 1 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void handle_simd_shift_fpint_conv(DisasContext *s, bool is_scalar, tcg_temp_free_i32(tcg_rmode); } -/* C3.6.9 AdvSIMD scalar shift by immediate +/* AdvSIMD scalar shift by immediate * 31 30 29 28 23 22 19 18 16 15 11 10 9 5 4 0 * +-----+---+-------------+------+------+--------+---+------+------+ * | 0 1 | U | 1 1 1 1 1 0 | immh | immb | opcode | 1 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_simd_scalar_shift_imm(DisasContext *s, uint32_t insn) } } -/* C3.6.10 AdvSIMD scalar three different +/* AdvSIMD scalar three different * 31 30 29 28 24 23 22 21 20 16 15 12 11 10 9 5 4 0 * +-----+---+-----------+------+---+------+--------+-----+------+------+ * | 0 1 | U | 1 1 1 1 0 | size | 1 | Rm | opcode | 0 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void handle_3same_float(DisasContext *s, int size, int elements, } } -/* C3.6.11 AdvSIMD scalar three same +/* AdvSIMD scalar three same * 31 30 29 28 24 23 22 21 20 16 15 11 10 9 5 4 0 * +-----+---+-----------+------+---+------+--------+---+------+------+ * | 0 1 | U | 1 1 1 1 0 | size | 1 | Rm | opcode | 1 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void handle_2misc_satacc(DisasContext *s, bool is_scalar, bool is_u, } } -/* C3.6.12 AdvSIMD scalar two reg misc +/* AdvSIMD scalar two reg misc * 31 30 29 28 24 23 22 21 17 16 12 11 10 9 5 4 0 * +-----+---+-----------+------+-----------+--------+-----+------+------+ * | 0 1 | U | 1 1 1 1 0 | size | 1 0 0 0 0 | opcode | 1 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void handle_vec_simd_shrn(DisasContext *s, bool is_q, } -/* C3.6.14 AdvSIMD shift by immediate +/* AdvSIMD shift by immediate * 31 30 29 28 23 22 19 18 16 15 11 10 9 5 4 0 * +---+---+---+-------------+------+------+--------+---+------+------+ * | 0 | Q | U | 0 1 1 1 1 0 | immh | immb | opcode | 1 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void handle_pmull_64(DisasContext *s, int is_q, int rd, int rn, int rm) tcg_temp_free_i64(tcg_res); } -/* C3.6.15 AdvSIMD three different +/* AdvSIMD three different * 31 30 29 28 24 23 22 21 20 16 15 12 11 10 9 5 4 0 * +---+---+---+-----------+------+---+------+--------+-----+------+------+ * | 0 | Q | U | 0 1 1 1 0 | size | 1 | Rm | opcode | 0 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_simd_3same_int(DisasContext *s, uint32_t insn) } } -/* C3.6.16 AdvSIMD three same +/* AdvSIMD three same * 31 30 29 28 24 23 22 21 20 16 15 11 10 9 5 4 0 * +---+---+---+-----------+------+---+------+--------+---+------+------+ * | 0 | Q | U | 0 1 1 1 0 | size | 1 | Rm | opcode | 1 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void handle_shll(DisasContext *s, bool is_q, int size, int rn, int rd) } } -/* C3.6.17 AdvSIMD two reg misc +/* AdvSIMD two reg misc * 31 30 29 28 24 23 22 21 17 16 12 11 10 9 5 4 0 * +---+---+---+-----------+------+-----------+--------+-----+------+------+ * | 0 | Q | U | 0 1 1 1 0 | size | 1 0 0 0 0 | opcode | 1 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_simd_two_reg_misc(DisasContext *s, uint32_t insn) } } -/* C3.6.13 AdvSIMD scalar x indexed element +/* AdvSIMD scalar x indexed element * 31 30 29 28 24 23 22 21 20 19 16 15 12 11 10 9 5 4 0 * +-----+---+-----------+------+---+---+------+-----+---+---+------+------+ * | 0 1 | U | 1 1 1 1 1 | size | L | M | Rm | opc | H | 0 | Rn | Rd | * +-----+---+-----------+------+---+---+------+-----+---+---+------+------+ - * C3.6.18 AdvSIMD vector x indexed element + * AdvSIMD vector x indexed element * 31 30 29 28 24 23 22 21 20 19 16 15 12 11 10 9 5 4 0 * +---+---+---+-----------+------+---+---+------+-----+---+---+------+------+ * | 0 | Q | U | 0 1 1 1 1 | size | L | M | Rm | opc | H | 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_simd_indexed(DisasContext *s, uint32_t insn) } } -/* C3.6.19 Crypto AES +/* Crypto AES * 31 24 23 22 21 17 16 12 11 10 9 5 4 0 * +-----------------+------+-----------+--------+-----+------+------+ * | 0 1 0 0 1 1 1 0 | size | 1 0 1 0 0 | opcode | 1 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_crypto_aes(DisasContext *s, uint32_t insn) tcg_temp_free_i32(tcg_decrypt); } -/* C3.6.20 Crypto three-reg SHA +/* Crypto three-reg SHA * 31 24 23 22 21 20 16 15 14 12 11 10 9 5 4 0 * +-----------------+------+---+------+---+--------+-----+------+------+ * | 0 1 0 1 1 1 1 0 | size | 0 | Rm | 0 | opcode | 0 0 | Rn | Rd | @@ -XXX,XX +XXX,XX @@ static void disas_crypto_three_reg_sha(DisasContext *s, uint32_t insn) tcg_temp_free_i32(tcg_rm_regno); } -/* C3.6.21 Crypto two-reg SHA +/* Crypto two-reg SHA * 31 24 23 22 21 17 16 12 11 10 9 5 4 0 * +-----------------+------+-----------+--------+-----+------+------+ * | 0 1 0 1 1 1 1 0 | size | 1 0 1 0 0 | opcode | 1 0 | Rn | Rd | -- 2.7.4 Update the static_ops functions to use new-style mmio rather than the legacy old_mmio functions. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505580378-9044-2-git-send-email-peter.maydell@linaro.org --- hw/arm/palm.c | 30 ++++++++++-------------------- 1 file changed, 10 insertions(+), 20 deletions(-) diff --git a/hw/arm/palm.c b/hw/arm/palm.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/palm.c +++ b/hw/arm/palm.c @@ -XXX,XX +XXX,XX @@ #include "exec/address-spaces.h" #include "cpu.h" -static uint32_t static_readb(void *opaque, hwaddr offset) +static uint64_t static_read(void *opaque, hwaddr offset, unsigned size) { - uint32_t *val = (uint32_t *) opaque; - return *val >> ((offset & 3) << 3); -} + uint32_t *val = (uint32_t *)opaque; + uint32_t sizemask = 7 >> size; -static uint32_t static_readh(void *opaque, hwaddr offset) -{ - uint32_t *val = (uint32_t *) opaque; - return *val >> ((offset & 1) << 3); -} - -static uint32_t static_readw(void *opaque, hwaddr offset) -{ - uint32_t *val = (uint32_t *) opaque; - return *val >> ((offset & 0) << 3); + return *val >> ((offset & sizemask) << 3); } -static void static_write(void *opaque, hwaddr offset, - uint32_t value) +static void static_write(void *opaque, hwaddr offset, uint64_t value, + unsigned size) { #ifdef SPY printf("%s: value %08lx written at " PA_FMT "\n", @@ -XXX,XX +XXX,XX @@ static void static_write(void *opaque, hwaddr offset, } static const MemoryRegionOps static_ops = { - .old_mmio = { - .read = { static_readb, static_readh, static_readw, }, - .write = { static_write, static_write, static_write, }, - }, + .read = static_read, + .write = static_write, + .valid.min_access_size = 1, + .valid.max_access_size = 4, .endianness = DEVICE_NATIVE_ENDIAN, }; -- 2.7.4 Drop the use of old_mmio in the omap2_gpio memory ops. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505580378-9044-3-git-send-email-peter.maydell@linaro.org --- hw/gpio/omap_gpio.c | 26 ++++++++++++-------------- 1 file changed, 12 insertions(+), 14 deletions(-) diff --git a/hw/gpio/omap_gpio.c b/hw/gpio/omap_gpio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/gpio/omap_gpio.c +++ b/hw/gpio/omap_gpio.c @@ -XXX,XX +XXX,XX @@ static void omap2_gpio_module_write(void *opaque, hwaddr addr, } } -static uint32_t omap2_gpio_module_readp(void *opaque, hwaddr addr) +static uint64_t omap2_gpio_module_readp(void *opaque, hwaddr addr, + unsigned size) { return omap2_gpio_module_read(opaque, addr & ~3) >> ((addr & 3) << 3); } static void omap2_gpio_module_writep(void *opaque, hwaddr addr, - uint32_t value) + uint64_t value, unsigned size) { uint32_t cur = 0; uint32_t mask = 0xffff; + if (size == 4) { + omap2_gpio_module_write(opaque, addr, value); + return; + } + switch (addr & ~3) { case 0x00: /* GPIO_REVISION */ case 0x14: /* GPIO_SYSSTATUS */ @@ -XXX,XX +XXX,XX @@ static void omap2_gpio_module_writep(void *opaque, hwaddr addr, } static const MemoryRegionOps omap2_gpio_module_ops = { - .old_mmio = { - .read = { - omap2_gpio_module_readp, - omap2_gpio_module_readp, - omap2_gpio_module_read, - }, - .write = { - omap2_gpio_module_writep, - omap2_gpio_module_writep, - omap2_gpio_module_write, - }, - }, + .read = omap2_gpio_module_readp, + .write = omap2_gpio_module_writep, + .valid.min_access_size = 1, + .valid.max_access_size = 4, .endianness = DEVICE_NATIVE_ENDIAN, }; -- 2.7.4 Don't use the old_mmio in the memory region ops struct. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505580378-9044-4-git-send-email-peter.maydell@linaro.org --- hw/timer/omap_synctimer.c | 35 +++++++++++++++++++++-------------- 1 file changed, 21 insertions(+), 14 deletions(-) diff --git a/hw/timer/omap_synctimer.c b/hw/timer/omap_synctimer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/omap_synctimer.c +++ b/hw/timer/omap_synctimer.c @@ -XXX,XX +XXX,XX @@ static uint32_t omap_synctimer_readh(void *opaque, hwaddr addr) } } -static void omap_synctimer_write(void *opaque, hwaddr addr, - uint32_t value) +static uint64_t omap_synctimer_readfn(void *opaque, hwaddr addr, + unsigned size) +{ + switch (size) { + case 1: + return omap_badwidth_read32(opaque, addr); + case 2: + return omap_synctimer_readh(opaque, addr); + case 4: + return omap_synctimer_readw(opaque, addr); + default: + g_assert_not_reached(); + } +} + +static void omap_synctimer_writefn(void *opaque, hwaddr addr, + uint64_t value, unsigned size) { OMAP_BAD_REG(addr); } static const MemoryRegionOps omap_synctimer_ops = { - .old_mmio = { - .read = { - omap_badwidth_read32, - omap_synctimer_readh, - omap_synctimer_readw, - }, - .write = { - omap_badwidth_write32, - omap_synctimer_write, - omap_synctimer_write, - }, - }, + .read = omap_synctimer_readfn, + .write = omap_synctimer_writefn, + .valid.min_access_size = 1, + .valid.max_access_size = 4, .endianness = DEVICE_NATIVE_ENDIAN, }; -- 2.7.4 Don't use the old_mmio struct in memory region ops. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505580378-9044-5-git-send-email-peter.maydell@linaro.org --- hw/timer/omap_gptimer.c | 49 +++++++++++++++++++++++++++++++++++++------------ 1 file changed, 37 insertions(+), 12 deletions(-) diff --git a/hw/timer/omap_gptimer.c b/hw/timer/omap_gptimer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/omap_gptimer.c +++ b/hw/timer/omap_gptimer.c @@ -XXX,XX +XXX,XX @@ static void omap_gp_timer_writeh(void *opaque, hwaddr addr, s->writeh = (uint16_t) value; } +static uint64_t omap_gp_timer_readfn(void *opaque, hwaddr addr, + unsigned size) +{ + switch (size) { + case 1: + return omap_badwidth_read32(opaque, addr); + case 2: + return omap_gp_timer_readh(opaque, addr); + case 4: + return omap_gp_timer_readw(opaque, addr); + default: + g_assert_not_reached(); + } +} + +static void omap_gp_timer_writefn(void *opaque, hwaddr addr, + uint64_t value, unsigned size) +{ + switch (size) { + case 1: + omap_badwidth_write32(opaque, addr, value); + break; + case 2: + omap_gp_timer_writeh(opaque, addr, value); + break; + case 4: + omap_gp_timer_write(opaque, addr, value); + break; + default: + g_assert_not_reached(); + } +} + static const MemoryRegionOps omap_gp_timer_ops = { - .old_mmio = { - .read = { - omap_badwidth_read32, - omap_gp_timer_readh, - omap_gp_timer_readw, - }, - .write = { - omap_badwidth_write32, - omap_gp_timer_writeh, - omap_gp_timer_write, - }, - }, + .read = omap_gp_timer_readfn, + .write = omap_gp_timer_writefn, + .valid.min_access_size = 1, + .valid.max_access_size = 4, .endianness = DEVICE_NATIVE_ENDIAN, }; -- 2.7.4 Don't use old_mmio in the memory region ops struct. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505580378-9044-6-git-send-email-peter.maydell@linaro.org --- hw/i2c/omap_i2c.c | 44 ++++++++++++++++++++++++++++++++------------ 1 file changed, 32 insertions(+), 12 deletions(-) diff --git a/hw/i2c/omap_i2c.c b/hw/i2c/omap_i2c.c index XXXXXXX..XXXXXXX 100644 --- a/hw/i2c/omap_i2c.c +++ b/hw/i2c/omap_i2c.c @@ -XXX,XX +XXX,XX @@ static void omap_i2c_writeb(void *opaque, hwaddr addr, } } +static uint64_t omap_i2c_readfn(void *opaque, hwaddr addr, + unsigned size) +{ + switch (size) { + case 2: + return omap_i2c_read(opaque, addr); + default: + return omap_badwidth_read16(opaque, addr); + } +} + +static void omap_i2c_writefn(void *opaque, hwaddr addr, + uint64_t value, unsigned size) +{ + switch (size) { + case 1: + /* Only the last fifo write can be 8 bit. */ + omap_i2c_writeb(opaque, addr, value); + break; + case 2: + omap_i2c_write(opaque, addr, value); + break; + default: + omap_badwidth_write16(opaque, addr, value); + break; + } +} + static const MemoryRegionOps omap_i2c_ops = { - .old_mmio = { - .read = { - omap_badwidth_read16, - omap_i2c_read, - omap_badwidth_read16, - }, - .write = { - omap_i2c_writeb, /* Only the last fifo write can be 8 bit. */ - omap_i2c_write, - omap_badwidth_write16, - }, - }, + .read = omap_i2c_readfn, + .write = omap_i2c_writefn, + .valid.min_access_size = 1, + .valid.max_access_size = 4, .endianness = DEVICE_NATIVE_ENDIAN, }; -- 2.7.4 Don't use old_mmio in the memory region ops struct. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 1505580378-9044-7-git-send-email-peter.maydell@linaro.org --- hw/arm/omap2.c | 49 +++++++++++++++++++++++++++++++++++++------------ 1 file changed, 37 insertions(+), 12 deletions(-) diff --git a/hw/arm/omap2.c b/hw/arm/omap2.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/omap2.c +++ b/hw/arm/omap2.c @@ -XXX,XX +XXX,XX @@ static void omap_sysctl_write(void *opaque, hwaddr addr, } } +static uint64_t omap_sysctl_readfn(void *opaque, hwaddr addr, + unsigned size) +{ + switch (size) { + case 1: + return omap_sysctl_read8(opaque, addr); + case 2: + return omap_badwidth_read32(opaque, addr); /* TODO */ + case 4: + return omap_sysctl_read(opaque, addr); + default: + g_assert_not_reached(); + } +} + +static void omap_sysctl_writefn(void *opaque, hwaddr addr, + uint64_t value, unsigned size) +{ + switch (size) { + case 1: + omap_sysctl_write8(opaque, addr, value); + break; + case 2: + omap_badwidth_write32(opaque, addr, value); /* TODO */ + break; + case 4: + omap_sysctl_write(opaque, addr, value); + break; + default: + g_assert_not_reached(); + } +} + static const MemoryRegionOps omap_sysctl_ops = { - .old_mmio = { - .read = { - omap_sysctl_read8, - omap_badwidth_read32, /* TODO */ - omap_sysctl_read, - }, - .write = { - omap_sysctl_write8, - omap_badwidth_write32, /* TODO */ - omap_sysctl_write, - }, - }, + .read = omap_sysctl_readfn, + .write = omap_sysctl_writefn, + .valid.min_access_size = 1, + .valid.max_access_size = 4, .endianness = DEVICE_NATIVE_ENDIAN, }; -- 2.7.4 From: Subbaraya Sundeep <sundeep.lkml@gmail.com> Modelled System Timer in Microsemi's Smartfusion2 Soc. Timer has two 32bit down counters and two interrupts. Signed-off-by: Subbaraya Sundeep <sundeep.lkml@gmail.com> Reviewed-by: Alistair Francis <alistair.francis@xilinx.com> Acked-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20170920201737.25723-2-f4bug@amsat.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/timer/Makefile.objs | 1 + include/hw/timer/mss-timer.h | 64 ++++++++++ hw/timer/mss-timer.c | 289 +++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 354 insertions(+) create mode 100644 include/hw/timer/mss-timer.h create mode 100644 hw/timer/mss-timer.c diff --git a/hw/timer/Makefile.objs b/hw/timer/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/Makefile.objs +++ b/hw/timer/Makefile.objs @@ -XXX,XX +XXX,XX @@ common-obj-$(CONFIG_ASPEED_SOC) += aspeed_timer.o common-obj-$(CONFIG_SUN4V_RTC) += sun4v-rtc.o common-obj-$(CONFIG_CMSDK_APB_TIMER) += cmsdk-apb-timer.o +common-obj-$(CONFIG_MSF2) += mss-timer.o diff --git a/include/hw/timer/mss-timer.h b/include/hw/timer/mss-timer.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/timer/mss-timer.h @@ -XXX,XX +XXX,XX @@ +/* + * Microsemi SmartFusion2 Timer. + * + * Copyright (c) 2017 Subbaraya Sundeep <sundeep.lkml@gmail.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#ifndef HW_MSS_TIMER_H +#define HW_MSS_TIMER_H + +#include "hw/sysbus.h" +#include "hw/ptimer.h" + +#define TYPE_MSS_TIMER "mss-timer" +#define MSS_TIMER(obj) OBJECT_CHECK(MSSTimerState, \ + (obj), TYPE_MSS_TIMER) + +/* + * There are two 32-bit down counting timers. + * Timers 1 and 2 can be concatenated into a single 64-bit Timer + * that operates either in Periodic mode or in One-shot mode. + * Writing 1 to the TIM64_MODE register bit 0 sets the Timers in 64-bit mode. + * In 64-bit mode, writing to the 32-bit registers has no effect. + * Similarly, in 32-bit mode, writing to the 64-bit mode registers + * has no effect. Only two 32-bit timers are supported currently. + */ +#define NUM_TIMERS 2 + +#define R_TIM1_MAX 6 + +struct Msf2Timer { + QEMUBH *bh; + ptimer_state *ptimer; + + uint32_t regs[R_TIM1_MAX]; + qemu_irq irq; +}; + +typedef struct MSSTimerState { + SysBusDevice parent_obj; + + MemoryRegion mmio; + uint32_t freq_hz; + struct Msf2Timer timers[NUM_TIMERS]; +} MSSTimerState; + +#endif /* HW_MSS_TIMER_H */ diff --git a/hw/timer/mss-timer.c b/hw/timer/mss-timer.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/timer/mss-timer.c @@ -XXX,XX +XXX,XX @@ +/* + * Block model of System timer present in + * Microsemi's SmartFusion2 and SmartFusion SoCs. + * + * Copyright (c) 2017 Subbaraya Sundeep <sundeep.lkml@gmail.com>. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "qemu/main-loop.h" +#include "qemu/log.h" +#include "hw/timer/mss-timer.h" + +#ifndef MSS_TIMER_ERR_DEBUG +#define MSS_TIMER_ERR_DEBUG 0 +#endif + +#define DB_PRINT_L(lvl, fmt, args...) do { \ + if (MSS_TIMER_ERR_DEBUG >= lvl) { \ + qemu_log("%s: " fmt "\n", __func__, ## args); \ + } \ +} while (0); + +#define DB_PRINT(fmt, args...) DB_PRINT_L(1, fmt, ## args) + +#define R_TIM_VAL 0 +#define R_TIM_LOADVAL 1 +#define R_TIM_BGLOADVAL 2 +#define R_TIM_CTRL 3 +#define R_TIM_RIS 4 +#define R_TIM_MIS 5 + +#define TIMER_CTRL_ENBL (1 << 0) +#define TIMER_CTRL_ONESHOT (1 << 1) +#define TIMER_CTRL_INTR (1 << 2) +#define TIMER_RIS_ACK (1 << 0) +#define TIMER_RST_CLR (1 << 6) +#define TIMER_MODE (1 << 0) + +static void timer_update_irq(struct Msf2Timer *st) +{ + bool isr, ier; + + isr = !!(st->regs[R_TIM_RIS] & TIMER_RIS_ACK); + ier = !!(st->regs[R_TIM_CTRL] & TIMER_CTRL_INTR); + qemu_set_irq(st->irq, (ier && isr)); +} + +static void timer_update(struct Msf2Timer *st) +{ + uint64_t count; + + if (!(st->regs[R_TIM_CTRL] & TIMER_CTRL_ENBL)) { + ptimer_stop(st->ptimer); + return; + } + + count = st->regs[R_TIM_LOADVAL]; + ptimer_set_limit(st->ptimer, count, 1); + ptimer_run(st->ptimer, 1); +} + +static uint64_t +timer_read(void *opaque, hwaddr offset, unsigned int size) +{ + MSSTimerState *t = opaque; + hwaddr addr; + struct Msf2Timer *st; + uint32_t ret = 0; + int timer = 0; + int isr; + int ier; + + addr = offset >> 2; + /* + * Two independent timers has same base address. + * Based on address passed figure out which timer is being used. + */ + if ((addr >= R_TIM1_MAX) && (addr < NUM_TIMERS * R_TIM1_MAX)) { + timer = 1; + addr -= R_TIM1_MAX; + } + + st = &t->timers[timer]; + + switch (addr) { + case R_TIM_VAL: + ret = ptimer_get_count(st->ptimer); + break; + + case R_TIM_MIS: + isr = !!(st->regs[R_TIM_RIS] & TIMER_RIS_ACK); + ier = !!(st->regs[R_TIM_CTRL] & TIMER_CTRL_INTR); + ret = ier & isr; + break; + + default: + if (addr < R_TIM1_MAX) { + ret = st->regs[addr]; + } else { + qemu_log_mask(LOG_GUEST_ERROR, + TYPE_MSS_TIMER": 64-bit mode not supported\n"); + return ret; + } + break; + } + + DB_PRINT("timer=%d 0x%" HWADDR_PRIx "=0x%" PRIx32, timer, offset, + ret); + return ret; +} + +static void +timer_write(void *opaque, hwaddr offset, + uint64_t val64, unsigned int size) +{ + MSSTimerState *t = opaque; + hwaddr addr; + struct Msf2Timer *st; + int timer = 0; + uint32_t value = val64; + + addr = offset >> 2; + /* + * Two independent timers has same base address. + * Based on addr passed figure out which timer is being used. + */ + if ((addr >= R_TIM1_MAX) && (addr < NUM_TIMERS * R_TIM1_MAX)) { + timer = 1; + addr -= R_TIM1_MAX; + } + + st = &t->timers[timer]; + + DB_PRINT("addr=0x%" HWADDR_PRIx " val=0x%" PRIx32 " (timer=%d)", offset, + value, timer); + + switch (addr) { + case R_TIM_CTRL: + st->regs[R_TIM_CTRL] = value; + timer_update(st); + break; + + case R_TIM_RIS: + if (value & TIMER_RIS_ACK) { + st->regs[R_TIM_RIS] &= ~TIMER_RIS_ACK; + } + break; + + case R_TIM_LOADVAL: + st->regs[R_TIM_LOADVAL] = value; + if (st->regs[R_TIM_CTRL] & TIMER_CTRL_ENBL) { + timer_update(st); + } + break; + + case R_TIM_BGLOADVAL: + st->regs[R_TIM_BGLOADVAL] = value; + st->regs[R_TIM_LOADVAL] = value; + break; + + case R_TIM_VAL: + case R_TIM_MIS: + break; + + default: + if (addr < R_TIM1_MAX) { + st->regs[addr] = value; + } else { + qemu_log_mask(LOG_GUEST_ERROR, + TYPE_MSS_TIMER": 64-bit mode not supported\n"); + return; + } + break; + } + timer_update_irq(st); +} + +static const MemoryRegionOps timer_ops = { + .read = timer_read, + .write = timer_write, + .endianness = DEVICE_NATIVE_ENDIAN, + .valid = { + .min_access_size = 1, + .max_access_size = 4 + } +}; + +static void timer_hit(void *opaque) +{ + struct Msf2Timer *st = opaque; + + st->regs[R_TIM_RIS] |= TIMER_RIS_ACK; + + if (!(st->regs[R_TIM_CTRL] & TIMER_CTRL_ONESHOT)) { + timer_update(st); + } + timer_update_irq(st); +} + +static void mss_timer_init(Object *obj) +{ + MSSTimerState *t = MSS_TIMER(obj); + int i; + + /* Init all the ptimers. */ + for (i = 0; i < NUM_TIMERS; i++) { + struct Msf2Timer *st = &t->timers[i]; + + st->bh = qemu_bh_new(timer_hit, st); + st->ptimer = ptimer_init(st->bh, PTIMER_POLICY_DEFAULT); + ptimer_set_freq(st->ptimer, t->freq_hz); + sysbus_init_irq(SYS_BUS_DEVICE(obj), &st->irq); + } + + memory_region_init_io(&t->mmio, OBJECT(t), &timer_ops, t, TYPE_MSS_TIMER, + NUM_TIMERS * R_TIM1_MAX * 4); + sysbus_init_mmio(SYS_BUS_DEVICE(obj), &t->mmio); +} + +static const VMStateDescription vmstate_timers = { + .name = "mss-timer-block", + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_PTIMER(ptimer, struct Msf2Timer), + VMSTATE_UINT32_ARRAY(regs, struct Msf2Timer, R_TIM1_MAX), + VMSTATE_END_OF_LIST() + } +}; + +static const VMStateDescription vmstate_mss_timer = { + .name = TYPE_MSS_TIMER, + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32(freq_hz, MSSTimerState), + VMSTATE_STRUCT_ARRAY(timers, MSSTimerState, NUM_TIMERS, 0, + vmstate_timers, struct Msf2Timer), + VMSTATE_END_OF_LIST() + } +}; + +static Property mss_timer_properties[] = { + /* Libero GUI shows 100Mhz as default for clocks */ + DEFINE_PROP_UINT32("clock-frequency", MSSTimerState, freq_hz, + 100 * 1000000), + DEFINE_PROP_END_OF_LIST(), +}; + +static void mss_timer_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->props = mss_timer_properties; + dc->vmsd = &vmstate_mss_timer; +} + +static const TypeInfo mss_timer_info = { + .name = TYPE_MSS_TIMER, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(MSSTimerState), + .instance_init = mss_timer_init, + .class_init = mss_timer_class_init, +}; + +static void mss_timer_register_types(void) +{ + type_register_static(&mss_timer_info); +} + +type_init(mss_timer_register_types) -- 2.7.4 From: Subbaraya Sundeep <sundeep.lkml@gmail.com> Added Sytem register block of Smartfusion2. This block has PLL registers which are accessed by guest. Signed-off-by: Subbaraya Sundeep <sundeep.lkml@gmail.com> Reviewed-by: Alistair Francis <alistair.francis@xilinx.com> Acked-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20170920201737.25723-3-f4bug@amsat.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/misc/Makefile.objs | 1 + include/hw/misc/msf2-sysreg.h | 77 ++++++++++++++++++++ hw/misc/msf2-sysreg.c | 160 ++++++++++++++++++++++++++++++++++++++++++ hw/misc/trace-events | 5 ++ 4 files changed, 243 insertions(+) create mode 100644 include/hw/misc/msf2-sysreg.h create mode 100644 hw/misc/msf2-sysreg.c diff --git a/hw/misc/Makefile.objs b/hw/misc/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/Makefile.objs +++ b/hw/misc/Makefile.objs @@ -XXX,XX +XXX,XX @@ obj-$(CONFIG_HYPERV_TESTDEV) += hyperv_testdev.o obj-$(CONFIG_AUX) += auxbus.o obj-$(CONFIG_ASPEED_SOC) += aspeed_scu.o aspeed_sdmc.o obj-y += mmio_interface.o +obj-$(CONFIG_MSF2) += msf2-sysreg.o diff --git a/include/hw/misc/msf2-sysreg.h b/include/hw/misc/msf2-sysreg.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/misc/msf2-sysreg.h @@ -XXX,XX +XXX,XX @@ +/* + * Microsemi SmartFusion2 SYSREG + * + * Copyright (c) 2017 Subbaraya Sundeep <sundeep.lkml@gmail.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#ifndef HW_MSF2_SYSREG_H +#define HW_MSF2_SYSREG_H + +#include "hw/sysbus.h" + +enum { + ESRAM_CR = 0x00 / 4, + ESRAM_MAX_LAT, + DDR_CR, + ENVM_CR, + ENVM_REMAP_BASE_CR, + ENVM_REMAP_FAB_CR, + CC_CR, + CC_REGION_CR, + CC_LOCK_BASE_ADDR_CR, + CC_FLUSH_INDX_CR, + DDRB_BUF_TIMER_CR, + DDRB_NB_ADDR_CR, + DDRB_NB_SIZE_CR, + DDRB_CR, + + SOFT_RESET_CR = 0x48 / 4, + M3_CR, + + GPIO_SYSRESET_SEL_CR = 0x58 / 4, + + MDDR_CR = 0x60 / 4, + + MSSDDR_PLL_STATUS_LOW_CR = 0x90 / 4, + MSSDDR_PLL_STATUS_HIGH_CR, + MSSDDR_FACC1_CR, + MSSDDR_FACC2_CR, + + MSSDDR_PLL_STATUS = 0x150 / 4, +}; + +#define MSF2_SYSREG_MMIO_SIZE 0x300 + +#define TYPE_MSF2_SYSREG "msf2-sysreg" +#define MSF2_SYSREG(obj) OBJECT_CHECK(MSF2SysregState, (obj), TYPE_MSF2_SYSREG) + +typedef struct MSF2SysregState { + SysBusDevice parent_obj; + + MemoryRegion iomem; + + uint8_t apb0div; + uint8_t apb1div; + + uint32_t regs[MSF2_SYSREG_MMIO_SIZE / 4]; +} MSF2SysregState; + +#endif /* HW_MSF2_SYSREG_H */ diff --git a/hw/misc/msf2-sysreg.c b/hw/misc/msf2-sysreg.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/misc/msf2-sysreg.c @@ -XXX,XX +XXX,XX @@ +/* + * System Register block model of Microsemi SmartFusion2. + * + * Copyright (c) 2017 Subbaraya Sundeep <sundeep.lkml@gmail.com> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "qapi/error.h" +#include "qemu/log.h" +#include "hw/misc/msf2-sysreg.h" +#include "qemu/error-report.h" +#include "trace.h" + +static inline int msf2_divbits(uint32_t div) +{ + int r = ctz32(div); + + return (div < 8) ? r : r + 1; +} + +static void msf2_sysreg_reset(DeviceState *d) +{ + MSF2SysregState *s = MSF2_SYSREG(d); + + s->regs[MSSDDR_PLL_STATUS_LOW_CR] = 0x021A2358; + s->regs[MSSDDR_PLL_STATUS] = 0x3; + s->regs[MSSDDR_FACC1_CR] = msf2_divbits(s->apb0div) << 5 | + msf2_divbits(s->apb1div) << 2; +} + +static uint64_t msf2_sysreg_read(void *opaque, hwaddr offset, + unsigned size) +{ + MSF2SysregState *s = opaque; + uint32_t ret = 0; + + offset >>= 2; + if (offset < ARRAY_SIZE(s->regs)) { + ret = s->regs[offset]; + trace_msf2_sysreg_read(offset << 2, ret); + } else { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: Bad offset 0x%08" HWADDR_PRIx "\n", __func__, + offset << 2); + } + + return ret; +} + +static void msf2_sysreg_write(void *opaque, hwaddr offset, + uint64_t val, unsigned size) +{ + MSF2SysregState *s = opaque; + uint32_t newval = val; + + offset >>= 2; + + switch (offset) { + case MSSDDR_PLL_STATUS: + trace_msf2_sysreg_write_pll_status(); + break; + + case ESRAM_CR: + case DDR_CR: + case ENVM_REMAP_BASE_CR: + if (newval != s->regs[offset]) { + qemu_log_mask(LOG_UNIMP, + TYPE_MSF2_SYSREG": remapping not supported\n"); + } + break; + + default: + if (offset < ARRAY_SIZE(s->regs)) { + trace_msf2_sysreg_write(offset << 2, newval, s->regs[offset]); + s->regs[offset] = newval; + } else { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: Bad offset 0x%08" HWADDR_PRIx "\n", __func__, + offset << 2); + } + break; + } +} + +static const MemoryRegionOps sysreg_ops = { + .read = msf2_sysreg_read, + .write = msf2_sysreg_write, + .endianness = DEVICE_NATIVE_ENDIAN, +}; + +static void msf2_sysreg_init(Object *obj) +{ + MSF2SysregState *s = MSF2_SYSREG(obj); + + memory_region_init_io(&s->iomem, obj, &sysreg_ops, s, TYPE_MSF2_SYSREG, + MSF2_SYSREG_MMIO_SIZE); + sysbus_init_mmio(SYS_BUS_DEVICE(obj), &s->iomem); +} + +static const VMStateDescription vmstate_msf2_sysreg = { + .name = TYPE_MSF2_SYSREG, + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32_ARRAY(regs, MSF2SysregState, MSF2_SYSREG_MMIO_SIZE / 4), + VMSTATE_END_OF_LIST() + } +}; + +static Property msf2_sysreg_properties[] = { + /* default divisors in Libero GUI */ + DEFINE_PROP_UINT8("apb0divisor", MSF2SysregState, apb0div, 2), + DEFINE_PROP_UINT8("apb1divisor", MSF2SysregState, apb1div, 2), + DEFINE_PROP_END_OF_LIST(), +}; + +static void msf2_sysreg_realize(DeviceState *dev, Error **errp) +{ + MSF2SysregState *s = MSF2_SYSREG(dev); + + if ((s->apb0div > 32 || !is_power_of_2(s->apb0div)) + || (s->apb1div > 32 || !is_power_of_2(s->apb1div))) { + error_setg(errp, "Invalid apb divisor value"); + error_append_hint(errp, "apb divisor must be a power of 2" + " and maximum value is 32\n"); + } +} + +static void msf2_sysreg_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->vmsd = &vmstate_msf2_sysreg; + dc->reset = msf2_sysreg_reset; + dc->props = msf2_sysreg_properties; + dc->realize = msf2_sysreg_realize; +} + +static const TypeInfo msf2_sysreg_info = { + .name = TYPE_MSF2_SYSREG, + .parent = TYPE_SYS_BUS_DEVICE, + .class_init = msf2_sysreg_class_init, + .instance_size = sizeof(MSF2SysregState), + .instance_init = msf2_sysreg_init, +}; + +static void msf2_sysreg_register_types(void) +{ + type_register_static(&msf2_sysreg_info); +} + +type_init(msf2_sysreg_register_types) diff --git a/hw/misc/trace-events b/hw/misc/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/trace-events +++ b/hw/misc/trace-events @@ -XXX,XX +XXX,XX @@ mps2_scc_reset(void) "MPS2 SCC: reset" mps2_scc_leds(char led7, char led6, char led5, char led4, char led3, char led2, char led1, char led0) "MPS2 SCC LEDs: %c%c%c%c%c%c%c%c" mps2_scc_cfg_write(unsigned function, unsigned device, uint32_t value) "MPS2 SCC config write: function %d device %d data 0x%" PRIx32 mps2_scc_cfg_read(unsigned function, unsigned device, uint32_t value) "MPS2 SCC config read: function %d device %d data 0x%" PRIx32 + +# hw/misc/msf2-sysreg.c +msf2_sysreg_write(uint64_t offset, uint32_t val, uint32_t prev) "msf2-sysreg write: addr 0x%08" HWADDR_PRIx " data 0x%" PRIx32 " prev 0x%" PRIx32 +msf2_sysreg_read(uint64_t offset, uint32_t val) "msf2-sysreg read: addr 0x%08" HWADDR_PRIx " data 0x%08" PRIx32 +msf2_sysreg_write_pll_status(void) "Invalid write to read only PLL status register" -- 2.7.4 From: Subbaraya Sundeep <sundeep.lkml@gmail.com> Modelled Microsemi's Smartfusion2 SPI controller. Signed-off-by: Subbaraya Sundeep <sundeep.lkml@gmail.com> Reviewed-by: Alistair Francis <alistair.francis@xilinx.com> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20170920201737.25723-4-f4bug@amsat.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/ssi/Makefile.objs | 1 + include/hw/ssi/mss-spi.h | 58 +++++++ hw/ssi/mss-spi.c | 404 +++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 463 insertions(+) create mode 100644 include/hw/ssi/mss-spi.h create mode 100644 hw/ssi/mss-spi.c diff --git a/hw/ssi/Makefile.objs b/hw/ssi/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/Makefile.objs +++ b/hw/ssi/Makefile.objs @@ -XXX,XX +XXX,XX @@ common-obj-$(CONFIG_XILINX_SPI) += xilinx_spi.o common-obj-$(CONFIG_XILINX_SPIPS) += xilinx_spips.o common-obj-$(CONFIG_ASPEED_SOC) += aspeed_smc.o common-obj-$(CONFIG_STM32F2XX_SPI) += stm32f2xx_spi.o +common-obj-$(CONFIG_MSF2) += mss-spi.o obj-$(CONFIG_OMAP) += omap_spi.o obj-$(CONFIG_IMX) += imx_spi.o diff --git a/include/hw/ssi/mss-spi.h b/include/hw/ssi/mss-spi.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/ssi/mss-spi.h @@ -XXX,XX +XXX,XX @@ +/* + * Microsemi SmartFusion2 SPI + * + * Copyright (c) 2017 Subbaraya Sundeep <sundeep.lkml@gmail.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#ifndef HW_MSS_SPI_H +#define HW_MSS_SPI_H + +#include "hw/sysbus.h" +#include "hw/ssi/ssi.h" +#include "qemu/fifo32.h" + +#define TYPE_MSS_SPI "mss-spi" +#define MSS_SPI(obj) OBJECT_CHECK(MSSSpiState, (obj), TYPE_MSS_SPI) + +#define R_SPI_MAX 16 + +typedef struct MSSSpiState { + SysBusDevice parent_obj; + + MemoryRegion mmio; + + qemu_irq irq; + + qemu_irq cs_line; + + SSIBus *spi; + + Fifo32 rx_fifo; + Fifo32 tx_fifo; + + int fifo_depth; + uint32_t frame_count; + bool enabled; + + uint32_t regs[R_SPI_MAX]; +} MSSSpiState; + +#endif /* HW_MSS_SPI_H */ diff --git a/hw/ssi/mss-spi.c b/hw/ssi/mss-spi.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/ssi/mss-spi.c @@ -XXX,XX +XXX,XX @@ +/* + * Block model of SPI controller present in + * Microsemi's SmartFusion2 and SmartFusion SoCs. + * + * Copyright (C) 2017 Subbaraya Sundeep <sundeep.lkml@gmail.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "hw/ssi/mss-spi.h" +#include "qemu/log.h" + +#ifndef MSS_SPI_ERR_DEBUG +#define MSS_SPI_ERR_DEBUG 0 +#endif + +#define DB_PRINT_L(lvl, fmt, args...) do { \ + if (MSS_SPI_ERR_DEBUG >= lvl) { \ + qemu_log("%s: " fmt "\n", __func__, ## args); \ + } \ +} while (0); + +#define DB_PRINT(fmt, args...) DB_PRINT_L(1, fmt, ## args) + +#define FIFO_CAPACITY 32 + +#define R_SPI_CONTROL 0 +#define R_SPI_DFSIZE 1 +#define R_SPI_STATUS 2 +#define R_SPI_INTCLR 3 +#define R_SPI_RX 4 +#define R_SPI_TX 5 +#define R_SPI_CLKGEN 6 +#define R_SPI_SS 7 +#define R_SPI_MIS 8 +#define R_SPI_RIS 9 + +#define S_TXDONE (1 << 0) +#define S_RXRDY (1 << 1) +#define S_RXCHOVRF (1 << 2) +#define S_RXFIFOFUL (1 << 4) +#define S_RXFIFOFULNXT (1 << 5) +#define S_RXFIFOEMP (1 << 6) +#define S_RXFIFOEMPNXT (1 << 7) +#define S_TXFIFOFUL (1 << 8) +#define S_TXFIFOFULNXT (1 << 9) +#define S_TXFIFOEMP (1 << 10) +#define S_TXFIFOEMPNXT (1 << 11) +#define S_FRAMESTART (1 << 12) +#define S_SSEL (1 << 13) +#define S_ACTIVE (1 << 14) + +#define C_ENABLE (1 << 0) +#define C_MODE (1 << 1) +#define C_INTRXDATA (1 << 4) +#define C_INTTXDATA (1 << 5) +#define C_INTRXOVRFLO (1 << 6) +#define C_SPS (1 << 26) +#define C_BIGFIFO (1 << 29) +#define C_RESET (1 << 31) + +#define FRAMESZ_MASK 0x1F +#define FMCOUNT_MASK 0x00FFFF00 +#define FMCOUNT_SHIFT 8 + +static void txfifo_reset(MSSSpiState *s) +{ + fifo32_reset(&s->tx_fifo); + + s->regs[R_SPI_STATUS] &= ~S_TXFIFOFUL; + s->regs[R_SPI_STATUS] |= S_TXFIFOEMP; +} + +static void rxfifo_reset(MSSSpiState *s) +{ + fifo32_reset(&s->rx_fifo); + + s->regs[R_SPI_STATUS] &= ~S_RXFIFOFUL; + s->regs[R_SPI_STATUS] |= S_RXFIFOEMP; +} + +static void set_fifodepth(MSSSpiState *s) +{ + unsigned int size = s->regs[R_SPI_DFSIZE] & FRAMESZ_MASK; + + if (size <= 8) { + s->fifo_depth = 32; + } else if (size <= 16) { + s->fifo_depth = 16; + } else if (size <= 32) { + s->fifo_depth = 8; + } else { + s->fifo_depth = 4; + } +} + +static void update_mis(MSSSpiState *s) +{ + uint32_t reg = s->regs[R_SPI_CONTROL]; + uint32_t tmp; + + /* + * form the Control register interrupt enable bits + * same as RIS, MIS and Interrupt clear registers for simplicity + */ + tmp = ((reg & C_INTRXOVRFLO) >> 4) | ((reg & C_INTRXDATA) >> 3) | + ((reg & C_INTTXDATA) >> 5); + s->regs[R_SPI_MIS] |= tmp & s->regs[R_SPI_RIS]; +} + +static void spi_update_irq(MSSSpiState *s) +{ + int irq; + + update_mis(s); + irq = !!(s->regs[R_SPI_MIS]); + + qemu_set_irq(s->irq, irq); +} + +static void mss_spi_reset(DeviceState *d) +{ + MSSSpiState *s = MSS_SPI(d); + + memset(s->regs, 0, sizeof s->regs); + s->regs[R_SPI_CONTROL] = 0x80000102; + s->regs[R_SPI_DFSIZE] = 0x4; + s->regs[R_SPI_STATUS] = S_SSEL | S_TXFIFOEMP | S_RXFIFOEMP; + s->regs[R_SPI_CLKGEN] = 0x7; + s->regs[R_SPI_RIS] = 0x0; + + s->fifo_depth = 4; + s->frame_count = 1; + s->enabled = false; + + rxfifo_reset(s); + txfifo_reset(s); +} + +static uint64_t +spi_read(void *opaque, hwaddr addr, unsigned int size) +{ + MSSSpiState *s = opaque; + uint32_t ret = 0; + + addr >>= 2; + switch (addr) { + case R_SPI_RX: + s->regs[R_SPI_STATUS] &= ~S_RXFIFOFUL; + s->regs[R_SPI_STATUS] &= ~S_RXCHOVRF; + ret = fifo32_pop(&s->rx_fifo); + if (fifo32_is_empty(&s->rx_fifo)) { + s->regs[R_SPI_STATUS] |= S_RXFIFOEMP; + } + break; + + case R_SPI_MIS: + update_mis(s); + ret = s->regs[R_SPI_MIS]; + break; + + default: + if (addr < ARRAY_SIZE(s->regs)) { + ret = s->regs[addr]; + } else { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: Bad offset 0x%" HWADDR_PRIx "\n", __func__, + addr * 4); + return ret; + } + break; + } + + DB_PRINT("addr=0x%" HWADDR_PRIx " = 0x%" PRIx32, addr * 4, ret); + spi_update_irq(s); + return ret; +} + +static void assert_cs(MSSSpiState *s) +{ + qemu_set_irq(s->cs_line, 0); +} + +static void deassert_cs(MSSSpiState *s) +{ + qemu_set_irq(s->cs_line, 1); +} + +static void spi_flush_txfifo(MSSSpiState *s) +{ + uint32_t tx; + uint32_t rx; + bool sps = !!(s->regs[R_SPI_CONTROL] & C_SPS); + + /* + * Chip Select(CS) is automatically controlled by this controller. + * If SPS bit is set in Control register then CS is asserted + * until all the frames set in frame count of Control register are + * transferred. If SPS is not set then CS pulses between frames. + * Note that Slave Select register specifies which of the CS line + * has to be controlled automatically by controller. Bits SS[7:1] are for + * masters in FPGA fabric since we model only Microcontroller subsystem + * of Smartfusion2 we control only one CS(SS[0]) line. + */ + while (!fifo32_is_empty(&s->tx_fifo) && s->frame_count) { + assert_cs(s); + + s->regs[R_SPI_STATUS] &= ~(S_TXDONE | S_RXRDY); + + tx = fifo32_pop(&s->tx_fifo); + DB_PRINT("data tx:0x%" PRIx32, tx); + rx = ssi_transfer(s->spi, tx); + DB_PRINT("data rx:0x%" PRIx32, rx); + + if (fifo32_num_used(&s->rx_fifo) == s->fifo_depth) { + s->regs[R_SPI_STATUS] |= S_RXCHOVRF; + s->regs[R_SPI_RIS] |= S_RXCHOVRF; + } else { + fifo32_push(&s->rx_fifo, rx); + s->regs[R_SPI_STATUS] &= ~S_RXFIFOEMP; + if (fifo32_num_used(&s->rx_fifo) == (s->fifo_depth - 1)) { + s->regs[R_SPI_STATUS] |= S_RXFIFOFULNXT; + } else if (fifo32_num_used(&s->rx_fifo) == s->fifo_depth) { + s->regs[R_SPI_STATUS] |= S_RXFIFOFUL; + } + } + s->frame_count--; + if (!sps) { + deassert_cs(s); + } + } + + if (!s->frame_count) { + s->frame_count = (s->regs[R_SPI_CONTROL] & FMCOUNT_MASK) >> + FMCOUNT_SHIFT; + deassert_cs(s); + s->regs[R_SPI_RIS] |= S_TXDONE | S_RXRDY; + s->regs[R_SPI_STATUS] |= S_TXDONE | S_RXRDY; + } +} + +static void spi_write(void *opaque, hwaddr addr, + uint64_t val64, unsigned int size) +{ + MSSSpiState *s = opaque; + uint32_t value = val64; + + DB_PRINT("addr=0x%" HWADDR_PRIx " =0x%" PRIx32, addr, value); + addr >>= 2; + + switch (addr) { + case R_SPI_TX: + /* adding to already full FIFO */ + if (fifo32_num_used(&s->tx_fifo) == s->fifo_depth) { + break; + } + s->regs[R_SPI_STATUS] &= ~S_TXFIFOEMP; + fifo32_push(&s->tx_fifo, value); + if (fifo32_num_used(&s->tx_fifo) == (s->fifo_depth - 1)) { + s->regs[R_SPI_STATUS] |= S_TXFIFOFULNXT; + } else if (fifo32_num_used(&s->tx_fifo) == s->fifo_depth) { + s->regs[R_SPI_STATUS] |= S_TXFIFOFUL; + } + if (s->enabled) { + spi_flush_txfifo(s); + } + break; + + case R_SPI_CONTROL: + s->regs[R_SPI_CONTROL] = value; + if (value & C_BIGFIFO) { + set_fifodepth(s); + } else { + s->fifo_depth = 4; + } + s->enabled = value & C_ENABLE; + s->frame_count = (value & FMCOUNT_MASK) >> FMCOUNT_SHIFT; + if (value & C_RESET) { + mss_spi_reset(DEVICE(s)); + } + break; + + case R_SPI_DFSIZE: + if (s->enabled) { + break; + } + s->regs[R_SPI_DFSIZE] = value; + break; + + case R_SPI_INTCLR: + s->regs[R_SPI_INTCLR] = value; + if (value & S_TXDONE) { + s->regs[R_SPI_RIS] &= ~S_TXDONE; + } + if (value & S_RXRDY) { + s->regs[R_SPI_RIS] &= ~S_RXRDY; + } + if (value & S_RXCHOVRF) { + s->regs[R_SPI_RIS] &= ~S_RXCHOVRF; + } + break; + + case R_SPI_MIS: + case R_SPI_STATUS: + case R_SPI_RIS: + qemu_log_mask(LOG_GUEST_ERROR, + "%s: Write to read only register 0x%" HWADDR_PRIx "\n", + __func__, addr * 4); + break; + + default: + if (addr < ARRAY_SIZE(s->regs)) { + s->regs[addr] = value; + } else { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: Bad offset 0x%" HWADDR_PRIx "\n", __func__, + addr * 4); + } + break; + } + + spi_update_irq(s); +} + +static const MemoryRegionOps spi_ops = { + .read = spi_read, + .write = spi_write, + .endianness = DEVICE_NATIVE_ENDIAN, + .valid = { + .min_access_size = 1, + .max_access_size = 4 + } +}; + +static void mss_spi_realize(DeviceState *dev, Error **errp) +{ + MSSSpiState *s = MSS_SPI(dev); + SysBusDevice *sbd = SYS_BUS_DEVICE(dev); + + s->spi = ssi_create_bus(dev, "spi"); + + sysbus_init_irq(sbd, &s->irq); + ssi_auto_connect_slaves(dev, &s->cs_line, s->spi); + sysbus_init_irq(sbd, &s->cs_line); + + memory_region_init_io(&s->mmio, OBJECT(s), &spi_ops, s, + TYPE_MSS_SPI, R_SPI_MAX * 4); + sysbus_init_mmio(sbd, &s->mmio); + + fifo32_create(&s->tx_fifo, FIFO_CAPACITY); + fifo32_create(&s->rx_fifo, FIFO_CAPACITY); +} + +static const VMStateDescription vmstate_mss_spi = { + .name = TYPE_MSS_SPI, + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_FIFO32(tx_fifo, MSSSpiState), + VMSTATE_FIFO32(rx_fifo, MSSSpiState), + VMSTATE_UINT32_ARRAY(regs, MSSSpiState, R_SPI_MAX), + VMSTATE_END_OF_LIST() + } +}; + +static void mss_spi_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->realize = mss_spi_realize; + dc->reset = mss_spi_reset; + dc->vmsd = &vmstate_mss_spi; +} + +static const TypeInfo mss_spi_info = { + .name = TYPE_MSS_SPI, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(MSSSpiState), + .class_init = mss_spi_class_init, +}; + +static void mss_spi_register_types(void) +{ + type_register_static(&mss_spi_info); +} + +type_init(mss_spi_register_types) -- 2.7.4 From: Subbaraya Sundeep <sundeep.lkml@gmail.com> Smartfusion2 SoC has hardened Microcontroller subsystem and flash based FPGA fabric. This patch adds support for Microcontroller subsystem in the SoC. Signed-off-by: Subbaraya Sundeep <sundeep.lkml@gmail.com> Reviewed-by: Alistair Francis <alistair.francis@xilinx.com> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20170920201737.25723-5-f4bug@amsat.org [PMD: drop cpu_model to directly use cpu type, check m3clk non null] Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/Makefile.objs | 1 + include/hw/arm/msf2-soc.h | 67 +++++++++++ hw/arm/msf2-soc.c | 238 ++++++++++++++++++++++++++++++++++++++++ default-configs/arm-softmmu.mak | 1 + 4 files changed, 307 insertions(+) create mode 100644 include/hw/arm/msf2-soc.h create mode 100644 hw/arm/msf2-soc.c diff --git a/hw/arm/Makefile.objs b/hw/arm/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Makefile.objs +++ b/hw/arm/Makefile.objs @@ -XXX,XX +XXX,XX @@ obj-$(CONFIG_FSL_IMX31) += fsl-imx31.o kzm.o obj-$(CONFIG_FSL_IMX6) += fsl-imx6.o sabrelite.o obj-$(CONFIG_ASPEED_SOC) += aspeed_soc.o aspeed.o obj-$(CONFIG_MPS2) += mps2.o +obj-$(CONFIG_MSF2) += msf2-soc.o diff --git a/include/hw/arm/msf2-soc.h b/include/hw/arm/msf2-soc.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/arm/msf2-soc.h @@ -XXX,XX +XXX,XX @@ +/* + * Microsemi Smartfusion2 SoC + * + * Copyright (c) 2017 Subbaraya Sundeep <sundeep.lkml@gmail.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#ifndef HW_ARM_MSF2_SOC_H +#define HW_ARM_MSF2_SOC_H + +#include "hw/arm/armv7m.h" +#include "hw/timer/mss-timer.h" +#include "hw/misc/msf2-sysreg.h" +#include "hw/ssi/mss-spi.h" + +#define TYPE_MSF2_SOC "msf2-soc" +#define MSF2_SOC(obj) OBJECT_CHECK(MSF2State, (obj), TYPE_MSF2_SOC) + +#define MSF2_NUM_SPIS 2 +#define MSF2_NUM_UARTS 2 + +/* + * System timer consists of two programmable 32-bit + * decrementing counters that generate individual interrupts to + * the Cortex-M3 processor + */ +#define MSF2_NUM_TIMERS 2 + +typedef struct MSF2State { + /*< private >*/ + SysBusDevice parent_obj; + /*< public >*/ + + ARMv7MState armv7m; + + char *cpu_type; + char *part_name; + uint64_t envm_size; + uint64_t esram_size; + + uint32_t m3clk; + uint8_t apb0div; + uint8_t apb1div; + + MSF2SysregState sysreg; + MSSTimerState timer; + MSSSpiState spi[MSF2_NUM_SPIS]; +} MSF2State; + +#endif diff --git a/hw/arm/msf2-soc.c b/hw/arm/msf2-soc.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/arm/msf2-soc.c @@ -XXX,XX +XXX,XX @@ +/* + * SmartFusion2 SoC emulation. + * + * Copyright (c) 2017 Subbaraya Sundeep <sundeep.lkml@gmail.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "qapi/error.h" +#include "qemu-common.h" +#include "hw/arm/arm.h" +#include "exec/address-spaces.h" +#include "hw/char/serial.h" +#include "hw/boards.h" +#include "sysemu/block-backend.h" +#include "qemu/cutils.h" +#include "hw/arm/msf2-soc.h" +#include "hw/misc/unimp.h" + +#define MSF2_TIMER_BASE 0x40004000 +#define MSF2_SYSREG_BASE 0x40038000 + +#define ENVM_BASE_ADDRESS 0x60000000 + +#define SRAM_BASE_ADDRESS 0x20000000 + +#define MSF2_ENVM_MAX_SIZE (512 * K_BYTE) + +/* + * eSRAM max size is 80k without SECDED(Single error correction and + * dual error detection) feature and 64k with SECDED. + * We do not support SECDED now. + */ +#define MSF2_ESRAM_MAX_SIZE (80 * K_BYTE) + +static const uint32_t spi_addr[MSF2_NUM_SPIS] = { 0x40001000 , 0x40011000 }; +static const uint32_t uart_addr[MSF2_NUM_UARTS] = { 0x40000000 , 0x40010000 }; + +static const int spi_irq[MSF2_NUM_SPIS] = { 2, 3 }; +static const int uart_irq[MSF2_NUM_UARTS] = { 10, 11 }; +static const int timer_irq[MSF2_NUM_TIMERS] = { 14, 15 }; + +static void m2sxxx_soc_initfn(Object *obj) +{ + MSF2State *s = MSF2_SOC(obj); + int i; + + object_initialize(&s->armv7m, sizeof(s->armv7m), TYPE_ARMV7M); + qdev_set_parent_bus(DEVICE(&s->armv7m), sysbus_get_default()); + + object_initialize(&s->sysreg, sizeof(s->sysreg), TYPE_MSF2_SYSREG); + qdev_set_parent_bus(DEVICE(&s->sysreg), sysbus_get_default()); + + object_initialize(&s->timer, sizeof(s->timer), TYPE_MSS_TIMER); + qdev_set_parent_bus(DEVICE(&s->timer), sysbus_get_default()); + + for (i = 0; i < MSF2_NUM_SPIS; i++) { + object_initialize(&s->spi[i], sizeof(s->spi[i]), + TYPE_MSS_SPI); + qdev_set_parent_bus(DEVICE(&s->spi[i]), sysbus_get_default()); + } +} + +static void m2sxxx_soc_realize(DeviceState *dev_soc, Error **errp) +{ + MSF2State *s = MSF2_SOC(dev_soc); + DeviceState *dev, *armv7m; + SysBusDevice *busdev; + Error *err = NULL; + int i; + + MemoryRegion *system_memory = get_system_memory(); + MemoryRegion *nvm = g_new(MemoryRegion, 1); + MemoryRegion *nvm_alias = g_new(MemoryRegion, 1); + MemoryRegion *sram = g_new(MemoryRegion, 1); + + memory_region_init_rom(nvm, NULL, "MSF2.eNVM", s->envm_size, + &error_fatal); + /* + * On power-on, the eNVM region 0x60000000 is automatically + * remapped to the Cortex-M3 processor executable region + * start address (0x0). We do not support remapping other eNVM, + * eSRAM and DDR regions by guest(via Sysreg) currently. + */ + memory_region_init_alias(nvm_alias, NULL, "MSF2.eNVM", + nvm, 0, s->envm_size); + + memory_region_add_subregion(system_memory, ENVM_BASE_ADDRESS, nvm); + memory_region_add_subregion(system_memory, 0, nvm_alias); + + memory_region_init_ram(sram, NULL, "MSF2.eSRAM", s->esram_size, + &error_fatal); + memory_region_add_subregion(system_memory, SRAM_BASE_ADDRESS, sram); + + armv7m = DEVICE(&s->armv7m); + qdev_prop_set_uint32(armv7m, "num-irq", 81); + qdev_prop_set_string(armv7m, "cpu-type", s->cpu_type); + object_property_set_link(OBJECT(&s->armv7m), OBJECT(get_system_memory()), + "memory", &error_abort); + object_property_set_bool(OBJECT(&s->armv7m), true, "realized", &err); + if (err != NULL) { + error_propagate(errp, err); + return; + } + + if (!s->m3clk) { + error_setg(errp, "Invalid m3clk value"); + error_append_hint(errp, "m3clk can not be zero\n"); + return; + } + system_clock_scale = NANOSECONDS_PER_SECOND / s->m3clk; + + for (i = 0; i < MSF2_NUM_UARTS; i++) { + if (serial_hds[i]) { + serial_mm_init(get_system_memory(), uart_addr[i], 2, + qdev_get_gpio_in(armv7m, uart_irq[i]), + 115200, serial_hds[i], DEVICE_NATIVE_ENDIAN); + } + } + + dev = DEVICE(&s->timer); + /* APB0 clock is the timer input clock */ + qdev_prop_set_uint32(dev, "clock-frequency", s->m3clk / s->apb0div); + object_property_set_bool(OBJECT(&s->timer), true, "realized", &err); + if (err != NULL) { + error_propagate(errp, err); + return; + } + busdev = SYS_BUS_DEVICE(dev); + sysbus_mmio_map(busdev, 0, MSF2_TIMER_BASE); + sysbus_connect_irq(busdev, 0, + qdev_get_gpio_in(armv7m, timer_irq[0])); + sysbus_connect_irq(busdev, 1, + qdev_get_gpio_in(armv7m, timer_irq[1])); + + dev = DEVICE(&s->sysreg); + qdev_prop_set_uint32(dev, "apb0divisor", s->apb0div); + qdev_prop_set_uint32(dev, "apb1divisor", s->apb1div); + object_property_set_bool(OBJECT(&s->sysreg), true, "realized", &err); + if (err != NULL) { + error_propagate(errp, err); + return; + } + busdev = SYS_BUS_DEVICE(dev); + sysbus_mmio_map(busdev, 0, MSF2_SYSREG_BASE); + + for (i = 0; i < MSF2_NUM_SPIS; i++) { + gchar *bus_name; + + object_property_set_bool(OBJECT(&s->spi[i]), true, "realized", &err); + if (err != NULL) { + error_propagate(errp, err); + return; + } + + sysbus_mmio_map(SYS_BUS_DEVICE(&s->spi[i]), 0, spi_addr[i]); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->spi[i]), 0, + qdev_get_gpio_in(armv7m, spi_irq[i])); + + /* Alias controller SPI bus to the SoC itself */ + bus_name = g_strdup_printf("spi%d", i); + object_property_add_alias(OBJECT(s), bus_name, + OBJECT(&s->spi[i]), "spi", + &error_abort); + g_free(bus_name); + } + + /* Below devices are not modelled yet. */ + create_unimplemented_device("i2c_0", 0x40002000, 0x1000); + create_unimplemented_device("dma", 0x40003000, 0x1000); + create_unimplemented_device("watchdog", 0x40005000, 0x1000); + create_unimplemented_device("i2c_1", 0x40012000, 0x1000); + create_unimplemented_device("gpio", 0x40013000, 0x1000); + create_unimplemented_device("hs-dma", 0x40014000, 0x1000); + create_unimplemented_device("can", 0x40015000, 0x1000); + create_unimplemented_device("rtc", 0x40017000, 0x1000); + create_unimplemented_device("apb_config", 0x40020000, 0x10000); + create_unimplemented_device("emac", 0x40041000, 0x1000); + create_unimplemented_device("usb", 0x40043000, 0x1000); +} + +static Property m2sxxx_soc_properties[] = { + /* + * part name specifies the type of SmartFusion2 device variant(this + * property is for information purpose only. + */ + DEFINE_PROP_STRING("cpu-type", MSF2State, cpu_type), + DEFINE_PROP_STRING("part-name", MSF2State, part_name), + DEFINE_PROP_UINT64("eNVM-size", MSF2State, envm_size, MSF2_ENVM_MAX_SIZE), + DEFINE_PROP_UINT64("eSRAM-size", MSF2State, esram_size, + MSF2_ESRAM_MAX_SIZE), + /* Libero GUI shows 100Mhz as default for clocks */ + DEFINE_PROP_UINT32("m3clk", MSF2State, m3clk, 100 * 1000000), + /* default divisors in Libero GUI */ + DEFINE_PROP_UINT8("apb0div", MSF2State, apb0div, 2), + DEFINE_PROP_UINT8("apb1div", MSF2State, apb1div, 2), + DEFINE_PROP_END_OF_LIST(), +}; + +static void m2sxxx_soc_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->realize = m2sxxx_soc_realize; + dc->props = m2sxxx_soc_properties; +} + +static const TypeInfo m2sxxx_soc_info = { + .name = TYPE_MSF2_SOC, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(MSF2State), + .instance_init = m2sxxx_soc_initfn, + .class_init = m2sxxx_soc_class_init, +}; + +static void m2sxxx_soc_types(void) +{ + type_register_static(&m2sxxx_soc_info); +} + +type_init(m2sxxx_soc_types) diff --git a/default-configs/arm-softmmu.mak b/default-configs/arm-softmmu.mak index XXXXXXX..XXXXXXX 100644 --- a/default-configs/arm-softmmu.mak +++ b/default-configs/arm-softmmu.mak @@ -XXX,XX +XXX,XX @@ CONFIG_ACPI=y CONFIG_SMBIOS=y CONFIG_ASPEED_SOC=y CONFIG_GPIO_KEY=y +CONFIG_MSF2=y -- 2.7.4 From: Subbaraya Sundeep <sundeep.lkml@gmail.com> Emulated Emcraft's Smartfusion2 System On Module starter kit. Signed-off-by: Subbaraya Sundeep <sundeep.lkml@gmail.com> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20170920201737.25723-6-f4bug@amsat.org [PMD: drop cpu_model to directly use cpu type] Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/Makefile.objs | 2 +- hw/arm/msf2-som.c | 105 +++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 106 insertions(+), 1 deletion(-) create mode 100644 hw/arm/msf2-som.c diff --git a/hw/arm/Makefile.objs b/hw/arm/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Makefile.objs +++ b/hw/arm/Makefile.objs @@ -XXX,XX +XXX,XX @@ obj-$(CONFIG_FSL_IMX31) += fsl-imx31.o kzm.o obj-$(CONFIG_FSL_IMX6) += fsl-imx6.o sabrelite.o obj-$(CONFIG_ASPEED_SOC) += aspeed_soc.o aspeed.o obj-$(CONFIG_MPS2) += mps2.o -obj-$(CONFIG_MSF2) += msf2-soc.o +obj-$(CONFIG_MSF2) += msf2-soc.o msf2-som.o diff --git a/hw/arm/msf2-som.c b/hw/arm/msf2-som.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/arm/msf2-som.c @@ -XXX,XX +XXX,XX @@ +/* + * SmartFusion2 SOM starter kit(from Emcraft) emulation. + * + * Copyright (c) 2017 Subbaraya Sundeep <sundeep.lkml@gmail.com> + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + */ + +#include "qemu/osdep.h" +#include "qapi/error.h" +#include "qemu/error-report.h" +#include "hw/boards.h" +#include "hw/arm/arm.h" +#include "exec/address-spaces.h" +#include "qemu/cutils.h" +#include "hw/arm/msf2-soc.h" +#include "cpu.h" + +#define DDR_BASE_ADDRESS 0xA0000000 +#define DDR_SIZE (64 * M_BYTE) + +#define M2S010_ENVM_SIZE (256 * K_BYTE) +#define M2S010_ESRAM_SIZE (64 * K_BYTE) + +static void emcraft_sf2_s2s010_init(MachineState *machine) +{ + DeviceState *dev; + DeviceState *spi_flash; + MSF2State *soc; + MachineClass *mc = MACHINE_GET_CLASS(machine); + DriveInfo *dinfo = drive_get_next(IF_MTD); + qemu_irq cs_line; + SSIBus *spi_bus; + MemoryRegion *sysmem = get_system_memory(); + MemoryRegion *ddr = g_new(MemoryRegion, 1); + + if (strcmp(machine->cpu_type, mc->default_cpu_type) != 0) { + error_report("This board can only be used with CPU %s", + mc->default_cpu_type); + } + + memory_region_init_ram(ddr, NULL, "ddr-ram", DDR_SIZE, + &error_fatal); + memory_region_add_subregion(sysmem, DDR_BASE_ADDRESS, ddr); + + dev = qdev_create(NULL, TYPE_MSF2_SOC); + qdev_prop_set_string(dev, "part-name", "M2S010"); + qdev_prop_set_string(dev, "cpu-type", mc->default_cpu_type); + + qdev_prop_set_uint64(dev, "eNVM-size", M2S010_ENVM_SIZE); + qdev_prop_set_uint64(dev, "eSRAM-size", M2S010_ESRAM_SIZE); + + /* + * CPU clock and peripheral clocks(APB0, APB1)are configurable + * in Libero. CPU clock is divided by APB0 and APB1 divisors for + * peripherals. Emcraft's SoM kit comes with these settings by default. + */ + qdev_prop_set_uint32(dev, "m3clk", 142 * 1000000); + qdev_prop_set_uint32(dev, "apb0div", 2); + qdev_prop_set_uint32(dev, "apb1div", 2); + + object_property_set_bool(OBJECT(dev), true, "realized", &error_fatal); + + soc = MSF2_SOC(dev); + + /* Attach SPI flash to SPI0 controller */ + spi_bus = (SSIBus *)qdev_get_child_bus(dev, "spi0"); + spi_flash = ssi_create_slave_no_init(spi_bus, "s25sl12801"); + qdev_prop_set_uint8(spi_flash, "spansion-cr2nv", 1); + if (dinfo) { + qdev_prop_set_drive(spi_flash, "drive", blk_by_legacy_dinfo(dinfo), + &error_fatal); + } + qdev_init_nofail(spi_flash); + cs_line = qdev_get_gpio_in_named(spi_flash, SSI_GPIO_CS, 0); + sysbus_connect_irq(SYS_BUS_DEVICE(&soc->spi[0]), 1, cs_line); + + armv7m_load_kernel(ARM_CPU(first_cpu), machine->kernel_filename, + soc->envm_size); +} + +static void emcraft_sf2_machine_init(MachineClass *mc) +{ + mc->desc = "SmartFusion2 SOM kit from Emcraft (M2S010)"; + mc->init = emcraft_sf2_s2s010_init; + mc->default_cpu_type = ARM_CPU_TYPE_NAME("cortex-m3"); +} + +DEFINE_MACHINE("emcraft-sf2", emcraft_sf2_machine_init) -- 2.7.4 Most of this is the Neon decodetree patches, followed by Edgar's versal cleanups. thanks -- PMM The following changes since commit 2ef486e76d64436be90f7359a3071fb2a56ce835: Merge remote-tracking branch 'remotes/marcel/tags/rdma-pull-request' into staging (2020-05-03 14:12:56 +0100) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20200504 for you to fetch changes up to 9aefc6cf9b73f66062d2f914a0136756e7a28211: target/arm: Move gen_ function typedefs to translate.h (2020-05-04 12:59:26 +0100) ---------------------------------------------------------------- target-arm queue: * Start of conversion of Neon insns to decodetree * versal board: support SD and RTC * Implement ARMv8.2-TTS2UXN * Make VQDMULL undefined when U=1 * Some minor code cleanups ---------------------------------------------------------------- Edgar E. Iglesias (11): hw/arm: versal: Remove inclusion of arm_gicv3_common.h hw/arm: versal: Move misplaced comment hw/arm: versal-virt: Fix typo xlnx-ve -> xlnx-versal hw/arm: versal: Embed the UARTs into the SoC type hw/arm: versal: Embed the GEMs into the SoC type hw/arm: versal: Embed the ADMAs into the SoC type hw/arm: versal: Embed the APUs into the SoC type hw/arm: versal: Add support for SD hw/arm: versal: Add support for the RTC hw/arm: versal-virt: Add support for SD hw/arm: versal-virt: Add support for the RTC Fredrik Strupe (1): target/arm: Make VQDMULL undefined when U=1 Peter Maydell (25): target/arm: Don't use a TLB for ARMMMUIdx_Stage2 target/arm: Use enum constant in get_phys_addr_lpae() call target/arm: Add new 's1_is_el0' argument to get_phys_addr_lpae() target/arm: Implement ARMv8.2-TTS2UXN target/arm: Use correct variable for setting 'max' cpu's ID_AA64DFR0 target/arm/translate-vfp.inc.c: Remove duplicate simd_r32 check target/arm: Don't allow Thumb Neon insns without FEATURE_NEON target/arm: Add stubs for AArch32 Neon decodetree target/arm: Convert VCMLA (vector) to decodetree target/arm: Convert VCADD (vector) to decodetree target/arm: Convert V[US]DOT (vector) to decodetree target/arm: Convert VFM[AS]L (vector) to decodetree target/arm: Convert VCMLA (scalar) to decodetree target/arm: Convert V[US]DOT (scalar) to decodetree target/arm: Convert VFM[AS]L (scalar) to decodetree target/arm: Convert Neon load/store multiple structures to decodetree target/arm: Convert Neon 'load single structure to all lanes' to decodetree target/arm: Convert Neon 'load/store single structure' to decodetree target/arm: Convert Neon 3-reg-same VADD/VSUB to decodetree target/arm: Convert Neon 3-reg-same logic ops to decodetree target/arm: Convert Neon 3-reg-same VMAX/VMIN to decodetree target/arm: Convert Neon 3-reg-same comparisons to decodetree target/arm: Convert Neon 3-reg-same VQADD/VQSUB to decodetree target/arm: Convert Neon 3-reg-same VMUL, VMLA, VMLS, VSHL to decodetree target/arm: Move gen_ function typedefs to translate.h Philippe Mathieu-Daudé (2): hw/arm/mps2-tz: Use TYPE_IOTKIT instead of hardcoded string target/arm: Use uint64_t for midr field in CPU state struct include/hw/arm/xlnx-versal.h | 31 +- target/arm/cpu-param.h | 2 +- target/arm/cpu.h | 38 ++- target/arm/translate-a64.h | 9 - target/arm/translate.h | 26 ++ target/arm/neon-dp.decode | 86 +++++ target/arm/neon-ls.decode | 52 +++ target/arm/neon-shared.decode | 66 ++++ hw/arm/mps2-tz.c | 2 +- hw/arm/xlnx-versal-virt.c | 74 ++++- hw/arm/xlnx-versal.c | 115 +++++-- target/arm/cpu.c | 3 +- target/arm/cpu64.c | 8 +- target/arm/helper.c | 183 ++++------ target/arm/translate-a64.c | 17 - target/arm/translate-neon.inc.c | 714 +++++++++++++++++++++++++++++++++++++++ target/arm/translate-vfp.inc.c | 6 - target/arm/translate.c | 716 +++------------------------------------- target/arm/Makefile.objs | 18 + 19 files changed, 1302 insertions(+), 864 deletions(-) create mode 100644 target/arm/neon-dp.decode create mode 100644 target/arm/neon-ls.decode create mode 100644 target/arm/neon-shared.decode create mode 100644 target/arm/translate-neon.inc.c From: Fredrik Strupe <fredrik@strupe.net> According to Arm ARM, VQDMULL is only valid when U=0, while having U=1 is unallocated. Signed-off-by: Fredrik Strupe <fredrik@strupe.net> Fixes: 695272dcb976 ("target-arm: Handle UNDEF cases for Neon 3-regs-different-widths") Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) {0, 0, 0, 0}, /* VMLSL */ {0, 0, 0, 9}, /* VQDMLSL */ {0, 0, 0, 0}, /* Integer VMULL */ - {0, 0, 0, 1}, /* VQDMULL */ + {0, 0, 0, 9}, /* VQDMULL */ {0, 0, 0, 0xa}, /* Polynomial VMULL */ {0, 0, 0, 7}, /* Reserved: always UNDEF */ }; -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> By using the TYPE_* definitions for devices, we can: - quickly find where devices are used with 'git-grep' - easily rename a device (one-line change). Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200428154650.21991-1-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/mps2-tz.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/arm/mps2-tz.c b/hw/arm/mps2-tz.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/mps2-tz.c +++ b/hw/arm/mps2-tz.c @@ -XXX,XX +XXX,XX @@ static void mps2tz_common_init(MachineState *machine) exit(EXIT_FAILURE); } - sysbus_init_child_obj(OBJECT(machine), "iotkit", &mms->iotkit, + sysbus_init_child_obj(OBJECT(machine), TYPE_IOTKIT, &mms->iotkit, sizeof(mms->iotkit), mmc->armsse_type); iotkitdev = DEVICE(&mms->iotkit); object_property_set_link(OBJECT(&mms->iotkit), OBJECT(system_memory), -- 2.20.1 We define ARMMMUIdx_Stage2 as being an MMU index which uses a QEMU TLB. However we never actually use the TLB -- all stage 2 lookups are done by direct calls to get_phys_addr_lpae() followed by a physical address load via address_space_ld*(). Remove Stage2 from the list of ARM MMU indexes which correspond to real core MMU indexes, and instead put it in the set of "NOTLB" ARM MMU indexes. This allows us to drop NB_MMU_MODES to 11. It also means we can safely add support for the ARMv8.3-TTS2UXN extension, which adds permission bits to the stage 2 descriptors which define execute permission separatel for EL0 and EL1; supporting that while keeping Stage2 in a QEMU TLB would require us to use separate TLBs for "Stage2 for an EL0 access" and "Stage2 for an EL1 access", which is a lot of extra complication given we aren't even using the QEMU TLB. In the process of updating the comment on our MMU index use, fix a couple of other minor errors: * NS EL2 EL2&0 was missing from the list in the comment * some text hadn't been updated from when we bumped NB_MMU_MODES above 8 Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200330210400.11724-2-peter.maydell@linaro.org --- target/arm/cpu-param.h | 2 +- target/arm/cpu.h | 21 +++++--- target/arm/helper.c | 112 ++++------------------------------------- 3 files changed, 27 insertions(+), 108 deletions(-) diff --git a/target/arm/cpu-param.h b/target/arm/cpu-param.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu-param.h +++ b/target/arm/cpu-param.h @@ -XXX,XX +XXX,XX @@ # define TARGET_PAGE_BITS_MIN 10 #endif -#define NB_MMU_MODES 12 +#define NB_MMU_MODES 11 #endif diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ bool write_cpustate_to_list(ARMCPU *cpu, bool kvm_sync); * handling via the TLB. The only way to do a stage 1 translation without * the immediate stage 2 translation is via the ATS or AT system insns, * which can be slow-pathed and always do a page table walk. + * The only use of stage 2 translations is either as part of an s1+2 + * lookup or when loading the descriptors during a stage 1 page table walk, + * and in both those cases we don't use the TLB. * 4. we can also safely fold together the "32 bit EL3" and "64 bit EL3" * translation regimes, because they map reasonably well to each other * and they can't both be active at the same time. @@ -XXX,XX +XXX,XX @@ bool write_cpustate_to_list(ARMCPU *cpu, bool kvm_sync); * NS EL1 EL1&0 stage 1+2 (aka NS PL1) * NS EL1 EL1&0 stage 1+2 +PAN * NS EL0 EL2&0 + * NS EL2 EL2&0 * NS EL2 EL2&0 +PAN * NS EL2 (aka NS PL2) * S EL0 EL1&0 (aka S PL0) * S EL1 EL1&0 (not used if EL3 is 32 bit) * S EL1 EL1&0 +PAN * S EL3 (aka S PL1) - * NS EL1&0 stage 2 * - * for a total of 12 different mmu_idx. + * for a total of 11 different mmu_idx. * * R profile CPUs have an MPU, but can use the same set of MMU indexes * as A profile. They only need to distinguish NS EL0 and NS EL1 (and @@ -XXX,XX +XXX,XX @@ bool write_cpustate_to_list(ARMCPU *cpu, bool kvm_sync); * are not quite the same -- different CPU types (most notably M profile * vs A/R profile) would like to use MMU indexes with different semantics, * but since we don't ever need to use all of those in a single CPU we - * can avoid setting NB_MMU_MODES to more than 8. The lower bits of + * can avoid having to set NB_MMU_MODES to "total number of A profile MMU + * modes + total number of M profile MMU modes". The lower bits of * ARMMMUIdx are the core TLB mmu index, and the higher bits are always * the same for any particular CPU. * Variables of type ARMMUIdx are always full values, and the core @@ -XXX,XX +XXX,XX @@ typedef enum ARMMMUIdx { ARMMMUIdx_SE10_1_PAN = 9 | ARM_MMU_IDX_A, ARMMMUIdx_SE3 = 10 | ARM_MMU_IDX_A, - ARMMMUIdx_Stage2 = 11 | ARM_MMU_IDX_A, - /* * These are not allocated TLBs and are used only for AT system * instructions or for the first stage of an S12 page table walk. @@ -XXX,XX +XXX,XX @@ typedef enum ARMMMUIdx { ARMMMUIdx_Stage1_E0 = 0 | ARM_MMU_IDX_NOTLB, ARMMMUIdx_Stage1_E1 = 1 | ARM_MMU_IDX_NOTLB, ARMMMUIdx_Stage1_E1_PAN = 2 | ARM_MMU_IDX_NOTLB, + /* + * Not allocated a TLB: used only for second stage of an S12 page + * table walk, or for descriptor loads during first stage of an S1 + * page table walk. Note that if we ever want to have a TLB for this + * then various TLB flush insns which currently are no-ops or flush + * only stage 1 MMU indexes will need to change to flush stage 2. + */ + ARMMMUIdx_Stage2 = 3 | ARM_MMU_IDX_NOTLB, /* * M-profile. @@ -XXX,XX +XXX,XX @@ typedef enum ARMMMUIdxBit { TO_CORE_BIT(SE10_1), TO_CORE_BIT(SE10_1_PAN), TO_CORE_BIT(SE3), - TO_CORE_BIT(Stage2), TO_CORE_BIT(MUser), TO_CORE_BIT(MPriv), diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ static void tlbiall_nsnh_write(CPUARMState *env, const ARMCPRegInfo *ri, tlb_flush_by_mmuidx(cs, ARMMMUIdxBit_E10_1 | ARMMMUIdxBit_E10_1_PAN | - ARMMMUIdxBit_E10_0 | - ARMMMUIdxBit_Stage2); + ARMMMUIdxBit_E10_0); } static void tlbiall_nsnh_is_write(CPUARMState *env, const ARMCPRegInfo *ri, @@ -XXX,XX +XXX,XX @@ static void tlbiall_nsnh_is_write(CPUARMState *env, const ARMCPRegInfo *ri, tlb_flush_by_mmuidx_all_cpus_synced(cs, ARMMMUIdxBit_E10_1 | ARMMMUIdxBit_E10_1_PAN | - ARMMMUIdxBit_E10_0 | - ARMMMUIdxBit_Stage2); + ARMMMUIdxBit_E10_0); } -static void tlbiipas2_write(CPUARMState *env, const ARMCPRegInfo *ri, - uint64_t value) -{ - /* Invalidate by IPA. This has to invalidate any structures that - * contain only stage 2 translation information, but does not need - * to apply to structures that contain combined stage 1 and stage 2 - * translation information. - * This must NOP if EL2 isn't implemented or SCR_EL3.NS is zero. - */ - CPUState *cs = env_cpu(env); - uint64_t pageaddr; - - if (!arm_feature(env, ARM_FEATURE_EL2) || !(env->cp15.scr_el3 & SCR_NS)) { - return; - } - - pageaddr = sextract64(value << 12, 0, 40); - - tlb_flush_page_by_mmuidx(cs, pageaddr, ARMMMUIdxBit_Stage2); -} - -static void tlbiipas2_is_write(CPUARMState *env, const ARMCPRegInfo *ri, - uint64_t value) -{ - CPUState *cs = env_cpu(env); - uint64_t pageaddr; - - if (!arm_feature(env, ARM_FEATURE_EL2) || !(env->cp15.scr_el3 & SCR_NS)) { - return; - } - - pageaddr = sextract64(value << 12, 0, 40); - - tlb_flush_page_by_mmuidx_all_cpus_synced(cs, pageaddr, - ARMMMUIdxBit_Stage2); -} static void tlbiall_hyp_write(CPUARMState *env, const ARMCPRegInfo *ri, uint64_t value) @@ -XXX,XX +XXX,XX @@ static void vttbr_write(CPUARMState *env, const ARMCPRegInfo *ri, tlb_flush_by_mmuidx(cs, ARMMMUIdxBit_E10_1 | ARMMMUIdxBit_E10_1_PAN | - ARMMMUIdxBit_E10_0 | - ARMMMUIdxBit_Stage2); + ARMMMUIdxBit_E10_0); raw_write(env, ri, value); } } @@ -XXX,XX +XXX,XX @@ static int alle1_tlbmask(CPUARMState *env) return ARMMMUIdxBit_SE10_1 | ARMMMUIdxBit_SE10_1_PAN | ARMMMUIdxBit_SE10_0; - } else if (arm_feature(env, ARM_FEATURE_EL2)) { - return ARMMMUIdxBit_E10_1 | - ARMMMUIdxBit_E10_1_PAN | - ARMMMUIdxBit_E10_0 | - ARMMMUIdxBit_Stage2; } else { return ARMMMUIdxBit_E10_1 | ARMMMUIdxBit_E10_1_PAN | @@ -XXX,XX +XXX,XX @@ static void tlbi_aa64_vae3is_write(CPUARMState *env, const ARMCPRegInfo *ri, ARMMMUIdxBit_SE3); } -static void tlbi_aa64_ipas2e1_write(CPUARMState *env, const ARMCPRegInfo *ri, - uint64_t value) -{ - /* Invalidate by IPA. This has to invalidate any structures that - * contain only stage 2 translation information, but does not need - * to apply to structures that contain combined stage 1 and stage 2 - * translation information. - * This must NOP if EL2 isn't implemented or SCR_EL3.NS is zero. - */ - ARMCPU *cpu = env_archcpu(env); - CPUState *cs = CPU(cpu); - uint64_t pageaddr; - - if (!arm_feature(env, ARM_FEATURE_EL2) || !(env->cp15.scr_el3 & SCR_NS)) { - return; - } - - pageaddr = sextract64(value << 12, 0, 48); - - tlb_flush_page_by_mmuidx(cs, pageaddr, ARMMMUIdxBit_Stage2); -} - -static void tlbi_aa64_ipas2e1is_write(CPUARMState *env, const ARMCPRegInfo *ri, - uint64_t value) -{ - CPUState *cs = env_cpu(env); - uint64_t pageaddr; - - if (!arm_feature(env, ARM_FEATURE_EL2) || !(env->cp15.scr_el3 & SCR_NS)) { - return; - } - - pageaddr = sextract64(value << 12, 0, 48); - - tlb_flush_page_by_mmuidx_all_cpus_synced(cs, pageaddr, - ARMMMUIdxBit_Stage2); -} - static CPAccessResult aa64_zva_access(CPUARMState *env, const ARMCPRegInfo *ri, bool isread) { @@ -XXX,XX +XXX,XX @@ static const ARMCPRegInfo v8_cp_reginfo[] = { .writefn = tlbi_aa64_vae1_write }, { .name = "TLBI_IPAS2E1IS", .state = ARM_CP_STATE_AA64, .opc0 = 1, .opc1 = 4, .crn = 8, .crm = 0, .opc2 = 1, - .access = PL2_W, .type = ARM_CP_NO_RAW, - .writefn = tlbi_aa64_ipas2e1is_write }, + .access = PL2_W, .type = ARM_CP_NOP }, { .name = "TLBI_IPAS2LE1IS", .state = ARM_CP_STATE_AA64, .opc0 = 1, .opc1 = 4, .crn = 8, .crm = 0, .opc2 = 5, - .access = PL2_W, .type = ARM_CP_NO_RAW, - .writefn = tlbi_aa64_ipas2e1is_write }, + .access = PL2_W, .type = ARM_CP_NOP }, { .name = "TLBI_ALLE1IS", .state = ARM_CP_STATE_AA64, .opc0 = 1, .opc1 = 4, .crn = 8, .crm = 3, .opc2 = 4, .access = PL2_W, .type = ARM_CP_NO_RAW, @@ -XXX,XX +XXX,XX @@ static const ARMCPRegInfo v8_cp_reginfo[] = { .writefn = tlbi_aa64_alle1is_write }, { .name = "TLBI_IPAS2E1", .state = ARM_CP_STATE_AA64, .opc0 = 1, .opc1 = 4, .crn = 8, .crm = 4, .opc2 = 1, - .access = PL2_W, .type = ARM_CP_NO_RAW, - .writefn = tlbi_aa64_ipas2e1_write }, + .access = PL2_W, .type = ARM_CP_NOP }, { .name = "TLBI_IPAS2LE1", .state = ARM_CP_STATE_AA64, .opc0 = 1, .opc1 = 4, .crn = 8, .crm = 4, .opc2 = 5, - .access = PL2_W, .type = ARM_CP_NO_RAW, - .writefn = tlbi_aa64_ipas2e1_write }, + .access = PL2_W, .type = ARM_CP_NOP }, { .name = "TLBI_ALLE1", .state = ARM_CP_STATE_AA64, .opc0 = 1, .opc1 = 4, .crn = 8, .crm = 7, .opc2 = 4, .access = PL2_W, .type = ARM_CP_NO_RAW, @@ -XXX,XX +XXX,XX @@ static const ARMCPRegInfo v8_cp_reginfo[] = { .writefn = tlbimva_hyp_is_write }, { .name = "TLBIIPAS2", .cp = 15, .opc1 = 4, .crn = 8, .crm = 4, .opc2 = 1, - .type = ARM_CP_NO_RAW, .access = PL2_W, - .writefn = tlbiipas2_write }, + .type = ARM_CP_NOP, .access = PL2_W }, { .name = "TLBIIPAS2IS", .cp = 15, .opc1 = 4, .crn = 8, .crm = 0, .opc2 = 1, - .type = ARM_CP_NO_RAW, .access = PL2_W, - .writefn = tlbiipas2_is_write }, + .type = ARM_CP_NOP, .access = PL2_W }, { .name = "TLBIIPAS2L", .cp = 15, .opc1 = 4, .crn = 8, .crm = 4, .opc2 = 5, - .type = ARM_CP_NO_RAW, .access = PL2_W, - .writefn = tlbiipas2_write }, + .type = ARM_CP_NOP, .access = PL2_W }, { .name = "TLBIIPAS2LIS", .cp = 15, .opc1 = 4, .crn = 8, .crm = 0, .opc2 = 5, - .type = ARM_CP_NO_RAW, .access = PL2_W, - .writefn = tlbiipas2_is_write }, + .type = ARM_CP_NOP, .access = PL2_W }, /* 32 bit cache operations */ { .name = "ICIALLUIS", .cp = 15, .opc1 = 0, .crn = 7, .crm = 1, .opc2 = 0, .type = ARM_CP_NOP, .access = PL1_W, .accessfn = aa64_cacheop_pou_access }, -- 2.20.1 The access_type argument to get_phys_addr_lpae() is an MMUAccessType; use the enum constant MMU_DATA_LOAD rather than a literal 0 when we call it in S1_ptw_translate(). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200330210400.11724-3-peter.maydell@linaro.org --- target/arm/helper.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ static hwaddr S1_ptw_translate(CPUARMState *env, ARMMMUIdx mmu_idx, pcacheattrs = &cacheattrs; } - ret = get_phys_addr_lpae(env, addr, 0, ARMMMUIdx_Stage2, &s2pa, - &txattrs, &s2prot, &s2size, fi, pcacheattrs); + ret = get_phys_addr_lpae(env, addr, MMU_DATA_LOAD, ARMMMUIdx_Stage2, + &s2pa, &txattrs, &s2prot, &s2size, fi, + pcacheattrs); if (ret) { assert(fi->type != ARMFault_None); fi->s2addr = addr; -- 2.20.1 For ARMv8.2-TTS2UXN, the stage 2 page table walk wants to know whether the stage 1 access is for EL0 or not, because whether exec permission is given can depend on whether this is an EL0 or EL1 access. Add a new argument to get_phys_addr_lpae() so the call sites can pass this information in. Since get_phys_addr_lpae() doesn't already have a doc comment, add one so we have a place to put the documentation of the semantics of the new s1_is_el0 argument. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200330210400.11724-4-peter.maydell@linaro.org --- target/arm/helper.c | 29 ++++++++++++++++++++++++++++- 1 file changed, 28 insertions(+), 1 deletion(-) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ static bool get_phys_addr_lpae(CPUARMState *env, target_ulong address, MMUAccessType access_type, ARMMMUIdx mmu_idx, + bool s1_is_el0, hwaddr *phys_ptr, MemTxAttrs *txattrs, int *prot, target_ulong *page_size_ptr, ARMMMUFaultInfo *fi, ARMCacheAttrs *cacheattrs); @@ -XXX,XX +XXX,XX @@ static hwaddr S1_ptw_translate(CPUARMState *env, ARMMMUIdx mmu_idx, } ret = get_phys_addr_lpae(env, addr, MMU_DATA_LOAD, ARMMMUIdx_Stage2, + false, &s2pa, &txattrs, &s2prot, &s2size, fi, pcacheattrs); if (ret) { @@ -XXX,XX +XXX,XX @@ static ARMVAParameters aa32_va_parameters(CPUARMState *env, uint32_t va, }; } +/** + * get_phys_addr_lpae: perform one stage of page table walk, LPAE format + * + * Returns false if the translation was successful. Otherwise, phys_ptr, attrs, + * prot and page_size may not be filled in, and the populated fsr value provides + * information on why the translation aborted, in the format of a long-format + * DFSR/IFSR fault register, with the following caveats: + * * the WnR bit is never set (the caller must do this). + * + * @env: CPUARMState + * @address: virtual address to get physical address for + * @access_type: MMU_DATA_LOAD, MMU_DATA_STORE or MMU_INST_FETCH + * @mmu_idx: MMU index indicating required translation regime + * @s1_is_el0: if @mmu_idx is ARMMMUIdx_Stage2 (so this is a stage 2 page table + * walk), must be true if this is stage 2 of a stage 1+2 walk for an + * EL0 access). If @mmu_idx is anything else, @s1_is_el0 is ignored. + * @phys_ptr: set to the physical address corresponding to the virtual address + * @attrs: set to the memory transaction attributes to use + * @prot: set to the permissions for the page containing phys_ptr + * @page_size_ptr: set to the size of the page containing phys_ptr + * @fi: set to fault info if the translation fails + * @cacheattrs: (if non-NULL) set to the cacheability/shareability attributes + */ static bool get_phys_addr_lpae(CPUARMState *env, target_ulong address, MMUAccessType access_type, ARMMMUIdx mmu_idx, + bool s1_is_el0, hwaddr *phys_ptr, MemTxAttrs *txattrs, int *prot, target_ulong *page_size_ptr, ARMMMUFaultInfo *fi, ARMCacheAttrs *cacheattrs) @@ -XXX,XX +XXX,XX @@ bool get_phys_addr(CPUARMState *env, target_ulong address, /* S1 is done. Now do S2 translation. */ ret = get_phys_addr_lpae(env, ipa, access_type, ARMMMUIdx_Stage2, + mmu_idx == ARMMMUIdx_E10_0, phys_ptr, attrs, &s2_prot, page_size, fi, cacheattrs != NULL ? &cacheattrs2 : NULL); @@ -XXX,XX +XXX,XX @@ bool get_phys_addr(CPUARMState *env, target_ulong address, } if (regime_using_lpae_format(env, mmu_idx)) { - return get_phys_addr_lpae(env, address, access_type, mmu_idx, + return get_phys_addr_lpae(env, address, access_type, mmu_idx, false, phys_ptr, attrs, prot, page_size, fi, cacheattrs); } else if (regime_sctlr(env, mmu_idx) & SCTLR_XP) { -- 2.20.1 The ARMv8.2-TTS2UXN feature extends the XN field in stage 2 translation table descriptors from just bit [54] to bits [54:53], allowing stage 2 to control execution permissions separately for EL0 and EL1. Implement the new semantics of the XN field and enable the feature for our 'max' CPU. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200330210400.11724-5-peter.maydell@linaro.org --- target/arm/cpu.h | 15 +++++++++++++++ target/arm/cpu.c | 1 + target/arm/cpu64.c | 2 ++ target/arm/helper.c | 37 +++++++++++++++++++++++++++++++------ 4 files changed, 49 insertions(+), 6 deletions(-) diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ static inline bool isar_feature_aa32_ccidx(const ARMISARegisters *id) return FIELD_EX32(id->id_mmfr4, ID_MMFR4, CCIDX) != 0; } +static inline bool isar_feature_aa32_tts2uxn(const ARMISARegisters *id) +{ + return FIELD_EX32(id->id_mmfr4, ID_MMFR4, XNX) != 0; +} + /* * 64-bit feature tests via id registers. */ @@ -XXX,XX +XXX,XX @@ static inline bool isar_feature_aa64_ccidx(const ARMISARegisters *id) return FIELD_EX64(id->id_aa64mmfr2, ID_AA64MMFR2, CCIDX) != 0; } +static inline bool isar_feature_aa64_tts2uxn(const ARMISARegisters *id) +{ + return FIELD_EX64(id->id_aa64mmfr1, ID_AA64MMFR1, XNX) != 0; +} + /* * Feature tests for "does this exist in either 32-bit or 64-bit?" */ @@ -XXX,XX +XXX,XX @@ static inline bool isar_feature_any_ccidx(const ARMISARegisters *id) return isar_feature_aa64_ccidx(id) || isar_feature_aa32_ccidx(id); } +static inline bool isar_feature_any_tts2uxn(const ARMISARegisters *id) +{ + return isar_feature_aa64_tts2uxn(id) || isar_feature_aa32_tts2uxn(id); +} + /* * Forward to the above feature tests given an ARMCPU pointer. */ diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void arm_max_initfn(Object *obj) t = FIELD_DP32(t, ID_MMFR4, HPDS, 1); /* AA32HPD */ t = FIELD_DP32(t, ID_MMFR4, AC2, 1); /* ACTLR2, HACTLR2 */ t = FIELD_DP32(t, ID_MMFR4, CNP, 1); /* TTCNP */ + t = FIELD_DP32(t, ID_MMFR4, XNX, 1); /* TTS2UXN */ cpu->isar.id_mmfr4 = t; } #endif diff --git a/target/arm/cpu64.c b/target/arm/cpu64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu64.c +++ b/target/arm/cpu64.c @@ -XXX,XX +XXX,XX @@ static void aarch64_max_initfn(Object *obj) t = FIELD_DP64(t, ID_AA64MMFR1, VH, 1); t = FIELD_DP64(t, ID_AA64MMFR1, PAN, 2); /* ATS1E1 */ t = FIELD_DP64(t, ID_AA64MMFR1, VMIDBITS, 2); /* VMID16 */ + t = FIELD_DP64(t, ID_AA64MMFR1, XNX, 1); /* TTS2UXN */ cpu->isar.id_aa64mmfr1 = t; t = cpu->isar.id_aa64mmfr2; @@ -XXX,XX +XXX,XX @@ static void aarch64_max_initfn(Object *obj) u = FIELD_DP32(u, ID_MMFR4, HPDS, 1); /* AA32HPD */ u = FIELD_DP32(u, ID_MMFR4, AC2, 1); /* ACTLR2, HACTLR2 */ u = FIELD_DP32(u, ID_MMFR4, CNP, 1); /* TTCNP */ + u = FIELD_DP32(u, ID_MMFR4, XNX, 1); /* TTS2UXN */ cpu->isar.id_mmfr4 = u; u = cpu->isar.id_aa64dfr0; diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ simple_ap_to_rw_prot(CPUARMState *env, ARMMMUIdx mmu_idx, int ap) * * @env: CPUARMState * @s2ap: The 2-bit stage2 access permissions (S2AP) - * @xn: XN (execute-never) bit + * @xn: XN (execute-never) bits + * @s1_is_el0: true if this is S2 of an S1+2 walk for EL0 */ -static int get_S2prot(CPUARMState *env, int s2ap, int xn) +static int get_S2prot(CPUARMState *env, int s2ap, int xn, bool s1_is_el0) { int prot = 0; @@ -XXX,XX +XXX,XX @@ static int get_S2prot(CPUARMState *env, int s2ap, int xn) if (s2ap & 2) { prot |= PAGE_WRITE; } - if (!xn) { - if (arm_el_is_aa64(env, 2) || prot & PAGE_READ) { + + if (cpu_isar_feature(any_tts2uxn, env_archcpu(env))) { + switch (xn) { + case 0: prot |= PAGE_EXEC; + break; + case 1: + if (s1_is_el0) { + prot |= PAGE_EXEC; + } + break; + case 2: + break; + case 3: + if (!s1_is_el0) { + prot |= PAGE_EXEC; + } + break; + default: + g_assert_not_reached(); + } + } else { + if (!extract32(xn, 1, 1)) { + if (arm_el_is_aa64(env, 2) || prot & PAGE_READ) { + prot |= PAGE_EXEC; + } } } return prot; @@ -XXX,XX +XXX,XX @@ static bool get_phys_addr_lpae(CPUARMState *env, target_ulong address, } ap = extract32(attrs, 4, 2); - xn = extract32(attrs, 12, 1); if (mmu_idx == ARMMMUIdx_Stage2) { ns = true; - *prot = get_S2prot(env, ap, xn); + xn = extract32(attrs, 11, 2); + *prot = get_S2prot(env, ap, xn, s1_is_el0); } else { ns = extract32(attrs, 3, 1); + xn = extract32(attrs, 12, 1); pxn = extract32(attrs, 11, 1); *prot = get_S1prot(env, mmu_idx, aarch64, ap, ns, xn, pxn); } -- 2.20.1 In aarch64_max_initfn() we update both 32-bit and 64-bit ID registers. The intended pattern is that for 64-bit ID registers we use FIELD_DP64 and the uint64_t 't' register, while 32-bit ID registers use FIELD_DP32 and the uint32_t 'u' register. For ID_AA64DFR0 we accidentally used 'u', meaning that the top 32 bits of this 64-bit ID register would end up always zero. Luckily at the moment that's what they should be anyway, so this bug has no visible effects. Use the right-sized variable. Fixes: 3bec78447a958d481991 Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Laurent Desnogues <laurent.desnogues@gmail.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200423110915.10527-1-peter.maydell@linaro.org --- target/arm/cpu64.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/target/arm/cpu64.c b/target/arm/cpu64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu64.c +++ b/target/arm/cpu64.c @@ -XXX,XX +XXX,XX @@ static void aarch64_max_initfn(Object *obj) u = FIELD_DP32(u, ID_MMFR4, XNX, 1); /* TTS2UXN */ cpu->isar.id_mmfr4 = u; - u = cpu->isar.id_aa64dfr0; - u = FIELD_DP64(u, ID_AA64DFR0, PMUVER, 5); /* v8.4-PMU */ - cpu->isar.id_aa64dfr0 = u; + t = cpu->isar.id_aa64dfr0; + t = FIELD_DP64(t, ID_AA64DFR0, PMUVER, 5); /* v8.4-PMU */ + cpu->isar.id_aa64dfr0 = t; u = cpu->isar.id_dfr0; u = FIELD_DP32(u, ID_DFR0, PERFMON, 5); /* v8.4-PMU */ -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> MIDR_EL1 is a 64-bit system register with the top 32-bit being RES0. Represent it in QEMU's ARMCPU struct with a uint64_t, not a uint32_t. This fixes an error when compiling with -Werror=conversion because we were manipulating the register value using a local uint64_t variable: target/arm/cpu64.c: In function ‘aarch64_max_initfn’: target/arm/cpu64.c:628:21: error: conversion from ‘uint64_t’ {aka ‘long unsigned int’} to ‘uint32_t’ {aka ‘unsigned int’} may change value [-Werror=conversion] 628 | cpu->midr = t; | ^ and future-proofs us against a possible future architecture change using some of the top 32 bits. Suggested-by: Laurent Desnogues <laurent.desnogues@gmail.com> Suggested-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Laurent Desnogues <laurent.desnogues@gmail.com> Message-id: 20200428172634.29707-1-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 2 +- target/arm/cpu.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ struct ARMCPU { uint64_t id_aa64dfr0; uint64_t id_aa64dfr1; } isar; - uint32_t midr; + uint64_t midr; uint32_t revidr; uint32_t reset_fpsid; uint32_t ctr; diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static const ARMCPUInfo arm_cpus[] = { static Property arm_cpu_properties[] = { DEFINE_PROP_BOOL("start-powered-off", ARMCPU, start_powered_off, false), DEFINE_PROP_UINT32("psci-conduit", ARMCPU, psci_conduit, 0), - DEFINE_PROP_UINT32("midr", ARMCPU, midr, 0), + DEFINE_PROP_UINT64("midr", ARMCPU, midr, 0), DEFINE_PROP_UINT64("mp-affinity", ARMCPU, mp_affinity, ARM64_AFFINITY_INVALID), DEFINE_PROP_INT32("node-id", ARMCPU, node_id, CPU_UNSET_NUMA_NODE_ID), -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Remove inclusion of arm_gicv3_common.h, this already gets included via xlnx-versal.h. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-2-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/xlnx-versal.c | 1 - 1 file changed, 1 deletion(-) diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ #include "hw/arm/boot.h" #include "kvm_arm.h" #include "hw/misc/unimp.h" -#include "hw/intc/arm_gicv3_common.h" #include "hw/arm/xlnx-versal.h" #include "hw/char/pl011.h" -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Move misplaced comment. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-3-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/xlnx-versal.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_apu_cpus(Versal *s) obj = object_new(XLNX_VERSAL_ACPU_TYPE); if (!obj) { - /* Secondary CPUs start in PSCI powered-down state */ error_report("Unable to create apu.cpu[%d] of type %s", i, XLNX_VERSAL_ACPU_TYPE); exit(EXIT_FAILURE); @@ -XXX,XX +XXX,XX @@ static void versal_create_apu_cpus(Versal *s) object_property_set_int(obj, s->cfg.psci_conduit, "psci-conduit", &error_abort); if (i) { + /* Secondary CPUs start in PSCI powered-down state */ object_property_set_bool(obj, true, "start-powered-off", &error_abort); } -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Fix typo xlnx-ve -> xlnx-versal. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-4-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/xlnx-versal-virt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hw/arm/xlnx-versal-virt.c b/hw/arm/xlnx-versal-virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal-virt.c +++ b/hw/arm/xlnx-versal-virt.c @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) psci_conduit = QEMU_PSCI_CONDUIT_SMC; } - sysbus_init_child_obj(OBJECT(machine), "xlnx-ve", &s->soc, + sysbus_init_child_obj(OBJECT(machine), "xlnx-versal", &s->soc, sizeof(s->soc), TYPE_XLNX_VERSAL); object_property_set_link(OBJECT(&s->soc), OBJECT(machine->ram), "ddr", &error_abort); -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Embed the UARTs into the SoC type. Suggested-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-5-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 3 ++- hw/arm/xlnx-versal.c | 12 ++++++------ 2 files changed, 8 insertions(+), 7 deletions(-) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "hw/sysbus.h" #include "hw/arm/boot.h" #include "hw/intc/arm_gicv3.h" +#include "hw/char/pl011.h" #define TYPE_XLNX_VERSAL "xlnx-versal" #define XLNX_VERSAL(obj) OBJECT_CHECK(Versal, (obj), TYPE_XLNX_VERSAL) @@ -XXX,XX +XXX,XX @@ typedef struct Versal { MemoryRegion mr_ocm; struct { - SysBusDevice *uart[XLNX_VERSAL_NR_UARTS]; + PL011State uart[XLNX_VERSAL_NR_UARTS]; SysBusDevice *gem[XLNX_VERSAL_NR_GEMS]; SysBusDevice *adma[XLNX_VERSAL_NR_ADMAS]; } iou; diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ #include "kvm_arm.h" #include "hw/misc/unimp.h" #include "hw/arm/xlnx-versal.h" -#include "hw/char/pl011.h" #define XLNX_VERSAL_ACPU_TYPE ARM_CPU_TYPE_NAME("cortex-a72") #define GEM_REVISION 0x40070106 @@ -XXX,XX +XXX,XX @@ static void versal_create_uarts(Versal *s, qemu_irq *pic) DeviceState *dev; MemoryRegion *mr; - dev = qdev_create(NULL, TYPE_PL011); - s->lpd.iou.uart[i] = SYS_BUS_DEVICE(dev); + sysbus_init_child_obj(OBJECT(s), name, + &s->lpd.iou.uart[i], sizeof(s->lpd.iou.uart[i]), + TYPE_PL011); + dev = DEVICE(&s->lpd.iou.uart[i]); qdev_prop_set_chr(dev, "chardev", serial_hd(i)); - object_property_add_child(OBJECT(s), name, OBJECT(dev), &error_fatal); qdev_init_nofail(dev); - mr = sysbus_mmio_get_region(s->lpd.iou.uart[i], 0); + mr = sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), 0); memory_region_add_subregion(&s->mr_ps, addrs[i], mr); - sysbus_connect_irq(s->lpd.iou.uart[i], 0, pic[irqs[i]]); + sysbus_connect_irq(SYS_BUS_DEVICE(dev), 0, pic[irqs[i]]); g_free(name); } } -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Embed the GEMs into the SoC type. Suggested-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-6-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 3 ++- hw/arm/xlnx-versal.c | 15 ++++++++------- 2 files changed, 10 insertions(+), 8 deletions(-) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "hw/arm/boot.h" #include "hw/intc/arm_gicv3.h" #include "hw/char/pl011.h" +#include "hw/net/cadence_gem.h" #define TYPE_XLNX_VERSAL "xlnx-versal" #define XLNX_VERSAL(obj) OBJECT_CHECK(Versal, (obj), TYPE_XLNX_VERSAL) @@ -XXX,XX +XXX,XX @@ typedef struct Versal { struct { PL011State uart[XLNX_VERSAL_NR_UARTS]; - SysBusDevice *gem[XLNX_VERSAL_NR_GEMS]; + CadenceGEMState gem[XLNX_VERSAL_NR_GEMS]; SysBusDevice *adma[XLNX_VERSAL_NR_ADMAS]; } iou; } lpd; diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_gems(Versal *s, qemu_irq *pic) DeviceState *dev; MemoryRegion *mr; - dev = qdev_create(NULL, "cadence_gem"); - s->lpd.iou.gem[i] = SYS_BUS_DEVICE(dev); - object_property_add_child(OBJECT(s), name, OBJECT(dev), &error_fatal); + sysbus_init_child_obj(OBJECT(s), name, + &s->lpd.iou.gem[i], sizeof(s->lpd.iou.gem[i]), + TYPE_CADENCE_GEM); + dev = DEVICE(&s->lpd.iou.gem[i]); if (nd->used) { qemu_check_nic_model(nd, "cadence_gem"); qdev_set_nic_properties(dev, nd); } - object_property_set_int(OBJECT(s->lpd.iou.gem[i]), + object_property_set_int(OBJECT(dev), 2, "num-priority-queues", &error_abort); - object_property_set_link(OBJECT(s->lpd.iou.gem[i]), + object_property_set_link(OBJECT(dev), OBJECT(&s->mr_ps), "dma", &error_abort); qdev_init_nofail(dev); - mr = sysbus_mmio_get_region(s->lpd.iou.gem[i], 0); + mr = sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), 0); memory_region_add_subregion(&s->mr_ps, addrs[i], mr); - sysbus_connect_irq(s->lpd.iou.gem[i], 0, pic[irqs[i]]); + sysbus_connect_irq(SYS_BUS_DEVICE(dev), 0, pic[irqs[i]]); g_free(name); } } -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Embed the ADMAs into the SoC type. Suggested-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-7-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 3 ++- hw/arm/xlnx-versal.c | 14 +++++++------- 2 files changed, 9 insertions(+), 8 deletions(-) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "hw/arm/boot.h" #include "hw/intc/arm_gicv3.h" #include "hw/char/pl011.h" +#include "hw/dma/xlnx-zdma.h" #include "hw/net/cadence_gem.h" #define TYPE_XLNX_VERSAL "xlnx-versal" @@ -XXX,XX +XXX,XX @@ typedef struct Versal { struct { PL011State uart[XLNX_VERSAL_NR_UARTS]; CadenceGEMState gem[XLNX_VERSAL_NR_GEMS]; - SysBusDevice *adma[XLNX_VERSAL_NR_ADMAS]; + XlnxZDMA adma[XLNX_VERSAL_NR_ADMAS]; } iou; } lpd; diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_admas(Versal *s, qemu_irq *pic) DeviceState *dev; MemoryRegion *mr; - dev = qdev_create(NULL, "xlnx.zdma"); - s->lpd.iou.adma[i] = SYS_BUS_DEVICE(dev); - object_property_set_int(OBJECT(s->lpd.iou.adma[i]), 128, "bus-width", - &error_abort); - object_property_add_child(OBJECT(s), name, OBJECT(dev), &error_fatal); + sysbus_init_child_obj(OBJECT(s), name, + &s->lpd.iou.adma[i], sizeof(s->lpd.iou.adma[i]), + TYPE_XLNX_ZDMA); + dev = DEVICE(&s->lpd.iou.adma[i]); + object_property_set_int(OBJECT(dev), 128, "bus-width", &error_abort); qdev_init_nofail(dev); - mr = sysbus_mmio_get_region(s->lpd.iou.adma[i], 0); + mr = sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), 0); memory_region_add_subregion(&s->mr_ps, MM_ADMA_CH0 + i * MM_ADMA_CH0_SIZE, mr); - sysbus_connect_irq(s->lpd.iou.adma[i], 0, pic[VERSAL_ADMA_IRQ_0 + i]); + sysbus_connect_irq(SYS_BUS_DEVICE(dev), 0, pic[VERSAL_ADMA_IRQ_0 + i]); g_free(name); } } -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Embed the APUs into the SoC type. Suggested-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-8-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 2 +- hw/arm/xlnx-versal-virt.c | 4 ++-- hw/arm/xlnx-versal.c | 19 +++++-------------- 3 files changed, 8 insertions(+), 17 deletions(-) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ typedef struct Versal { struct { struct { MemoryRegion mr; - ARMCPU *cpu[XLNX_VERSAL_NR_ACPUS]; + ARMCPU cpu[XLNX_VERSAL_NR_ACPUS]; GICv3State gic; } apu; } fpd; diff --git a/hw/arm/xlnx-versal-virt.c b/hw/arm/xlnx-versal-virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal-virt.c +++ b/hw/arm/xlnx-versal-virt.c @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) s->binfo.get_dtb = versal_virt_get_dtb; s->binfo.modify_dtb = versal_virt_modify_dtb; if (machine->kernel_filename) { - arm_load_kernel(s->soc.fpd.apu.cpu[0], machine, &s->binfo); + arm_load_kernel(&s->soc.fpd.apu.cpu[0], machine, &s->binfo); } else { - AddressSpace *as = arm_boot_address_space(s->soc.fpd.apu.cpu[0], + AddressSpace *as = arm_boot_address_space(&s->soc.fpd.apu.cpu[0], &s->binfo); /* Some boot-loaders (e.g u-boot) don't like blobs at address 0 (NULL). * Offset things by 4K. */ diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_apu_cpus(Versal *s) for (i = 0; i < ARRAY_SIZE(s->fpd.apu.cpu); i++) { Object *obj; - char *name; - - obj = object_new(XLNX_VERSAL_ACPU_TYPE); - if (!obj) { - error_report("Unable to create apu.cpu[%d] of type %s", - i, XLNX_VERSAL_ACPU_TYPE); - exit(EXIT_FAILURE); - } - - name = g_strdup_printf("apu-cpu[%d]", i); - object_property_add_child(OBJECT(s), name, obj, &error_fatal); - g_free(name); + object_initialize_child(OBJECT(s), "apu-cpu[*]", + &s->fpd.apu.cpu[i], sizeof(s->fpd.apu.cpu[i]), + XLNX_VERSAL_ACPU_TYPE, &error_abort, NULL); + obj = OBJECT(&s->fpd.apu.cpu[i]); object_property_set_int(obj, s->cfg.psci_conduit, "psci-conduit", &error_abort); if (i) { @@ -XXX,XX +XXX,XX @@ static void versal_create_apu_cpus(Versal *s) object_property_set_link(obj, OBJECT(&s->fpd.apu.mr), "memory", &error_abort); object_property_set_bool(obj, true, "realized", &error_fatal); - s->fpd.apu.cpu[i] = ARM_CPU(obj); } } @@ -XXX,XX +XXX,XX @@ static void versal_create_apu_gic(Versal *s, qemu_irq *pic) } for (i = 0; i < nr_apu_cpus; i++) { - DeviceState *cpudev = DEVICE(s->fpd.apu.cpu[i]); + DeviceState *cpudev = DEVICE(&s->fpd.apu.cpu[i]); int ppibase = XLNX_VERSAL_NR_IRQS + i * GIC_INTERNAL + GIC_NR_SGIS; qemu_irq maint_irq; int ti; -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Add support for SD. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-9-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 12 ++++++++++++ hw/arm/xlnx-versal.c | 31 +++++++++++++++++++++++++++++++ 2 files changed, 43 insertions(+) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "hw/sysbus.h" #include "hw/arm/boot.h" +#include "hw/sd/sdhci.h" #include "hw/intc/arm_gicv3.h" #include "hw/char/pl011.h" #include "hw/dma/xlnx-zdma.h" @@ -XXX,XX +XXX,XX @@ #define XLNX_VERSAL_NR_UARTS 2 #define XLNX_VERSAL_NR_GEMS 2 #define XLNX_VERSAL_NR_ADMAS 8 +#define XLNX_VERSAL_NR_SDS 2 #define XLNX_VERSAL_NR_IRQS 192 typedef struct Versal { @@ -XXX,XX +XXX,XX @@ typedef struct Versal { } iou; } lpd; + /* The Platform Management Controller subsystem. */ + struct { + struct { + SDHCIState sd[XLNX_VERSAL_NR_SDS]; + } iou; + } pmc; + struct { MemoryRegion *mr_ddr; uint32_t psci_conduit; @@ -XXX,XX +XXX,XX @@ typedef struct Versal { #define VERSAL_GEM1_IRQ_0 58 #define VERSAL_GEM1_WAKE_IRQ_0 59 #define VERSAL_ADMA_IRQ_0 60 +#define VERSAL_SD0_IRQ_0 126 /* Architecturally reserved IRQs suitable for virtualization. */ #define VERSAL_RSVD_IRQ_FIRST 111 @@ -XXX,XX +XXX,XX @@ typedef struct Versal { #define MM_FPD_CRF 0xfd1a0000U #define MM_FPD_CRF_SIZE 0x140000 +#define MM_PMC_SD0 0xf1040000U +#define MM_PMC_SD0_SIZE 0x10000 #define MM_PMC_CRP 0xf1260000U #define MM_PMC_CRP_SIZE 0x10000 #endif diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_admas(Versal *s, qemu_irq *pic) } } +#define SDHCI_CAPABILITIES 0x280737ec6481 /* Same as on ZynqMP. */ +static void versal_create_sds(Versal *s, qemu_irq *pic) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(s->pmc.iou.sd); i++) { + DeviceState *dev; + MemoryRegion *mr; + + sysbus_init_child_obj(OBJECT(s), "sd[*]", + &s->pmc.iou.sd[i], sizeof(s->pmc.iou.sd[i]), + TYPE_SYSBUS_SDHCI); + dev = DEVICE(&s->pmc.iou.sd[i]); + + object_property_set_uint(OBJECT(dev), + 3, "sd-spec-version", &error_fatal); + object_property_set_uint(OBJECT(dev), SDHCI_CAPABILITIES, "capareg", + &error_fatal); + object_property_set_uint(OBJECT(dev), UHS_I, "uhs", &error_fatal); + qdev_init_nofail(dev); + + mr = sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), 0); + memory_region_add_subregion(&s->mr_ps, + MM_PMC_SD0 + i * MM_PMC_SD0_SIZE, mr); + + sysbus_connect_irq(SYS_BUS_DEVICE(dev), 0, + pic[VERSAL_SD0_IRQ_0 + i * 2]); + } +} + /* This takes the board allocated linear DDR memory and creates aliases * for each split DDR range/aperture on the Versal address map. */ @@ -XXX,XX +XXX,XX @@ static void versal_realize(DeviceState *dev, Error **errp) versal_create_uarts(s, pic); versal_create_gems(s, pic); versal_create_admas(s, pic); + versal_create_sds(s, pic); versal_map_ddr(s); versal_unimp(s); -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> hw/arm: versal: Add support for the RTC. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-10-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/xlnx-versal.h | 8 ++++++++ hw/arm/xlnx-versal.c | 21 +++++++++++++++++++++ 2 files changed, 29 insertions(+) diff --git a/include/hw/arm/xlnx-versal.h b/include/hw/arm/xlnx-versal.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/xlnx-versal.h +++ b/include/hw/arm/xlnx-versal.h @@ -XXX,XX +XXX,XX @@ #include "hw/char/pl011.h" #include "hw/dma/xlnx-zdma.h" #include "hw/net/cadence_gem.h" +#include "hw/rtc/xlnx-zynqmp-rtc.h" #define TYPE_XLNX_VERSAL "xlnx-versal" #define XLNX_VERSAL(obj) OBJECT_CHECK(Versal, (obj), TYPE_XLNX_VERSAL) @@ -XXX,XX +XXX,XX @@ typedef struct Versal { struct { SDHCIState sd[XLNX_VERSAL_NR_SDS]; } iou; + + XlnxZynqMPRTC rtc; } pmc; struct { @@ -XXX,XX +XXX,XX @@ typedef struct Versal { #define VERSAL_GEM1_IRQ_0 58 #define VERSAL_GEM1_WAKE_IRQ_0 59 #define VERSAL_ADMA_IRQ_0 60 +#define VERSAL_RTC_APB_ERR_IRQ 121 #define VERSAL_SD0_IRQ_0 126 +#define VERSAL_RTC_ALARM_IRQ 142 +#define VERSAL_RTC_SECONDS_IRQ 143 /* Architecturally reserved IRQs suitable for virtualization. */ #define VERSAL_RSVD_IRQ_FIRST 111 @@ -XXX,XX +XXX,XX @@ typedef struct Versal { #define MM_PMC_SD0_SIZE 0x10000 #define MM_PMC_CRP 0xf1260000U #define MM_PMC_CRP_SIZE 0x10000 +#define MM_PMC_RTC 0xf12a0000 +#define MM_PMC_RTC_SIZE 0x10000 #endif diff --git a/hw/arm/xlnx-versal.c b/hw/arm/xlnx-versal.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal.c +++ b/hw/arm/xlnx-versal.c @@ -XXX,XX +XXX,XX @@ static void versal_create_sds(Versal *s, qemu_irq *pic) } } +static void versal_create_rtc(Versal *s, qemu_irq *pic) +{ + SysBusDevice *sbd; + MemoryRegion *mr; + + sysbus_init_child_obj(OBJECT(s), "rtc", &s->pmc.rtc, sizeof(s->pmc.rtc), + TYPE_XLNX_ZYNQMP_RTC); + sbd = SYS_BUS_DEVICE(&s->pmc.rtc); + qdev_init_nofail(DEVICE(sbd)); + + mr = sysbus_mmio_get_region(sbd, 0); + memory_region_add_subregion(&s->mr_ps, MM_PMC_RTC, mr); + + /* + * TODO: Connect the ALARM and SECONDS interrupts once our RTC model + * supports them. + */ + sysbus_connect_irq(sbd, 1, pic[VERSAL_RTC_APB_ERR_IRQ]); +} + /* This takes the board allocated linear DDR memory and creates aliases * for each split DDR range/aperture on the Versal address map. */ @@ -XXX,XX +XXX,XX @@ static void versal_realize(DeviceState *dev, Error **errp) versal_create_gems(s, pic); versal_create_admas(s, pic); versal_create_sds(s, pic); + versal_create_rtc(s, pic); versal_map_ddr(s); versal_unimp(s); -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Add support for SD. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-11-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/xlnx-versal-virt.c | 46 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 46 insertions(+) diff --git a/hw/arm/xlnx-versal-virt.c b/hw/arm/xlnx-versal-virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal-virt.c +++ b/hw/arm/xlnx-versal-virt.c @@ -XXX,XX +XXX,XX @@ #include "hw/arm/sysbus-fdt.h" #include "hw/arm/fdt.h" #include "cpu.h" +#include "hw/qdev-properties.h" #include "hw/arm/xlnx-versal.h" #define TYPE_XLNX_VERSAL_VIRT_MACHINE MACHINE_TYPE_NAME("xlnx-versal-virt") @@ -XXX,XX +XXX,XX @@ static void fdt_add_zdma_nodes(VersalVirt *s) } } +static void fdt_add_sd_nodes(VersalVirt *s) +{ + const char clocknames[] = "clk_xin\0clk_ahb"; + const char compat[] = "arasan,sdhci-8.9a"; + int i; + + for (i = ARRAY_SIZE(s->soc.pmc.iou.sd) - 1; i >= 0; i--) { + uint64_t addr = MM_PMC_SD0 + MM_PMC_SD0_SIZE * i; + char *name = g_strdup_printf("/sdhci@%" PRIx64, addr); + + qemu_fdt_add_subnode(s->fdt, name); + + qemu_fdt_setprop_cells(s->fdt, name, "clocks", + s->phandle.clk_25Mhz, s->phandle.clk_25Mhz); + qemu_fdt_setprop(s->fdt, name, "clock-names", + clocknames, sizeof(clocknames)); + qemu_fdt_setprop_cells(s->fdt, name, "interrupts", + GIC_FDT_IRQ_TYPE_SPI, VERSAL_SD0_IRQ_0 + i * 2, + GIC_FDT_IRQ_FLAGS_LEVEL_HI); + qemu_fdt_setprop_sized_cells(s->fdt, name, "reg", + 2, addr, 2, MM_PMC_SD0_SIZE); + qemu_fdt_setprop(s->fdt, name, "compatible", compat, sizeof(compat)); + g_free(name); + } +} + static void fdt_nop_memory_nodes(void *fdt, Error **errp) { Error *err = NULL; @@ -XXX,XX +XXX,XX @@ static void create_virtio_regions(VersalVirt *s) } } +static void sd_plugin_card(SDHCIState *sd, DriveInfo *di) +{ + BlockBackend *blk = di ? blk_by_legacy_dinfo(di) : NULL; + DeviceState *card; + + card = qdev_create(qdev_get_child_bus(DEVICE(sd), "sd-bus"), TYPE_SD_CARD); + object_property_add_child(OBJECT(sd), "card[*]", OBJECT(card), + &error_fatal); + qdev_prop_set_drive(card, "drive", blk, &error_fatal); + object_property_set_bool(OBJECT(card), true, "realized", &error_fatal); +} + static void versal_virt_init(MachineState *machine) { VersalVirt *s = XLNX_VERSAL_VIRT_MACHINE(machine); int psci_conduit = QEMU_PSCI_CONDUIT_DISABLED; + int i; /* * If the user provides an Operating System to be loaded, we expect them @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) fdt_add_gic_nodes(s); fdt_add_timer_nodes(s); fdt_add_zdma_nodes(s); + fdt_add_sd_nodes(s); fdt_add_cpu_nodes(s, psci_conduit); fdt_add_clk_node(s, "/clk125", 125000000, s->phandle.clk_125Mhz); fdt_add_clk_node(s, "/clk25", 25000000, s->phandle.clk_25Mhz); @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) memory_region_add_subregion_overlap(get_system_memory(), 0, &s->soc.fpd.apu.mr, 0); + /* Plugin SD cards. */ + for (i = 0; i < ARRAY_SIZE(s->soc.pmc.iou.sd); i++) { + sd_plugin_card(&s->soc.pmc.iou.sd[i], drive_get_next(IF_SD)); + } + s->binfo.ram_size = machine->ram_size; s->binfo.loader_start = 0x0; s->binfo.get_dtb = versal_virt_get_dtb; -- 2.20.1 From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com> Add support for the RTC. Signed-off-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Luc Michel <luc.michel@greensocs.com> Message-id: 20200427181649.26851-12-edgar.iglesias@gmail.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/xlnx-versal-virt.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/hw/arm/xlnx-versal-virt.c b/hw/arm/xlnx-versal-virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-versal-virt.c +++ b/hw/arm/xlnx-versal-virt.c @@ -XXX,XX +XXX,XX @@ static void fdt_add_sd_nodes(VersalVirt *s) } } +static void fdt_add_rtc_node(VersalVirt *s) +{ + const char compat[] = "xlnx,zynqmp-rtc"; + const char interrupt_names[] = "alarm\0sec"; + char *name = g_strdup_printf("/rtc@%x", MM_PMC_RTC); + + qemu_fdt_add_subnode(s->fdt, name); + + qemu_fdt_setprop_cells(s->fdt, name, "interrupts", + GIC_FDT_IRQ_TYPE_SPI, VERSAL_RTC_ALARM_IRQ, + GIC_FDT_IRQ_FLAGS_LEVEL_HI, + GIC_FDT_IRQ_TYPE_SPI, VERSAL_RTC_SECONDS_IRQ, + GIC_FDT_IRQ_FLAGS_LEVEL_HI); + qemu_fdt_setprop(s->fdt, name, "interrupt-names", + interrupt_names, sizeof(interrupt_names)); + qemu_fdt_setprop_sized_cells(s->fdt, name, "reg", + 2, MM_PMC_RTC, 2, MM_PMC_RTC_SIZE); + qemu_fdt_setprop(s->fdt, name, "compatible", compat, sizeof(compat)); + g_free(name); +} + static void fdt_nop_memory_nodes(void *fdt, Error **errp) { Error *err = NULL; @@ -XXX,XX +XXX,XX @@ static void versal_virt_init(MachineState *machine) fdt_add_timer_nodes(s); fdt_add_zdma_nodes(s); fdt_add_sd_nodes(s); + fdt_add_rtc_node(s); fdt_add_cpu_nodes(s, psci_conduit); fdt_add_clk_node(s, "/clk125", 125000000, s->phandle.clk_125Mhz); fdt_add_clk_node(s, "/clk25", 25000000, s->phandle.clk_25Mhz); -- 2.20.1 Somewhere along theline we accidentally added a duplicate "using D16-D31 when they don't exist" check to do_vfm_dp() (probably an artifact of a patchseries rebase). Remove it. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200430181003.21682-2-peter.maydell@linaro.org --- target/arm/translate-vfp.inc.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/target/arm/translate-vfp.inc.c b/target/arm/translate-vfp.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-vfp.inc.c +++ b/target/arm/translate-vfp.inc.c @@ -XXX,XX +XXX,XX @@ static bool do_vfm_dp(DisasContext *s, arg_VFMA_dp *a, bool neg_n, bool neg_d) return false; } - /* UNDEF accesses to D16-D31 if they don't exist. */ - if (!dc_isar_feature(aa32_simd_r32, s) && - ((a->vd | a->vn | a->vm) & 0x10)) { - return false; - } - if (!vfp_access_check(s)) { return true; } -- 2.20.1 We were accidentally permitting decode of Thumb Neon insns even if the CPU didn't have the FEATURE_NEON bit set, because the feature check was being done before the call to disas_neon_data_insn() and disas_neon_ls_insn() in the Arm decoder but was omitted from the Thumb decoder. Push the feature bit check down into the called functions so it is done for both Arm and Thumb encodings. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200430181003.21682-3-peter.maydell@linaro.org --- target/arm/translate.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_ls_insn(DisasContext *s, uint32_t insn) TCGv_i32 tmp2; TCGv_i64 tmp64; + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return 1; + } + /* FIXME: this access check should not take precedence over UNDEF * for invalid encodings; we will generate incorrect syndrome information * for attempts to execute invalid vfp/neon encodings with FP disabled. @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) TCGv_ptr ptr1, ptr2, ptr3; TCGv_i64 tmp64; + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return 1; + } + /* FIXME: this access check should not take precedence over UNDEF * for invalid encodings; we will generate incorrect syndrome information * for attempts to execute invalid vfp/neon encodings with FP disabled. @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) if (((insn >> 25) & 7) == 1) { /* NEON Data processing. */ - if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { - goto illegal_op; - } - if (disas_neon_data_insn(s, insn)) { goto illegal_op; } @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) } if ((insn & 0x0f100000) == 0x04000000) { /* NEON load/store. */ - if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { - goto illegal_op; - } - if (disas_neon_ls_insn(s, insn)) { goto illegal_op; } -- 2.20.1 Add the infrastructure for building and invoking a decodetree decoder for the AArch32 Neon encodings. At the moment the new decoder covers nothing, so we always fall back to the existing hand-written decode. We follow the same pattern we did for the VFP decodetree conversion (commit 78e138bc1f672c145ef6ace74617d and following): code that deals with Neon will be moving gradually out to translate-neon.vfp.inc, which we #include into translate.c. In order to share the decode files between A32 and T32, we split Neon into 3 parts: * data-processing * load-store * 'shared' encodings The first two groups of instructions have similar but not identical A32 and T32 encodings, so we need to manually transform the T32 encoding into the A32 one before calling the decoder; the third group covers the Neon instructions which are identical in A32 and T32. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-4-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 29 ++++++++++++++++++++++++++ target/arm/neon-ls.decode | 29 ++++++++++++++++++++++++++ target/arm/neon-shared.decode | 27 +++++++++++++++++++++++++ target/arm/translate-neon.inc.c | 32 +++++++++++++++++++++++++++++ target/arm/translate.c | 36 +++++++++++++++++++++++++++++++-- target/arm/Makefile.objs | 18 +++++++++++++++++ 6 files changed, 169 insertions(+), 2 deletions(-) create mode 100644 target/arm/neon-dp.decode create mode 100644 target/arm/neon-ls.decode create mode 100644 target/arm/neon-shared.decode create mode 100644 target/arm/translate-neon.inc.c diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ +# AArch32 Neon data-processing instruction descriptions +# +# Copyright (c) 2020 Linaro, Ltd +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. + +# +# This file is processed by scripts/decodetree.py +# + +# Encodings for Neon data processing instructions where the T32 encoding +# is a simple transformation of the A32 encoding. +# More specifically, this file covers instructions where the A32 encoding is +# 0b1111_001p_qqqq_qqqq_qqqq_qqqq_qqqq_qqqq +# and the T32 encoding is +# 0b111p_1111_qqqq_qqqq_qqqq_qqqq_qqqq_qqqq +# This file works on the A32 encoding only; calling code for T32 has to +# transform the insn into the A32 version first. diff --git a/target/arm/neon-ls.decode b/target/arm/neon-ls.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/neon-ls.decode @@ -XXX,XX +XXX,XX @@ +# AArch32 Neon load/store instruction descriptions +# +# Copyright (c) 2020 Linaro, Ltd +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. + +# +# This file is processed by scripts/decodetree.py +# + +# Encodings for Neon load/store instructions where the T32 encoding +# is a simple transformation of the A32 encoding. +# More specifically, this file covers instructions where the A32 encoding is +# 0b1111_0100_xxx0_xxxx_xxxx_xxxx_xxxx_xxxx +# and the T32 encoding is +# 0b1111_1001_xxx0_xxxx_xxxx_xxxx_xxxx_xxxx +# This file works on the A32 encoding only; calling code for T32 has to +# transform the insn into the A32 version first. diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ +# AArch32 Neon instruction descriptions +# +# Copyright (c) 2020 Linaro, Ltd +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. + +# +# This file is processed by scripts/decodetree.py +# + +# Encodings for Neon instructions whose encoding is the same for +# both A32 and T32. + +# More specifically, this covers: +# 2reg scalar ext: 0b1111_1110_xxxx_xxxx_xxxx_1x0x_xxxx_xxxx +# 3same ext: 0b1111_110x_xxxx_xxxx_xxxx_1x0x_xxxx_xxxx diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ +/* + * ARM translation: AArch32 Neon instructions + * + * Copyright (c) 2003 Fabrice Bellard + * Copyright (c) 2005-2007 CodeSourcery + * Copyright (c) 2007 OpenedHand, Ltd. + * Copyright (c) 2020 Linaro, Ltd. + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, see <http://www.gnu.org/licenses/>. + */ + +/* + * This file is intended to be included from translate.c; it uses + * some macros and definitions provided by that file. + * It might be possible to convert it to a standalone .c file eventually. + */ + +/* Include the generated Neon decoder */ +#include "decode-neon-dp.inc.c" +#include "decode-neon-ls.inc.c" +#include "decode-neon-shared.inc.c" diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static TCGv_ptr vfp_reg_ptr(bool dp, int reg) #define ARM_CP_RW_BIT (1 << 20) -/* Include the VFP decoder */ +/* Include the VFP and Neon decoders */ #include "translate-vfp.inc.c" +#include "translate-neon.inc.c" static inline void iwmmxt_load_reg(TCGv_i64 var, int reg) { @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) /* Unconditional instructions. */ /* TODO: Perhaps merge these into one decodetree output file. */ if (disas_a32_uncond(s, insn) || - disas_vfp_uncond(s, insn)) { + disas_vfp_uncond(s, insn) || + disas_neon_dp(s, insn) || + disas_neon_ls(s, insn) || + disas_neon_shared(s, insn)) { return; } /* fall back to legacy decoder */ @@ -XXX,XX +XXX,XX @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn) ARCH(6T2); } + if ((insn & 0xef000000) == 0xef000000) { + /* + * T32 encodings 0b111p_1111_qqqq_qqqq_qqqq_qqqq_qqqq_qqqq + * transform into + * A32 encodings 0b1111_001p_qqqq_qqqq_qqqq_qqqq_qqqq_qqqq + */ + uint32_t a32_insn = (insn & 0xe2ffffff) | + ((insn & (1 << 28)) >> 4) | (1 << 28); + + if (disas_neon_dp(s, a32_insn)) { + return; + } + } + + if ((insn & 0xff100000) == 0xf9000000) { + /* + * T32 encodings 0b1111_1001_ppp0_qqqq_qqqq_qqqq_qqqq_qqqq + * transform into + * A32 encodings 0b1111_0100_ppp0_qqqq_qqqq_qqqq_qqqq_qqqq + */ + uint32_t a32_insn = (insn & 0x00ffffff) | 0xf4000000; + + if (disas_neon_ls(s, a32_insn)) { + return; + } + } + /* * TODO: Perhaps merge these into one decodetree output file. * Note disas_vfp is written for a32 with cond field in the @@ -XXX,XX +XXX,XX @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn) */ if (disas_t32(s, insn) || disas_vfp_uncond(s, insn) || + disas_neon_shared(s, insn) || ((insn >> 28) == 0xe && disas_vfp(s, insn))) { return; } diff --git a/target/arm/Makefile.objs b/target/arm/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/target/arm/Makefile.objs +++ b/target/arm/Makefile.objs @@ -XXX,XX +XXX,XX @@ target/arm/decode-sve.inc.c: $(SRC_PATH)/target/arm/sve.decode $(DECODETREE) $(PYTHON) $(DECODETREE) --decode disas_sve -o $@ $<,\ "GEN", $(TARGET_DIR)$@) +target/arm/decode-neon-shared.inc.c: $(SRC_PATH)/target/arm/neon-shared.decode $(DECODETREE) + $(call quiet-command,\ + $(PYTHON) $(DECODETREE) --static-decode disas_neon_shared -o $@ $<,\ + "GEN", $(TARGET_DIR)$@) + +target/arm/decode-neon-dp.inc.c: $(SRC_PATH)/target/arm/neon-dp.decode $(DECODETREE) + $(call quiet-command,\ + $(PYTHON) $(DECODETREE) --static-decode disas_neon_dp -o $@ $<,\ + "GEN", $(TARGET_DIR)$@) + +target/arm/decode-neon-ls.inc.c: $(SRC_PATH)/target/arm/neon-ls.decode $(DECODETREE) + $(call quiet-command,\ + $(PYTHON) $(DECODETREE) --static-decode disas_neon_ls -o $@ $<,\ + "GEN", $(TARGET_DIR)$@) + target/arm/decode-vfp.inc.c: $(SRC_PATH)/target/arm/vfp.decode $(DECODETREE) $(call quiet-command,\ $(PYTHON) $(DECODETREE) --static-decode disas_vfp -o $@ $<,\ @@ -XXX,XX +XXX,XX @@ target/arm/decode-t16.inc.c: $(SRC_PATH)/target/arm/t16.decode $(DECODETREE) "GEN", $(TARGET_DIR)$@) target/arm/translate-sve.o: target/arm/decode-sve.inc.c +target/arm/translate.o: target/arm/decode-neon-shared.inc.c +target/arm/translate.o: target/arm/decode-neon-dp.inc.c +target/arm/translate.o: target/arm/decode-neon-ls.inc.c target/arm/translate.o: target/arm/decode-vfp.inc.c target/arm/translate.o: target/arm/decode-vfp-uncond.inc.c target/arm/translate.o: target/arm/decode-a32.inc.c -- 2.20.1 Convert the VCMLA (vector) insns in the 3same extension group to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-5-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 11 ++++++++++ target/arm/translate-neon.inc.c | 37 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 11 +--------- 3 files changed, 49 insertions(+), 10 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ # More specifically, this covers: # 2reg scalar ext: 0b1111_1110_xxxx_xxxx_xxxx_1x0x_xxxx_xxxx # 3same ext: 0b1111_110x_xxxx_xxxx_xxxx_1x0x_xxxx_xxxx + +# VFP/Neon register fields; same as vfp.decode +%vm_dp 5:1 0:4 +%vm_sp 0:4 5:1 +%vn_dp 7:1 16:4 +%vn_sp 16:4 7:1 +%vd_dp 22:1 12:4 +%vd_sp 12:4 22:1 + +VCMLA 1111 110 rot:2 . 1 size:1 .... .... 1000 . q:1 . 0 .... \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ #include "decode-neon-dp.inc.c" #include "decode-neon-ls.inc.c" #include "decode-neon-shared.inc.c" + +static bool trans_VCMLA(DisasContext *s, arg_VCMLA *a) +{ + int opr_sz; + TCGv_ptr fpst; + gen_helper_gvec_3_ptr *fn_gvec_ptr; + + if (!dc_isar_feature(aa32_vcma, s) + || (!a->size && !dc_isar_feature(aa32_fp16_arith, s))) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vn | a->vm) & 0x10)) { + return false; + } + + if ((a->vn | a->vm | a->vd) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + opr_sz = (1 + a->q) * 8; + fpst = get_fpstatus_ptr(1); + fn_gvec_ptr = a->size ? gen_helper_gvec_fcmlas : gen_helper_gvec_fcmlah; + tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), + vfp_reg_offset(1, a->vn), + vfp_reg_offset(1, a->vm), + fpst, opr_sz, opr_sz, a->rot, + fn_gvec_ptr); + tcg_temp_free_ptr(fpst); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_insn_3same_ext(DisasContext *s, uint32_t insn) bool is_long = false, q = extract32(insn, 6, 1); bool ptr_is_env = false; - if ((insn & 0xfe200f10) == 0xfc200800) { - /* VCMLA -- 1111 110R R.1S .... .... 1000 ...0 .... */ - int size = extract32(insn, 20, 1); - data = extract32(insn, 23, 2); /* rot */ - if (!dc_isar_feature(aa32_vcma, s) - || (!size && !dc_isar_feature(aa32_fp16_arith, s))) { - return 1; - } - fn_gvec_ptr = size ? gen_helper_gvec_fcmlas : gen_helper_gvec_fcmlah; - } else if ((insn & 0xfea00f10) == 0xfc800800) { + if ((insn & 0xfea00f10) == 0xfc800800) { /* VCADD -- 1111 110R 1.0S .... .... 1000 ...0 .... */ int size = extract32(insn, 20, 1); data = extract32(insn, 24, 1); /* rot */ -- 2.20.1 Convert the VCADD (vector) insns to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-6-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 3 +++ target/arm/translate-neon.inc.c | 37 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 11 +--------- 3 files changed, 41 insertions(+), 10 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ VCMLA 1111 110 rot:2 . 1 size:1 .... .... 1000 . q:1 . 0 .... \ vm=%vm_dp vn=%vn_dp vd=%vd_dp + +VCADD 1111 110 rot:1 1 . 0 size:1 .... .... 1000 . q:1 . 0 .... \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VCMLA(DisasContext *s, arg_VCMLA *a) tcg_temp_free_ptr(fpst); return true; } + +static bool trans_VCADD(DisasContext *s, arg_VCADD *a) +{ + int opr_sz; + TCGv_ptr fpst; + gen_helper_gvec_3_ptr *fn_gvec_ptr; + + if (!dc_isar_feature(aa32_vcma, s) + || (!a->size && !dc_isar_feature(aa32_fp16_arith, s))) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vn | a->vm) & 0x10)) { + return false; + } + + if ((a->vn | a->vm | a->vd) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + opr_sz = (1 + a->q) * 8; + fpst = get_fpstatus_ptr(1); + fn_gvec_ptr = a->size ? gen_helper_gvec_fcadds : gen_helper_gvec_fcaddh; + tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), + vfp_reg_offset(1, a->vn), + vfp_reg_offset(1, a->vm), + fpst, opr_sz, opr_sz, a->rot, + fn_gvec_ptr); + tcg_temp_free_ptr(fpst); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_insn_3same_ext(DisasContext *s, uint32_t insn) bool is_long = false, q = extract32(insn, 6, 1); bool ptr_is_env = false; - if ((insn & 0xfea00f10) == 0xfc800800) { - /* VCADD -- 1111 110R 1.0S .... .... 1000 ...0 .... */ - int size = extract32(insn, 20, 1); - data = extract32(insn, 24, 1); /* rot */ - if (!dc_isar_feature(aa32_vcma, s) - || (!size && !dc_isar_feature(aa32_fp16_arith, s))) { - return 1; - } - fn_gvec_ptr = size ? gen_helper_gvec_fcadds : gen_helper_gvec_fcaddh; - } else if ((insn & 0xfeb00f00) == 0xfc200d00) { + if ((insn & 0xfeb00f00) == 0xfc200d00) { /* V[US]DOT -- 1111 1100 0.10 .... .... 1101 .Q.U .... */ bool u = extract32(insn, 4, 1); if (!dc_isar_feature(aa32_dp, s)) { -- 2.20.1 Convert the V[US]DOT (vector) insns to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-7-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 4 ++++ target/arm/translate-neon.inc.c | 32 ++++++++++++++++++++++++++++++++ target/arm/translate.c | 9 +-------- 3 files changed, 37 insertions(+), 8 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ VCMLA 1111 110 rot:2 . 1 size:1 .... .... 1000 . q:1 . 0 .... \ VCADD 1111 110 rot:1 1 . 0 size:1 .... .... 1000 . q:1 . 0 .... \ vm=%vm_dp vn=%vn_dp vd=%vd_dp + +# VUDOT and VSDOT +VDOT 1111 110 00 . 10 .... .... 1101 . q:1 . u:1 .... \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VCADD(DisasContext *s, arg_VCADD *a) tcg_temp_free_ptr(fpst); return true; } + +static bool trans_VDOT(DisasContext *s, arg_VDOT *a) +{ + int opr_sz; + gen_helper_gvec_3 *fn_gvec; + + if (!dc_isar_feature(aa32_dp, s)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vn | a->vm) & 0x10)) { + return false; + } + + if ((a->vn | a->vm | a->vd) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + opr_sz = (1 + a->q) * 8; + fn_gvec = a->u ? gen_helper_gvec_udot_b : gen_helper_gvec_sdot_b; + tcg_gen_gvec_3_ool(vfp_reg_offset(1, a->vd), + vfp_reg_offset(1, a->vn), + vfp_reg_offset(1, a->vm), + opr_sz, opr_sz, 0, fn_gvec); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_insn_3same_ext(DisasContext *s, uint32_t insn) bool is_long = false, q = extract32(insn, 6, 1); bool ptr_is_env = false; - if ((insn & 0xfeb00f00) == 0xfc200d00) { - /* V[US]DOT -- 1111 1100 0.10 .... .... 1101 .Q.U .... */ - bool u = extract32(insn, 4, 1); - if (!dc_isar_feature(aa32_dp, s)) { - return 1; - } - fn_gvec = u ? gen_helper_gvec_udot_b : gen_helper_gvec_sdot_b; - } else if ((insn & 0xff300f10) == 0xfc200810) { + if ((insn & 0xff300f10) == 0xfc200810) { /* VFM[AS]L -- 1111 1100 S.10 .... .... 1000 .Q.1 .... */ int is_s = extract32(insn, 23, 1); if (!dc_isar_feature(aa32_fhm, s)) { -- 2.20.1 Convert the VFM[AS]L (vector) insns to decodetree. This is the last insn in the legacy decoder for the 3same_ext group, so we can delete the legacy decoder function for the group entirely. Note that in disas_thumb2_insn() the parts of this encoding space where the decodetree decoder returns false will correctly be directed to illegal_op by the "(insn & (1 << 28))" check so they won't fall into disas_coproc_insn() by mistake. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-8-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 6 +++ target/arm/translate-neon.inc.c | 31 +++++++++++ target/arm/translate.c | 92 +-------------------------------- 3 files changed, 38 insertions(+), 91 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ VCADD 1111 110 rot:1 1 . 0 size:1 .... .... 1000 . q:1 . 0 .... \ # VUDOT and VSDOT VDOT 1111 110 00 . 10 .... .... 1101 . q:1 . u:1 .... \ vm=%vm_dp vn=%vn_dp vd=%vd_dp + +# VFM[AS]L +VFML 1111 110 0 s:1 . 10 .... .... 1000 . 0 . 1 .... \ + vm=%vm_sp vn=%vn_sp vd=%vd_dp q=0 +VFML 1111 110 0 s:1 . 10 .... .... 1000 . 1 . 1 .... \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp q=1 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VDOT(DisasContext *s, arg_VDOT *a) opr_sz, opr_sz, 0, fn_gvec); return true; } + +static bool trans_VFML(DisasContext *s, arg_VFML *a) +{ + int opr_sz; + + if (!dc_isar_feature(aa32_fhm, s)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + (a->vd & 0x10)) { + return false; + } + + if (a->vd & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + opr_sz = (1 + a->q) * 8; + tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), + vfp_reg_offset(a->q, a->vn), + vfp_reg_offset(a->q, a->vm), + cpu_env, opr_sz, opr_sz, a->s, /* is_2 == 0 */ + gen_helper_gvec_fmlal_a32); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) return 0; } -/* Advanced SIMD three registers of the same length extension. - * 31 25 23 22 20 16 12 11 10 9 8 3 0 - * +---------------+-----+---+-----+----+----+---+----+---+----+---------+----+ - * | 1 1 1 1 1 1 0 | op1 | D | op2 | Vn | Vd | 1 | o3 | 0 | o4 | N Q M U | Vm | - * +---------------+-----+---+-----+----+----+---+----+---+----+---------+----+ - */ -static int disas_neon_insn_3same_ext(DisasContext *s, uint32_t insn) -{ - gen_helper_gvec_3 *fn_gvec = NULL; - gen_helper_gvec_3_ptr *fn_gvec_ptr = NULL; - int rd, rn, rm, opr_sz; - int data = 0; - int off_rn, off_rm; - bool is_long = false, q = extract32(insn, 6, 1); - bool ptr_is_env = false; - - if ((insn & 0xff300f10) == 0xfc200810) { - /* VFM[AS]L -- 1111 1100 S.10 .... .... 1000 .Q.1 .... */ - int is_s = extract32(insn, 23, 1); - if (!dc_isar_feature(aa32_fhm, s)) { - return 1; - } - is_long = true; - data = is_s; /* is_2 == 0 */ - fn_gvec_ptr = gen_helper_gvec_fmlal_a32; - ptr_is_env = true; - } else { - return 1; - } - - VFP_DREG_D(rd, insn); - if (rd & q) { - return 1; - } - if (q || !is_long) { - VFP_DREG_N(rn, insn); - VFP_DREG_M(rm, insn); - if ((rn | rm) & q & !is_long) { - return 1; - } - off_rn = vfp_reg_offset(1, rn); - off_rm = vfp_reg_offset(1, rm); - } else { - rn = VFP_SREG_N(insn); - rm = VFP_SREG_M(insn); - off_rn = vfp_reg_offset(0, rn); - off_rm = vfp_reg_offset(0, rm); - } - - if (s->fp_excp_el) { - gen_exception_insn(s, s->pc_curr, EXCP_UDEF, - syn_simd_access_trap(1, 0xe, false), s->fp_excp_el); - return 0; - } - if (!s->vfp_enabled) { - return 1; - } - - opr_sz = (1 + q) * 8; - if (fn_gvec_ptr) { - TCGv_ptr ptr; - if (ptr_is_env) { - ptr = cpu_env; - } else { - ptr = get_fpstatus_ptr(1); - } - tcg_gen_gvec_3_ptr(vfp_reg_offset(1, rd), off_rn, off_rm, ptr, - opr_sz, opr_sz, data, fn_gvec_ptr); - if (!ptr_is_env) { - tcg_temp_free_ptr(ptr); - } - } else { - tcg_gen_gvec_3_ool(vfp_reg_offset(1, rd), off_rn, off_rm, - opr_sz, opr_sz, data, fn_gvec); - } - return 0; -} - /* Advanced SIMD two registers and a scalar extension. * 31 24 23 22 20 16 12 11 10 9 8 3 0 * +-----------------+----+---+----+----+----+---+----+---+----+---------+----+ @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) } } } - } else if ((insn & 0x0e000a00) == 0x0c000800 - && arm_dc_feature(s, ARM_FEATURE_V8)) { - if (disas_neon_insn_3same_ext(s, insn)) { - goto illegal_op; - } - return; } else if ((insn & 0x0f000a00) == 0x0e000800 && arm_dc_feature(s, ARM_FEATURE_V8)) { if (disas_neon_insn_2reg_scalar_ext(s, insn)) { @@ -XXX,XX +XXX,XX @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn) } break; } - if ((insn & 0xfe000a00) == 0xfc000800 + if ((insn & 0xff000a00) == 0xfe000800 && arm_dc_feature(s, ARM_FEATURE_V8)) { /* The Thumb2 and ARM encodings are identical. */ - if (disas_neon_insn_3same_ext(s, insn)) { - goto illegal_op; - } - } else if ((insn & 0xff000a00) == 0xfe000800 - && arm_dc_feature(s, ARM_FEATURE_V8)) { - /* The Thumb2 and ARM encodings are identical. */ if (disas_neon_insn_2reg_scalar_ext(s, insn)) { goto illegal_op; } -- 2.20.1 Convert VCMLA (scalar) in the 2reg-scalar-ext group to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-9-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 5 +++++ target/arm/translate-neon.inc.c | 40 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 26 +-------------------- 3 files changed, 46 insertions(+), 25 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ VFML 1111 110 0 s:1 . 10 .... .... 1000 . 0 . 1 .... \ vm=%vm_sp vn=%vn_sp vd=%vd_dp q=0 VFML 1111 110 0 s:1 . 10 .... .... 1000 . 1 . 1 .... \ vm=%vm_dp vn=%vn_dp vd=%vd_dp q=1 + +VCMLA_scalar 1111 1110 0 . rot:2 .... .... 1000 . q:1 index:1 0 vm:4 \ + vn=%vn_dp vd=%vd_dp size=0 +VCMLA_scalar 1111 1110 1 . rot:2 .... .... 1000 . q:1 . 0 .... \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp size=1 index=0 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VFML(DisasContext *s, arg_VFML *a) gen_helper_gvec_fmlal_a32); return true; } + +static bool trans_VCMLA_scalar(DisasContext *s, arg_VCMLA_scalar *a) +{ + gen_helper_gvec_3_ptr *fn_gvec_ptr; + int opr_sz; + TCGv_ptr fpst; + + if (!dc_isar_feature(aa32_vcma, s)) { + return false; + } + if (a->size == 0 && !dc_isar_feature(aa32_fp16_arith, s)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vn | a->vm) & 0x10)) { + return false; + } + + if ((a->vd | a->vn) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + fn_gvec_ptr = (a->size ? gen_helper_gvec_fcmlas_idx + : gen_helper_gvec_fcmlah_idx); + opr_sz = (1 + a->q) * 8; + fpst = get_fpstatus_ptr(1); + tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), + vfp_reg_offset(1, a->vn), + vfp_reg_offset(1, a->vm), + fpst, opr_sz, opr_sz, + (a->index << 2) | a->rot, fn_gvec_ptr); + tcg_temp_free_ptr(fpst); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_insn_2reg_scalar_ext(DisasContext *s, uint32_t insn) bool is_long = false, q = extract32(insn, 6, 1); bool ptr_is_env = false; - if ((insn & 0xff000f10) == 0xfe000800) { - /* VCMLA (indexed) -- 1111 1110 S.RR .... .... 1000 ...0 .... */ - int rot = extract32(insn, 20, 2); - int size = extract32(insn, 23, 1); - int index; - - if (!dc_isar_feature(aa32_vcma, s)) { - return 1; - } - if (size == 0) { - if (!dc_isar_feature(aa32_fp16_arith, s)) { - return 1; - } - /* For fp16, rm is just Vm, and index is M. */ - rm = extract32(insn, 0, 4); - index = extract32(insn, 5, 1); - } else { - /* For fp32, rm is the usual M:Vm, and index is 0. */ - VFP_DREG_M(rm, insn); - index = 0; - } - data = (index << 2) | rot; - fn_gvec_ptr = (size ? gen_helper_gvec_fcmlas_idx - : gen_helper_gvec_fcmlah_idx); - } else if ((insn & 0xffb00f00) == 0xfe200d00) { + if ((insn & 0xffb00f00) == 0xfe200d00) { /* V[US]DOT -- 1111 1110 0.10 .... .... 1101 .Q.U .... */ int u = extract32(insn, 4, 1); -- 2.20.1 Convert the V[US]DOT (scalar) insns in the 2reg-scalar-ext group to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-10-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 3 +++ target/arm/translate-neon.inc.c | 35 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 13 +----------- 3 files changed, 39 insertions(+), 12 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ VCMLA_scalar 1111 1110 0 . rot:2 .... .... 1000 . q:1 index:1 0 vm:4 \ vn=%vn_dp vd=%vd_dp size=0 VCMLA_scalar 1111 1110 1 . rot:2 .... .... 1000 . q:1 . 0 .... \ vm=%vm_dp vn=%vn_dp vd=%vd_dp size=1 index=0 + +VDOT_scalar 1111 1110 0 . 10 .... .... 1101 . q:1 index:1 u:1 rm:4 \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VCMLA_scalar(DisasContext *s, arg_VCMLA_scalar *a) tcg_temp_free_ptr(fpst); return true; } + +static bool trans_VDOT_scalar(DisasContext *s, arg_VDOT_scalar *a) +{ + gen_helper_gvec_3 *fn_gvec; + int opr_sz; + TCGv_ptr fpst; + + if (!dc_isar_feature(aa32_dp, s)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vn) & 0x10)) { + return false; + } + + if ((a->vd | a->vn) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + fn_gvec = a->u ? gen_helper_gvec_udot_idx_b : gen_helper_gvec_sdot_idx_b; + opr_sz = (1 + a->q) * 8; + fpst = get_fpstatus_ptr(1); + tcg_gen_gvec_3_ool(vfp_reg_offset(1, a->vd), + vfp_reg_offset(1, a->vn), + vfp_reg_offset(1, a->rm), + opr_sz, opr_sz, a->index, fn_gvec); + tcg_temp_free_ptr(fpst); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_insn_2reg_scalar_ext(DisasContext *s, uint32_t insn) bool is_long = false, q = extract32(insn, 6, 1); bool ptr_is_env = false; - if ((insn & 0xffb00f00) == 0xfe200d00) { - /* V[US]DOT -- 1111 1110 0.10 .... .... 1101 .Q.U .... */ - int u = extract32(insn, 4, 1); - - if (!dc_isar_feature(aa32_dp, s)) { - return 1; - } - fn_gvec = u ? gen_helper_gvec_udot_idx_b : gen_helper_gvec_sdot_idx_b; - /* rm is just Vm, and index is M. */ - data = extract32(insn, 5, 1); /* index */ - rm = extract32(insn, 0, 4); - } else if ((insn & 0xffa00f10) == 0xfe000810) { + if ((insn & 0xffa00f10) == 0xfe000810) { /* VFM[AS]L -- 1111 1110 0.0S .... .... 1000 .Q.1 .... */ int is_s = extract32(insn, 20, 1); int vm20 = extract32(insn, 0, 3); -- 2.20.1 Convert the VFM[AS]L (scalar) insns in the 2reg-scalar-ext group to decodetree. These are the last ones in the group so we can remove all the legacy decode for the group. Note that in disas_thumb2_insn() the parts of this encoding space where the decodetree decoder returns false will correctly be directed to illegal_op by the "(insn & (1 << 28))" check so they won't fall into disas_coproc_insn() by mistake. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-11-peter.maydell@linaro.org --- target/arm/neon-shared.decode | 7 +++ target/arm/translate-neon.inc.c | 32 ++++++++++ target/arm/translate.c | 107 +------------------------------- 3 files changed, 40 insertions(+), 106 deletions(-) diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -XXX,XX +XXX,XX @@ VCMLA_scalar 1111 1110 1 . rot:2 .... .... 1000 . q:1 . 0 .... \ VDOT_scalar 1111 1110 0 . 10 .... .... 1101 . q:1 index:1 u:1 rm:4 \ vm=%vm_dp vn=%vn_dp vd=%vd_dp + +%vfml_scalar_q0_rm 0:3 5:1 +%vfml_scalar_q1_index 5:1 3:1 +VFML_scalar 1111 1110 0 . 0 s:1 .... .... 1000 . 0 . 1 index:1 ... \ + rm=%vfml_scalar_q0_rm vn=%vn_sp vd=%vd_dp q=0 +VFML_scalar 1111 1110 0 . 0 s:1 .... .... 1000 . 1 . 1 . rm:3 \ + index=%vfml_scalar_q1_index vn=%vn_dp vd=%vd_dp q=1 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VDOT_scalar(DisasContext *s, arg_VDOT_scalar *a) tcg_temp_free_ptr(fpst); return true; } + +static bool trans_VFML_scalar(DisasContext *s, arg_VFML_scalar *a) +{ + int opr_sz; + + if (!dc_isar_feature(aa32_fhm, s)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd & 0x10) || (a->q && (a->vn & 0x10)))) { + return false; + } + + if (a->vd & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + opr_sz = (1 + a->q) * 8; + tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), + vfp_reg_offset(a->q, a->vn), + vfp_reg_offset(a->q, a->rm), + cpu_env, opr_sz, opr_sz, + (a->index << 2) | a->s, /* is_2 == 0 */ + gen_helper_gvec_fmlal_idx_a32); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_dsp_insn(DisasContext *s, uint32_t insn) } #define VFP_REG_SHR(x, n) (((n) > 0) ? (x) >> (n) : (x) << -(n)) -#define VFP_SREG(insn, bigbit, smallbit) \ - ((VFP_REG_SHR(insn, bigbit - 1) & 0x1e) | (((insn) >> (smallbit)) & 1)) #define VFP_DREG(reg, insn, bigbit, smallbit) do { \ if (dc_isar_feature(aa32_simd_r32, s)) { \ reg = (((insn) >> (bigbit)) & 0x0f) \ @@ -XXX,XX +XXX,XX @@ static int disas_dsp_insn(DisasContext *s, uint32_t insn) reg = ((insn) >> (bigbit)) & 0x0f; \ }} while (0) -#define VFP_SREG_D(insn) VFP_SREG(insn, 12, 22) #define VFP_DREG_D(reg, insn) VFP_DREG(reg, insn, 12, 22) -#define VFP_SREG_N(insn) VFP_SREG(insn, 16, 7) #define VFP_DREG_N(reg, insn) VFP_DREG(reg, insn, 16, 7) -#define VFP_SREG_M(insn) VFP_SREG(insn, 0, 5) #define VFP_DREG_M(reg, insn) VFP_DREG(reg, insn, 0, 5) static void gen_neon_dup_low16(TCGv_i32 var) @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) return 0; } -/* Advanced SIMD two registers and a scalar extension. - * 31 24 23 22 20 16 12 11 10 9 8 3 0 - * +-----------------+----+---+----+----+----+---+----+---+----+---------+----+ - * | 1 1 1 1 1 1 1 0 | o1 | D | o2 | Vn | Vd | 1 | o3 | 0 | o4 | N Q M U | Vm | - * +-----------------+----+---+----+----+----+---+----+---+----+---------+----+ - * - */ - -static int disas_neon_insn_2reg_scalar_ext(DisasContext *s, uint32_t insn) -{ - gen_helper_gvec_3 *fn_gvec = NULL; - gen_helper_gvec_3_ptr *fn_gvec_ptr = NULL; - int rd, rn, rm, opr_sz, data; - int off_rn, off_rm; - bool is_long = false, q = extract32(insn, 6, 1); - bool ptr_is_env = false; - - if ((insn & 0xffa00f10) == 0xfe000810) { - /* VFM[AS]L -- 1111 1110 0.0S .... .... 1000 .Q.1 .... */ - int is_s = extract32(insn, 20, 1); - int vm20 = extract32(insn, 0, 3); - int vm3 = extract32(insn, 3, 1); - int m = extract32(insn, 5, 1); - int index; - - if (!dc_isar_feature(aa32_fhm, s)) { - return 1; - } - if (q) { - rm = vm20; - index = m * 2 + vm3; - } else { - rm = vm20 * 2 + m; - index = vm3; - } - is_long = true; - data = (index << 2) | is_s; /* is_2 == 0 */ - fn_gvec_ptr = gen_helper_gvec_fmlal_idx_a32; - ptr_is_env = true; - } else { - return 1; - } - - VFP_DREG_D(rd, insn); - if (rd & q) { - return 1; - } - if (q || !is_long) { - VFP_DREG_N(rn, insn); - if (rn & q & !is_long) { - return 1; - } - off_rn = vfp_reg_offset(1, rn); - off_rm = vfp_reg_offset(1, rm); - } else { - rn = VFP_SREG_N(insn); - off_rn = vfp_reg_offset(0, rn); - off_rm = vfp_reg_offset(0, rm); - } - if (s->fp_excp_el) { - gen_exception_insn(s, s->pc_curr, EXCP_UDEF, - syn_simd_access_trap(1, 0xe, false), s->fp_excp_el); - return 0; - } - if (!s->vfp_enabled) { - return 1; - } - - opr_sz = (1 + q) * 8; - if (fn_gvec_ptr) { - TCGv_ptr ptr; - if (ptr_is_env) { - ptr = cpu_env; - } else { - ptr = get_fpstatus_ptr(1); - } - tcg_gen_gvec_3_ptr(vfp_reg_offset(1, rd), off_rn, off_rm, ptr, - opr_sz, opr_sz, data, fn_gvec_ptr); - if (!ptr_is_env) { - tcg_temp_free_ptr(ptr); - } - } else { - tcg_gen_gvec_3_ool(vfp_reg_offset(1, rd), off_rn, off_rm, - opr_sz, opr_sz, data, fn_gvec); - } - return 0; -} - static int disas_coproc_insn(DisasContext *s, uint32_t insn) { int cpnum, is64, crn, crm, opc1, opc2, isread, rt, rt2; @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) } } } - } else if ((insn & 0x0f000a00) == 0x0e000800 - && arm_dc_feature(s, ARM_FEATURE_V8)) { - if (disas_neon_insn_2reg_scalar_ext(s, insn)) { - goto illegal_op; - } - return; } goto illegal_op; } @@ -XXX,XX +XXX,XX @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn) } break; } - if ((insn & 0xff000a00) == 0xfe000800 - && arm_dc_feature(s, ARM_FEATURE_V8)) { - /* The Thumb2 and ARM encodings are identical. */ - if (disas_neon_insn_2reg_scalar_ext(s, insn)) { - goto illegal_op; - } - } else if (((insn >> 24) & 3) == 3) { + if (((insn >> 24) & 3) == 3) { /* Translate into the equivalent ARM encoding. */ insn = (insn & 0xe2ffffff) | ((insn & (1 << 28)) >> 4) | (1 << 28); if (disas_neon_data_insn(s, insn)) { -- 2.20.1 Convert the Neon "load/store multiple structures" insns to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-12-peter.maydell@linaro.org --- target/arm/neon-ls.decode | 7 ++ target/arm/translate-neon.inc.c | 124 ++++++++++++++++++++++++++++++++ target/arm/translate.c | 91 +---------------------- 3 files changed, 133 insertions(+), 89 deletions(-) diff --git a/target/arm/neon-ls.decode b/target/arm/neon-ls.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-ls.decode +++ b/target/arm/neon-ls.decode @@ -XXX,XX +XXX,XX @@ # 0b1111_1001_xxx0_xxxx_xxxx_xxxx_xxxx_xxxx # This file works on the A32 encoding only; calling code for T32 has to # transform the insn into the A32 version first. + +%vd_dp 22:1 12:4 + +# Neon load/store multiple structures + +VLDST_multiple 1111 0100 0 . l:1 0 rn:4 .... itype:4 size:2 align:2 rm:4 \ + vd=%vd_dp diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VFML_scalar(DisasContext *s, arg_VFML_scalar *a) gen_helper_gvec_fmlal_idx_a32); return true; } + +static struct { + int nregs; + int interleave; + int spacing; +} const neon_ls_element_type[11] = { + {1, 4, 1}, + {1, 4, 2}, + {4, 1, 1}, + {2, 2, 2}, + {1, 3, 1}, + {1, 3, 2}, + {3, 1, 1}, + {1, 1, 1}, + {1, 2, 1}, + {1, 2, 2}, + {2, 1, 1} +}; + +static void gen_neon_ldst_base_update(DisasContext *s, int rm, int rn, + int stride) +{ + if (rm != 15) { + TCGv_i32 base; + + base = load_reg(s, rn); + if (rm == 13) { + tcg_gen_addi_i32(base, base, stride); + } else { + TCGv_i32 index; + index = load_reg(s, rm); + tcg_gen_add_i32(base, base, index); + tcg_temp_free_i32(index); + } + store_reg(s, rn, base); + } +} + +static bool trans_VLDST_multiple(DisasContext *s, arg_VLDST_multiple *a) +{ + /* Neon load/store multiple structures */ + int nregs, interleave, spacing, reg, n; + MemOp endian = s->be_data; + int mmu_idx = get_mem_index(s); + int size = a->size; + TCGv_i64 tmp64; + TCGv_i32 addr, tmp; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist */ + if (!dc_isar_feature(aa32_simd_r32, s) && (a->vd & 0x10)) { + return false; + } + if (a->itype > 10) { + return false; + } + /* Catch UNDEF cases for bad values of align field */ + switch (a->itype & 0xc) { + case 4: + if (a->align >= 2) { + return false; + } + break; + case 8: + if (a->align == 3) { + return false; + } + break; + default: + break; + } + nregs = neon_ls_element_type[a->itype].nregs; + interleave = neon_ls_element_type[a->itype].interleave; + spacing = neon_ls_element_type[a->itype].spacing; + if (size == 3 && (interleave | spacing) != 1) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + /* For our purposes, bytes are always little-endian. */ + if (size == 0) { + endian = MO_LE; + } + /* + * Consecutive little-endian elements from a single register + * can be promoted to a larger little-endian operation. + */ + if (interleave == 1 && endian == MO_LE) { + size = 3; + } + tmp64 = tcg_temp_new_i64(); + addr = tcg_temp_new_i32(); + tmp = tcg_const_i32(1 << size); + load_reg_var(s, addr, a->rn); + for (reg = 0; reg < nregs; reg++) { + for (n = 0; n < 8 >> size; n++) { + int xs; + for (xs = 0; xs < interleave; xs++) { + int tt = a->vd + reg + spacing * xs; + + if (a->l) { + gen_aa32_ld_i64(s, tmp64, addr, mmu_idx, endian | size); + neon_store_element64(tt, n, size, tmp64); + } else { + neon_load_element64(tmp64, tt, n, size); + gen_aa32_st_i64(s, tmp64, addr, mmu_idx, endian | size); + } + tcg_gen_add_i32(addr, addr, tmp); + } + } + } + tcg_temp_free_i32(addr); + tcg_temp_free_i32(tmp); + tcg_temp_free_i64(tmp64); + + gen_neon_ldst_base_update(s, a->rm, a->rn, nregs * interleave * 8); + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static void gen_neon_trn_u16(TCGv_i32 t0, TCGv_i32 t1) } -static struct { - int nregs; - int interleave; - int spacing; -} const neon_ls_element_type[11] = { - {1, 4, 1}, - {1, 4, 2}, - {4, 1, 1}, - {2, 2, 2}, - {1, 3, 1}, - {1, 3, 2}, - {3, 1, 1}, - {1, 1, 1}, - {1, 2, 1}, - {1, 2, 2}, - {2, 1, 1} -}; - /* Translate a NEON load/store element instruction. Return nonzero if the instruction is invalid. */ static int disas_neon_ls_insn(DisasContext *s, uint32_t insn) { int rd, rn, rm; - int op; int nregs; - int interleave; - int spacing; int stride; int size; int reg; int load; - int n; int vec_size; - int mmu_idx; - MemOp endian; TCGv_i32 addr; TCGv_i32 tmp; - TCGv_i32 tmp2; - TCGv_i64 tmp64; if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { return 1; @@ -XXX,XX +XXX,XX @@ static int disas_neon_ls_insn(DisasContext *s, uint32_t insn) rn = (insn >> 16) & 0xf; rm = insn & 0xf; load = (insn & (1 << 21)) != 0; - endian = s->be_data; - mmu_idx = get_mem_index(s); if ((insn & (1 << 23)) == 0) { - /* Load store all elements. */ - op = (insn >> 8) & 0xf; - size = (insn >> 6) & 3; - if (op > 10) - return 1; - /* Catch UNDEF cases for bad values of align field */ - switch (op & 0xc) { - case 4: - if (((insn >> 5) & 1) == 1) { - return 1; - } - break; - case 8: - if (((insn >> 4) & 3) == 3) { - return 1; - } - break; - default: - break; - } - nregs = neon_ls_element_type[op].nregs; - interleave = neon_ls_element_type[op].interleave; - spacing = neon_ls_element_type[op].spacing; - if (size == 3 && (interleave | spacing) != 1) { - return 1; - } - /* For our purposes, bytes are always little-endian. */ - if (size == 0) { - endian = MO_LE; - } - /* Consecutive little-endian elements from a single register - * can be promoted to a larger little-endian operation. - */ - if (interleave == 1 && endian == MO_LE) { - size = 3; - } - tmp64 = tcg_temp_new_i64(); - addr = tcg_temp_new_i32(); - tmp2 = tcg_const_i32(1 << size); - load_reg_var(s, addr, rn); - for (reg = 0; reg < nregs; reg++) { - for (n = 0; n < 8 >> size; n++) { - int xs; - for (xs = 0; xs < interleave; xs++) { - int tt = rd + reg + spacing * xs; - - if (load) { - gen_aa32_ld_i64(s, tmp64, addr, mmu_idx, endian | size); - neon_store_element64(tt, n, size, tmp64); - } else { - neon_load_element64(tmp64, tt, n, size); - gen_aa32_st_i64(s, tmp64, addr, mmu_idx, endian | size); - } - tcg_gen_add_i32(addr, addr, tmp2); - } - } - } - tcg_temp_free_i32(addr); - tcg_temp_free_i32(tmp2); - tcg_temp_free_i64(tmp64); - stride = nregs * interleave * 8; + /* Load store all elements -- handled already by decodetree */ + return 1; } else { size = (insn >> 10) & 3; if (size == 3) { -- 2.20.1 Convert the Neon "load single structure to all lanes" insns to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-13-peter.maydell@linaro.org --- target/arm/neon-ls.decode | 5 +++ target/arm/translate-neon.inc.c | 73 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 55 +------------------------ 3 files changed, 80 insertions(+), 53 deletions(-) diff --git a/target/arm/neon-ls.decode b/target/arm/neon-ls.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-ls.decode +++ b/target/arm/neon-ls.decode @@ -XXX,XX +XXX,XX @@ VLDST_multiple 1111 0100 0 . l:1 0 rn:4 .... itype:4 size:2 align:2 rm:4 \ vd=%vd_dp + +# Neon load single element to all lanes + +VLD_all_lanes 1111 0100 1 . 1 0 rn:4 .... 11 n:2 size:2 t:1 a:1 rm:4 \ + vd=%vd_dp diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VLDST_multiple(DisasContext *s, arg_VLDST_multiple *a) gen_neon_ldst_base_update(s, a->rm, a->rn, nregs * interleave * 8); return true; } + +static bool trans_VLD_all_lanes(DisasContext *s, arg_VLD_all_lanes *a) +{ + /* Neon load single structure to all lanes */ + int reg, stride, vec_size; + int vd = a->vd; + int size = a->size; + int nregs = a->n + 1; + TCGv_i32 addr, tmp; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist */ + if (!dc_isar_feature(aa32_simd_r32, s) && (a->vd & 0x10)) { + return false; + } + + if (size == 3) { + if (nregs != 4 || a->a == 0) { + return false; + } + /* For VLD4 size == 3 a == 1 means 32 bits at 16 byte alignment */ + size = 2; + } + if (nregs == 1 && a->a == 1 && size == 0) { + return false; + } + if (nregs == 3 && a->a == 1) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + /* + * VLD1 to all lanes: T bit indicates how many Dregs to write. + * VLD2/3/4 to all lanes: T bit indicates register stride. + */ + stride = a->t ? 2 : 1; + vec_size = nregs == 1 ? stride * 8 : 8; + + tmp = tcg_temp_new_i32(); + addr = tcg_temp_new_i32(); + load_reg_var(s, addr, a->rn); + for (reg = 0; reg < nregs; reg++) { + gen_aa32_ld_i32(s, tmp, addr, get_mem_index(s), + s->be_data | size); + if ((vd & 1) && vec_size == 16) { + /* + * We cannot write 16 bytes at once because the + * destination is unaligned. + */ + tcg_gen_gvec_dup_i32(size, neon_reg_offset(vd, 0), + 8, 8, tmp); + tcg_gen_gvec_mov(0, neon_reg_offset(vd + 1, 0), + neon_reg_offset(vd, 0), 8, 8); + } else { + tcg_gen_gvec_dup_i32(size, neon_reg_offset(vd, 0), + vec_size, vec_size, tmp); + } + tcg_gen_addi_i32(addr, addr, 1 << size); + vd += stride; + } + tcg_temp_free_i32(tmp); + tcg_temp_free_i32(addr); + + gen_neon_ldst_base_update(s, a->rm, a->rn, (1 << size) * nregs); + + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_ls_insn(DisasContext *s, uint32_t insn) int size; int reg; int load; - int vec_size; TCGv_i32 addr; TCGv_i32 tmp; @@ -XXX,XX +XXX,XX @@ static int disas_neon_ls_insn(DisasContext *s, uint32_t insn) } else { size = (insn >> 10) & 3; if (size == 3) { - /* Load single element to all lanes. */ - int a = (insn >> 4) & 1; - if (!load) { - return 1; - } - size = (insn >> 6) & 3; - nregs = ((insn >> 8) & 3) + 1; - - if (size == 3) { - if (nregs != 4 || a == 0) { - return 1; - } - /* For VLD4 size==3 a == 1 means 32 bits at 16 byte alignment */ - size = 2; - } - if (nregs == 1 && a == 1 && size == 0) { - return 1; - } - if (nregs == 3 && a == 1) { - return 1; - } - addr = tcg_temp_new_i32(); - load_reg_var(s, addr, rn); - - /* VLD1 to all lanes: bit 5 indicates how many Dregs to write. - * VLD2/3/4 to all lanes: bit 5 indicates register stride. - */ - stride = (insn & (1 << 5)) ? 2 : 1; - vec_size = nregs == 1 ? stride * 8 : 8; - - tmp = tcg_temp_new_i32(); - for (reg = 0; reg < nregs; reg++) { - gen_aa32_ld_i32(s, tmp, addr, get_mem_index(s), - s->be_data | size); - if ((rd & 1) && vec_size == 16) { - /* We cannot write 16 bytes at once because the - * destination is unaligned. - */ - tcg_gen_gvec_dup_i32(size, neon_reg_offset(rd, 0), - 8, 8, tmp); - tcg_gen_gvec_mov(0, neon_reg_offset(rd + 1, 0), - neon_reg_offset(rd, 0), 8, 8); - } else { - tcg_gen_gvec_dup_i32(size, neon_reg_offset(rd, 0), - vec_size, vec_size, tmp); - } - tcg_gen_addi_i32(addr, addr, 1 << size); - rd += stride; - } - tcg_temp_free_i32(tmp); - tcg_temp_free_i32(addr); - stride = (1 << size) * nregs; + /* Load single element to all lanes -- handled by decodetree */ + return 1; } else { /* Single element. */ int idx = (insn >> 4) & 0xf; -- 2.20.1 Convert the Neon "load/store single structure to one lane" insns to decodetree. As this is the last set of insns in the neon load/store group, we can remove the whole disas_neon_ls_insn() function. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-14-peter.maydell@linaro.org --- target/arm/neon-ls.decode | 11 +++ target/arm/translate-neon.inc.c | 89 +++++++++++++++++++ target/arm/translate.c | 147 -------------------------------- 3 files changed, 100 insertions(+), 147 deletions(-) diff --git a/target/arm/neon-ls.decode b/target/arm/neon-ls.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-ls.decode +++ b/target/arm/neon-ls.decode @@ -XXX,XX +XXX,XX @@ VLDST_multiple 1111 0100 0 . l:1 0 rn:4 .... itype:4 size:2 align:2 rm:4 \ VLD_all_lanes 1111 0100 1 . 1 0 rn:4 .... 11 n:2 size:2 t:1 a:1 rm:4 \ vd=%vd_dp + +# Neon load/store single structure to one lane +%imm1_5_p1 5:1 !function=plus1 +%imm1_6_p1 6:1 !function=plus1 + +VLDST_single 1111 0100 1 . l:1 0 rn:4 .... 00 n:2 reg_idx:3 align:1 rm:4 \ + vd=%vd_dp size=0 stride=1 +VLDST_single 1111 0100 1 . l:1 0 rn:4 .... 01 n:2 reg_idx:2 align:2 rm:4 \ + vd=%vd_dp size=1 stride=%imm1_5_p1 +VLDST_single 1111 0100 1 . l:1 0 rn:4 .... 10 n:2 reg_idx:1 align:3 rm:4 \ + vd=%vd_dp size=2 stride=%imm1_6_p1 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ * It might be possible to convert it to a standalone .c file eventually. */ +static inline int plus1(DisasContext *s, int x) +{ + return x + 1; +} + /* Include the generated Neon decoder */ #include "decode-neon-dp.inc.c" #include "decode-neon-ls.inc.c" @@ -XXX,XX +XXX,XX @@ static bool trans_VLD_all_lanes(DisasContext *s, arg_VLD_all_lanes *a) return true; } + +static bool trans_VLDST_single(DisasContext *s, arg_VLDST_single *a) +{ + /* Neon load/store single structure to one lane */ + int reg; + int nregs = a->n + 1; + int vd = a->vd; + TCGv_i32 addr, tmp; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist */ + if (!dc_isar_feature(aa32_simd_r32, s) && (a->vd & 0x10)) { + return false; + } + + /* Catch the UNDEF cases. This is unavoidably a bit messy. */ + switch (nregs) { + case 1: + if (((a->align & (1 << a->size)) != 0) || + (a->size == 2 && ((a->align & 3) == 1 || (a->align & 3) == 2))) { + return false; + } + break; + case 3: + if ((a->align & 1) != 0) { + return false; + } + /* fall through */ + case 2: + if (a->size == 2 && (a->align & 2) != 0) { + return false; + } + break; + case 4: + if ((a->size == 2) && ((a->align & 3) == 3)) { + return false; + } + break; + default: + abort(); + } + if ((vd + a->stride * (nregs - 1)) > 31) { + /* + * Attempts to write off the end of the register file are + * UNPREDICTABLE; we choose to UNDEF because otherwise we would + * access off the end of the array that holds the register data. + */ + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + tmp = tcg_temp_new_i32(); + addr = tcg_temp_new_i32(); + load_reg_var(s, addr, a->rn); + /* + * TODO: if we implemented alignment exceptions, we should check + * addr against the alignment encoded in a->align here. + */ + for (reg = 0; reg < nregs; reg++) { + if (a->l) { + gen_aa32_ld_i32(s, tmp, addr, get_mem_index(s), + s->be_data | a->size); + neon_store_element(vd, a->reg_idx, a->size, tmp); + } else { /* Store */ + neon_load_element(tmp, vd, a->reg_idx, a->size); + gen_aa32_st_i32(s, tmp, addr, get_mem_index(s), + s->be_data | a->size); + } + vd += a->stride; + tcg_gen_addi_i32(addr, addr, 1 << a->size); + } + tcg_temp_free_i32(addr); + tcg_temp_free_i32(tmp); + + gen_neon_ldst_base_update(s, a->rm, a->rn, (1 << a->size) * nregs); + + return true; +} diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static void gen_neon_trn_u16(TCGv_i32 t0, TCGv_i32 t1) tcg_temp_free_i32(rd); } - -/* Translate a NEON load/store element instruction. Return nonzero if the - instruction is invalid. */ -static int disas_neon_ls_insn(DisasContext *s, uint32_t insn) -{ - int rd, rn, rm; - int nregs; - int stride; - int size; - int reg; - int load; - TCGv_i32 addr; - TCGv_i32 tmp; - - if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { - return 1; - } - - /* FIXME: this access check should not take precedence over UNDEF - * for invalid encodings; we will generate incorrect syndrome information - * for attempts to execute invalid vfp/neon encodings with FP disabled. - */ - if (s->fp_excp_el) { - gen_exception_insn(s, s->pc_curr, EXCP_UDEF, - syn_simd_access_trap(1, 0xe, false), s->fp_excp_el); - return 0; - } - - if (!s->vfp_enabled) - return 1; - VFP_DREG_D(rd, insn); - rn = (insn >> 16) & 0xf; - rm = insn & 0xf; - load = (insn & (1 << 21)) != 0; - if ((insn & (1 << 23)) == 0) { - /* Load store all elements -- handled already by decodetree */ - return 1; - } else { - size = (insn >> 10) & 3; - if (size == 3) { - /* Load single element to all lanes -- handled by decodetree */ - return 1; - } else { - /* Single element. */ - int idx = (insn >> 4) & 0xf; - int reg_idx; - switch (size) { - case 0: - reg_idx = (insn >> 5) & 7; - stride = 1; - break; - case 1: - reg_idx = (insn >> 6) & 3; - stride = (insn & (1 << 5)) ? 2 : 1; - break; - case 2: - reg_idx = (insn >> 7) & 1; - stride = (insn & (1 << 6)) ? 2 : 1; - break; - default: - abort(); - } - nregs = ((insn >> 8) & 3) + 1; - /* Catch the UNDEF cases. This is unavoidably a bit messy. */ - switch (nregs) { - case 1: - if (((idx & (1 << size)) != 0) || - (size == 2 && ((idx & 3) == 1 || (idx & 3) == 2))) { - return 1; - } - break; - case 3: - if ((idx & 1) != 0) { - return 1; - } - /* fall through */ - case 2: - if (size == 2 && (idx & 2) != 0) { - return 1; - } - break; - case 4: - if ((size == 2) && ((idx & 3) == 3)) { - return 1; - } - break; - default: - abort(); - } - if ((rd + stride * (nregs - 1)) > 31) { - /* Attempts to write off the end of the register file - * are UNPREDICTABLE; we choose to UNDEF because otherwise - * the neon_load_reg() would write off the end of the array. - */ - return 1; - } - tmp = tcg_temp_new_i32(); - addr = tcg_temp_new_i32(); - load_reg_var(s, addr, rn); - for (reg = 0; reg < nregs; reg++) { - if (load) { - gen_aa32_ld_i32(s, tmp, addr, get_mem_index(s), - s->be_data | size); - neon_store_element(rd, reg_idx, size, tmp); - } else { /* Store */ - neon_load_element(tmp, rd, reg_idx, size); - gen_aa32_st_i32(s, tmp, addr, get_mem_index(s), - s->be_data | size); - } - rd += stride; - tcg_gen_addi_i32(addr, addr, 1 << size); - } - tcg_temp_free_i32(addr); - tcg_temp_free_i32(tmp); - stride = nregs * (1 << size); - } - } - if (rm != 15) { - TCGv_i32 base; - - base = load_reg(s, rn); - if (rm == 13) { - tcg_gen_addi_i32(base, base, stride); - } else { - TCGv_i32 index; - index = load_reg(s, rm); - tcg_gen_add_i32(base, base, index); - tcg_temp_free_i32(index); - } - store_reg(s, rn, base); - } - return 0; -} - static inline void gen_neon_narrow(int size, TCGv_i32 dest, TCGv_i64 src) { switch (size) { @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) } return; } - if ((insn & 0x0f100000) == 0x04000000) { - /* NEON load/store. */ - if (disas_neon_ls_insn(s, insn)) { - goto illegal_op; - } - return; - } if ((insn & 0x0e000f00) == 0x0c000100) { if (arm_dc_feature(s, ARM_FEATURE_IWMMXT)) { /* iWMMXt register transfer. */ @@ -XXX,XX +XXX,XX @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn) } break; case 12: - if ((insn & 0x01100000) == 0x01000000) { - if (disas_neon_ls_insn(s, insn)) { - goto illegal_op; - } - break; - } goto illegal_op; default: illegal_op: -- 2.20.1 Convert the Neon 3-reg-same VADD and VSUB insns to decodetree. Note that we don't need the neon_3r_sizes[op] check here because all size values are OK for VADD and VSUB; we'll add this when we convert the first insn that has size restrictions. For this we need one of the GVecGen*Fn typedefs currently in translate-a64.h; move them all to translate.h as a block so they are visible to the 32-bit decoder. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-15-peter.maydell@linaro.org --- target/arm/translate-a64.h | 9 -------- target/arm/translate.h | 9 ++++++++ target/arm/neon-dp.decode | 17 +++++++++++++++ target/arm/translate-neon.inc.c | 38 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 14 ++++-------- 5 files changed, 68 insertions(+), 19 deletions(-) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ static inline int vec_full_reg_size(DisasContext *s) bool disas_sve(DisasContext *, uint32_t); -/* Note that the gvec expanders operate on offsets + sizes. */ -typedef void GVecGen2Fn(unsigned, uint32_t, uint32_t, uint32_t, uint32_t); -typedef void GVecGen2iFn(unsigned, uint32_t, uint32_t, int64_t, - uint32_t, uint32_t); -typedef void GVecGen3Fn(unsigned, uint32_t, uint32_t, - uint32_t, uint32_t, uint32_t); -typedef void GVecGen4Fn(unsigned, uint32_t, uint32_t, uint32_t, - uint32_t, uint32_t, uint32_t); - #endif /* TARGET_ARM_TRANSLATE_A64_H */ diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ void gen_sshl_i64(TCGv_i64 d, TCGv_i64 a, TCGv_i64 b); #define dc_isar_feature(name, ctx) \ ({ DisasContext *ctx_ = (ctx); isar_feature_##name(ctx_->isar); }) +/* Note that the gvec expanders operate on offsets + sizes. */ +typedef void GVecGen2Fn(unsigned, uint32_t, uint32_t, uint32_t, uint32_t); +typedef void GVecGen2iFn(unsigned, uint32_t, uint32_t, int64_t, + uint32_t, uint32_t); +typedef void GVecGen3Fn(unsigned, uint32_t, uint32_t, + uint32_t, uint32_t, uint32_t); +typedef void GVecGen4Fn(unsigned, uint32_t, uint32_t, uint32_t, + uint32_t, uint32_t, uint32_t); + #endif /* TARGET_ARM_TRANSLATE_H */ diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ # # This file is processed by scripts/decodetree.py # +# VFP/Neon register fields; same as vfp.decode +%vm_dp 5:1 0:4 +%vn_dp 7:1 16:4 +%vd_dp 22:1 12:4 # Encodings for Neon data processing instructions where the T32 encoding # is a simple transformation of the A32 encoding. @@ -XXX,XX +XXX,XX @@ # 0b111p_1111_qqqq_qqqq_qqqq_qqqq_qqqq_qqqq # This file works on the A32 encoding only; calling code for T32 has to # transform the insn into the A32 version first. + +###################################################################### +# 3-reg-same grouping: +# 1111 001 U 0 D sz:2 Vn:4 Vd:4 opc:4 N Q M op Vm:4 +###################################################################### + +&3same vm vn vd q size + +@3same .... ... . . . size:2 .... .... .... . q:1 . . .... \ + &3same vm=%vm_dp vn=%vn_dp vd=%vd_dp + +VADD_3s 1111 001 0 0 . .. .... .... 1000 . . . 0 .... @3same +VSUB_3s 1111 001 1 0 . .. .... .... 1000 . . . 0 .... @3same diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VLDST_single(DisasContext *s, arg_VLDST_single *a) return true; } + +static bool do_3same(DisasContext *s, arg_3same *a, GVecGen3Fn fn) +{ + int vec_size = a->q ? 16 : 8; + int rd_ofs = neon_reg_offset(a->vd, 0); + int rn_ofs = neon_reg_offset(a->vn, 0); + int rm_ofs = neon_reg_offset(a->vm, 0); + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vn | a->vm) & 0x10)) { + return false; + } + + if ((a->vn | a->vm | a->vd) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + fn(a->size, rd_ofs, rn_ofs, rm_ofs, vec_size, vec_size); + return true; +} + +#define DO_3SAME(INSN, FUNC) \ + static bool trans_##INSN##_3s(DisasContext *s, arg_3same *a) \ + { \ + return do_3same(s, a, FUNC); \ + } + +DO_3SAME(VADD, tcg_gen_gvec_add) +DO_3SAME(VSUB, tcg_gen_gvec_sub) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) } return 0; - case NEON_3R_VADD_VSUB: - if (u) { - tcg_gen_gvec_sub(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } else { - tcg_gen_gvec_add(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } - return 0; - case NEON_3R_VQADD: tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc), rn_ofs, rm_ofs, vec_size, vec_size, @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) tcg_gen_gvec_3(rd_ofs, rm_ofs, rn_ofs, vec_size, vec_size, u ? &ushl_op[size] : &sshl_op[size]); return 0; + + case NEON_3R_VADD_VSUB: + /* Already handled by decodetree */ + return 1; } if (size == 3) { -- 2.20.1 Convert the Neon logic ops in the 3-reg-same grouping to decodetree. Note that for the logic ops the 'size' field forms part of their decode and the actual operations are always bitwise. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-16-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 12 +++++++++++ target/arm/translate-neon.inc.c | 19 +++++++++++++++++ target/arm/translate.c | 38 +-------------------------------- 3 files changed, 32 insertions(+), 37 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ @3same .... ... . . . size:2 .... .... .... . q:1 . . .... \ &3same vm=%vm_dp vn=%vn_dp vd=%vd_dp +@3same_logic .... ... . . . .. .... .... .... . q:1 .. .... \ + &3same vm=%vm_dp vn=%vn_dp vd=%vd_dp size=0 + +VAND_3s 1111 001 0 0 . 00 .... .... 0001 ... 1 .... @3same_logic +VBIC_3s 1111 001 0 0 . 01 .... .... 0001 ... 1 .... @3same_logic +VORR_3s 1111 001 0 0 . 10 .... .... 0001 ... 1 .... @3same_logic +VORN_3s 1111 001 0 0 . 11 .... .... 0001 ... 1 .... @3same_logic +VEOR_3s 1111 001 1 0 . 00 .... .... 0001 ... 1 .... @3same_logic +VBSL_3s 1111 001 1 0 . 01 .... .... 0001 ... 1 .... @3same_logic +VBIT_3s 1111 001 1 0 . 10 .... .... 0001 ... 1 .... @3same_logic +VBIF_3s 1111 001 1 0 . 11 .... .... 0001 ... 1 .... @3same_logic + VADD_3s 1111 001 0 0 . .. .... .... 1000 . . . 0 .... @3same VSUB_3s 1111 001 1 0 . .. .... .... 1000 . . . 0 .... @3same diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool do_3same(DisasContext *s, arg_3same *a, GVecGen3Fn fn) DO_3SAME(VADD, tcg_gen_gvec_add) DO_3SAME(VSUB, tcg_gen_gvec_sub) +DO_3SAME(VAND, tcg_gen_gvec_and) +DO_3SAME(VBIC, tcg_gen_gvec_andc) +DO_3SAME(VORR, tcg_gen_gvec_or) +DO_3SAME(VORN, tcg_gen_gvec_orc) +DO_3SAME(VEOR, tcg_gen_gvec_xor) + +/* These insns are all gvec_bitsel but with the inputs in various orders. */ +#define DO_3SAME_BITSEL(INSN, O1, O2, O3) \ + static void gen_##INSN##_3s(unsigned vece, uint32_t rd_ofs, \ + uint32_t rn_ofs, uint32_t rm_ofs, \ + uint32_t oprsz, uint32_t maxsz) \ + { \ + tcg_gen_gvec_bitsel(vece, rd_ofs, O1, O2, O3, oprsz, maxsz); \ + } \ + DO_3SAME(INSN, gen_##INSN##_3s) + +DO_3SAME_BITSEL(VBSL, rd_ofs, rn_ofs, rm_ofs) +DO_3SAME_BITSEL(VBIT, rm_ofs, rn_ofs, rd_ofs) +DO_3SAME_BITSEL(VBIF, rm_ofs, rd_ofs, rn_ofs) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) } return 1; - case NEON_3R_LOGIC: /* Logic ops. */ - switch ((u << 2) | size) { - case 0: /* VAND */ - tcg_gen_gvec_and(0, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - break; - case 1: /* VBIC */ - tcg_gen_gvec_andc(0, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - break; - case 2: /* VORR */ - tcg_gen_gvec_or(0, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - break; - case 3: /* VORN */ - tcg_gen_gvec_orc(0, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - break; - case 4: /* VEOR */ - tcg_gen_gvec_xor(0, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - break; - case 5: /* VBSL */ - tcg_gen_gvec_bitsel(MO_8, rd_ofs, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - break; - case 6: /* VBIT */ - tcg_gen_gvec_bitsel(MO_8, rd_ofs, rm_ofs, rn_ofs, rd_ofs, - vec_size, vec_size); - break; - case 7: /* VBIF */ - tcg_gen_gvec_bitsel(MO_8, rd_ofs, rm_ofs, rd_ofs, rn_ofs, - vec_size, vec_size); - break; - } - return 0; - case NEON_3R_VQADD: tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc), rn_ofs, rm_ofs, vec_size, vec_size, @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) return 0; case NEON_3R_VADD_VSUB: + case NEON_3R_LOGIC: /* Already handled by decodetree */ return 1; } -- 2.20.1 Convert the Neon 3-reg-same VMAX and VMIN insns to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-17-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 5 +++++ target/arm/translate-neon.inc.c | 14 ++++++++++++++ target/arm/translate.c | 21 ++------------------- 3 files changed, 21 insertions(+), 19 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ VBSL_3s 1111 001 1 0 . 01 .... .... 0001 ... 1 .... @3same_logic VBIT_3s 1111 001 1 0 . 10 .... .... 0001 ... 1 .... @3same_logic VBIF_3s 1111 001 1 0 . 11 .... .... 0001 ... 1 .... @3same_logic +VMAX_S_3s 1111 001 0 0 . .. .... .... 0110 . . . 0 .... @3same +VMAX_U_3s 1111 001 1 0 . .. .... .... 0110 . . . 0 .... @3same +VMIN_S_3s 1111 001 0 0 . .. .... .... 0110 . . . 1 .... @3same +VMIN_U_3s 1111 001 1 0 . .. .... .... 0110 . . . 1 .... @3same + VADD_3s 1111 001 0 0 . .. .... .... 1000 . . . 0 .... @3same VSUB_3s 1111 001 1 0 . .. .... .... 1000 . . . 0 .... @3same diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ DO_3SAME(VEOR, tcg_gen_gvec_xor) DO_3SAME_BITSEL(VBSL, rd_ofs, rn_ofs, rm_ofs) DO_3SAME_BITSEL(VBIT, rm_ofs, rn_ofs, rd_ofs) DO_3SAME_BITSEL(VBIF, rm_ofs, rd_ofs, rn_ofs) + +#define DO_3SAME_NO_SZ_3(INSN, FUNC) \ + static bool trans_##INSN##_3s(DisasContext *s, arg_3same *a) \ + { \ + if (a->size == 3) { \ + return false; \ + } \ + return do_3same(s, a, FUNC); \ + } + +DO_3SAME_NO_SZ_3(VMAX_S, tcg_gen_gvec_smax) +DO_3SAME_NO_SZ_3(VMAX_U, tcg_gen_gvec_umax) +DO_3SAME_NO_SZ_3(VMIN_S, tcg_gen_gvec_smin) +DO_3SAME_NO_SZ_3(VMIN_U, tcg_gen_gvec_umin) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) rd_ofs, rn_ofs, rm_ofs, vec_size, vec_size); return 0; - case NEON_3R_VMAX: - if (u) { - tcg_gen_gvec_umax(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } else { - tcg_gen_gvec_smax(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } - return 0; - case NEON_3R_VMIN: - if (u) { - tcg_gen_gvec_umin(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } else { - tcg_gen_gvec_smin(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } - return 0; - case NEON_3R_VSHL: /* Note the operation is vshl vd,vm,vn */ tcg_gen_gvec_3(rd_ofs, rm_ofs, rn_ofs, vec_size, vec_size, @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_3R_VADD_VSUB: case NEON_3R_LOGIC: + case NEON_3R_VMAX: + case NEON_3R_VMIN: /* Already handled by decodetree */ return 1; } -- 2.20.1 Convert the Neon comparison ops in the 3-reg-same grouping to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-18-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 8 ++++++++ target/arm/translate-neon.inc.c | 22 ++++++++++++++++++++++ target/arm/translate.c | 23 +++-------------------- 3 files changed, 33 insertions(+), 20 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ VBSL_3s 1111 001 1 0 . 01 .... .... 0001 ... 1 .... @3same_logic VBIT_3s 1111 001 1 0 . 10 .... .... 0001 ... 1 .... @3same_logic VBIF_3s 1111 001 1 0 . 11 .... .... 0001 ... 1 .... @3same_logic +VCGT_S_3s 1111 001 0 0 . .. .... .... 0011 . . . 0 .... @3same +VCGT_U_3s 1111 001 1 0 . .. .... .... 0011 . . . 0 .... @3same +VCGE_S_3s 1111 001 0 0 . .. .... .... 0011 . . . 1 .... @3same +VCGE_U_3s 1111 001 1 0 . .. .... .... 0011 . . . 1 .... @3same + VMAX_S_3s 1111 001 0 0 . .. .... .... 0110 . . . 0 .... @3same VMAX_U_3s 1111 001 1 0 . .. .... .... 0110 . . . 0 .... @3same VMIN_S_3s 1111 001 0 0 . .. .... .... 0110 . . . 1 .... @3same @@ -XXX,XX +XXX,XX @@ VMIN_U_3s 1111 001 1 0 . .. .... .... 0110 . . . 1 .... @3same VADD_3s 1111 001 0 0 . .. .... .... 1000 . . . 0 .... @3same VSUB_3s 1111 001 1 0 . .. .... .... 1000 . . . 0 .... @3same + +VTST_3s 1111 001 0 0 . .. .... .... 1000 . . . 1 .... @3same +VCEQ_3s 1111 001 1 0 . .. .... .... 1000 . . . 1 .... @3same diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ DO_3SAME_NO_SZ_3(VMAX_S, tcg_gen_gvec_smax) DO_3SAME_NO_SZ_3(VMAX_U, tcg_gen_gvec_umax) DO_3SAME_NO_SZ_3(VMIN_S, tcg_gen_gvec_smin) DO_3SAME_NO_SZ_3(VMIN_U, tcg_gen_gvec_umin) + +#define DO_3SAME_CMP(INSN, COND) \ + static void gen_##INSN##_3s(unsigned vece, uint32_t rd_ofs, \ + uint32_t rn_ofs, uint32_t rm_ofs, \ + uint32_t oprsz, uint32_t maxsz) \ + { \ + tcg_gen_gvec_cmp(COND, vece, rd_ofs, rn_ofs, rm_ofs, oprsz, maxsz); \ + } \ + DO_3SAME_NO_SZ_3(INSN, gen_##INSN##_3s) + +DO_3SAME_CMP(VCGT_S, TCG_COND_GT) +DO_3SAME_CMP(VCGT_U, TCG_COND_GTU) +DO_3SAME_CMP(VCGE_S, TCG_COND_GE) +DO_3SAME_CMP(VCGE_U, TCG_COND_GEU) +DO_3SAME_CMP(VCEQ, TCG_COND_EQ) + +static void gen_VTST_3s(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, + uint32_t rm_ofs, uint32_t oprsz, uint32_t maxsz) +{ + tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, oprsz, maxsz, &cmtst_op[vece]); +} +DO_3SAME_NO_SZ_3(VTST, gen_VTST_3s) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) u ? &mls_op[size] : &mla_op[size]); return 0; - case NEON_3R_VTST_VCEQ: - if (u) { /* VCEQ */ - tcg_gen_gvec_cmp(TCG_COND_EQ, size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } else { /* VTST */ - tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size, &cmtst_op[size]); - } - return 0; - - case NEON_3R_VCGT: - tcg_gen_gvec_cmp(u ? TCG_COND_GTU : TCG_COND_GT, size, - rd_ofs, rn_ofs, rm_ofs, vec_size, vec_size); - return 0; - - case NEON_3R_VCGE: - tcg_gen_gvec_cmp(u ? TCG_COND_GEU : TCG_COND_GE, size, - rd_ofs, rn_ofs, rm_ofs, vec_size, vec_size); - return 0; - case NEON_3R_VSHL: /* Note the operation is vshl vd,vm,vn */ tcg_gen_gvec_3(rd_ofs, rm_ofs, rn_ofs, vec_size, vec_size, @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_3R_LOGIC: case NEON_3R_VMAX: case NEON_3R_VMIN: + case NEON_3R_VTST_VCEQ: + case NEON_3R_VCGT: + case NEON_3R_VCGE: /* Already handled by decodetree */ return 1; } -- 2.20.1 Convert the Neon VQADD/VQSUB insns in the 3-reg-same grouping to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-19-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 6 ++++++ target/arm/translate-neon.inc.c | 15 +++++++++++++++ target/arm/translate.c | 14 ++------------ 3 files changed, 23 insertions(+), 12 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ @3same .... ... . . . size:2 .... .... .... . q:1 . . .... \ &3same vm=%vm_dp vn=%vn_dp vd=%vd_dp +VQADD_S_3s 1111 001 0 0 . .. .... .... 0000 . . . 1 .... @3same +VQADD_U_3s 1111 001 1 0 . .. .... .... 0000 . . . 1 .... @3same + @3same_logic .... ... . . . .. .... .... .... . q:1 .. .... \ &3same vm=%vm_dp vn=%vn_dp vd=%vd_dp size=0 @@ -XXX,XX +XXX,XX @@ VBSL_3s 1111 001 1 0 . 01 .... .... 0001 ... 1 .... @3same_logic VBIT_3s 1111 001 1 0 . 10 .... .... 0001 ... 1 .... @3same_logic VBIF_3s 1111 001 1 0 . 11 .... .... 0001 ... 1 .... @3same_logic +VQSUB_S_3s 1111 001 0 0 . .. .... .... 0010 . . . 1 .... @3same +VQSUB_U_3s 1111 001 1 0 . .. .... .... 0010 . . . 1 .... @3same + VCGT_S_3s 1111 001 0 0 . .. .... .... 0011 . . . 0 .... @3same VCGT_U_3s 1111 001 1 0 . .. .... .... 0011 . . . 0 .... @3same VCGE_S_3s 1111 001 0 0 . .. .... .... 0011 . . . 1 .... @3same diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static void gen_VTST_3s(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, oprsz, maxsz, &cmtst_op[vece]); } DO_3SAME_NO_SZ_3(VTST, gen_VTST_3s) + +#define DO_3SAME_GVEC4(INSN, OPARRAY) \ + static void gen_##INSN##_3s(unsigned vece, uint32_t rd_ofs, \ + uint32_t rn_ofs, uint32_t rm_ofs, \ + uint32_t oprsz, uint32_t maxsz) \ + { \ + tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc), \ + rn_ofs, rm_ofs, oprsz, maxsz, &OPARRAY[vece]); \ + } \ + DO_3SAME(INSN, gen_##INSN##_3s) + +DO_3SAME_GVEC4(VQADD_S, sqadd_op) +DO_3SAME_GVEC4(VQADD_U, uqadd_op) +DO_3SAME_GVEC4(VQSUB_S, sqsub_op) +DO_3SAME_GVEC4(VQSUB_U, uqsub_op) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) } return 1; - case NEON_3R_VQADD: - tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc), - rn_ofs, rm_ofs, vec_size, vec_size, - (u ? uqadd_op : sqadd_op) + size); - return 0; - - case NEON_3R_VQSUB: - tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc), - rn_ofs, rm_ofs, vec_size, vec_size, - (u ? uqsub_op : sqsub_op) + size); - return 0; - case NEON_3R_VMUL: /* VMUL */ if (u) { /* Polynomial case allows only P8. */ @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_3R_VTST_VCEQ: case NEON_3R_VCGT: case NEON_3R_VCGE: + case NEON_3R_VQADD: + case NEON_3R_VQSUB: /* Already handled by decodetree */ return 1; } -- 2.20.1 Convert the Neon VMUL, VMLA, VMLS and VSHL insns in the 3-reg-same grouping to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-20-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 9 +++++++ target/arm/translate-neon.inc.c | 44 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 28 +++------------------ 3 files changed, 56 insertions(+), 25 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ VCGT_U_3s 1111 001 1 0 . .. .... .... 0011 . . . 0 .... @3same VCGE_S_3s 1111 001 0 0 . .. .... .... 0011 . . . 1 .... @3same VCGE_U_3s 1111 001 1 0 . .. .... .... 0011 . . . 1 .... @3same +VSHL_S_3s 1111 001 0 0 . .. .... .... 0100 . . . 0 .... @3same +VSHL_U_3s 1111 001 1 0 . .. .... .... 0100 . . . 0 .... @3same + VMAX_S_3s 1111 001 0 0 . .. .... .... 0110 . . . 0 .... @3same VMAX_U_3s 1111 001 1 0 . .. .... .... 0110 . . . 0 .... @3same VMIN_S_3s 1111 001 0 0 . .. .... .... 0110 . . . 1 .... @3same @@ -XXX,XX +XXX,XX @@ VSUB_3s 1111 001 1 0 . .. .... .... 1000 . . . 0 .... @3same VTST_3s 1111 001 0 0 . .. .... .... 1000 . . . 1 .... @3same VCEQ_3s 1111 001 1 0 . .. .... .... 1000 . . . 1 .... @3same + +VMLA_3s 1111 001 0 0 . .. .... .... 1001 . . . 0 .... @3same +VMLS_3s 1111 001 1 0 . .. .... .... 1001 . . . 0 .... @3same + +VMUL_3s 1111 001 0 0 . .. .... .... 1001 . . . 1 .... @3same +VMUL_p_3s 1111 001 1 0 . .. .... .... 1001 . . . 1 .... @3same diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ DO_3SAME_NO_SZ_3(VMAX_S, tcg_gen_gvec_smax) DO_3SAME_NO_SZ_3(VMAX_U, tcg_gen_gvec_umax) DO_3SAME_NO_SZ_3(VMIN_S, tcg_gen_gvec_smin) DO_3SAME_NO_SZ_3(VMIN_U, tcg_gen_gvec_umin) +DO_3SAME_NO_SZ_3(VMUL, tcg_gen_gvec_mul) #define DO_3SAME_CMP(INSN, COND) \ static void gen_##INSN##_3s(unsigned vece, uint32_t rd_ofs, \ @@ -XXX,XX +XXX,XX @@ DO_3SAME_GVEC4(VQADD_S, sqadd_op) DO_3SAME_GVEC4(VQADD_U, uqadd_op) DO_3SAME_GVEC4(VQSUB_S, sqsub_op) DO_3SAME_GVEC4(VQSUB_U, uqsub_op) + +static void gen_VMUL_p_3s(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, + uint32_t rm_ofs, uint32_t oprsz, uint32_t maxsz) +{ + tcg_gen_gvec_3_ool(rd_ofs, rn_ofs, rm_ofs, oprsz, maxsz, + 0, gen_helper_gvec_pmul_b); +} + +static bool trans_VMUL_p_3s(DisasContext *s, arg_3same *a) +{ + if (a->size != 0) { + return false; + } + return do_3same(s, a, gen_VMUL_p_3s); +} + +#define DO_3SAME_GVEC3_NO_SZ_3(INSN, OPARRAY) \ + static void gen_##INSN##_3s(unsigned vece, uint32_t rd_ofs, \ + uint32_t rn_ofs, uint32_t rm_ofs, \ + uint32_t oprsz, uint32_t maxsz) \ + { \ + tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, \ + oprsz, maxsz, &OPARRAY[vece]); \ + } \ + DO_3SAME_NO_SZ_3(INSN, gen_##INSN##_3s) + + +DO_3SAME_GVEC3_NO_SZ_3(VMLA, mla_op) +DO_3SAME_GVEC3_NO_SZ_3(VMLS, mls_op) + +#define DO_3SAME_GVEC3_SHIFT(INSN, OPARRAY) \ + static void gen_##INSN##_3s(unsigned vece, uint32_t rd_ofs, \ + uint32_t rn_ofs, uint32_t rm_ofs, \ + uint32_t oprsz, uint32_t maxsz) \ + { \ + /* Note the operation is vshl vd,vm,vn */ \ + tcg_gen_gvec_3(rd_ofs, rm_ofs, rn_ofs, \ + oprsz, maxsz, &OPARRAY[vece]); \ + } \ + DO_3SAME(INSN, gen_##INSN##_3s) + +DO_3SAME_GVEC3_SHIFT(VSHL_S, sshl_op) +DO_3SAME_GVEC3_SHIFT(VSHL_U, ushl_op) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) } return 1; - case NEON_3R_VMUL: /* VMUL */ - if (u) { - /* Polynomial case allows only P8. */ - if (size != 0) { - return 1; - } - tcg_gen_gvec_3_ool(rd_ofs, rn_ofs, rm_ofs, vec_size, vec_size, - 0, gen_helper_gvec_pmul_b); - } else { - tcg_gen_gvec_mul(size, rd_ofs, rn_ofs, rm_ofs, - vec_size, vec_size); - } - return 0; - - case NEON_3R_VML: /* VMLA, VMLS */ - tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, vec_size, vec_size, - u ? &mls_op[size] : &mla_op[size]); - return 0; - - case NEON_3R_VSHL: - /* Note the operation is vshl vd,vm,vn */ - tcg_gen_gvec_3(rd_ofs, rm_ofs, rn_ofs, vec_size, vec_size, - u ? &ushl_op[size] : &sshl_op[size]); - return 0; - case NEON_3R_VADD_VSUB: case NEON_3R_LOGIC: case NEON_3R_VMAX: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_3R_VCGE: case NEON_3R_VQADD: case NEON_3R_VQSUB: + case NEON_3R_VMUL: + case NEON_3R_VML: + case NEON_3R_VSHL: /* Already handled by decodetree */ return 1; } -- 2.20.1 We're going to want at least some of the NeonGen* typedefs for the refactored 32-bit Neon decoder, so move them all to translate.h since it makes more sense to keep them in one group. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200430181003.21682-23-peter.maydell@linaro.org --- target/arm/translate.h | 17 +++++++++++++++++ target/arm/translate-a64.c | 17 ----------------- 2 files changed, 17 insertions(+), 17 deletions(-) diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ typedef void GVecGen3Fn(unsigned, uint32_t, uint32_t, typedef void GVecGen4Fn(unsigned, uint32_t, uint32_t, uint32_t, uint32_t, uint32_t, uint32_t); +/* Function prototype for gen_ functions for calling Neon helpers */ +typedef void NeonGenOneOpEnvFn(TCGv_i32, TCGv_ptr, TCGv_i32); +typedef void NeonGenTwoOpFn(TCGv_i32, TCGv_i32, TCGv_i32); +typedef void NeonGenTwoOpEnvFn(TCGv_i32, TCGv_ptr, TCGv_i32, TCGv_i32); +typedef void NeonGenTwo64OpFn(TCGv_i64, TCGv_i64, TCGv_i64); +typedef void NeonGenTwo64OpEnvFn(TCGv_i64, TCGv_ptr, TCGv_i64, TCGv_i64); +typedef void NeonGenNarrowFn(TCGv_i32, TCGv_i64); +typedef void NeonGenNarrowEnvFn(TCGv_i32, TCGv_ptr, TCGv_i64); +typedef void NeonGenWidenFn(TCGv_i64, TCGv_i32); +typedef void NeonGenTwoSingleOPFn(TCGv_i32, TCGv_i32, TCGv_i32, TCGv_ptr); +typedef void NeonGenTwoDoubleOPFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGv_ptr); +typedef void NeonGenOneOpFn(TCGv_i64, TCGv_i64); +typedef void CryptoTwoOpFn(TCGv_ptr, TCGv_ptr); +typedef void CryptoThreeOpIntFn(TCGv_ptr, TCGv_ptr, TCGv_i32); +typedef void CryptoThreeOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr); +typedef void AtomicThreeOpFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGArg, MemOp); + #endif /* TARGET_ARM_TRANSLATE_H */ diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ typedef struct AArch64DecodeTable { AArch64DecodeFn *disas_fn; } AArch64DecodeTable; -/* Function prototype for gen_ functions for calling Neon helpers */ -typedef void NeonGenOneOpEnvFn(TCGv_i32, TCGv_ptr, TCGv_i32); -typedef void NeonGenTwoOpFn(TCGv_i32, TCGv_i32, TCGv_i32); -typedef void NeonGenTwoOpEnvFn(TCGv_i32, TCGv_ptr, TCGv_i32, TCGv_i32); -typedef void NeonGenTwo64OpFn(TCGv_i64, TCGv_i64, TCGv_i64); -typedef void NeonGenTwo64OpEnvFn(TCGv_i64, TCGv_ptr, TCGv_i64, TCGv_i64); -typedef void NeonGenNarrowFn(TCGv_i32, TCGv_i64); -typedef void NeonGenNarrowEnvFn(TCGv_i32, TCGv_ptr, TCGv_i64); -typedef void NeonGenWidenFn(TCGv_i64, TCGv_i32); -typedef void NeonGenTwoSingleOPFn(TCGv_i32, TCGv_i32, TCGv_i32, TCGv_ptr); -typedef void NeonGenTwoDoubleOPFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGv_ptr); -typedef void NeonGenOneOpFn(TCGv_i64, TCGv_i64); -typedef void CryptoTwoOpFn(TCGv_ptr, TCGv_ptr); -typedef void CryptoThreeOpIntFn(TCGv_ptr, TCGv_ptr, TCGv_i32); -typedef void CryptoThreeOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr); -typedef void AtomicThreeOpFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGArg, MemOp); - /* initialize TCG globals. */ void a64_translate_init(void) { -- 2.20.1