diff options
Diffstat (limited to 'drivers/gpu/drm/xe')
25 files changed, 279 insertions, 179 deletions
diff --git a/drivers/gpu/drm/xe/Kconfig b/drivers/gpu/drm/xe/Kconfig index 7d7995196702..5c2f459a2925 100644 --- a/drivers/gpu/drm/xe/Kconfig +++ b/drivers/gpu/drm/xe/Kconfig @@ -53,7 +53,7 @@ config DRM_XE config DRM_XE_DISPLAY bool "Enable display support" depends on DRM_XE && DRM_XE=m && HAS_IOPORT - select FB_IOMEM_HELPERS + select FB_IOMEM_HELPERS if DRM_FBDEV_EMULATION select I2C select I2C_ALGOBIT default y diff --git a/drivers/gpu/drm/xe/instructions/xe_gpu_commands.h b/drivers/gpu/drm/xe/instructions/xe_gpu_commands.h index a255946b6f77..8cfcd3360896 100644 --- a/drivers/gpu/drm/xe/instructions/xe_gpu_commands.h +++ b/drivers/gpu/drm/xe/instructions/xe_gpu_commands.h @@ -41,6 +41,7 @@ #define GFX_OP_PIPE_CONTROL(len) ((0x3<<29)|(0x3<<27)|(0x2<<24)|((len)-2)) +#define PIPE_CONTROL0_L3_READ_ONLY_CACHE_INVALIDATE BIT(10) /* gen12 */ #define PIPE_CONTROL0_HDC_PIPELINE_FLUSH BIT(9) /* gen12 */ #define PIPE_CONTROL_COMMAND_CACHE_INVALIDATE (1<<29) diff --git a/drivers/gpu/drm/xe/regs/xe_engine_regs.h b/drivers/gpu/drm/xe/regs/xe_engine_regs.h index 4f372dc2cb89..fb8ec317b6ee 100644 --- a/drivers/gpu/drm/xe/regs/xe_engine_regs.h +++ b/drivers/gpu/drm/xe/regs/xe_engine_regs.h @@ -130,6 +130,10 @@ #define RING_EXECLIST_STATUS_LO(base) XE_REG((base) + 0x234) #define RING_EXECLIST_STATUS_HI(base) XE_REG((base) + 0x234 + 4) +#define RING_IDLEDLY(base) XE_REG((base) + 0x23c) +#define INHIBIT_SWITCH_UNTIL_PREEMPTED REG_BIT(31) +#define IDLE_DELAY REG_GENMASK(20, 0) + #define RING_CONTEXT_CONTROL(base) XE_REG((base) + 0x244, XE_REG_OPTION_MASKED) #define CTX_CTRL_PXP_ENABLE REG_BIT(10) #define CTX_CTRL_OAC_CONTEXT_ENABLE REG_BIT(8) diff --git a/drivers/gpu/drm/xe/xe_device.c b/drivers/gpu/drm/xe/xe_device.c index 5d79b439dd62..00191227bc95 100644 --- a/drivers/gpu/drm/xe/xe_device.c +++ b/drivers/gpu/drm/xe/xe_device.c @@ -53,6 +53,7 @@ #include "xe_pxp.h" #include "xe_query.h" #include "xe_shrinker.h" +#include "xe_survivability_mode.h" #include "xe_sriov.h" #include "xe_tile.h" #include "xe_ttm_stolen_mgr.h" @@ -705,8 +706,20 @@ int xe_device_probe_early(struct xe_device *xe) sriov_update_device_info(xe); err = xe_pcode_probe_early(xe); - if (err) - return err; + if (err) { + int save_err = err; + + /* + * Try to leave device in survivability mode if device is + * possible, but still return the previous error for error + * propagation + */ + err = xe_survivability_mode_enable(xe); + if (err) + return err; + + return save_err; + } err = wait_for_lmem_ready(xe); if (err) diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h index 72ef0b6fc425..9f8667ebba85 100644 --- a/drivers/gpu/drm/xe/xe_device_types.h +++ b/drivers/gpu/drm/xe/xe_device_types.h @@ -585,6 +585,7 @@ struct xe_device { INTEL_DRAM_DDR5, INTEL_DRAM_LPDDR5, INTEL_DRAM_GDDR, + INTEL_DRAM_GDDR_ECC, } type; u8 num_qgv_points; u8 num_psf_gv_points; diff --git a/drivers/gpu/drm/xe/xe_dma_buf.c b/drivers/gpu/drm/xe/xe_dma_buf.c index f67803e15a0e..f7a20264ea33 100644 --- a/drivers/gpu/drm/xe/xe_dma_buf.c +++ b/drivers/gpu/drm/xe/xe_dma_buf.c @@ -145,10 +145,7 @@ static void xe_dma_buf_unmap(struct dma_buf_attachment *attach, struct sg_table *sgt, enum dma_data_direction dir) { - struct dma_buf *dma_buf = attach->dmabuf; - struct xe_bo *bo = gem_to_xe_bo(dma_buf->priv); - - if (!xe_bo_is_vram(bo)) { + if (sg_page(sgt->sgl)) { dma_unmap_sgtable(attach->dev, sgt, dir, 0); sg_free_table(sgt); kfree(sgt); diff --git a/drivers/gpu/drm/xe/xe_eu_stall.c b/drivers/gpu/drm/xe/xe_eu_stall.c index 88a92baf5c95..f2bb9168967c 100644 --- a/drivers/gpu/drm/xe/xe_eu_stall.c +++ b/drivers/gpu/drm/xe/xe_eu_stall.c @@ -222,13 +222,7 @@ int xe_eu_stall_init(struct xe_gt *gt) goto exit_free; } - ret = devm_add_action_or_reset(xe->drm.dev, xe_eu_stall_fini, gt); - if (ret) - goto exit_destroy; - - return 0; -exit_destroy: - destroy_workqueue(gt->eu_stall->buf_ptr_poll_wq); + return devm_add_action_or_reset(xe->drm.dev, xe_eu_stall_fini, gt); exit_free: mutex_destroy(>->eu_stall->stream_lock); kfree(gt->eu_stall); diff --git a/drivers/gpu/drm/xe/xe_execlist.c b/drivers/gpu/drm/xe/xe_execlist.c index 9fbed1a2fcc6..788f56b066b6 100644 --- a/drivers/gpu/drm/xe/xe_execlist.c +++ b/drivers/gpu/drm/xe/xe_execlist.c @@ -297,7 +297,7 @@ err: void xe_execlist_port_destroy(struct xe_execlist_port *port) { - del_timer(&port->irq_fail); + timer_delete(&port->irq_fail); /* Prevent an interrupt while we're destroying */ spin_lock_irq(>_to_xe(port->hwe->gt)->irq.lock); diff --git a/drivers/gpu/drm/xe/xe_gt_clock.c b/drivers/gpu/drm/xe/xe_gt_clock.c index 2a958c92d8ea..4f011d1573c6 100644 --- a/drivers/gpu/drm/xe/xe_gt_clock.c +++ b/drivers/gpu/drm/xe/xe_gt_clock.c @@ -16,35 +16,47 @@ #include "xe_macros.h" #include "xe_mmio.h" -static u32 get_crystal_clock_freq(u32 rpm_config_reg) +#define f19_2_mhz 19200000 +#define f24_mhz 24000000 +#define f25_mhz 25000000 +#define f38_4_mhz 38400000 +#define ts_base_83 83333 +#define ts_base_52 52083 +#define ts_base_80 80000 + +static void read_crystal_clock(struct xe_gt *gt, u32 rpm_config_reg, u32 *freq, + u32 *timestamp_base) { - const u32 f19_2_mhz = 19200000; - const u32 f24_mhz = 24000000; - const u32 f25_mhz = 25000000; - const u32 f38_4_mhz = 38400000; u32 crystal_clock = REG_FIELD_GET(RPM_CONFIG0_CRYSTAL_CLOCK_FREQ_MASK, rpm_config_reg); switch (crystal_clock) { case RPM_CONFIG0_CRYSTAL_CLOCK_FREQ_24_MHZ: - return f24_mhz; + *freq = f24_mhz; + *timestamp_base = ts_base_83; + return; case RPM_CONFIG0_CRYSTAL_CLOCK_FREQ_19_2_MHZ: - return f19_2_mhz; + *freq = f19_2_mhz; + *timestamp_base = ts_base_52; + return; case RPM_CONFIG0_CRYSTAL_CLOCK_FREQ_38_4_MHZ: - return f38_4_mhz; + *freq = f38_4_mhz; + *timestamp_base = ts_base_52; + return; case RPM_CONFIG0_CRYSTAL_CLOCK_FREQ_25_MHZ: - return f25_mhz; + *freq = f25_mhz; + *timestamp_base = ts_base_80; + return; default: - XE_WARN_ON("NOT_POSSIBLE"); - return 0; + xe_gt_warn(gt, "Invalid crystal clock frequency: %u", crystal_clock); + *freq = 0; + *timestamp_base = 0; + return; } } -int xe_gt_clock_init(struct xe_gt *gt) +static void check_ctc_mode(struct xe_gt *gt) { - u32 c0 = xe_mmio_read32(>->mmio, RPM_CONFIG0); - u32 freq = 0; - /* * CTC_MODE[0] = 1 is definitely not supported for Xe2 and later * platforms. In theory it could be a valid setting for pre-Xe2 @@ -57,8 +69,18 @@ int xe_gt_clock_init(struct xe_gt *gt) */ if (xe_mmio_read32(>->mmio, CTC_MODE) & CTC_SOURCE_DIVIDE_LOGIC) xe_gt_warn(gt, "CTC_MODE[0] is set; this is unexpected and undocumented\n"); +} + +int xe_gt_clock_init(struct xe_gt *gt) +{ + u32 freq; + u32 c0; + + if (!IS_SRIOV_VF(gt_to_xe(gt))) + check_ctc_mode(gt); - freq = get_crystal_clock_freq(c0); + c0 = xe_mmio_read32(>->mmio, RPM_CONFIG0); + read_crystal_clock(gt, c0, &freq, >->info.timestamp_base); /* * Now figure out how the command stream's timestamp diff --git a/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c b/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c index 03072e094991..084cbdeba8ea 100644 --- a/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c +++ b/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c @@ -322,6 +322,13 @@ int xe_gt_tlb_invalidation_ggtt(struct xe_gt *gt) return 0; } +/* + * Ensure that roundup_pow_of_two(length) doesn't overflow. + * Note that roundup_pow_of_two() operates on unsigned long, + * not on u64. + */ +#define MAX_RANGE_TLB_INVALIDATION_LENGTH (rounddown_pow_of_two(ULONG_MAX)) + /** * xe_gt_tlb_invalidation_range - Issue a TLB invalidation on this GT for an * address range @@ -346,6 +353,7 @@ int xe_gt_tlb_invalidation_range(struct xe_gt *gt, struct xe_device *xe = gt_to_xe(gt); #define MAX_TLB_INVALIDATION_LEN 7 u32 action[MAX_TLB_INVALIDATION_LEN]; + u64 length = end - start; int len = 0; xe_gt_assert(gt, fence); @@ -358,11 +366,11 @@ int xe_gt_tlb_invalidation_range(struct xe_gt *gt, action[len++] = XE_GUC_ACTION_TLB_INVALIDATION; action[len++] = 0; /* seqno, replaced in send_tlb_invalidation */ - if (!xe->info.has_range_tlb_invalidation) { + if (!xe->info.has_range_tlb_invalidation || + length > MAX_RANGE_TLB_INVALIDATION_LENGTH) { action[len++] = MAKE_INVAL_OP(XE_GUC_TLB_INVAL_FULL); } else { u64 orig_start = start; - u64 length = end - start; u64 align; if (length < SZ_4K) diff --git a/drivers/gpu/drm/xe/xe_gt_types.h b/drivers/gpu/drm/xe/xe_gt_types.h index e3cfb026ac88..7def0959da35 100644 --- a/drivers/gpu/drm/xe/xe_gt_types.h +++ b/drivers/gpu/drm/xe/xe_gt_types.h @@ -121,6 +121,8 @@ struct xe_gt { enum xe_gt_type type; /** @info.reference_clock: clock frequency */ u32 reference_clock; + /** @info.timestamp_base: GT timestamp base */ + u32 timestamp_base; /** * @info.engine_mask: mask of engines present on GT. Some of * them may be reserved in runtime and not available for user. diff --git a/drivers/gpu/drm/xe/xe_guc_ads.c b/drivers/gpu/drm/xe/xe_guc_ads.c index e7c9e095a19f..7031542a70ce 100644 --- a/drivers/gpu/drm/xe/xe_guc_ads.c +++ b/drivers/gpu/drm/xe/xe_guc_ads.c @@ -490,24 +490,52 @@ static void fill_engine_enable_masks(struct xe_gt *gt, engine_enable_mask(gt, XE_ENGINE_CLASS_OTHER)); } -static void guc_prep_golden_lrc_null(struct xe_guc_ads *ads) +/* + * Write the offsets corresponding to the golden LRCs. The actual data is + * populated later by guc_golden_lrc_populate() + */ +static void guc_golden_lrc_init(struct xe_guc_ads *ads) { struct xe_device *xe = ads_to_xe(ads); + struct xe_gt *gt = ads_to_gt(ads); struct iosys_map info_map = IOSYS_MAP_INIT_OFFSET(ads_to_map(ads), offsetof(struct __guc_ads_blob, system_info)); - u8 guc_class; + size_t alloc_size, real_size; + u32 addr_ggtt, offset; + int class; + + offset = guc_ads_golden_lrc_offset(ads); + addr_ggtt = xe_bo_ggtt_addr(ads->bo) + offset; + + for (class = 0; class < XE_ENGINE_CLASS_MAX; ++class) { + u8 guc_class; + + guc_class = xe_engine_class_to_guc_class(class); - for (guc_class = 0; guc_class <= GUC_MAX_ENGINE_CLASSES; ++guc_class) { if (!info_map_read(xe, &info_map, engine_enabled_masks[guc_class])) continue; + real_size = xe_gt_lrc_size(gt, class); + alloc_size = PAGE_ALIGN(real_size); + + /* + * This interface is slightly confusing. We need to pass the + * base address of the full golden context and the size of just + * the engine state, which is the section of the context image + * that starts after the execlists LRC registers. This is + * required to allow the GuC to restore just the engine state + * when a watchdog reset occurs. + * We calculate the engine state size by removing the size of + * what comes before it in the context image (which is identical + * on all engines). + */ ads_blob_write(ads, ads.eng_state_size[guc_class], - guc_ads_golden_lrc_size(ads) - - xe_lrc_skip_size(xe)); + real_size - xe_lrc_skip_size(xe)); ads_blob_write(ads, ads.golden_context_lrca[guc_class], - xe_bo_ggtt_addr(ads->bo) + - guc_ads_golden_lrc_offset(ads)); + addr_ggtt); + + addr_ggtt += alloc_size; } } @@ -857,7 +885,7 @@ void xe_guc_ads_populate_minimal(struct xe_guc_ads *ads) xe_map_memset(ads_to_xe(ads), ads_to_map(ads), 0, 0, ads->bo->size); guc_policies_init(ads); - guc_prep_golden_lrc_null(ads); + guc_golden_lrc_init(ads); guc_mapping_table_init_invalid(gt, &info_map); guc_doorbell_init(ads); @@ -883,7 +911,7 @@ void xe_guc_ads_populate(struct xe_guc_ads *ads) guc_policies_init(ads); fill_engine_enable_masks(gt, &info_map); guc_mmio_reg_state_init(ads); - guc_prep_golden_lrc_null(ads); + guc_golden_lrc_init(ads); guc_mapping_table_init(gt, &info_map); guc_capture_prep_lists(ads); guc_doorbell_init(ads); @@ -903,18 +931,22 @@ void xe_guc_ads_populate(struct xe_guc_ads *ads) guc_ads_private_data_offset(ads)); } -static void guc_populate_golden_lrc(struct xe_guc_ads *ads) +/* + * After the golden LRC's are recorded for each engine class by the first + * submission, copy them to the ADS, as initialized earlier by + * guc_golden_lrc_init(). + */ +static void guc_golden_lrc_populate(struct xe_guc_ads *ads) { struct xe_device *xe = ads_to_xe(ads); struct xe_gt *gt = ads_to_gt(ads); struct iosys_map info_map = IOSYS_MAP_INIT_OFFSET(ads_to_map(ads), offsetof(struct __guc_ads_blob, system_info)); size_t total_size = 0, alloc_size, real_size; - u32 addr_ggtt, offset; + u32 offset; int class; offset = guc_ads_golden_lrc_offset(ads); - addr_ggtt = xe_bo_ggtt_addr(ads->bo) + offset; for (class = 0; class < XE_ENGINE_CLASS_MAX; ++class) { u8 guc_class; @@ -931,26 +963,9 @@ static void guc_populate_golden_lrc(struct xe_guc_ads *ads) alloc_size = PAGE_ALIGN(real_size); total_size += alloc_size; - /* - * This interface is slightly confusing. We need to pass the - * base address of the full golden context and the size of just - * the engine state, which is the section of the context image - * that starts after the execlists LRC registers. This is - * required to allow the GuC to restore just the engine state - * when a watchdog reset occurs. - * We calculate the engine state size by removing the size of - * what comes before it in the context image (which is identical - * on all engines). - */ - ads_blob_write(ads, ads.eng_state_size[guc_class], - real_size - xe_lrc_skip_size(xe)); - ads_blob_write(ads, ads.golden_context_lrca[guc_class], - addr_ggtt); - xe_map_memcpy_to(xe, ads_to_map(ads), offset, gt->default_lrc[class], real_size); - addr_ggtt += alloc_size; offset += alloc_size; } @@ -959,7 +974,7 @@ static void guc_populate_golden_lrc(struct xe_guc_ads *ads) void xe_guc_ads_populate_post_load(struct xe_guc_ads *ads) { - guc_populate_golden_lrc(ads); + guc_golden_lrc_populate(ads); } static int guc_ads_action_update_policies(struct xe_guc_ads *ads, u32 policy_offset) diff --git a/drivers/gpu/drm/xe/xe_guc_pc.c b/drivers/gpu/drm/xe/xe_guc_pc.c index 85215313976c..43b1192ba61c 100644 --- a/drivers/gpu/drm/xe/xe_guc_pc.c +++ b/drivers/gpu/drm/xe/xe_guc_pc.c @@ -1070,6 +1070,7 @@ int xe_guc_pc_start(struct xe_guc_pc *pc) if (wait_for_pc_state(pc, SLPC_GLOBAL_STATE_RUNNING, SLPC_RESET_EXTENDED_TIMEOUT_MS)) { xe_gt_err(gt, "GuC PC Start failed: Dynamic GT frequency control and GT sleep states are now disabled.\n"); + ret = -EIO; goto out; } diff --git a/drivers/gpu/drm/xe/xe_hmm.c b/drivers/gpu/drm/xe/xe_hmm.c index c3cc0fa105e8..57b71956ddf4 100644 --- a/drivers/gpu/drm/xe/xe_hmm.c +++ b/drivers/gpu/drm/xe/xe_hmm.c @@ -19,29 +19,6 @@ static u64 xe_npages_in_range(unsigned long start, unsigned long end) return (end - start) >> PAGE_SHIFT; } -/** - * xe_mark_range_accessed() - mark a range is accessed, so core mm - * have such information for memory eviction or write back to - * hard disk - * @range: the range to mark - * @write: if write to this range, we mark pages in this range - * as dirty - */ -static void xe_mark_range_accessed(struct hmm_range *range, bool write) -{ - struct page *page; - u64 i, npages; - - npages = xe_npages_in_range(range->start, range->end); - for (i = 0; i < npages; i++) { - page = hmm_pfn_to_page(range->hmm_pfns[i]); - if (write) - set_page_dirty_lock(page); - - mark_page_accessed(page); - } -} - static int xe_alloc_sg(struct xe_device *xe, struct sg_table *st, struct hmm_range *range, struct rw_semaphore *notifier_sem) { @@ -331,7 +308,6 @@ int xe_hmm_userptr_populate_range(struct xe_userptr_vma *uvma, if (ret) goto out_unlock; - xe_mark_range_accessed(&hmm_range, write); userptr->sg = &userptr->sgt; xe_hmm_userptr_set_mapped(uvma); userptr->notifier_seq = hmm_range.notifier_seq; diff --git a/drivers/gpu/drm/xe/xe_hw_engine.c b/drivers/gpu/drm/xe/xe_hw_engine.c index 223b95de388c..93241fd0a4ba 100644 --- a/drivers/gpu/drm/xe/xe_hw_engine.c +++ b/drivers/gpu/drm/xe/xe_hw_engine.c @@ -8,7 +8,9 @@ #include <linux/nospec.h> #include <drm/drm_managed.h> +#include <drm/drm_print.h> #include <uapi/drm/xe_drm.h> +#include <generated/xe_wa_oob.h> #include "regs/xe_engine_regs.h" #include "regs/xe_gt_regs.h" @@ -21,6 +23,7 @@ #include "xe_gsc.h" #include "xe_gt.h" #include "xe_gt_ccs_mode.h" +#include "xe_gt_clock.h" #include "xe_gt_printk.h" #include "xe_gt_mcr.h" #include "xe_gt_topology.h" @@ -386,12 +389,6 @@ xe_hw_engine_setup_default_lrc_state(struct xe_hw_engine *hwe) blit_cctl_val, XE_RTP_ACTION_FLAG(ENGINE_BASE))) }, - /* Use Fixed slice CCS mode */ - { XE_RTP_NAME("RCU_MODE_FIXED_SLICE_CCS_MODE"), - XE_RTP_RULES(FUNC(xe_hw_engine_match_fixed_cslice_mode)), - XE_RTP_ACTIONS(FIELD_SET(RCU_MODE, RCU_MODE_FIXED_SLICE_CCS_MODE, - RCU_MODE_FIXED_SLICE_CCS_MODE)) - }, /* Disable WMTP if HW doesn't support it */ { XE_RTP_NAME("DISABLE_WMTP_ON_UNSUPPORTED_HW"), XE_RTP_RULES(FUNC(xe_rtp_cfeg_wmtp_disabled)), @@ -458,6 +455,12 @@ hw_engine_setup_default_state(struct xe_hw_engine *hwe) XE_RTP_ACTIONS(SET(CSFE_CHICKEN1(0), CS_PRIORITY_MEM_READ, XE_RTP_ACTION_FLAG(ENGINE_BASE))) }, + /* Use Fixed slice CCS mode */ + { XE_RTP_NAME("RCU_MODE_FIXED_SLICE_CCS_MODE"), + XE_RTP_RULES(FUNC(xe_hw_engine_match_fixed_cslice_mode)), + XE_RTP_ACTIONS(FIELD_SET(RCU_MODE, RCU_MODE_FIXED_SLICE_CCS_MODE, + RCU_MODE_FIXED_SLICE_CCS_MODE)) + }, }; xe_rtp_process_to_sr(&ctx, engine_entries, ARRAY_SIZE(engine_entries), &hwe->reg_sr); @@ -564,6 +567,33 @@ static void hw_engine_init_early(struct xe_gt *gt, struct xe_hw_engine *hwe, xe_reg_whitelist_process_engine(hwe); } +static void adjust_idledly(struct xe_hw_engine *hwe) +{ + struct xe_gt *gt = hwe->gt; + u32 idledly, maxcnt; + u32 idledly_units_ps = 8 * gt->info.timestamp_base; + u32 maxcnt_units_ns = 640; + bool inhibit_switch = 0; + + if (!IS_SRIOV_VF(gt_to_xe(hwe->gt)) && XE_WA(gt, 16023105232)) { + idledly = xe_mmio_read32(>->mmio, RING_IDLEDLY(hwe->mmio_base)); + maxcnt = xe_mmio_read32(>->mmio, RING_PWRCTX_MAXCNT(hwe->mmio_base)); + + inhibit_switch = idledly & INHIBIT_SWITCH_UNTIL_PREEMPTED; + idledly = REG_FIELD_GET(IDLE_DELAY, idledly); + idledly = DIV_ROUND_CLOSEST(idledly * idledly_units_ps, 1000); + maxcnt = REG_FIELD_GET(IDLE_WAIT_TIME, maxcnt); + maxcnt *= maxcnt_units_ns; + + if (xe_gt_WARN_ON(gt, idledly >= maxcnt || inhibit_switch)) { + idledly = DIV_ROUND_CLOSEST(((maxcnt - 1) * maxcnt_units_ns), + idledly_units_ps); + idledly = DIV_ROUND_CLOSEST(idledly, 1000); + xe_mmio_write32(>->mmio, RING_IDLEDLY(hwe->mmio_base), idledly); + } + } +} + static int hw_engine_init(struct xe_gt *gt, struct xe_hw_engine *hwe, enum xe_hw_engine_id id) { @@ -604,6 +634,9 @@ static int hw_engine_init(struct xe_gt *gt, struct xe_hw_engine *hwe, if (xe->info.has_usm && hwe->class == XE_ENGINE_CLASS_COPY) gt->usm.reserved_bcs_instance = hwe->instance; + /* Ensure IDLEDLY is lower than MAXCNT */ + adjust_idledly(hwe); + return devm_add_action_or_reset(xe->drm.dev, hw_engine_fini, hwe); err_hwsp: diff --git a/drivers/gpu/drm/xe/xe_hw_engine_class_sysfs.c b/drivers/gpu/drm/xe/xe_hw_engine_class_sysfs.c index b53e8d2accdb..a440442b4d72 100644 --- a/drivers/gpu/drm/xe/xe_hw_engine_class_sysfs.c +++ b/drivers/gpu/drm/xe/xe_hw_engine_class_sysfs.c @@ -32,14 +32,61 @@ bool xe_hw_engine_timeout_in_range(u64 timeout, u64 min, u64 max) return timeout >= min && timeout <= max; } -static void kobj_xe_hw_engine_release(struct kobject *kobj) +static void xe_hw_engine_sysfs_kobj_release(struct kobject *kobj) { kfree(kobj); } +static ssize_t xe_hw_engine_class_sysfs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct xe_device *xe = kobj_to_xe(kobj); + struct kobj_attribute *kattr; + ssize_t ret = -EIO; + + kattr = container_of(attr, struct kobj_attribute, attr); + if (kattr->show) { + xe_pm_runtime_get(xe); + ret = kattr->show(kobj, kattr, buf); + xe_pm_runtime_put(xe); + } + + return ret; +} + +static ssize_t xe_hw_engine_class_sysfs_attr_store(struct kobject *kobj, + struct attribute *attr, + const char *buf, + size_t count) +{ + struct xe_device *xe = kobj_to_xe(kobj); + struct kobj_attribute *kattr; + ssize_t ret = -EIO; + + kattr = container_of(attr, struct kobj_attribute, attr); + if (kattr->store) { + xe_pm_runtime_get(xe); + ret = kattr->store(kobj, kattr, buf, count); + xe_pm_runtime_put(xe); + } + + return ret; +} + +static const struct sysfs_ops xe_hw_engine_class_sysfs_ops = { + .show = xe_hw_engine_class_sysfs_attr_show, + .store = xe_hw_engine_class_sysfs_attr_store, +}; + static const struct kobj_type kobj_xe_hw_engine_type = { - .release = kobj_xe_hw_engine_release, - .sysfs_ops = &kobj_sysfs_ops + .release = xe_hw_engine_sysfs_kobj_release, + .sysfs_ops = &xe_hw_engine_class_sysfs_ops, +}; + +static const struct kobj_type kobj_xe_hw_engine_type_def = { + .release = xe_hw_engine_sysfs_kobj_release, + .sysfs_ops = &kobj_sysfs_ops, }; static ssize_t job_timeout_max_store(struct kobject *kobj, @@ -543,7 +590,7 @@ static int xe_add_hw_engine_class_defaults(struct xe_device *xe, if (!kobj) return -ENOMEM; - kobject_init(kobj, &kobj_xe_hw_engine_type); + kobject_init(kobj, &kobj_xe_hw_engine_type_def); err = kobject_add(kobj, parent, "%s", ".defaults"); if (err) goto err_object; @@ -559,57 +606,6 @@ err_object: return err; } -static void xe_hw_engine_sysfs_kobj_release(struct kobject *kobj) -{ - kfree(kobj); -} - -static ssize_t xe_hw_engine_class_sysfs_attr_show(struct kobject *kobj, - struct attribute *attr, - char *buf) -{ - struct xe_device *xe = kobj_to_xe(kobj); - struct kobj_attribute *kattr; - ssize_t ret = -EIO; - - kattr = container_of(attr, struct kobj_attribute, attr); - if (kattr->show) { - xe_pm_runtime_get(xe); - ret = kattr->show(kobj, kattr, buf); - xe_pm_runtime_put(xe); - } - - return ret; -} - -static ssize_t xe_hw_engine_class_sysfs_attr_store(struct kobject *kobj, - struct attribute *attr, - const char *buf, - size_t count) -{ - struct xe_device *xe = kobj_to_xe(kobj); - struct kobj_attribute *kattr; - ssize_t ret = -EIO; - - kattr = container_of(attr, struct kobj_attribute, attr); - if (kattr->store) { - xe_pm_runtime_get(xe); - ret = kattr->store(kobj, kattr, buf, count); - xe_pm_runtime_put(xe); - } - - return ret; -} - -static const struct sysfs_ops xe_hw_engine_class_sysfs_ops = { - .show = xe_hw_engine_class_sysfs_attr_show, - .store = xe_hw_engine_class_sysfs_attr_store, -}; - -static const struct kobj_type xe_hw_engine_sysfs_kobj_type = { - .release = xe_hw_engine_sysfs_kobj_release, - .sysfs_ops = &xe_hw_engine_class_sysfs_ops, -}; static void hw_engine_class_sysfs_fini(void *arg) { @@ -640,7 +636,7 @@ int xe_hw_engine_class_sysfs_init(struct xe_gt *gt) if (!kobj) return -ENOMEM; - kobject_init(kobj, &xe_hw_engine_sysfs_kobj_type); + kobject_init(kobj, &kobj_xe_hw_engine_type); err = kobject_add(kobj, gt->sysfs, "engines"); if (err) diff --git a/drivers/gpu/drm/xe/xe_migrate.c b/drivers/gpu/drm/xe/xe_migrate.c index df4282c71bf0..5a3e89022c38 100644 --- a/drivers/gpu/drm/xe/xe_migrate.c +++ b/drivers/gpu/drm/xe/xe_migrate.c @@ -1177,7 +1177,7 @@ err: err_sync: /* Sync partial copies if any. FIXME: job_mutex? */ if (fence) { - dma_fence_wait(m->fence, false); + dma_fence_wait(fence, false); dma_fence_put(fence); } @@ -1547,7 +1547,7 @@ void xe_migrate_wait(struct xe_migrate *m) static u32 pte_update_cmd_size(u64 size) { u32 num_dword; - u64 entries = DIV_ROUND_UP(size, XE_PAGE_SIZE); + u64 entries = DIV_U64_ROUND_UP(size, XE_PAGE_SIZE); XE_WARN_ON(size > MAX_PREEMPTDISABLE_TRANSFER); /* @@ -1558,7 +1558,7 @@ static u32 pte_update_cmd_size(u64 size) * 2 dword for the page table's physical location * 2*n dword for value of pte to fill (each pte entry is 2 dwords) */ - num_dword = (1 + 2) * DIV_ROUND_UP(entries, 0x1ff); + num_dword = (1 + 2) * DIV_U64_ROUND_UP(entries, 0x1ff); num_dword += entries * 2; return num_dword; diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c index da9679c8cf26..818f023166d5 100644 --- a/drivers/gpu/drm/xe/xe_pci.c +++ b/drivers/gpu/drm/xe/xe_pci.c @@ -803,16 +803,14 @@ static int xe_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent) return err; err = xe_device_probe_early(xe); - - /* - * In Boot Survivability mode, no drm card is exposed and driver is - * loaded with bare minimum to allow for firmware to be flashed through - * mei. If early probe fails, check if survivability mode is flagged by - * HW to be enabled. In that case enable it and return success. - */ if (err) { - if (xe_survivability_mode_required(xe) && - xe_survivability_mode_enable(xe)) + /* + * In Boot Survivability mode, no drm card is exposed and driver + * is loaded with bare minimum to allow for firmware to be + * flashed through mei. If early probe failed, but it managed to + * enable survivability mode, return success. + */ + if (xe_survivability_mode_is_enabled(xe)) return 0; return err; diff --git a/drivers/gpu/drm/xe/xe_pxp_debugfs.c b/drivers/gpu/drm/xe/xe_pxp_debugfs.c index ccfbacf08efc..525a2f6bb076 100644 --- a/drivers/gpu/drm/xe/xe_pxp_debugfs.c +++ b/drivers/gpu/drm/xe/xe_pxp_debugfs.c @@ -66,9 +66,18 @@ static int pxp_terminate(struct seq_file *m, void *data) { struct xe_pxp *pxp = node_to_pxp(m->private); struct drm_printer p = drm_seq_file_printer(m); + int ready = xe_pxp_get_readiness_status(pxp); - if (!xe_pxp_is_enabled(pxp)) - return -ENODEV; + if (ready < 0) + return ready; /* disabled or error occurred */ + else if (!ready) + return -EBUSY; /* init still in progress */ + + /* no need for a termination if PXP is not active */ + if (pxp->status != XE_PXP_ACTIVE) { + drm_printf(&p, "PXP not active\n"); + return 0; + } /* simulate a termination interrupt */ spin_lock_irq(&pxp->xe->irq.lock); diff --git a/drivers/gpu/drm/xe/xe_ring_ops.c b/drivers/gpu/drm/xe/xe_ring_ops.c index 917fc16de866..a7582b097ae6 100644 --- a/drivers/gpu/drm/xe/xe_ring_ops.c +++ b/drivers/gpu/drm/xe/xe_ring_ops.c @@ -137,7 +137,8 @@ emit_pipe_control(u32 *dw, int i, u32 bit_group_0, u32 bit_group_1, u32 offset, static int emit_pipe_invalidate(u32 mask_flags, bool invalidate_tlb, u32 *dw, int i) { - u32 flags = PIPE_CONTROL_CS_STALL | + u32 flags0 = 0; + u32 flags1 = PIPE_CONTROL_CS_STALL | PIPE_CONTROL_COMMAND_CACHE_INVALIDATE | PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE | PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE | @@ -148,11 +149,15 @@ static int emit_pipe_invalidate(u32 mask_flags, bool invalidate_tlb, u32 *dw, PIPE_CONTROL_STORE_DATA_INDEX; if (invalidate_tlb) - flags |= PIPE_CONTROL_TLB_INVALIDATE; + flags1 |= PIPE_CONTROL_TLB_INVALIDATE; - flags &= ~mask_flags; + flags1 &= ~mask_flags; - return emit_pipe_control(dw, i, 0, flags, LRC_PPHWSP_FLUSH_INVAL_SCRATCH_ADDR, 0); + if (flags1 & PIPE_CONTROL_VF_CACHE_INVALIDATE) + flags0 |= PIPE_CONTROL0_L3_READ_ONLY_CACHE_INVALIDATE; + + return emit_pipe_control(dw, i, flags0, flags1, + LRC_PPHWSP_FLUSH_INVAL_SCRATCH_ADDR, 0); } static int emit_store_imm_ppgtt_posted(u64 addr, u64 value, diff --git a/drivers/gpu/drm/xe/xe_survivability_mode.c b/drivers/gpu/drm/xe/xe_survivability_mode.c index d939ce70e6fa..cb813b337fd3 100644 --- a/drivers/gpu/drm/xe/xe_survivability_mode.c +++ b/drivers/gpu/drm/xe/xe_survivability_mode.c @@ -155,13 +155,21 @@ static int enable_survivability_mode(struct pci_dev *pdev) if (ret) return ret; + /* Make sure xe_heci_gsc_init() knows about survivability mode */ + survivability->mode = true; + ret = xe_heci_gsc_init(xe); - if (ret) + if (ret) { + /* + * But if it fails, device can't enter survivability + * so move it back for correct error handling + */ + survivability->mode = false; return ret; + } xe_vsec_init(xe); - survivability->mode = true; dev_err(dev, "In Survivability Mode\n"); return 0; @@ -178,15 +186,16 @@ bool xe_survivability_mode_is_enabled(struct xe_device *xe) return xe->survivability.mode; } -/** - * xe_survivability_mode_required - checks if survivability mode is required - * @xe: xe device instance +/* + * survivability_mode_requested - check if it's possible to enable + * survivability mode and that was requested by firmware * - * This function reads the boot status from Pcode + * This function reads the boot status from Pcode. * - * Return: true if boot status indicates failure, false otherwise + * Return: true if platform support is available and boot status indicates + * failure, false otherwise. */ -bool xe_survivability_mode_required(struct xe_device *xe) +static bool survivability_mode_requested(struct xe_device *xe) { struct xe_survivability *survivability = &xe->survivability; struct xe_mmio *mmio = xe_root_tile_mmio(xe); @@ -208,7 +217,8 @@ bool xe_survivability_mode_required(struct xe_device *xe) * * Initialize survivability information and enable survivability mode * - * Return: 0 for success, negative error code otherwise. + * Return: 0 if survivability mode is enabled or not requested; negative error + * code otherwise. */ int xe_survivability_mode_enable(struct xe_device *xe) { @@ -216,6 +226,9 @@ int xe_survivability_mode_enable(struct xe_device *xe) struct xe_survivability_info *info; struct pci_dev *pdev = to_pci_dev(xe->drm.dev); + if (!survivability_mode_requested(xe)) + return 0; + survivability->size = MAX_SCRATCH_MMIO; info = devm_kcalloc(xe->drm.dev, survivability->size, sizeof(*info), diff --git a/drivers/gpu/drm/xe/xe_survivability_mode.h b/drivers/gpu/drm/xe/xe_survivability_mode.h index f4df5f9025ce..d7e64885570d 100644 --- a/drivers/gpu/drm/xe/xe_survivability_mode.h +++ b/drivers/gpu/drm/xe/xe_survivability_mode.h @@ -12,6 +12,5 @@ struct xe_device; int xe_survivability_mode_enable(struct xe_device *xe); bool xe_survivability_mode_is_enabled(struct xe_device *xe); -bool xe_survivability_mode_required(struct xe_device *xe); #endif /* _XE_SURVIVABILITY_MODE_H_ */ diff --git a/drivers/gpu/drm/xe/xe_svm.c b/drivers/gpu/drm/xe/xe_svm.c index 3e829c87d7b4..f8c128524d9f 100644 --- a/drivers/gpu/drm/xe/xe_svm.c +++ b/drivers/gpu/drm/xe/xe_svm.c @@ -696,11 +696,14 @@ retry: list_for_each_entry(block, blocks, link) block->private = vr; + xe_bo_get(bo); err = drm_gpusvm_migrate_to_devmem(&vm->svm.gpusvm, &range->base, &bo->devmem_allocation, ctx); - xe_bo_unlock(bo); if (err) - xe_bo_put(bo); /* Creation ref */ + xe_svm_devmem_release(&bo->devmem_allocation); + + xe_bo_unlock(bo); + xe_bo_put(bo); unlock: mmap_read_unlock(mm); diff --git a/drivers/gpu/drm/xe/xe_wa.c b/drivers/gpu/drm/xe/xe_wa.c index a25afb757f70..24f644c0a673 100644 --- a/drivers/gpu/drm/xe/xe_wa.c +++ b/drivers/gpu/drm/xe/xe_wa.c @@ -622,6 +622,12 @@ static const struct xe_rtp_entry_sr engine_was[] = { FUNC(xe_rtp_match_first_render_or_compute)), XE_RTP_ACTIONS(SET(TDL_TSL_CHICKEN, RES_CHK_SPR_DIS)) }, + { XE_RTP_NAME("16023105232"), + XE_RTP_RULES(MEDIA_VERSION_RANGE(1301, 3000), OR, + GRAPHICS_VERSION_RANGE(2001, 3001)), + XE_RTP_ACTIONS(SET(RING_PSMI_CTL(0), RC_SEMA_IDLE_MSG_DISABLE, + XE_RTP_ACTION_FLAG(ENGINE_BASE))) + }, }; static const struct xe_rtp_entry_sr lrc_was[] = { diff --git a/drivers/gpu/drm/xe/xe_wa_oob.rules b/drivers/gpu/drm/xe/xe_wa_oob.rules index e0c5fa460487..9b9e176992a8 100644 --- a/drivers/gpu/drm/xe/xe_wa_oob.rules +++ b/drivers/gpu/drm/xe/xe_wa_oob.rules @@ -32,8 +32,10 @@ GRAPHICS_VERSION(3001) 14022293748 GRAPHICS_VERSION(2001) GRAPHICS_VERSION(2004) + GRAPHICS_VERSION_RANGE(3000, 3001) 22019794406 GRAPHICS_VERSION(2001) GRAPHICS_VERSION(2004) + GRAPHICS_VERSION_RANGE(3000, 3001) 22019338487 MEDIA_VERSION(2000) GRAPHICS_VERSION(2001) MEDIA_VERSION(3000), MEDIA_STEP(A0, B0), FUNC(xe_rtp_match_not_sriov_vf) @@ -53,3 +55,5 @@ no_media_l3 MEDIA_VERSION(3000) GRAPHICS_VERSION_RANGE(1270, 1274) 1508761755 GRAPHICS_VERSION(1255) GRAPHICS_VERSION(1260), GRAPHICS_STEP(A0, B0) +16023105232 GRAPHICS_VERSION_RANGE(2001, 3001) + MEDIA_VERSION_RANGE(1301, 3000) |