Return-Path: Received: from mga11.intel.com (mga11.intel.com [192.55.52.93]) by gabe.freedesktop.org (Postfix) with ESMTPS id 204DF6E8C6 for ; Tue, 9 Nov 2021 23:04:06 +0000 (UTC) From: Matthew Brost Date: Tue, 9 Nov 2021 14:59:28 -0800 Message-Id: <20211109225928.23711-8-matthew.brost@intel.com> In-Reply-To: <20211109225928.23711-1-matthew.brost@intel.com> References: <20211109225928.23711-1-matthew.brost@intel.com> MIME-Version: 1.0 Content-Type: text/x-diff; charset=UTF-8 Content-Transfer-Encoding: 8bit Subject: [igt-dev] [PATCH 7/7] REVIEW: Full tree diff against internal List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: igt-dev-bounces@lists.freedesktop.org Sender: "igt-dev" To: igt-dev@lists.freedesktop.org List-ID: Auto-generated diff between internal..ct_backport --- .../drm/i915/gt/uc/abi/guc_communication_ctb_abi.h | 9 +- .../i915/gt/uc/abi/guc_communication_mmio_abi.h | 8 +- drivers/gpu/drm/i915/gt/uc/abi/guc_errors_abi.h | 1 + drivers/gpu/drm/i915/gt/uc/intel_guc.c | 41 ++--- drivers/gpu/drm/i915/gt/uc/intel_guc.h | 6 - drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c | 167 ++++++++++++--------- drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h | 8 + drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c | 8 +- drivers/gpu/drm/i915/i915_params.c | 2 +- 9 files changed, 135 insertions(+), 115 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/uc/abi/guc_communication_ctb_abi.h b/drivers/gpu/drm/i915/gt/uc/abi/guc_communication_ctb_abi.h index 92660726c094..99e1fad5ca20 100644 --- a/drivers/gpu/drm/i915/gt/uc/abi/guc_communication_ctb_abi.h +++ b/drivers/gpu/drm/i915/gt/uc/abi/guc_communication_ctb_abi.h @@ -14,7 +14,7 @@ /** * DOC: CT Buffer * - * TBD + * Circular buffer used to send `CTB Message`_ */ /** @@ -37,8 +37,6 @@ * | | | - _`GUC_CTB_STATUS_OVERFLOW` = 1 (head/tail too large) | * | | | - _`GUC_CTB_STATUS_UNDERFLOW` = 2 (truncated message) | * | | | - _`GUC_CTB_STATUS_MISMATCH` = 4 (head/tail modified) | - * | | | - _`GUC_CTB_STATUS_NO_BACKCHANNEL` = 8 | - * | | | - _`GUC_CTB_STATUS_MALFORMED_MSG` = 16 | * +---+-------+--------------------------------------------------------------+ * |...| | RESERVED = MBZ | * +---+-------+--------------------------------------------------------------+ @@ -54,8 +52,6 @@ struct guc_ct_buffer_desc { #define GUC_CTB_STATUS_OVERFLOW (1 << 0) #define GUC_CTB_STATUS_UNDERFLOW (1 << 1) #define GUC_CTB_STATUS_MISMATCH (1 << 2) -#define GUC_CTB_STATUS_NO_BACKCHANNEL (1 << 3) -#define GUC_CTB_STATUS_MALFORMED_MSG (1 << 4) u32 reserved[13]; } __packed; static_assert(sizeof(struct guc_ct_buffer_desc) == 64); @@ -83,7 +79,8 @@ static_assert(sizeof(struct guc_ct_buffer_desc) == 64); * +---+-------+--------------------------------------------------------------+ */ -#define GUC_CTB_MSG_MIN_LEN 1u +#define GUC_CTB_HDR_LEN 1u +#define GUC_CTB_MSG_MIN_LEN GUC_CTB_HDR_LEN #define GUC_CTB_MSG_MAX_LEN 256u #define GUC_CTB_MSG_0_FENCE (0xffff << 16) #define GUC_CTB_MSG_0_FORMAT (0xf << 12) diff --git a/drivers/gpu/drm/i915/gt/uc/abi/guc_communication_mmio_abi.h b/drivers/gpu/drm/i915/gt/uc/abi/guc_communication_mmio_abi.h index 3f9039e3ef9d..bbf1ddb77434 100644 --- a/drivers/gpu/drm/i915/gt/uc/abi/guc_communication_mmio_abi.h +++ b/drivers/gpu/drm/i915/gt/uc/abi/guc_communication_mmio_abi.h @@ -18,17 +18,17 @@ * registers, is directly written into those scratch registers. * * For Gen9+, there are 16 software scratch registers 0xC180-0xC1B8, - * but no H2G command takes more than 8 parameters and the GuC firmware - * itself uses an 8-element array to store the H2G message. + * but no H2G command takes more than 4 parameters and the GuC firmware + * itself uses an 4-element array to store the H2G message. * * For Gen11+, there are additional 4 registers 0x190240-0x19024C, which - * are, regardless on lower count, preffered over legacy ones. + * are, regardless on lower count, preferred over legacy ones. * * The MMIO based communication is mainly used during driver initialization * phase to setup the `CTB based communication`_ that will be used afterwards. */ -#define GUC_MAX_MMIO_MSG_LEN 8 +#define GUC_MAX_MMIO_MSG_LEN 4 /** * DOC: MMIO HXG Message diff --git a/drivers/gpu/drm/i915/gt/uc/abi/guc_errors_abi.h b/drivers/gpu/drm/i915/gt/uc/abi/guc_errors_abi.h index b1c79a2b814d..8085fb181274 100644 --- a/drivers/gpu/drm/i915/gt/uc/abi/guc_errors_abi.h +++ b/drivers/gpu/drm/i915/gt/uc/abi/guc_errors_abi.h @@ -7,6 +7,7 @@ #define _ABI_GUC_ERRORS_ABI_H enum intel_guc_response_status { + INTEL_GUC_RESPONSE_STATUS_SUCCESS = 0x0, INTEL_GUC_RESPONSE_NOT_SUPPORTED = 0x20, INTEL_GUC_RESPONSE_NO_ATTRIBUTE_TABLE = 0x201, INTEL_GUC_RESPONSE_NO_DECRYPTION_KEY = 0x202, diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc.c b/drivers/gpu/drm/i915/gt/uc/intel_guc.c index 7e63463fe83a..e2df29fa0099 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc.c @@ -106,14 +106,12 @@ static void gen9_enable_guc_interrupts(struct intel_guc *guc) struct intel_gt *gt = guc_to_gt(guc); assert_rpm_wakelock_held(>->i915->runtime_pm); + guc->interrupts.enabled = true; spin_lock_irq(gt->irq_lock); - if (!guc->interrupts.enabled) { - WARN_ON_ONCE(intel_uncore_read(gt->uncore, GEN8_GT_IIR(2)) & - gt->pm_guc_events); - guc->interrupts.enabled = true; - gen6_gt_pm_enable_irq(gt, gt->pm_guc_events); - } + WARN_ON_ONCE(intel_uncore_read(gt->uncore, GEN8_GT_IIR(2)) & + gt->pm_guc_events); + gen6_gt_pm_enable_irq(gt, gt->pm_guc_events); spin_unlock_irq(gt->irq_lock); } @@ -122,9 +120,9 @@ static void gen9_disable_guc_interrupts(struct intel_guc *guc) struct intel_gt *gt = guc_to_gt(guc); assert_rpm_wakelock_held(>->i915->runtime_pm); + guc->interrupts.enabled = false; spin_lock_irq(gt->irq_lock); - guc->interrupts.enabled = false; gen6_gt_pm_disable_irq(gt, gt->pm_guc_events); @@ -172,14 +170,13 @@ static void gen11_enable_fake_interrupts(struct intel_guc *guc) { struct intel_gt *gt = guc_to_gt(guc); + guc->interrupts.enabled = true; + spin_lock_irq(gt->irq_lock); - if (!guc->interrupts.enabled) { - WARN_ON_ONCE(gen11_gt_reset_one_iir(gt, 0, GEN11_GUC)); + WARN_ON_ONCE(gen11_gt_reset_one_iir(gt, 0, GEN11_GUC)); - /* FIXME: Connect timer to GT PM */ - fake_int_timer_start(gt); - guc->interrupts.enabled = true; - } + /* FIXME: Connect timer to GT PM */ + fake_int_timer_start(gt); spin_unlock_irq(gt->irq_lock); } @@ -187,11 +184,9 @@ static void gen11_disable_fake_interrupts(struct intel_guc *guc) { struct intel_gt *gt = guc_to_gt(guc); - if (!guc->interrupts.enabled) - return; + guc->interrupts.enabled = false; spin_lock_irq(gt->irq_lock); - guc->interrupts.enabled = false; fake_int_timer_stop(gt); @@ -203,8 +198,6 @@ static void gen11_disable_fake_interrupts(struct intel_guc *guc) void intel_guc_init_fake_interrupts(struct intel_guc *guc) { - GEM_BUG_ON(guc->interrupts.enabled); - guc->interrupts.enable = gen11_enable_fake_interrupts; guc->interrupts.disable = gen11_disable_fake_interrupts; } @@ -213,11 +206,10 @@ static void gen11_enable_guc_interrupts(struct intel_guc *guc) { struct intel_gt *gt = guc_to_gt(guc); + guc->interrupts.enabled = true; + spin_lock_irq(gt->irq_lock); - if (!guc->interrupts.enabled) { - __gen11_reset_guc_interrupts(gt); - guc->interrupts.enabled = true; - } + __gen11_reset_guc_interrupts(gt); spin_unlock_irq(gt->irq_lock); } @@ -225,12 +217,7 @@ static void gen11_disable_guc_interrupts(struct intel_guc *guc) { struct intel_gt *gt = guc_to_gt(guc); - if (!guc->interrupts.enabled) - return; - - spin_lock_irq(gt->irq_lock); guc->interrupts.enabled = false; - spin_unlock_irq(gt->irq_lock); intel_synchronize_irq(gt->i915); gen11_reset_guc_interrupts(guc); diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc.h b/drivers/gpu/drm/i915/gt/uc/intel_guc.h index 3e7fe30861b7..590365152e57 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc.h +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc.h @@ -155,12 +155,6 @@ inline int intel_guc_send(struct intel_guc *guc, const u32 *action, u32 len) return intel_guc_ct_send(&guc->ct, action, len, NULL, 0, 0); } -#define INTEL_GUC_SEND_NB BIT(31) -#define INTEL_GUC_SEND_G2H_DW_SHIFT 0 -#define INTEL_GUC_SEND_G2H_DW_MASK (0xff << INTEL_GUC_SEND_G2H_DW_SHIFT) -#define MAKE_SEND_FLAGS(len) \ - ({GEM_BUG_ON(!FIELD_FIT(INTEL_GUC_SEND_G2H_DW_MASK, len)); \ - (FIELD_PREP(INTEL_GUC_SEND_G2H_DW_MASK, len) | INTEL_GUC_SEND_NB);}) static inline int intel_guc_send_nb(struct intel_guc *guc, const u32 *action, u32 len, u32 g2h_len_dw) diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c index 8d1549c63600..494c66df8359 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c @@ -11,6 +11,7 @@ #include "intel_guc_ct.h" #include "intel_pagefault.h" #include "gt/intel_gt.h" +#include "gem/i915_gem_lmem.h" #include "gt/iov/intel_iov_event.h" #include "gt/iov/intel_iov_fault.h" #include "gt/iov/intel_iov_relay.h" @@ -82,7 +83,7 @@ static inline struct drm_device *ct_to_drm(struct intel_guc_ct *ct) * Size of each `CT Buffer`_ must be multiple of 4K. * We don't expect too many messages in flight at any time, unless we are * using the GuC submission. In that case each request requires a minimum - * 16 bytes which gives us a maximum 256 queue'd requests. Hopefully this + * 2 dwords which gives us a maximum 256 queue'd requests. Hopefully this * enough space to avoid backpressure on the driver. We increase the size * of the receive buffer (relative to the send) to ensure a G2H response * CTB has a landing spot. @@ -92,8 +93,6 @@ static inline struct drm_device *ct_to_drm(struct intel_guc_ct *ct) #define CTB_G2H_BUFFER_SIZE (4 * CTB_H2G_BUFFER_SIZE) #define G2H_ROOM_BUFFER_SIZE (PAGE_SIZE) -#define MAX_US_STALL_CTB 1000000 - struct ct_request { struct list_head link; u32 fence; @@ -112,7 +111,7 @@ enum { CTB_SEND = 0, CTB_RECV = 1 }; enum { CTB_OWNER_HOST = 0 }; -static void ct_receive_tasklet_func(unsigned long data); +static void ct_receive_tasklet_func(struct tasklet_struct *t); static void ct_incoming_request_worker_func(struct work_struct *w); /** @@ -128,7 +127,7 @@ void intel_guc_ct_init_early(struct intel_guc_ct *ct) INIT_LIST_HEAD(&ct->requests.incoming); INIT_WORK(&ct->dead_ct_worker, ct_dead_ct_worker_func); INIT_WORK(&ct->requests.worker, ct_incoming_request_worker_func); - tasklet_init(&ct->receive_tasklet, ct_receive_tasklet_func, (unsigned long)ct); + tasklet_setup(&ct->receive_tasklet, ct_receive_tasklet_func); init_waitqueue_head(&ct->wq); } @@ -354,7 +353,6 @@ int intel_guc_ct_enable(struct intel_guc_ct *ct) if (unlikely(err)) goto err_out; - ct->requests.last_fence = 1; ct->enabled = true; ct->stall_time = KTIME_MAX; ct->dead_ct_reported = false; @@ -394,14 +392,24 @@ static u32 ct_get_next_fence(struct intel_guc_ct *ct) return ++ct->requests.last_fence; } -static void write_barrier(struct intel_guc_ct *ct) { +static void write_barrier(struct intel_guc_ct *ct) +{ struct intel_guc *guc = ct_to_guc(ct); struct intel_gt *gt = guc_to_gt(guc); if (i915_gem_object_is_lmem(guc->ct.vma->obj)) { GEM_BUG_ON(guc->send_regs.fw_domains); + /* + * This register is used by the i915 and GuC for MMIO based + * communication. Once we are in this code CTBs are the only + * method the i915 uses to communicate with the GuC so it is + * safe to write to this register (a value of 0 is NOP for MMIO + * communication). If we ever start mixing CTBs and MMIOs a new + * register will have to be chosen. + */ intel_uncore_write_fw(gt->uncore, GEN11_SOFT_SCRATCH(0), 0); } else { + /* wmb() sufficient for a barrier if in smem */ wmb(); } } @@ -417,21 +425,27 @@ static int ct_write(struct intel_guc_ct *ct, u32 size = ctb->size; u32 header; u32 hxg; + u32 type; u32 *cmds = ctb->cmds; unsigned int i; if (!I915_SELFTEST_ONLY(ct_to_guc(ct)->deadlock_expected)) { - if (unlikely(ctb->broken)) - return -EIO; - if (unlikely(desc->status)) goto corrupted; } + GEM_BUG_ON(tail > size); + #ifdef CONFIG_DRM_I915_DEBUG_GUC - if (unlikely((desc->tail | desc->head) >= size)) { - CT_ERROR(ct, "Invalid offsets head=%u tail=%u (size=%u)\n", - desc->head, desc->tail, size); + if (unlikely(tail != READ_ONCE(desc->tail))) { + CT_ERROR(ct, "Tail was modified %u != %u\n", + desc->tail, tail); + desc->status |= GUC_CTB_STATUS_MISMATCH; + goto corrupted; + } + if (unlikely(READ_ONCE(desc->head) >= size)) { + CT_ERROR(ct, "Invalid head offset %u >= %u)\n", + desc->head, size); desc->status |= GUC_CTB_STATUS_OVERFLOW; goto corrupted; } @@ -439,7 +453,8 @@ static int ct_write(struct intel_guc_ct *ct, /* * dw0: CT header (including fence) - * dw1: HXG header + * dw1: HXG header (including action code) + * dw2+: action data */ header = FIELD_PREP(GUC_CTB_MSG_0_FORMAT, GUC_CTB_FORMAT_HXG) | FIELD_PREP(GUC_CTB_MSG_0_NUM_DWORDS, len) | @@ -454,13 +469,11 @@ static int ct_write(struct intel_guc_ct *ct, } #endif - hxg = (flags & INTEL_GUC_SEND_NB) ? - (FIELD_PREP(GUC_HXG_MSG_0_TYPE, GUC_HXG_TYPE_EVENT) | - FIELD_PREP(GUC_HXG_EVENT_MSG_0_ACTION | - GUC_HXG_EVENT_MSG_0_DATA0, action[0])) : - (FIELD_PREP(GUC_HXG_MSG_0_TYPE, GUC_HXG_TYPE_REQUEST) | - FIELD_PREP(GUC_HXG_REQUEST_MSG_0_ACTION | - GUC_HXG_REQUEST_MSG_0_DATA0, action[0])); + type = (flags & INTEL_GUC_CT_SEND_NB) ? GUC_HXG_TYPE_EVENT : + GUC_HXG_TYPE_REQUEST; + hxg = FIELD_PREP(GUC_HXG_MSG_0_TYPE, type) | + FIELD_PREP(GUC_HXG_EVENT_MSG_0_ACTION | + GUC_HXG_EVENT_MSG_0_DATA0, action[0]); CT_DEBUG(ct, "writing (tail %u) %*ph %*ph %*ph\n", tail, 4, &header, 4, &hxg, 4 * (len - 1), &action[1]); @@ -479,14 +492,17 @@ static int ct_write(struct intel_guc_ct *ct, /* * make sure H2G buffer update and LRC tail update (if this triggering a - * submission) are visable before updating the descriptor tail + * submission) are visible before updating the descriptor tail */ write_barrier(ct); - /* now update descriptor */ + /* update local copies */ ctb->tail = tail; + GEM_BUG_ON(atomic_read(&ctb->space) < len + GUC_CTB_HDR_LEN); + atomic_sub(len + GUC_CTB_HDR_LEN, &ctb->space); + + /* now update descriptor */ WRITE_ONCE(desc->tail, tail); - atomic_sub(len + 1, &ctb->space); return 0; @@ -501,7 +517,7 @@ static int ct_write(struct intel_guc_ct *ct, queue_work(system_unbound_wq, &ct->dead_ct_worker); } ctb->broken = true; - return -EIO; + return -EPIPE; } /** @@ -522,25 +538,24 @@ static int ct_write(struct intel_guc_ct *ct, static int wait_for_ct_request_update(struct intel_guc_ct *ct, struct ct_request *req, u32 *status) { - long timeout; int err; /* * Fast commands should complete in less than 10us, so sample quickly * up to that length of time, then switch to a slower sleep-wait loop. - * No GuC command should ever take longer than 10ms. - * - * However, there might be other CT requests in flight before this one, - * so use @CONFIG_DRM_I915_HEARTBEAT_INTERVAL as backup timeout value. + * No GuC command should ever take longer than 10ms but many GuC + * commands can be inflight at time, so use a 1s timeout on the slower + * sleep-wait loop. */ - timeout = max(10, CONFIG_DRM_I915_HEARTBEAT_INTERVAL); - +#define GUC_CTB_RESPONSE_TIMEOUT_SHORT_MS 10 +#define GUC_CTB_RESPONSE_TIMEOUT_LONG_MS 1000 +#define done \ + (FIELD_GET(GUC_HXG_MSG_0_ORIGIN, READ_ONCE(req->status)) == \ + GUC_HXG_ORIGIN_GUC) intel_boost_fake_int_timer(ct_to_gt(ct), true); - -#define done (FIELD_GET(GUC_HXG_MSG_0_ORIGIN, READ_ONCE(req->status)) == GUC_HXG_ORIGIN_GUC) - err = wait_for_us(done, 10); + err = wait_for_us(done, GUC_CTB_RESPONSE_TIMEOUT_SHORT_MS); if (err) - err = wait_for(done, timeout * PRESI_GET_MULTIPLIER(1)); + err = wait_for(done, GUC_CTB_RESPONSE_TIMEOUT_LONG_MS * PRESI_GET_MULTIPLIER(1)); #undef done if (unlikely(err)) @@ -552,25 +567,32 @@ static int wait_for_ct_request_update(struct intel_guc_ct *ct, return err; } -#define PRESI_CTB_TIMEOUT \ - (((s64)MAX_US_STALL_CTB) * PRESI_GET_MULTIPLIER(MAX_US_STALL_CTB)) +#define GUC_CTB_TIMEOUT_MS 1500 static inline bool ct_deadlocked(struct intel_guc_ct *ct) { - bool ret = ktime_us_delta(ktime_get(), ct->stall_time) > - PRESI_CTB_TIMEOUT; + long timeout = GUC_CTB_TIMEOUT_MS * PRESI_GET_MULTIPLIER(1); + bool ret = ktime_ms_delta(ktime_get(), ct->stall_time) > timeout; if (unlikely(ret)) { + struct guc_ct_buffer_desc *send = ct->ctbs.send.desc; + struct guc_ct_buffer_desc *recv = ct->ctbs.send.desc; + /* * CI doesn't like error messages, demote to debug if deadlock was * intentionally hit. */ - if (I915_SELFTEST_ONLY(ct_to_guc(ct)->deadlock_expected)) - CT_DEBUG(ct, "CT deadlocked\n"); - else { - CT_ERROR(ct, "CT deadlocked\n"); + if (I915_SELFTEST_ONLY(ct_to_guc(ct)->deadlock_expected)) { + CT_DEBUG(ct, "Communication stalled for %lld ms, desc status=%#x,%#x\n", + ktime_ms_delta(ktime_get(), ct->stall_time), + send->status, recv->status); + } else { + CT_ERROR(ct, "Communication stalled for %lld ms, desc status=%#x,%#x\n", + ktime_ms_delta(ktime_get(), ct->stall_time), + send->status, recv->status); ct->dead_ct_reason |= 1 << CT_DEAD_DEADLOCK; queue_work(system_unbound_wq, &ct->dead_ct_worker); } + ct->ctbs.send.broken = true; } return ret; @@ -600,17 +622,18 @@ static inline void g2h_reserve_space(struct intel_guc_ct *ct, u32 g2h_len_dw) static inline bool h2g_has_room(struct intel_guc_ct *ct, u32 len_dw) { struct intel_guc_ct_buffer *ctb = &ct->ctbs.send; + struct guc_ct_buffer_desc *desc = ctb->desc; u32 head; u32 space; if (atomic_read(&ctb->space) >= len_dw) return true; - head = READ_ONCE(ctb->desc->head); + head = READ_ONCE(desc->head); if (unlikely(head > ctb->size)) { - CT_ERROR(ct, "Corrupted descriptor head=%u tail=%u size=%u\n", - ctb->desc->head, ctb->desc->tail, ctb->size); - ctb->desc->status |= GUC_CTB_STATUS_OVERFLOW; + CT_ERROR(ct, "Invalid head offset %u >= %u)\n", + head, ctb->size); + desc->status |= GUC_CTB_STATUS_OVERFLOW; ctb->broken = true; ct->dead_ct_reason |= 1 << CT_DEAD_H2G_HAS_ROOM; queue_work(system_unbound_wq, &ct->dead_ct_worker); @@ -632,7 +655,7 @@ static int has_room_nb(struct intel_guc_ct *ct, u32 h2g_dw, u32 g2h_dw) ct->stall_time = ktime_get(); if (unlikely(ct_deadlocked(ct))) - return -EIO; + return -EPIPE; else return -EBUSY; } @@ -641,7 +664,7 @@ static int has_room_nb(struct intel_guc_ct *ct, u32 h2g_dw, u32 g2h_dw) return 0; } -#define G2H_LEN_DW(flags) FIELD_GET(INTEL_GUC_SEND_G2H_DW_MASK, flags) +#define G2H_LEN_DW(flags) FIELD_GET(INTEL_GUC_CT_SEND_G2H_DW_MASK, flags) static int ct_send_nb(struct intel_guc_ct *ct, const u32 *action, u32 len, @@ -655,7 +678,7 @@ static int ct_send_nb(struct intel_guc_ct *ct, spin_lock_irqsave(&ctb->lock, spin_flags); - ret = has_room_nb(ct, len + 1, g2h_len_dw); + ret = has_room_nb(ct, len + GUC_CTB_HDR_LEN, g2h_len_dw); if (unlikely(ret)) goto out; @@ -683,6 +706,7 @@ static int ct_send(struct intel_guc_ct *ct, struct intel_guc_ct_buffer *ctb = &ct->ctbs.send; struct ct_request request; unsigned long flags; + unsigned int sleep_period_ms = 1; bool send_again; u32 fence; int err; @@ -703,15 +727,18 @@ static int ct_send(struct intel_guc_ct *ct, */ retry: spin_lock_irqsave(&ctb->lock, flags); - if (unlikely(!h2g_has_room(ct, len + 1))) { + if (unlikely(!h2g_has_room(ct, len + GUC_CTB_HDR_LEN))) { if (ct->stall_time == KTIME_MAX) ct->stall_time = ktime_get(); spin_unlock_irqrestore(&ctb->lock, flags); if (unlikely(ct_deadlocked(ct))) - return -EIO; + return -EPIPE; + + if (msleep_interruptible(sleep_period_ms)) + return -EINTR; + sleep_period_ms = sleep_period_ms << 1; - cond_resched(); goto retry; } @@ -800,7 +827,10 @@ int intel_guc_ct_send(struct intel_guc_ct *ct, const u32 *action, u32 len, return -ENODEV; } - if (flags & INTEL_GUC_SEND_NB) + if (unlikely(ct->ctbs.send.broken)) + return -EPIPE; + + if (flags & INTEL_GUC_CT_SEND_NB) return ct_send_nb(ct, action, len, flags); ret = ct_send(ct, action, len, response_buf, response_buf_size, &status); @@ -840,7 +870,7 @@ static int ct_read(struct intel_guc_ct *ct, struct ct_incoming_msg **msg) struct intel_guc_ct_buffer *ctb = &ct->ctbs.recv; struct guc_ct_buffer_desc *desc = ctb->desc; u32 head = ctb->head; - u32 tail = desc->tail; + u32 tail = READ_ONCE(desc->tail); u32 size = ctb->size; u32 *cmds = ctb->cmds; s32 available; @@ -854,21 +884,22 @@ static int ct_read(struct intel_guc_ct *ct, struct ct_incoming_msg **msg) if (unlikely(desc->status)) goto corrupted; + GEM_BUG_ON(head > size); + #ifdef CONFIG_DRM_I915_DEBUG_GUC - if (unlikely((desc->tail | desc->head) >= size)) { - CT_ERROR(ct, "Invalid offsets head=%u tail=%u (size=%u)\n", - head, tail, size); - desc->status |= GUC_CTB_STATUS_OVERFLOW; + if (unlikely(head != READ_ONCE(desc->head))) { + CT_ERROR(ct, "Head was modified %u != %u\n", + desc->head, head); + desc->status |= GUC_CTB_STATUS_MISMATCH; goto corrupted; } -#else - if (unlikely((tail | ctb->head) >= size)) { - CT_ERROR(ct, "Invalid offsets head=%u tail=%u (size=%u)\n", - head, tail, size); +#endif + if (unlikely(tail >= size)) { + CT_ERROR(ct, "Invalid tail offset %u >= %u)\n", + tail, size); desc->status |= GUC_CTB_STATUS_OVERFLOW; goto corrupted; } -#endif /* tail == head condition indicates empty */ available = tail - head; @@ -918,7 +949,9 @@ static int ct_read(struct intel_guc_ct *ct, struct ct_incoming_msg **msg) } CT_DEBUG(ct, "received %*ph\n", 4 * len, (*msg)->msg); + /* update local copies */ ctb->head = head; + /* now update descriptor */ WRITE_ONCE(desc->head, head); @@ -1280,9 +1313,9 @@ static void ct_try_receive_message(struct intel_guc_ct *ct) tasklet_hi_schedule(&ct->receive_tasklet); } -static void ct_receive_tasklet_func(unsigned long data) +static void ct_receive_tasklet_func(struct tasklet_struct *t) { - struct intel_guc_ct *ct = (struct intel_guc_ct *)data; + struct intel_guc_ct *ct = from_tasklet(ct, t, receive_tasklet); ct_try_receive_message(ct); } diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h index 1fbd89049edc..de3cde968108 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h @@ -110,6 +110,14 @@ static inline bool intel_guc_ct_enabled(const struct intel_guc_ct *ct) return ct->enabled; } +#define INTEL_GUC_CT_SEND_NB BIT(31) +#define INTEL_GUC_CT_SEND_G2H_DW_SHIFT 0 +#define INTEL_GUC_CT_SEND_G2H_DW_MASK (0xff << INTEL_GUC_CT_SEND_G2H_DW_SHIFT) +#define MAKE_SEND_FLAGS(len) ({ \ + typeof(len) len_ = (len); \ + GEM_BUG_ON(!FIELD_FIT(INTEL_GUC_CT_SEND_G2H_DW_MASK, len_)); \ + (FIELD_PREP(INTEL_GUC_CT_SEND_G2H_DW_MASK, len_) | INTEL_GUC_CT_SEND_NB); \ +}) int intel_guc_ct_send(struct intel_guc_ct *ct, const u32 *action, u32 len, u32 *response_buf, u32 response_buf_size, u32 flags); void intel_guc_ct_event_handler(struct intel_guc_ct *ct); diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c index cb808c86eed5..65bb61e0abea 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c @@ -1065,7 +1065,7 @@ static int gse_dequeue_one_context(struct guc_submit_engine *gse) ret = tasklet_register_context(gse, last); if (unlikely(ret == -EINPROGRESS)) goto blk_tasklet; - else if (unlikely(ret == -EIO)) + else if (unlikely(ret == -EPIPE)) goto deadlk; else if (unlikely(ret == -EBUSY)) goto schedule_tasklet; @@ -1089,7 +1089,7 @@ static int gse_dequeue_one_context(struct guc_submit_engine *gse) add_request: ret = gse_add_request(gse, last); - if (unlikely(ret == -EIO)) + if (unlikely(ret == -EPIPE)) goto deadlk; else if (ret == -EBUSY) goto schedule_tasklet; @@ -1896,7 +1896,7 @@ static int gse_bypass_tasklet_submit(struct guc_submit_engine *gse, ret = gse_add_request(gse, rq); } - if (unlikely(ret == -EIO)) + if (unlikely(ret == -EPIPE)) disable_submission(gse->guc); return ret; @@ -3616,7 +3616,7 @@ static int guc_request_alloc(struct i915_request *rq) if (context_needs_lrc_desc_pin(ce, !!ret)) { ret = guc_lrc_desc_pin(ce, true); if (unlikely(ret)) { /* unwind */ - if (ret == -EIO) { /* GPU will be reset */ + if (ret == -EPIPE) { /* GPU will be reset */ disable_submission(guc); goto out; } diff --git a/drivers/gpu/drm/i915/i915_params.c b/drivers/gpu/drm/i915/i915_params.c index b03637c8ca49..054fdfa486b7 100644 --- a/drivers/gpu/drm/i915/i915_params.c +++ b/drivers/gpu/drm/i915/i915_params.c @@ -179,7 +179,7 @@ i915_param_named_unsafe(edp_vswing, int, 0400, i915_param_named_unsafe(enable_guc, int, 0400, "Enable GuC load for GuC submission and/or HuC load. " "Required functionality can be selected using bitmask values. " - "(-1=auto, 0=disable [default], 1=GuC submission, 2=HuC load, " + "(-1=auto [default], 0=disable, 1=GuC submission, 2=HuC load, " "4=SR-IOV PF)"); i915_param_named_unsafe(guc_feature_flags, uint, 0400, -- git-pile 0.98-dev