6.12.17-rc1 review

@@ -41,9 +41,12 @@ static inline unsigned long arch_calc_vm
 	 * backed by tags-capable memory. The vm_flags may be overridden by a
 	 * filesystem supporting MTE (RAM-based).
 	 */
-	if (system_supports_mte() &&
-	    ((flags & MAP_ANONYMOUS) || shmem_file(file)))
-		return VM_MTE_ALLOWED;
+	if (system_supports_mte()) {
+		if ((flags & MAP_ANONYMOUS) && !(flags & MAP_HUGETLB))
+			return VM_MTE_ALLOWED;
+		if (shmem_file(file))
+			return VM_MTE_ALLOWED;
+	}
 
 	return 0;
 }

index 90aef2627ca27..64eaf74fbebc8 100644
@@ -545,8 +545,7 @@ static noinline int add_ra_bio_pages(struct inode *inode,
 		 * subpage::readers and to unlock the page.
 		 */
 		if (fs_info->sectorsize < PAGE_SIZE)
-			btrfs_subpage_start_reader(fs_info, folio, cur,
-						   add_size);
+			btrfs_folio_set_writer_lock(fs_info, folio, cur, add_size);
 		folio_put(folio);
 		cur += add_size;
 	}

index 8222ae6f29af5..5d6b3b812593d 100644
@@ -438,7 +438,7 @@ static void end_folio_read(struct folio *folio, bool uptodate, u64 start, u32 le
 	if (!btrfs_is_subpage(fs_info, folio->mapping))
 		folio_unlock(folio);
 	else
-		btrfs_subpage_end_reader(fs_info, folio, start, len);
+		btrfs_folio_end_writer_lock(fs_info, folio, start, len);
 }
 
 /*

@@ -495,7 +495,7 @@ static void begin_folio_read(struct btrfs_fs_info *fs_info, struct folio *folio)
 		return;
 
 	ASSERT(folio_test_private(folio));
-	btrfs_subpage_start_reader(fs_info, folio, folio_pos(folio), PAGE_SIZE);
+	btrfs_folio_set_writer_lock(fs_info, folio, folio_pos(folio), PAGE_SIZE);
 }
 
 /*

@@ -2507,12 +2507,6 @@ static bool folio_range_has_eb(struct btrfs_fs_info *fs_info, struct folio *foli
 		subpage = folio_get_private(folio);
 		if (atomic_read(&subpage->eb_refs))
 			return true;
-		/*
-		 * Even there is no eb refs here, we may still have
-		 * end_folio_read() call relying on page::private.
-		 */
-		if (atomic_read(&subpage->readers))
-			return true;
 	}
 	return false;
 }

76 lines not on this page · raw message

index 6289d6f65b87d..8488ea057b30b 100644
@@ -45,14 +45,6 @@ enum {
 struct btrfs_subpage {
 	/* Common members for both data and metadata pages */
 	spinlock_t lock;
-	/*
-	 * Both data and metadata needs to track how many readers are for the
-	 * page.
-	 * Data relies on @readers to unlock the page when last reader finished.
-	 * While metadata doesn't need page unlock, it needs to prevent
-	 * page::private get cleared before the last end_page_read().
-	 */
-	atomic_t readers;
 	union {
 		/*
 		 * Structures only used by metadata

@@ -95,11 +87,6 @@ void btrfs_free_subpage(struct btrfs_subpage *subpage);
 void btrfs_folio_inc_eb_refs(const struct btrfs_fs_info *fs_info, struct folio *folio);
 void btrfs_folio_dec_eb_refs(const struct btrfs_fs_info *fs_info, struct folio *folio);
 
-void btrfs_subpage_start_reader(const struct btrfs_fs_info *fs_info,
-				struct folio *folio, u64 start, u32 len);
-void btrfs_subpage_end_reader(const struct btrfs_fs_info *fs_info,
-			      struct folio *folio, u64 start, u32 len);
-
 void btrfs_folio_end_writer_lock(const struct btrfs_fs_info *fs_info,
 				 struct folio *folio, u64 start, u32 len);
 void btrfs_folio_set_writer_lock(const struct btrfs_fs_info *fs_info,

index 64eaf74fbebc8..40332ab62f101 100644
@@ -545,7 +545,7 @@ static noinline int add_ra_bio_pages(struct inode *inode,
 		 * subpage::readers and to unlock the page.
 		 */
 		if (fs_info->sectorsize < PAGE_SIZE)
-			btrfs_folio_set_writer_lock(fs_info, folio, cur, add_size);
+			btrfs_folio_set_lock(fs_info, folio, cur, add_size);
 		folio_put(folio);
 		cur += add_size;
 	}

index 5d6b3b812593d..5a1bde8cc8b64 100644
@@ -190,7 +190,7 @@ static void process_one_folio(struct btrfs_fs_info *fs_info,
 		btrfs_folio_clamp_clear_writeback(fs_info, folio, start, len);
 
 	if (folio != locked_folio && (page_ops & PAGE_UNLOCK))
-		btrfs_folio_end_writer_lock(fs_info, folio, start, len);
+		btrfs_folio_end_lock(fs_info, folio, start, len);
 }
 
 static void __process_folios_contig(struct address_space *mapping,

@@ -276,7 +276,7 @@ static noinline int lock_delalloc_folios(struct inode *inode,
 			range_start = max_t(u64, folio_pos(folio), start);
 			range_len = min_t(u64, folio_pos(folio) + folio_size(folio),
 					  end + 1) - range_start;
-			btrfs_folio_set_writer_lock(fs_info, folio, range_start, range_len);
+			btrfs_folio_set_lock(fs_info, folio, range_start, range_len);
 
 			processed_end = range_start + range_len - 1;
 		}

@@ -438,7 +438,7 @@ static void end_folio_read(struct folio *folio, bool uptodate, u64 start, u32 le
 	if (!btrfs_is_subpage(fs_info, folio->mapping))
 		folio_unlock(folio);
 	else
-		btrfs_folio_end_writer_lock(fs_info, folio, start, len);
+		btrfs_folio_end_lock(fs_info, folio, start, len);
 }
 
 /*

@@ -495,7 +495,7 @@ static void begin_folio_read(struct btrfs_fs_info *fs_info, struct folio *folio)
 		return;
 
 	ASSERT(folio_test_private(folio));
-	btrfs_folio_set_writer_lock(fs_info, folio, folio_pos(folio), PAGE_SIZE);
+	btrfs_folio_set_lock(fs_info, folio, folio_pos(folio), PAGE_SIZE);
 }
 
 /*

@@ -1187,7 +1187,7 @@ static noinline_for_stack int writepage_delalloc(struct btrfs_inode *inode,
 	for_each_set_bit(bit, &bio_ctrl->submit_bitmap, fs_info->sectors_per_page) {
 		u64 start = page_start + (bit << fs_info->sectorsize_bits);
 
-		btrfs_folio_set_writer_lock(fs_info, folio, start, fs_info->sectorsize);
+		btrfs_folio_set_lock(fs_info, folio, start, fs_info->sectorsize);
 	}
 
 	/* Lock all (subpage) delalloc ranges inside the folio first. */

@@ -1523,7 +1523,7 @@ static int extent_writepage(struct folio *folio, struct btrfs_bio_ctrl *bio_ctrl
 	 * Only unlock ranges that are submitted. As there can be some async
 	 * submitted ranges inside the folio.
 	 */
-	btrfs_folio_end_writer_lock_bitmap(fs_info, folio, bio_ctrl->submit_bitmap);
+	btrfs_folio_end_lock_bitmap(fs_info, folio, bio_ctrl->submit_bitmap);
 	ASSERT(ret <= 0);
 	return ret;
 }

@@ -2280,7 +2280,7 @@ void extent_write_locked_range(struct inode *inode, const struct folio *locked_f
 						       cur, cur_len, !ret);
 			mapping_set_error(mapping, ret);
 		}
-		btrfs_folio_end_writer_lock(fs_info, folio, cur, cur_len);
+		btrfs_folio_end_lock(fs_info, folio, cur, cur_len);
 		if (ret < 0)
 			found_error = true;
 next_page:

118 lines not on this page · raw message

index 8488ea057b30b..44fff1f4eac48 100644
@@ -54,8 +54,12 @@ struct btrfs_subpage {
 		 */
 		atomic_t eb_refs;
 
-		/* Structures only used by data */
-		atomic_t writers;
+		/*
+		 * Structures only used by data,
+		 *
+		 * How many sectors inside the page is locked.
+		 */
+		atomic_t nr_locked;
 	};
 	unsigned long bitmaps[];
 };

@@ -87,12 +91,12 @@ void btrfs_free_subpage(struct btrfs_subpage *subpage);
 void btrfs_folio_inc_eb_refs(const struct btrfs_fs_info *fs_info, struct folio *folio);
 void btrfs_folio_dec_eb_refs(const struct btrfs_fs_info *fs_info, struct folio *folio);
 
-void btrfs_folio_end_writer_lock(const struct btrfs_fs_info *fs_info,
-				 struct folio *folio, u64 start, u32 len);
-void btrfs_folio_set_writer_lock(const struct btrfs_fs_info *fs_info,
-				 struct folio *folio, u64 start, u32 len);
-void btrfs_folio_end_writer_lock_bitmap(const struct btrfs_fs_info *fs_info,
-					struct folio *folio, unsigned long bitmap);
+void btrfs_folio_end_lock(const struct btrfs_fs_info *fs_info,
+			  struct folio *folio, u64 start, u32 len);
+void btrfs_folio_set_lock(const struct btrfs_fs_info *fs_info,
+			  struct folio *folio, u64 start, u32 len);
+void btrfs_folio_end_lock_bitmap(const struct btrfs_fs_info *fs_info,
+				 struct folio *folio, unsigned long bitmap);
 /*
  * Template for subpage related operations.
  *

index 5a1bde8cc8b64..e8f882f949051 100644
@@ -1467,15 +1467,15 @@ static noinline_for_stack int extent_writepage_io(struct btrfs_inode *inode,
  */
 static int extent_writepage(struct folio *folio, struct btrfs_bio_ctrl *bio_ctrl)
 {
-	struct inode *inode = folio->mapping->host;
-	struct btrfs_fs_info *fs_info = inode_to_fs_info(inode);
+	struct btrfs_inode *inode = BTRFS_I(folio->mapping->host);
+	struct btrfs_fs_info *fs_info = inode->root->fs_info;
 	const u64 page_start = folio_pos(folio);
 	int ret;
 	size_t pg_offset;
-	loff_t i_size = i_size_read(inode);
+	loff_t i_size = i_size_read(&inode->vfs_inode);
 	unsigned long end_index = i_size >> PAGE_SHIFT;
 
-	trace_extent_writepage(folio, inode, bio_ctrl->wbc);
+	trace_extent_writepage(folio, &inode->vfs_inode, bio_ctrl->wbc);
 
 	WARN_ON(!folio_test_locked(folio));
 

@@ -1499,13 +1499,13 @@ static int extent_writepage(struct folio *folio, struct btrfs_bio_ctrl *bio_ctrl
 	if (ret < 0)
 		goto done;
 
-	ret = writepage_delalloc(BTRFS_I(inode), folio, bio_ctrl);
+	ret = writepage_delalloc(inode, folio, bio_ctrl);
 	if (ret == 1)
 		return 0;
 	if (ret)
 		goto done;
 
-	ret = extent_writepage_io(BTRFS_I(inode), folio, folio_pos(folio),
+	ret = extent_writepage_io(inode, folio, folio_pos(folio),
 				  PAGE_SIZE, bio_ctrl, i_size);
 	if (ret == 1)
 		return 0;

@@ -1514,7 +1514,7 @@ static int extent_writepage(struct folio *folio, struct btrfs_bio_ctrl *bio_ctrl
 
 done:
 	if (ret) {
-		btrfs_mark_ordered_io_finished(BTRFS_I(inode), folio,
+		btrfs_mark_ordered_io_finished(inode, folio,
 					       page_start, PAGE_SIZE, !ret);
 		mapping_set_error(folio->mapping, ret);
 	}

index e8f882f949051..15cbb2a865e5e 100644
@@ -1145,14 +1145,19 @@ static bool find_next_delalloc_bitmap(struct folio *folio,
 }
 
 /*
- * helper for extent_writepage(), doing all of the delayed allocation setup.
+ * Do all of the delayed allocation setup.
  *
- * This returns 1 if btrfs_run_delalloc_range function did all the work required
- * to write the page (copy into inline extent).  In this case the IO has
- * been started and the page is already unlocked.
+ * Return >0 if all the dirty blocks are submitted async (compression) or inlined.
+ * The @folio should no longer be touched (treat it as already unlocked).
  *
- * This returns 0 if all went well (page still locked)
- * This returns < 0 if there were errors (page still locked)
+ * Return 0 if there is still dirty block that needs to be submitted through
+ * extent_writepage_io().
+ * bio_ctrl->submit_bitmap will indicate which blocks of the folio should be
+ * submitted, and @folio is still kept locked.
+ *
+ * Return <0 if there is any error hit.
+ * Any allocated ordered extent range covering this folio will be marked
+ * finished (IOERR), and @folio is still kept locked.
  */
 static noinline_for_stack int writepage_delalloc(struct btrfs_inode *inode,
 						 struct folio *folio,

@@ -1170,6 +1175,16 @@ static noinline_for_stack int writepage_delalloc(struct btrfs_inode *inode,
 	 * last delalloc end.
 	 */
 	u64 last_delalloc_end = 0;
+	/*
+	 * The range end (exclusive) of the last successfully finished delalloc
+	 * range.
+	 * Any range covered by ordered extent must either be manually marked
+	 * finished (error handling), or has IO submitted (and finish the
+	 * ordered extent normally).
+	 *
+	 * This records the end of ordered extent cleanup if we hit an error.
+	 */
+	u64 last_finished_delalloc_end = page_start;
 	u64 delalloc_start = page_start;
 	u64 delalloc_end = page_end;
 	u64 delalloc_to_write = 0;

@@ -1238,11 +1253,19 @@ static noinline_for_stack int writepage_delalloc(struct btrfs_inode *inode,
 			found_len = last_delalloc_end + 1 - found_start;
 
 		if (ret >= 0) {
+			/*
+			 * Some delalloc range may be created by previous folios.
+			 * Thus we still need to clean up this range during error
+			 * handling.
+			 */
+			last_finished_delalloc_end = found_start;
 			/* No errors hit so far, run the current delalloc range. */
 			ret = btrfs_run_delalloc_range(inode, folio,
 						       found_start,
 						       found_start + found_len - 1,
 						       wbc);
+			if (ret >= 0)
+				last_finished_delalloc_end = found_start + found_len;
 		} else {
 			/*
 			 * We've hit an error during previous delalloc range,

@@ -1277,8 +1300,22 @@ static noinline_for_stack int writepage_delalloc(struct btrfs_inode *inode,
 
 		delalloc_start = found_start + found_len;
 	}
-	if (ret < 0)
+	/*
+	 * It's possible we had some ordered extents created before we hit
+	 * an error, cleanup non-async successfully created delalloc ranges.
+	 */
+	if (unlikely(ret < 0)) {
+		unsigned int bitmap_size = min(
+				(last_finished_delalloc_end - page_start) >>
+				fs_info->sectorsize_bits,
+				fs_info->sectors_per_page);
+
+		for_each_set_bit(bit, &bio_ctrl->submit_bitmap, bitmap_size)
+			btrfs_mark_ordered_io_finished(inode, folio,
+				page_start + (bit << fs_info->sectorsize_bits),
+				fs_info->sectorsize, false);
 		return ret;
+	}
 out:
 	if (last_delalloc_end)
 		delalloc_end = last_delalloc_end;

@@ -1512,13 +1549,13 @@ static int extent_writepage(struct folio *folio, struct btrfs_bio_ctrl *bio_ctrl
 
 	bio_ctrl->wbc->nr_to_write--;
 
-done:
-	if (ret) {
+	if (ret)
 		btrfs_mark_ordered_io_finished(inode, folio,
 					       page_start, PAGE_SIZE, !ret);
-		mapping_set_error(folio->mapping, ret);
-	}
 
+done:
+	if (ret < 0)
+		mapping_set_error(folio->mapping, ret);
 	/*
 	 * Only unlock ranges that are submitted. As there can be some async
 	 * submitted ranges inside the folio.

index f7e7d864f4144..5b842276573e8 100644
@@ -2419,8 +2419,7 @@ int btrfs_run_delalloc_range(struct btrfs_inode *inode, struct folio *locked_fol
 
 out:
 	if (ret < 0)
-		btrfs_cleanup_ordered_extents(inode, locked_folio, start,
-					      end - start + 1);
+		btrfs_cleanup_ordered_extents(inode, NULL, start, end - start + 1);
 	return ret;
 }
 

index 15cbb2a865e5e..660a5b9c08e9e 100644
@@ -1431,6 +1431,7 @@ static noinline_for_stack int extent_writepage_io(struct btrfs_inode *inode,
 	struct btrfs_fs_info *fs_info = inode->root->fs_info;
 	unsigned long range_bitmap = 0;
 	bool submitted_io = false;
+	bool error = false;
 	const u64 folio_start = folio_pos(folio);
 	u64 cur;
 	int bit;

@@ -1473,11 +1474,26 @@ static noinline_for_stack int extent_writepage_io(struct btrfs_inode *inode,
 			break;
 		}
 		ret = submit_one_sector(inode, folio, cur, bio_ctrl, i_size);
-		if (ret < 0)
-			goto out;
+		if (unlikely(ret < 0)) {
+			/*
+			 * bio_ctrl may contain a bio crossing several folios.
+			 * Submit it immediately so that the bio has a chance
+			 * to finish normally, other than marked as error.
+			 */
+			submit_one_bio(bio_ctrl);
+			/*
+			 * Failed to grab the extent map which should be very rare.
+			 * Since there is no bio submitted to finish the ordered
+			 * extent, we have to manually finish this sector.
+			 */
+			btrfs_mark_ordered_io_finished(inode, folio, cur,
+						       fs_info->sectorsize, false);
+			error = true;
+			continue;
+		}
 		submitted_io = true;
 	}
-out:
+
 	/*
 	 * If we didn't submitted any sector (>= i_size), folio dirty get
 	 * cleared but PAGECACHE_TAG_DIRTY is not cleared (only cleared

@@ -1485,8 +1501,11 @@ static noinline_for_stack int extent_writepage_io(struct btrfs_inode *inode,
 	 *
 	 * Here we set writeback and clear for the range. If the full folio
 	 * is no longer dirty then we clear the PAGECACHE_TAG_DIRTY tag.
+	 *
+	 * If we hit any error, the corresponding sector will still be dirty
+	 * thus no need to clear PAGECACHE_TAG_DIRTY.
 	 */
-	if (!submitted_io) {
+	if (!submitted_io && !error) {
 		btrfs_folio_set_writeback(fs_info, folio, start, len);
 		btrfs_folio_clear_writeback(fs_info, folio, start, len);
 	}

@@ -1506,7 +1525,6 @@ static int extent_writepage(struct folio *folio, struct btrfs_bio_ctrl *bio_ctrl
 {
 	struct btrfs_inode *inode = BTRFS_I(folio->mapping->host);
 	struct btrfs_fs_info *fs_info = inode->root->fs_info;
-	const u64 page_start = folio_pos(folio);
 	int ret;
 	size_t pg_offset;
 	loff_t i_size = i_size_read(&inode->vfs_inode);

@@ -1549,10 +1567,6 @@ static int extent_writepage(struct folio *folio, struct btrfs_bio_ctrl *bio_ctrl
 
 	bio_ctrl->wbc->nr_to_write--;
 
-	if (ret)
-		btrfs_mark_ordered_io_finished(inode, folio,
-					       page_start, PAGE_SIZE, !ret);
-
 done:
 	if (ret < 0)
 		mapping_set_error(folio->mapping, ret);

@@ -2312,11 +2326,8 @@ void extent_write_locked_range(struct inode *inode, const struct folio *locked_f
 		if (ret == 1)
 			goto next_page;
 
-		if (ret) {
-			btrfs_mark_ordered_io_finished(BTRFS_I(inode), folio,
-						       cur, cur_len, !ret);
+		if (ret)
 			mapping_set_error(mapping, ret);
-		}
 		btrfs_folio_end_lock(fs_info, folio, cur, cur_len);
 		if (ret < 0)
 			found_error = true;

index 1b8ef9856422a..441abc4f4afd9 100644
@@ -117,7 +117,6 @@ bool kvm_irq_delivery_to_apic_fast(struct kvm *kvm, struct kvm_lapic *src,
 		struct kvm_lapic_irq *irq, int *r, struct dest_map *dest_map);
 void kvm_apic_send_ipi(struct kvm_lapic *apic, u32 icr_low, u32 icr_high);
 
-u64 kvm_get_apic_base(struct kvm_vcpu *vcpu);
 int kvm_set_apic_base(struct kvm_vcpu *vcpu, struct msr_data *msr_info);
 int kvm_apic_get_state(struct kvm_vcpu *vcpu, struct kvm_lapic_state *s);
 int kvm_apic_set_state(struct kvm_vcpu *vcpu, struct kvm_lapic_state *s);

index 0846e3af5f6c5..36bedf235340c 100644
@@ -667,14 +667,9 @@ static void drop_user_return_notifiers(void)
 		kvm_on_user_return(&msrs->urn);
 }
 
-u64 kvm_get_apic_base(struct kvm_vcpu *vcpu)
-{
-	return vcpu->arch.apic_base;
-}
-
 enum lapic_mode kvm_get_apic_mode(struct kvm_vcpu *vcpu)
 {
-	return kvm_apic_mode(kvm_get_apic_base(vcpu));
+	return kvm_apic_mode(vcpu->arch.apic_base);
 }
 EXPORT_SYMBOL_GPL(kvm_get_apic_mode);
 

@@ -4314,7 +4309,7 @@ int kvm_get_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 		msr_info->data = 1 << 24;
 		break;
 	case MSR_IA32_APICBASE:
-		msr_info->data = kvm_get_apic_base(vcpu);
+		msr_info->data = vcpu->arch.apic_base;
 		break;
 	case APIC_BASE_MSR ... APIC_BASE_MSR + 0xff:
 		return kvm_x2apic_msr_read(vcpu, msr_info->index, &msr_info->data);

@@ -10159,7 +10154,7 @@ static void post_kvm_run_save(struct kvm_vcpu *vcpu)
 
 	kvm_run->if_flag = kvm_x86_call(get_if_flag)(vcpu);
 	kvm_run->cr8 = kvm_get_cr8(vcpu);
-	kvm_run->apic_base = kvm_get_apic_base(vcpu);
+	kvm_run->apic_base = vcpu->arch.apic_base;
 
 	kvm_run->ready_for_interrupt_injection =
 		pic_in_kernel(vcpu->kvm) ||

@@ -11718,7 +11713,7 @@ static void __get_sregs_common(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
 	sregs->cr4 = kvm_read_cr4(vcpu);
 	sregs->cr8 = kvm_get_cr8(vcpu);
 	sregs->efer = vcpu->arch.efer;
-	sregs->apic_base = kvm_get_apic_base(vcpu);
+	sregs->apic_base = vcpu->arch.apic_base;
 }
 
 static void __get_sregs(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)

index 441abc4f4afd9..fc4bd36d44cfc 100644
@@ -120,7 +120,6 @@ void kvm_apic_send_ipi(struct kvm_lapic *apic, u32 icr_low, u32 icr_high);
 int kvm_set_apic_base(struct kvm_vcpu *vcpu, struct msr_data *msr_info);
 int kvm_apic_get_state(struct kvm_vcpu *vcpu, struct kvm_lapic_state *s);
 int kvm_apic_set_state(struct kvm_vcpu *vcpu, struct kvm_lapic_state *s);
-enum lapic_mode kvm_get_apic_mode(struct kvm_vcpu *vcpu);
 int kvm_lapic_find_highest_irr(struct kvm_vcpu *vcpu);
 
 u64 kvm_get_lapic_tscdeadline_msr(struct kvm_vcpu *vcpu);

@@ -270,6 +269,11 @@ static inline enum lapic_mode kvm_apic_mode(u64 apic_base)
 	return apic_base & (MSR_IA32_APICBASE_ENABLE | X2APIC_ENABLE);
 }
 
+static inline enum lapic_mode kvm_get_apic_mode(struct kvm_vcpu *vcpu)
+{
+	return kvm_apic_mode(vcpu->arch.apic_base);
+}
+
 static inline u8 kvm_xapic_id(struct kvm_lapic *apic)
 {
 	return kvm_lapic_get_reg(apic, APIC_ID) >> 24;

index 36bedf235340c..b67a2f46e40b0 100644
@@ -667,12 +667,6 @@ static void drop_user_return_notifiers(void)
 		kvm_on_user_return(&msrs->urn);
 }
 
-enum lapic_mode kvm_get_apic_mode(struct kvm_vcpu *vcpu)
-{
-	return kvm_apic_mode(vcpu->arch.apic_base);
-}
-EXPORT_SYMBOL_GPL(kvm_get_apic_mode);
-
 int kvm_set_apic_base(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 {
 	enum lapic_mode old_mode = kvm_get_apic_mode(vcpu);

index 375bbb9600d3c..1a8148dec4afe 100644
@@ -816,6 +816,17 @@ static inline void apic_clear_isr(int vec, struct kvm_lapic *apic)
 	}
 }
 
+void kvm_apic_update_hwapic_isr(struct kvm_vcpu *vcpu)
+{
+	struct kvm_lapic *apic = vcpu->arch.apic;
+
+	if (WARN_ON_ONCE(!lapic_in_kernel(vcpu)) || !apic->apicv_active)
+		return;
+
+	kvm_x86_call(hwapic_isr_update)(vcpu, apic_find_highest_isr(apic));
+}
+EXPORT_SYMBOL_GPL(kvm_apic_update_hwapic_isr);
+
 int kvm_lapic_find_highest_irr(struct kvm_vcpu *vcpu)
 {
 	/* This may race with setting of irr in __apic_accept_irq() and

index fc4bd36d44cfc..3aa599db77968 100644
@@ -120,6 +120,7 @@ void kvm_apic_send_ipi(struct kvm_lapic *apic, u32 icr_low, u32 icr_high);
 int kvm_set_apic_base(struct kvm_vcpu *vcpu, struct msr_data *msr_info);
 int kvm_apic_get_state(struct kvm_vcpu *vcpu, struct kvm_lapic_state *s);
 int kvm_apic_set_state(struct kvm_vcpu *vcpu, struct kvm_lapic_state *s);
+void kvm_apic_update_hwapic_isr(struct kvm_vcpu *vcpu);
 int kvm_lapic_find_highest_irr(struct kvm_vcpu *vcpu);
 
 u64 kvm_get_lapic_tscdeadline_msr(struct kvm_vcpu *vcpu);

index 931a7361c30f2..22bee8a711442 100644
@@ -5043,6 +5043,11 @@ void nested_vmx_vmexit(struct kvm_vcpu *vcpu, u32 vm_exit_reason,
 		kvm_make_request(KVM_REQ_APICV_UPDATE, vcpu);
 	}
 
+	if (vmx->nested.update_vmcs01_hwapic_isr) {
+		vmx->nested.update_vmcs01_hwapic_isr = false;
+		kvm_apic_update_hwapic_isr(vcpu);
+	}
+
 	if ((vm_exit_reason != -1) &&
 	    (enable_shadow_vmcs || nested_vmx_is_evmptr12_valid(vmx)))
 		vmx->nested.need_vmcs12_to_shadow_sync = true;

index f06d443ec3c68..1af30e3472cdd 100644
@@ -6858,6 +6858,27 @@ void vmx_hwapic_isr_update(struct kvm_vcpu *vcpu, int max_isr)
 	u16 status;
 	u8 old;
 
+	/*
+	 * If L2 is active, defer the SVI update until vmcs01 is loaded, as SVI
+	 * is only relevant for if and only if Virtual Interrupt Delivery is
+	 * enabled in vmcs12, and if VID is enabled then L2 EOIs affect L2's
+	 * vAPIC, not L1's vAPIC.  KVM must update vmcs01 on the next nested
+	 * VM-Exit, otherwise L1 with run with a stale SVI.
+	 */
+	if (is_guest_mode(vcpu)) {
+		/*
+		 * KVM is supposed to forward intercepted L2 EOIs to L1 if VID
+		 * is enabled in vmcs12; as above, the EOIs affect L2's vAPIC.
+		 * Note, userspace can stuff state while L2 is active; assert
+		 * that VID is disabled if and only if the vCPU is in KVM_RUN
+		 * to avoid false positives if userspace is setting APIC state.
+		 */
+		WARN_ON_ONCE(vcpu->wants_to_run &&
+			     nested_cpu_has_vid(get_vmcs12(vcpu)));
+		to_vmx(vcpu)->nested.update_vmcs01_hwapic_isr = true;
+		return;
+	}
+
 	if (max_isr == -1)
 		max_isr = 0;
 

index 2325f773a20be..41bf59bbc6426 100644
@@ -176,6 +176,7 @@ struct nested_vmx {
 	bool reload_vmcs01_apic_access_page;
 	bool update_vmcs01_cpu_dirty_logging;
 	bool update_vmcs01_apicv_status;
+	bool update_vmcs01_hwapic_isr;
 
 	/*
 	 * Enlightened VMCS has been enabled. It does not mean that L1 has to

index e93df3d6222e6..bc123f1884da3 100644
@@ -50,12 +50,13 @@
 #include "link.h"
 
 #include "logger_types.h"
+
+
+#include "yellow_carp_offset.h"
 #undef DC_LOGGER
 #define DC_LOGGER \
 	clk_mgr->base.base.ctx->logger
 
-#include "yellow_carp_offset.h"
-
 #define regCLK1_CLK_PLL_REQ			0x0237
 #define regCLK1_CLK_PLL_REQ_BASE_IDX		0
 

index 29eff386505ab..91d872d6d392b 100644
@@ -53,9 +53,6 @@
 
 
 #include "logger_types.h"
-#undef DC_LOGGER
-#define DC_LOGGER \
-	clk_mgr->base.base.ctx->logger
 
 
 #define MAX_INSTANCE                                        7

@@ -77,6 +74,9 @@ static const struct IP_BASE CLK_BASE = { { { { 0x00016C00, 0x02401800, 0, 0, 0,
 					{ { 0x0001B200, 0x0242DC00, 0, 0, 0, 0, 0, 0 } },
 					{ { 0x0001B400, 0x0242E000, 0, 0, 0, 0, 0, 0 } } } };
 
+#undef DC_LOGGER
+#define DC_LOGGER \
+	clk_mgr->base.base.ctx->logger
 #define regCLK1_CLK_PLL_REQ			0x0237
 #define regCLK1_CLK_PLL_REQ_BASE_IDX		0
 

index 3bd0d46c17010..bbdc39ae57b9d 100644
@@ -55,6 +55,7 @@
 #define DC_LOGGER \
 	clk_mgr->base.base.ctx->logger
 
+
 #define regCLK1_CLK_PLL_REQ			0x0237
 #define regCLK1_CLK_PLL_REQ_BASE_IDX		0
 

index d78c8ec4de79e..885e749cdc6e9 100644
@@ -51,9 +51,10 @@
 #include "dc_dmub_srv.h"
 #include "gpio_service_interface.h"
 
+#define DC_TRACE_LEVEL_MESSAGE(...) /* do nothing */
+
 #define DC_LOGGER \
 	link->ctx->logger
-#define DC_TRACE_LEVEL_MESSAGE(...) /* do nothing */
 
 #ifndef MAX
 #define MAX(X, Y) ((X) > (Y) ? (X) : (Y))

index 6fc00d63b2857..3328529774cb7 100644
@@ -567,11 +567,10 @@ static __poll_t xe_oa_poll(struct file *file, poll_table *wait)
 	return ret;
 }
 
-static int xe_oa_submit_bb(struct xe_oa_stream *stream, struct xe_bb *bb)
+static struct dma_fence *xe_oa_submit_bb(struct xe_oa_stream *stream, struct xe_bb *bb)
 {
 	struct xe_sched_job *job;
 	struct dma_fence *fence;
-	long timeout;
 	int err = 0;
 
 	/* Kernel configuration is issued on stream->k_exec_q, not stream->exec_q */

@@ -585,14 +584,9 @@ static int xe_oa_submit_bb(struct xe_oa_stream *stream, struct xe_bb *bb)
 	fence = dma_fence_get(&job->drm.s_fence->finished);
 	xe_sched_job_push(job);
 
-	timeout = dma_fence_wait_timeout(fence, false, HZ);
-	dma_fence_put(fence);
-	if (timeout < 0)
-		err = timeout;
-	else if (!timeout)
-		err = -ETIME;
+	return fence;
 exit:
-	return err;
+	return ERR_PTR(err);
 }
 
 static void write_cs_mi_lri(struct xe_bb *bb, const struct xe_oa_reg *reg_data, u32 n_regs)

@@ -656,6 +650,7 @@ static void xe_oa_store_flex(struct xe_oa_stream *stream, struct xe_lrc *lrc,
 static int xe_oa_modify_ctx_image(struct xe_oa_stream *stream, struct xe_lrc *lrc,
 				  const struct flex *flex, u32 count)
 {
+	struct dma_fence *fence;
 	struct xe_bb *bb;
 	int err;
 

@@ -667,7 +662,16 @@ static int xe_oa_modify_ctx_image(struct xe_oa_stream *stream, struct xe_lrc *lr
 
 	xe_oa_store_flex(stream, lrc, bb, flex, count);
 
-	err = xe_oa_submit_bb(stream, bb);
+	fence = xe_oa_submit_bb(stream, bb);
+	if (IS_ERR(fence)) {
+		err = PTR_ERR(fence);
+		goto free_bb;
+	}
+	xe_bb_free(bb, fence);
+	dma_fence_put(fence);
+
+	return 0;
+free_bb:
 	xe_bb_free(bb, NULL);
 exit:
 	return err;

@@ -675,6 +679,7 @@ static int xe_oa_modify_ctx_image(struct xe_oa_stream *stream, struct xe_lrc *lr
 
 static int xe_oa_load_with_lri(struct xe_oa_stream *stream, struct xe_oa_reg *reg_lri)
 {
+	struct dma_fence *fence;
 	struct xe_bb *bb;
 	int err;
 

@@ -686,7 +691,16 @@ static int xe_oa_load_with_lri(struct xe_oa_stream *stream, struct xe_oa_reg *re
 
 	write_cs_mi_lri(bb, reg_lri, 1);
 
-	err = xe_oa_submit_bb(stream, bb);
+	fence = xe_oa_submit_bb(stream, bb);
+	if (IS_ERR(fence)) {
+		err = PTR_ERR(fence);
+		goto free_bb;
+	}
+	xe_bb_free(bb, fence);
+	dma_fence_put(fence);
+
+	return 0;
+free_bb:
 	xe_bb_free(bb, NULL);
 exit:
 	return err;

@@ -914,15 +928,32 @@ static int xe_oa_emit_oa_config(struct xe_oa_stream *stream, struct xe_oa_config
 {
 #define NOA_PROGRAM_ADDITIONAL_DELAY_US 500
 	struct xe_oa_config_bo *oa_bo;
-	int err, us = NOA_PROGRAM_ADDITIONAL_DELAY_US;
+	int err = 0, us = NOA_PROGRAM_ADDITIONAL_DELAY_US;
+	struct dma_fence *fence;
+	long timeout;
 
+	/* Emit OA configuration batch */
 	oa_bo = xe_oa_alloc_config_buffer(stream, config);
 	if (IS_ERR(oa_bo)) {
 		err = PTR_ERR(oa_bo);
 		goto exit;
 	}
 
-	err = xe_oa_submit_bb(stream, oa_bo->bb);
+	fence = xe_oa_submit_bb(stream, oa_bo->bb);
+	if (IS_ERR(fence)) {
+		err = PTR_ERR(fence);
+		goto exit;
+	}
+
+	/* Wait till all previous batches have executed */
+	timeout = dma_fence_wait_timeout(fence, false, 5 * HZ);
+	dma_fence_put(fence);
+	if (timeout < 0)
+		err = timeout;
+	else if (!timeout)
+		err = -ETIME;
+	if (err)
+		drm_dbg(&stream->oa->xe->drm, "dma_fence_wait_timeout err %d\n", err);
 
 	/* Additional empirical delay needed for NOA programming after registers are written */
 	usleep_range(us, 2 * us);

index 3328529774cb7..20d279ed3c382 100644
@@ -36,6 +36,7 @@
 #include "xe_pm.h"
 #include "xe_sched_job.h"
 #include "xe_sriov.h"
+#include "xe_sync.h"
 
 #define DEFAULT_POLL_FREQUENCY_HZ 200
 #define DEFAULT_POLL_PERIOD_NS (NSEC_PER_SEC / DEFAULT_POLL_FREQUENCY_HZ)

@@ -70,6 +71,7 @@ struct flex {
 };
 
 struct xe_oa_open_param {
+	struct xe_file *xef;
 	u32 oa_unit_id;
 	bool sample;
 	u32 metric_set;

@@ -81,6 +83,9 @@ struct xe_oa_open_param {
 	struct xe_exec_queue *exec_q;
 	struct xe_hw_engine *hwe;
 	bool no_preempt;
+	struct drm_xe_sync __user *syncs_user;
+	int num_syncs;
+	struct xe_sync_entry *syncs;
 };
 
 struct xe_oa_config_bo {

@@ -1393,6 +1398,9 @@ static int xe_oa_stream_init(struct xe_oa_stream *stream,
 	stream->period_exponent = param->period_exponent;
 	stream->no_preempt = param->no_preempt;
 
+	stream->num_syncs = param->num_syncs;
+	stream->syncs = param->syncs;
+
 	/*
 	 * For Xe2+, when overrun mode is enabled, there are no partial reports at the end
 	 * of buffer, making the OA buffer effectively a non-power-of-2 size circular

@@ -1743,6 +1751,20 @@ static int xe_oa_set_no_preempt(struct xe_oa *oa, u64 value,
 	return 0;
 }
 
+static int xe_oa_set_prop_num_syncs(struct xe_oa *oa, u64 value,
+				    struct xe_oa_open_param *param)
+{
+	param->num_syncs = value;
+	return 0;
+}
+
+static int xe_oa_set_prop_syncs_user(struct xe_oa *oa, u64 value,
+				     struct xe_oa_open_param *param)
+{
+	param->syncs_user = u64_to_user_ptr(value);
+	return 0;
+}
+
 typedef int (*xe_oa_set_property_fn)(struct xe_oa *oa, u64 value,
 				     struct xe_oa_open_param *param);
 static const xe_oa_set_property_fn xe_oa_set_property_funcs[] = {

@@ -1755,6 +1777,8 @@ static const xe_oa_set_property_fn xe_oa_set_property_funcs[] = {
 	[DRM_XE_OA_PROPERTY_EXEC_QUEUE_ID] = xe_oa_set_prop_exec_queue_id,
 	[DRM_XE_OA_PROPERTY_OA_ENGINE_INSTANCE] = xe_oa_set_prop_engine_instance,
 	[DRM_XE_OA_PROPERTY_NO_PREEMPT] = xe_oa_set_no_preempt,
+	[DRM_XE_OA_PROPERTY_NUM_SYNCS] = xe_oa_set_prop_num_syncs,
+	[DRM_XE_OA_PROPERTY_SYNCS] = xe_oa_set_prop_syncs_user,
 };
 
 static int xe_oa_user_ext_set_property(struct xe_oa *oa, u64 extension,

@@ -1814,6 +1838,49 @@ static int xe_oa_user_extensions(struct xe_oa *oa, u64 extension, int ext_number
 	return 0;
 }
 
+static int xe_oa_parse_syncs(struct xe_oa *oa, struct xe_oa_open_param *param)
+{
+	int ret, num_syncs, num_ufence = 0;
+
+	if (param->num_syncs && !param->syncs_user) {
+		drm_dbg(&oa->xe->drm, "num_syncs specified without sync array\n");
+		ret = -EINVAL;
+		goto exit;
+	}
+
+	if (param->num_syncs) {
+		param->syncs = kcalloc(param->num_syncs, sizeof(*param->syncs), GFP_KERNEL);
+		if (!param->syncs) {
+			ret = -ENOMEM;
+			goto exit;
+		}
+	}
+
+	for (num_syncs = 0; num_syncs < param->num_syncs; num_syncs++) {
+		ret = xe_sync_entry_parse(oa->xe, param->xef, &param->syncs[num_syncs],
+					  &param->syncs_user[num_syncs], 0);
+		if (ret)
+			goto err_syncs;
+
+		if (xe_sync_is_ufence(&param->syncs[num_syncs]))
+			num_ufence++;
+	}
+
+	if (XE_IOCTL_DBG(oa->xe, num_ufence > 1)) {
+		ret = -EINVAL;
+		goto err_syncs;
+	}
+
+	return 0;
+
+err_syncs:
+	while (num_syncs--)
+		xe_sync_entry_cleanup(&param->syncs[num_syncs]);
+	kfree(param->syncs);
+exit:
+	return ret;
+}
+
 /**
  * xe_oa_stream_open_ioctl - Opens an OA stream
  * @dev: @drm_device

@@ -1839,6 +1906,7 @@ int xe_oa_stream_open_ioctl(struct drm_device *dev, u64 data, struct drm_file *f
 		return -ENODEV;
 	}
 
+	param.xef = xef;
 	ret = xe_oa_user_extensions(oa, data, 0, &param);
 	if (ret)
 		return ret;

@@ -1907,11 +1975,24 @@ int xe_oa_stream_open_ioctl(struct drm_device *dev, u64 data, struct drm_file *f
 		drm_dbg(&oa->xe->drm, "Using periodic sampling freq %lld Hz\n", oa_freq_hz);
 	}
 
+	ret = xe_oa_parse_syncs(oa, &param);
+	if (ret)
+		goto err_exec_q;
+
 	mutex_lock(&param.hwe->gt->oa.gt_lock);
 	ret = xe_oa_stream_open_ioctl_locked(oa, &param);
 	mutex_unlock(&param.hwe->gt->oa.gt_lock);
+	if (ret < 0)
+		goto err_sync_cleanup;
+
+	return ret;
+
+err_sync_cleanup:
+	while (param.num_syncs--)
+		xe_sync_entry_cleanup(&param.syncs[param.num_syncs]);
+	kfree(param.syncs);
 err_exec_q:
-	if (ret < 0 && param.exec_q)
+	if (param.exec_q)
 		xe_exec_queue_put(param.exec_q);
 	return ret;
 }

index 8862eca73fbe3..99f4b2d4bdcf6 100644
@@ -238,5 +238,11 @@ struct xe_oa_stream {
 
 	/** @no_preempt: Whether preemption and timeslicing is disabled for stream exec_q */
 	u32 no_preempt;
+
+	/** @num_syncs: size of @syncs array */
+	u32 num_syncs;
+
+	/** @syncs: syncs to wait on and to signal */
+	struct xe_sync_entry *syncs;
 };
 #endif

index 1c96375bd7df7..6fec5d1a1eb44 100644
@@ -679,7 +679,7 @@ static int query_oa_units(struct xe_device *xe,
 			du->oa_unit_id = u->oa_unit_id;
 			du->oa_unit_type = u->type;
 			du->oa_timestamp_freq = xe_oa_timestamp_frequency(gt);
-			du->capabilities = DRM_XE_OA_CAPS_BASE;
+			du->capabilities = DRM_XE_OA_CAPS_BASE | DRM_XE_OA_CAPS_SYNCS;
 
 			j = 0;
 			for_each_hw_engine(hwe, gt, hwe_id) {

index c4182e95a6195..4a8a4a63e99ca 100644
@@ -1485,6 +1485,7 @@ struct drm_xe_oa_unit {
 	/** @capabilities: OA capabilities bit-mask */
 	__u64 capabilities;
 #define DRM_XE_OA_CAPS_BASE		(1 << 0)
+#define DRM_XE_OA_CAPS_SYNCS		(1 << 1)
 
 	/** @oa_timestamp_freq: OA timestamp freq */
 	__u64 oa_timestamp_freq;

@@ -1634,6 +1635,22 @@ enum drm_xe_oa_property_id {
 	 * to be disabled for the stream exec queue.
 	 */
 	DRM_XE_OA_PROPERTY_NO_PREEMPT,
+
+	/**
+	 * @DRM_XE_OA_PROPERTY_NUM_SYNCS: Number of syncs in the sync array
+	 * specified in @DRM_XE_OA_PROPERTY_SYNCS
+	 */
+	DRM_XE_OA_PROPERTY_NUM_SYNCS,
+
+	/**
+	 * @DRM_XE_OA_PROPERTY_SYNCS: Pointer to struct @drm_xe_sync array
+	 * with array size specified via @DRM_XE_OA_PROPERTY_NUM_SYNCS. OA
+	 * configuration will wait till input fences signal. Output fences
+	 * will signal after the new OA configuration takes effect. For
+	 * @DRM_XE_SYNC_TYPE_USER_FENCE, @addr is a user pointer, similar
+	 * to the VM bind case.
+	 */
+	DRM_XE_OA_PROPERTY_SYNCS,
 };
 
 /**

index 20d279ed3c382..1bfc4b58b5c17 100644
@@ -42,6 +42,11 @@
 #define DEFAULT_POLL_PERIOD_NS (NSEC_PER_SEC / DEFAULT_POLL_FREQUENCY_HZ)
 #define XE_OA_UNIT_INVALID U32_MAX
 
+enum xe_oa_submit_deps {
+	XE_OA_SUBMIT_NO_DEPS,
+	XE_OA_SUBMIT_ADD_DEPS,
+};
+
 struct xe_oa_reg {
 	struct xe_reg addr;
 	u32 value;

@@ -572,7 +577,8 @@ static __poll_t xe_oa_poll(struct file *file, poll_table *wait)
 	return ret;
 }
 
-static struct dma_fence *xe_oa_submit_bb(struct xe_oa_stream *stream, struct xe_bb *bb)
+static struct dma_fence *xe_oa_submit_bb(struct xe_oa_stream *stream, enum xe_oa_submit_deps deps,
+					 struct xe_bb *bb)
 {
 	struct xe_sched_job *job;
 	struct dma_fence *fence;

@@ -585,11 +591,22 @@ static struct dma_fence *xe_oa_submit_bb(struct xe_oa_stream *stream, struct xe_
 		goto exit;
 	}
 
+	if (deps == XE_OA_SUBMIT_ADD_DEPS) {
+		for (int i = 0; i < stream->num_syncs && !err; i++)
+			err = xe_sync_entry_add_deps(&stream->syncs[i], job);
+		if (err) {
+			drm_dbg(&stream->oa->xe->drm, "xe_sync_entry_add_deps err %d\n", err);
+			goto err_put_job;
+		}
+	}
+
 	xe_sched_job_arm(job);
 	fence = dma_fence_get(&job->drm.s_fence->finished);
 	xe_sched_job_push(job);
 
 	return fence;
+err_put_job:
+	xe_sched_job_put(job);
 exit:
 	return ERR_PTR(err);
 }

@@ -667,7 +684,7 @@ static int xe_oa_modify_ctx_image(struct xe_oa_stream *stream, struct xe_lrc *lr
 
 	xe_oa_store_flex(stream, lrc, bb, flex, count);
 
-	fence = xe_oa_submit_bb(stream, bb);
+	fence = xe_oa_submit_bb(stream, XE_OA_SUBMIT_NO_DEPS, bb);
 	if (IS_ERR(fence)) {
 		err = PTR_ERR(fence);
 		goto free_bb;

@@ -696,7 +713,7 @@ static int xe_oa_load_with_lri(struct xe_oa_stream *stream, struct xe_oa_reg *re
 
 	write_cs_mi_lri(bb, reg_lri, 1);
 
-	fence = xe_oa_submit_bb(stream, bb);
+	fence = xe_oa_submit_bb(stream, XE_OA_SUBMIT_NO_DEPS, bb);
 	if (IS_ERR(fence)) {
 		err = PTR_ERR(fence);
 		goto free_bb;

@@ -944,7 +961,7 @@ static int xe_oa_emit_oa_config(struct xe_oa_stream *stream, struct xe_oa_config
 		goto exit;
 	}
 
-	fence = xe_oa_submit_bb(stream, oa_bo->bb);
+	fence = xe_oa_submit_bb(stream, XE_OA_SUBMIT_ADD_DEPS, oa_bo->bb);
 	if (IS_ERR(fence)) {
 		err = PTR_ERR(fence);
 		goto exit;

index 1bfc4b58b5c17..e6744422dee49 100644
@@ -69,12 +69,6 @@ struct xe_oa_config {
 	struct rcu_head rcu;
 };
 
-struct flex {
-	struct xe_reg reg;
-	u32 offset;
-	u32 value;
-};
-
 struct xe_oa_open_param {
 	struct xe_file *xef;
 	u32 oa_unit_id;

@@ -577,19 +571,38 @@ static __poll_t xe_oa_poll(struct file *file, poll_table *wait)
 	return ret;
 }
 
+static void xe_oa_lock_vma(struct xe_exec_queue *q)
+{
+	if (q->vm) {
+		down_read(&q->vm->lock);
+		xe_vm_lock(q->vm, false);
+	}
+}
+
+static void xe_oa_unlock_vma(struct xe_exec_queue *q)
+{
+	if (q->vm) {
+		xe_vm_unlock(q->vm);
+		up_read(&q->vm->lock);
+	}
+}
+
 static struct dma_fence *xe_oa_submit_bb(struct xe_oa_stream *stream, enum xe_oa_submit_deps deps,
 					 struct xe_bb *bb)
 {
+	struct xe_exec_queue *q = stream->exec_q ?: stream->k_exec_q;
 	struct xe_sched_job *job;
 	struct dma_fence *fence;
 	int err = 0;
 
-	/* Kernel configuration is issued on stream->k_exec_q, not stream->exec_q */
-	job = xe_bb_create_job(stream->k_exec_q, bb);
+	xe_oa_lock_vma(q);
+
+	job = xe_bb_create_job(q, bb);
 	if (IS_ERR(job)) {
 		err = PTR_ERR(job);
 		goto exit;
 	}
+	job->ggtt = true;
 
 	if (deps == XE_OA_SUBMIT_ADD_DEPS) {
 		for (int i = 0; i < stream->num_syncs && !err; i++)

@@ -604,10 +617,13 @@ static struct dma_fence *xe_oa_submit_bb(struct xe_oa_stream *stream, enum xe_oa
 	fence = dma_fence_get(&job->drm.s_fence->finished);
 	xe_sched_job_push(job);
 
+	xe_oa_unlock_vma(q);
+
 	return fence;
 err_put_job:
 	xe_sched_job_put(job);
 exit:
+	xe_oa_unlock_vma(q);
 	return ERR_PTR(err);
 }
 

@@ -655,63 +671,19 @@ static void xe_oa_free_configs(struct xe_oa_stream *stream)
 		free_oa_config_bo(oa_bo);
 }
 
-static void xe_oa_store_flex(struct xe_oa_stream *stream, struct xe_lrc *lrc,
-			     struct xe_bb *bb, const struct flex *flex, u32 count)
-{
-	u32 offset = xe_bo_ggtt_addr(lrc->bo);
-
-	do {
-		bb->cs[bb->len++] = MI_STORE_DATA_IMM | MI_SDI_GGTT | MI_SDI_NUM_DW(1);
-		bb->cs[bb->len++] = offset + flex->offset * sizeof(u32);
-		bb->cs[bb->len++] = 0;
-		bb->cs[bb->len++] = flex->value;
-
-	} while (flex++, --count);
-}
-
-static int xe_oa_modify_ctx_image(struct xe_oa_stream *stream, struct xe_lrc *lrc,
-				  const struct flex *flex, u32 count)
-{
-	struct dma_fence *fence;
-	struct xe_bb *bb;
-	int err;
-
-	bb = xe_bb_new(stream->gt, 4 * count, false);
-	if (IS_ERR(bb)) {
-		err = PTR_ERR(bb);
-		goto exit;
-	}
-
-	xe_oa_store_flex(stream, lrc, bb, flex, count);
-
-	fence = xe_oa_submit_bb(stream, XE_OA_SUBMIT_NO_DEPS, bb);
-	if (IS_ERR(fence)) {
-		err = PTR_ERR(fence);
-		goto free_bb;
-	}
-	xe_bb_free(bb, fence);
-	dma_fence_put(fence);
-
-	return 0;
-free_bb:
-	xe_bb_free(bb, NULL);
-exit:
-	return err;
-}
-
-static int xe_oa_load_with_lri(struct xe_oa_stream *stream, struct xe_oa_reg *reg_lri)
+static int xe_oa_load_with_lri(struct xe_oa_stream *stream, struct xe_oa_reg *reg_lri, u32 count)
 {
 	struct dma_fence *fence;
 	struct xe_bb *bb;
 	int err;
 
-	bb = xe_bb_new(stream->gt, 3, false);
+	bb = xe_bb_new(stream->gt, 2 * count + 1, false);
 	if (IS_ERR(bb)) {
 		err = PTR_ERR(bb);
 		goto exit;
 	}
 
-	write_cs_mi_lri(bb, reg_lri, 1);
+	write_cs_mi_lri(bb, reg_lri, count);
 
 	fence = xe_oa_submit_bb(stream, XE_OA_SUBMIT_NO_DEPS, bb);
 	if (IS_ERR(fence)) {

@@ -731,70 +703,54 @@ static int xe_oa_load_with_lri(struct xe_oa_stream *stream, struct xe_oa_reg *re
 static int xe_oa_configure_oar_context(struct xe_oa_stream *stream, bool enable)
 {
 	const struct xe_oa_format *format = stream->oa_buffer.format;
-	struct xe_lrc *lrc = stream->exec_q->lrc[0];
-	u32 regs_offset = xe_lrc_regs_offset(lrc) / sizeof(u32);
 	u32 oacontrol = __format_to_oactrl(format, OAR_OACONTROL_COUNTER_SEL_MASK) |
 		(enable ? OAR_OACONTROL_COUNTER_ENABLE : 0);
 
-	struct flex regs_context[] = {
+	struct xe_oa_reg reg_lri[] = {
 		{
 			OACTXCONTROL(stream->hwe->mmio_base),
-			stream->oa->ctx_oactxctrl_offset[stream->hwe->class] + 1,
 			enable ? OA_COUNTER_RESUME : 0,
 		},
+		{
+			OAR_OACONTROL,
+			oacontrol,
+		},
 		{
 			RING_CONTEXT_CONTROL(stream->hwe->mmio_base),
-			regs_offset + CTX_CONTEXT_CONTROL,
-			_MASKED_BIT_ENABLE(CTX_CTRL_OAC_CONTEXT_ENABLE),
+			_MASKED_FIELD(CTX_CTRL_OAC_CONTEXT_ENABLE,
+				      enable ? CTX_CTRL_OAC_CONTEXT_ENABLE : 0)
 		},
 	};
-	struct xe_oa_reg reg_lri = { OAR_OACONTROL, oacontrol };
-	int err;
-
-	/* Modify stream hwe context image with regs_context */
-	err = xe_oa_modify_ctx_image(stream, stream->exec_q->lrc[0],
-				     regs_context, ARRAY_SIZE(regs_context));
-	if (err)
-		return err;
 
-	/* Apply reg_lri using LRI */
-	return xe_oa_load_with_lri(stream, &reg_lri);
+	return xe_oa_load_with_lri(stream, reg_lri, ARRAY_SIZE(reg_lri));
 }
 
 static int xe_oa_configure_oac_context(struct xe_oa_stream *stream, bool enable)
 {
 	const struct xe_oa_format *format = stream->oa_buffer.format;
-	struct xe_lrc *lrc = stream->exec_q->lrc[0];
-	u32 regs_offset = xe_lrc_regs_offset(lrc) / sizeof(u32);
 	u32 oacontrol = __format_to_oactrl(format, OAR_OACONTROL_COUNTER_SEL_MASK) |
 		(enable ? OAR_OACONTROL_COUNTER_ENABLE : 0);
-	struct flex regs_context[] = {
+	struct xe_oa_reg reg_lri[] = {
 		{
 			OACTXCONTROL(stream->hwe->mmio_base),
-			stream->oa->ctx_oactxctrl_offset[stream->hwe->class] + 1,
 			enable ? OA_COUNTER_RESUME : 0,
 		},
+		{
+			OAC_OACONTROL,
+			oacontrol
+		},
 		{
 			RING_CONTEXT_CONTROL(stream->hwe->mmio_base),
-			regs_offset + CTX_CONTEXT_CONTROL,
-			_MASKED_BIT_ENABLE(CTX_CTRL_OAC_CONTEXT_ENABLE) |
+			_MASKED_FIELD(CTX_CTRL_OAC_CONTEXT_ENABLE,
+				      enable ? CTX_CTRL_OAC_CONTEXT_ENABLE : 0) |
 			_MASKED_FIELD(CTX_CTRL_RUN_ALONE, enable ? CTX_CTRL_RUN_ALONE : 0),
 		},
 	};
-	struct xe_oa_reg reg_lri = { OAC_OACONTROL, oacontrol };
-	int err;
 
 	/* Set ccs select to enable programming of OAC_OACONTROL */
 	xe_mmio_write32(stream->gt, __oa_regs(stream)->oa_ctrl, __oa_ccs_select(stream));
 
-	/* Modify stream hwe context image with regs_context */
-	err = xe_oa_modify_ctx_image(stream, stream->exec_q->lrc[0],
-				     regs_context, ARRAY_SIZE(regs_context));
-	if (err)
-		return err;
-
-	/* Apply reg_lri using LRI */
-	return xe_oa_load_with_lri(stream, &reg_lri);
+	return xe_oa_load_with_lri(stream, reg_lri, ARRAY_SIZE(reg_lri));
 }
 
 static int xe_oa_configure_oa_context(struct xe_oa_stream *stream, bool enable)

@@ -1933,8 +1889,8 @@ int xe_oa_stream_open_ioctl(struct drm_device *dev, u64 data, struct drm_file *f
 		if (XE_IOCTL_DBG(oa->xe, !param.exec_q))
 			return -ENOENT;
 
-		if (param.exec_q->width > 1)
-			drm_dbg(&oa->xe->drm, "exec_q->width > 1, programming only exec_q->lrc[0]\n");
+		if (XE_IOCTL_DBG(oa->xe, param.exec_q->width > 1))
+			return -EOPNOTSUPP;
 	}
 
 	/*

index 0be4f489d3e12..9f327f27c0726 100644
@@ -221,7 +221,10 @@ static int emit_pipe_imm_ggtt(u32 addr, u32 value, bool stall_only, u32 *dw,
 
 static u32 get_ppgtt_flag(struct xe_sched_job *job)
 {
-	return job->q->vm ? BIT(8) : 0;
+	if (job->q->vm && !job->ggtt)
+		return BIT(8);
+
+	return 0;
 }
 
 static int emit_copy_timestamp(struct xe_lrc *lrc, u32 *dw, int i)

index 0d3f76fb05cea..c207361bf43e1 100644
@@ -57,6 +57,8 @@ struct xe_sched_job {
 	u32 migrate_flush_flags;
 	/** @ring_ops_flush_tlb: The ring ops need to flush TLB before payload. */
 	bool ring_ops_flush_tlb;
+	/** @ggtt: mapped in ggtt. */
+	bool ggtt;
 	/** @ptrs: per instance pointers. */
 	struct xe_job_ptrs ptrs[];
 };

index fe08c983d5bb4..9ff72a5a13eb3 100644
@@ -1394,8 +1394,6 @@ static noinline_for_stack int extent_writepage_io(struct btrfs_inode *inode,
 			goto out;
 		submitted_io = true;
 	}
-
-	btrfs_folio_assert_not_dirty(fs_info, folio, start, len);
 out:
 	/*
 	 * If we didn't submitted any sector (>= i_size), folio dirty get

index 9ff72a5a13eb3..ba34b92d48c2f 100644
@@ -1105,6 +1105,45 @@ int btrfs_read_folio(struct file *file, struct folio *folio)
 	return ret;
 }
 
+static void set_delalloc_bitmap(struct folio *folio, unsigned long *delalloc_bitmap,
+				u64 start, u32 len)
+{
+	struct btrfs_fs_info *fs_info = folio_to_fs_info(folio);
+	const u64 folio_start = folio_pos(folio);
+	unsigned int start_bit;
+	unsigned int nbits;
+
+	ASSERT(start >= folio_start && start + len <= folio_start + PAGE_SIZE);
+	start_bit = (start - folio_start) >> fs_info->sectorsize_bits;
+	nbits = len >> fs_info->sectorsize_bits;
+	ASSERT(bitmap_test_range_all_zero(delalloc_bitmap, start_bit, nbits));
+	bitmap_set(delalloc_bitmap, start_bit, nbits);
+}
+
+static bool find_next_delalloc_bitmap(struct folio *folio,
+				      unsigned long *delalloc_bitmap, u64 start,
+				      u64 *found_start, u32 *found_len)
+{
+	struct btrfs_fs_info *fs_info = folio_to_fs_info(folio);
+	const u64 folio_start = folio_pos(folio);
+	const unsigned int bitmap_size = fs_info->sectors_per_page;
+	unsigned int start_bit;
+	unsigned int first_zero;
+	unsigned int first_set;
+
+	ASSERT(start >= folio_start && start < folio_start + PAGE_SIZE);
+
+	start_bit = (start - folio_start) >> fs_info->sectorsize_bits;
+	first_set = find_next_bit(delalloc_bitmap, bitmap_size, start_bit);
+	if (first_set >= bitmap_size)
+		return false;
+
+	*found_start = folio_start + (first_set << fs_info->sectorsize_bits);
+	first_zero = find_next_zero_bit(delalloc_bitmap, bitmap_size, first_set);
+	*found_len = (first_zero - first_set) << fs_info->sectorsize_bits;
+	return true;
+}
+
 /*
  * helper for extent_writepage(), doing all of the delayed allocation setup.
  *

@@ -1124,6 +1163,7 @@ static noinline_for_stack int writepage_delalloc(struct btrfs_inode *inode,
 	const bool is_subpage = btrfs_is_subpage(fs_info, folio->mapping);
 	const u64 page_start = folio_pos(folio);
 	const u64 page_end = page_start + folio_size(folio) - 1;
+	unsigned long delalloc_bitmap = 0;
 	/*
 	 * Save the last found delalloc end. As the delalloc end can go beyond
 	 * page boundary, thus we cannot rely on subpage bitmap to locate the

@@ -1151,6 +1191,8 @@ static noinline_for_stack int writepage_delalloc(struct btrfs_inode *inode,
 			delalloc_start = delalloc_end + 1;
 			continue;
 		}
+		set_delalloc_bitmap(folio, &delalloc_bitmap, delalloc_start,
+				    min(delalloc_end, page_end) + 1 - delalloc_start);
 		btrfs_folio_set_writer_lock(fs_info, folio, delalloc_start,
 					    min(delalloc_end, page_end) + 1 -
 					    delalloc_start);

@@ -1178,7 +1220,7 @@ static noinline_for_stack int writepage_delalloc(struct btrfs_inode *inode,
 			found_len = last_delalloc_end + 1 - found_start;
 			found = true;
 		} else {
-			found = btrfs_subpage_find_writer_locked(fs_info, folio,
+			found = find_next_delalloc_bitmap(folio, &delalloc_bitmap,
 					delalloc_start, &found_start, &found_len);
 		}
 		if (!found)

53 lines not on this page · raw message

index cdb554e0d215e..197ec6c0b07b2 100644
@@ -108,10 +108,6 @@ void btrfs_folio_set_writer_lock(const struct btrfs_fs_info *fs_info,
 				 struct folio *folio, u64 start, u32 len);
 void btrfs_folio_end_writer_lock_bitmap(const struct btrfs_fs_info *fs_info,
 					struct folio *folio, unsigned long bitmap);
-bool btrfs_subpage_find_writer_locked(const struct btrfs_fs_info *fs_info,
-				      struct folio *folio, u64 search_start,
-				      u64 *found_start_ret, u32 *found_len_ret);
-
 /*
  * Template for subpage related operations.
  *

index ba34b92d48c2f..8222ae6f29af5 100644
@@ -1174,6 +1174,7 @@ static noinline_for_stack int writepage_delalloc(struct btrfs_inode *inode,
 	u64 delalloc_end = page_end;
 	u64 delalloc_to_write = 0;
 	int ret = 0;
+	int bit;
 
 	/* Save the dirty bitmap as our submission bitmap will be a subset of it. */
 	if (btrfs_is_subpage(fs_info, inode->vfs_inode.i_mapping)) {

@@ -1183,6 +1184,12 @@ static noinline_for_stack int writepage_delalloc(struct btrfs_inode *inode,
 		bio_ctrl->submit_bitmap = 1;
 	}
 
+	for_each_set_bit(bit, &bio_ctrl->submit_bitmap, fs_info->sectors_per_page) {
+		u64 start = page_start + (bit << fs_info->sectorsize_bits);
+
+		btrfs_folio_set_writer_lock(fs_info, folio, start, fs_info->sectorsize);
+	}
+
 	/* Lock all (subpage) delalloc ranges inside the folio first. */
 	while (delalloc_start < page_end) {
 		delalloc_end = page_end;

@@ -1193,9 +1200,6 @@ static noinline_for_stack int writepage_delalloc(struct btrfs_inode *inode,
 		}
 		set_delalloc_bitmap(folio, &delalloc_bitmap, delalloc_start,
 				    min(delalloc_end, page_end) + 1 - delalloc_start);
-		btrfs_folio_set_writer_lock(fs_info, folio, delalloc_start,
-					    min(delalloc_end, page_end) + 1 -
-					    delalloc_start);
 		last_delalloc_end = delalloc_end;
 		delalloc_start = delalloc_end + 1;
 	}

index c4950e04f481a..99341e98bbcc7 100644
@@ -295,26 +295,6 @@ static void btrfs_subpage_clamp_range(struct folio *folio, u64 *start, u32 *len)
 			     orig_start + orig_len) - *start;
 }
 
-static void btrfs_subpage_start_writer(const struct btrfs_fs_info *fs_info,
-				       struct folio *folio, u64 start, u32 len)
-{
-	struct btrfs_subpage *subpage = folio_get_private(folio);
-	const int start_bit = subpage_calc_start_bit(fs_info, folio, locked, start, len);
-	const int nbits = (len >> fs_info->sectorsize_bits);
-	unsigned long flags;
-	int ret;
-
-	btrfs_subpage_assert(fs_info, folio, start, len);
-
-	spin_lock_irqsave(&subpage->lock, flags);
-	ASSERT(atomic_read(&subpage->readers) == 0);
-	ASSERT(bitmap_test_range_all_zero(subpage->bitmaps, start_bit, nbits));
-	bitmap_set(subpage->bitmaps, start_bit, nbits);
-	ret = atomic_add_return(nbits, &subpage->writers);
-	ASSERT(ret == nbits);
-	spin_unlock_irqrestore(&subpage->lock, flags);
-}
-
 static bool btrfs_subpage_end_and_test_writer(const struct btrfs_fs_info *fs_info,
 					      struct folio *folio, u64 start, u32 len)
 {

@@ -351,33 +331,6 @@ static bool btrfs_subpage_end_and_test_writer(const struct btrfs_fs_info *fs_inf
 	return last;
 }
 
-/*
- * Lock a folio for delalloc page writeback.
- *
- * Return -EAGAIN if the page is not properly initialized.
- * Return 0 with the page locked, and writer counter updated.
- *
- * Even with 0 returned, the page still need extra check to make sure
- * it's really the correct page, as the caller is using
- * filemap_get_folios_contig(), which can race with page invalidating.
- */
-int btrfs_folio_start_writer_lock(const struct btrfs_fs_info *fs_info,
-				  struct folio *folio, u64 start, u32 len)
-{
-	if (unlikely(!fs_info) || !btrfs_is_subpage(fs_info, folio->mapping)) {
-		folio_lock(folio);
-		return 0;
-	}
-	folio_lock(folio);
-	if (!folio_test_private(folio) || !folio_get_private(folio)) {
-		folio_unlock(folio);
-		return -EAGAIN;
-	}
-	btrfs_subpage_clamp_range(folio, &start, &len);
-	btrfs_subpage_start_writer(fs_info, folio, start, len);
-	return 0;
-}
-
 /*
  * Handle different locked folios:
  *

index 197ec6c0b07b2..6289d6f65b87d 100644
@@ -100,8 +100,6 @@ void btrfs_subpage_start_reader(const struct btrfs_fs_info *fs_info,
 void btrfs_subpage_end_reader(const struct btrfs_fs_info *fs_info,
 			      struct folio *folio, u64 start, u32 len);
 
-int btrfs_folio_start_writer_lock(const struct btrfs_fs_info *fs_info,
-				  struct folio *folio, u64 start, u32 len);
 void btrfs_folio_end_writer_lock(const struct btrfs_fs_info *fs_info,
 				 struct folio *folio, u64 start, u32 len);
 void btrfs_folio_set_writer_lock(const struct btrfs_fs_info *fs_info,

index d8a4cdbb5495d..7d0d8852ce8d2 100644
@@ -89,7 +89,7 @@
 #define mmCLK1_CLK4_ALLOW_DS 0x16EA8
 #define mmCLK1_CLK5_ALLOW_DS 0x16EB1
 
-#define mmCLK5_spll_field_8 0x1B04B
+#define mmCLK5_spll_field_8 0x1B24B
 #define mmDENTIST_DISPCLK_CNTL 0x0124
 #define regDENTIST_DISPCLK_CNTL 0x0064
 #define regDENTIST_DISPCLK_CNTL_BASE_IDX 1

index c3699105656a7..70f1a46d07c5e 100644
@@ -416,7 +416,7 @@ static void pcim_intx_restore(struct device *dev, void *data)
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct pcim_intx_devres *res = data;
 
-	pci_intx_unmanaged(pdev, res->orig_intx);
+	pci_intx(pdev, res->orig_intx);
 }
 
 static struct pcim_intx_devres *get_or_create_intx_devres(struct device *dev)

@@ -453,7 +453,7 @@ int pcim_intx(struct pci_dev *pdev, int enable)
 		return -ENOMEM;
 
 	res->orig_intx = !enable;
-	pci_intx_unmanaged(pdev, enable);
+	pci_intx(pdev, enable);
 
 	return 0;
 }

60 lines not on this page · raw message

index 6ef32a8d146b1..74114acbb07fb 100644
@@ -1353,7 +1353,6 @@ int __must_check pcim_set_mwi(struct pci_dev *dev);
 int pci_try_set_mwi(struct pci_dev *dev);
 void pci_clear_mwi(struct pci_dev *dev);
 void pci_disable_parity(struct pci_dev *dev);
-void pci_intx_unmanaged(struct pci_dev *pdev, int enable);
 void pci_intx(struct pci_dev *dev, int enable);
 bool pci_check_and_mask_intx(struct pci_dev *dev);
 bool pci_check_and_unmask_intx(struct pci_dev *dev);

index 70f1a46d07c5e..643f85849ef64 100644
@@ -419,19 +419,12 @@ static void pcim_intx_restore(struct device *dev, void *data)
 	pci_intx(pdev, res->orig_intx);
 }
 
-static struct pcim_intx_devres *get_or_create_intx_devres(struct device *dev)
+static void save_orig_intx(struct pci_dev *pdev, struct pcim_intx_devres *res)
 {
-	struct pcim_intx_devres *res;
-
-	res = devres_find(dev, pcim_intx_restore, NULL, NULL);
-	if (res)
-		return res;
+	u16 pci_command;
 
-	res = devres_alloc(pcim_intx_restore, sizeof(*res), GFP_KERNEL);
-	if (res)
-		devres_add(dev, res);
-
-	return res;
+	pci_read_config_word(pdev, PCI_COMMAND, &pci_command);
+	res->orig_intx = !(pci_command & PCI_COMMAND_INTX_DISABLE);
 }
 
 /**

@@ -447,12 +440,23 @@ static struct pcim_intx_devres *get_or_create_intx_devres(struct device *dev)
 int pcim_intx(struct pci_dev *pdev, int enable)
 {
 	struct pcim_intx_devres *res;
+	struct device *dev = &pdev->dev;
 
-	res = get_or_create_intx_devres(&pdev->dev);
-	if (!res)
-		return -ENOMEM;
+	/*
+	 * pcim_intx() must only restore the INTx value that existed before the
+	 * driver was loaded, i.e., before it called pcim_intx() for the
+	 * first time.
+	 */
+	res = devres_find(dev, pcim_intx_restore, NULL, NULL);
+	if (!res) {
+		res = devres_alloc(pcim_intx_restore, sizeof(*res), GFP_KERNEL);
+		if (!res)
+			return -ENOMEM;
+
+		save_orig_intx(pdev, res);
+		devres_add(dev, res);
+	}
 
-	res->orig_intx = !enable;
 	pci_intx(pdev, enable);
 
 	return 0;

index ede6165e09d90..b2b6d89f06537 100644
@@ -25,7 +25,6 @@
 #define FW_SHAVE_NN_MAX_SIZE	SZ_2M
 #define FW_RUNTIME_MIN_ADDR	(FW_GLOBAL_MEM_START)
 #define FW_RUNTIME_MAX_ADDR	(FW_GLOBAL_MEM_END - FW_SHARED_MEM_SIZE)
-#define FW_VERSION_HEADER_SIZE	SZ_4K
 #define FW_FILE_IMAGE_OFFSET	(VPU_FW_HEADER_SIZE + FW_VERSION_HEADER_SIZE)
 
 #define WATCHDOG_MSS_REDIRECT	32

@@ -191,8 +190,10 @@ static int ivpu_fw_parse(struct ivpu_device *vdev)
 	ivpu_dbg(vdev, FW_BOOT, "Header version: 0x%x, format 0x%x\n",
 		 fw_hdr->header_version, fw_hdr->image_format);
 
-	ivpu_info(vdev, "Firmware: %s, version: %s", fw->name,
-		  (const char *)fw_hdr + VPU_FW_HEADER_SIZE);
+	if (!scnprintf(fw->version, sizeof(fw->version), "%s", fw->file->data + VPU_FW_HEADER_SIZE))
+		ivpu_warn(vdev, "Missing firmware version\n");
+
+	ivpu_info(vdev, "Firmware: %s, version: %s\n", fw->name, fw->version);
 
 	if (IVPU_FW_CHECK_API_COMPAT(vdev, fw_hdr, BOOT, 3))
 		return -EINVAL;

index 40d9d17be3f52..5e8eb608b70f1 100644
@@ -1,11 +1,14 @@
 /* SPDX-License-Identifier: GPL-2.0-only */
 /*
- * Copyright (C) 2020-2023 Intel Corporation
+ * Copyright (C) 2020-2024 Intel Corporation
  */
 
 #ifndef __IVPU_FW_H__
 #define __IVPU_FW_H__
 
+#define FW_VERSION_HEADER_SIZE	SZ_4K
+#define FW_VERSION_STR_SIZE	SZ_256
+
 struct ivpu_device;
 struct ivpu_bo;
 struct vpu_boot_params;

@@ -13,6 +16,7 @@ struct vpu_boot_params;
 struct ivpu_fw_info {
 	const struct firmware *file;
 	const char *name;
+	char version[FW_VERSION_STR_SIZE];
 	struct ivpu_bo *mem;
 	struct ivpu_bo *mem_shave_nn;
 	struct ivpu_bo *mem_log_crit;

new file mode 100644
index 0000000000000..16ad0c30818cc
@@ -0,0 +1,39 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2020-2024 Intel Corporation
+ */
+
+#include <linux/devcoredump.h>
+#include <linux/firmware.h>
+
+#include "ivpu_coredump.h"
+#include "ivpu_fw.h"
+#include "ivpu_gem.h"
+#include "vpu_boot_api.h"
+
+#define CRASH_DUMP_HEADER "Intel NPU crash dump"
+#define CRASH_DUMP_HEADERS_SIZE SZ_4K
+
+void ivpu_dev_coredump(struct ivpu_device *vdev)
+{
+	struct drm_print_iterator pi = {};
+	struct drm_printer p;
+	size_t coredump_size;
+	char *coredump;
+
+	coredump_size = CRASH_DUMP_HEADERS_SIZE + FW_VERSION_HEADER_SIZE +
+			ivpu_bo_size(vdev->fw->mem_log_crit) + ivpu_bo_size(vdev->fw->mem_log_verb);
+	coredump = vmalloc(coredump_size);
+	if (!coredump)
+		return;
+
+	pi.data = coredump;
+	pi.remain = coredump_size;
+	p = drm_coredump_printer(&pi);
+
+	drm_printf(&p, "%s\n", CRASH_DUMP_HEADER);
+	drm_printf(&p, "FW version: %s\n", vdev->fw->version);
+	ivpu_fw_log_print(vdev, false, &p);
+
+	dev_coredumpv(vdev->drm.dev, coredump, pi.offset, GFP_KERNEL);
+}

new file mode 100644
index 0000000000000..8efb09d024411
@@ -0,0 +1,25 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (C) 2020-2024 Intel Corporation
+ */
+
+#ifndef __IVPU_COREDUMP_H__
+#define __IVPU_COREDUMP_H__
+
+#include <drm/drm_print.h>
+
+#include "ivpu_drv.h"
+#include "ivpu_fw_log.h"
+
+#ifdef CONFIG_DEV_COREDUMP
+void ivpu_dev_coredump(struct ivpu_device *vdev);
+#else
+static inline void ivpu_dev_coredump(struct ivpu_device *vdev)
+{
+	struct drm_printer p = drm_info_printer(vdev->drm.dev);
+
+	ivpu_fw_log_print(vdev, false, &p);
+}
+#endif
+
+#endif /* __IVPU_COREDUMP_H__ */

index c91400ecf9265..38b4158f52784 100644
@@ -14,7 +14,7 @@
 #include <drm/drm_ioctl.h>
 #include <drm/drm_prime.h>
 
-#include "vpu_boot_api.h"
+#include "ivpu_coredump.h"
 #include "ivpu_debugfs.h"
 #include "ivpu_drv.h"
 #include "ivpu_fw.h"

@@ -29,6 +29,7 @@
 #include "ivpu_ms.h"
 #include "ivpu_pm.h"
 #include "ivpu_sysfs.h"
+#include "vpu_boot_api.h"
 
 #ifndef DRIVER_VERSION_STR
 #define DRIVER_VERSION_STR __stringify(DRM_IVPU_DRIVER_MAJOR) "." \

@@ -382,7 +383,7 @@ int ivpu_boot(struct ivpu_device *vdev)
 		ivpu_err(vdev, "Failed to boot the firmware: %d\n", ret);
 		ivpu_hw_diagnose_failure(vdev);
 		ivpu_mmu_evtq_dump(vdev);
-		ivpu_fw_log_dump(vdev);
+		ivpu_dev_coredump(vdev);
 		return ret;
 	}
 

index 0b2573f6f3151..4b390a99699d6 100644
@@ -8,8 +8,6 @@
 
 #include <linux/types.h>
 
-#include <drm/drm_print.h>
-
 #include "ivpu_drv.h"
 
 #define IVPU_FW_LOG_DEFAULT 0

@@ -28,11 +26,5 @@ extern unsigned int ivpu_log_level;
 void ivpu_fw_log_print(struct ivpu_device *vdev, bool only_new_msgs, struct drm_printer *p);
 void ivpu_fw_log_clear(struct ivpu_device *vdev);
 
-static inline void ivpu_fw_log_dump(struct ivpu_device *vdev)
-{
-	struct drm_printer p = drm_info_printer(vdev->drm.dev);
-
-	ivpu_fw_log_print(vdev, false, &p);
-}
 
 #endif /* __IVPU_FW_LOG_H__ */

index ef9a4ba18cb8a..0110f5ee7d069 100644
@@ -9,17 +9,18 @@
 #include <linux/pm_runtime.h>
 #include <linux/reboot.h>
 
-#include "vpu_boot_api.h"
+#include "ivpu_coredump.h"
 #include "ivpu_drv.h"
-#include "ivpu_hw.h"
 #include "ivpu_fw.h"
 #include "ivpu_fw_log.h"
+#include "ivpu_hw.h"
 #include "ivpu_ipc.h"
 #include "ivpu_job.h"
 #include "ivpu_jsm_msg.h"
 #include "ivpu_mmu.h"
 #include "ivpu_ms.h"
 #include "ivpu_pm.h"
+#include "vpu_boot_api.h"
 
 static bool ivpu_disable_recovery;
 module_param_named_unsafe(disable_recovery, ivpu_disable_recovery, bool, 0644);

@@ -123,7 +124,7 @@ static void ivpu_pm_recovery_work(struct work_struct *work)
 	if (ret)
 		ivpu_err(vdev, "Failed to resume NPU: %d\n", ret);
 
-	ivpu_fw_log_dump(vdev);
+	ivpu_dev_coredump(vdev);
 
 	atomic_inc(&vdev->pm->reset_counter);
 	atomic_set(&vdev->pm->reset_pending, 1);

@@ -262,7 +263,7 @@ int ivpu_pm_runtime_suspend_cb(struct device *dev)
 	if (!is_idle || ret_d0i3) {
 		ivpu_err(vdev, "Forcing cold boot due to previous errors\n");
 		atomic_inc(&vdev->pm->reset_counter);
-		ivpu_fw_log_dump(vdev);
+		ivpu_dev_coredump(vdev);
 		ivpu_pm_prepare_cold_boot(vdev);
 	} else {
 		ivpu_pm_prepare_warm_boot(vdev);

index 63f13b697eed7..2b30cc2e9272e 100644
@@ -152,6 +152,7 @@ struct ivpu_device {
 		int tdr;
 		int autosuspend;
 		int d0i3_entry_msg;
+		int state_dump_msg;
 	} timeout;
 };
 

index e69c0613513f1..08b3cef58fd2d 100644
@@ -89,12 +89,14 @@ static void timeouts_init(struct ivpu_device *vdev)
 		vdev->timeout.tdr = 2000000;
 		vdev->timeout.autosuspend = -1;
 		vdev->timeout.d0i3_entry_msg = 500;
+		vdev->timeout.state_dump_msg = 10;
 	} else if (ivpu_is_simics(vdev)) {
 		vdev->timeout.boot = 50;
 		vdev->timeout.jsm = 500;
 		vdev->timeout.tdr = 10000;
 		vdev->timeout.autosuspend = -1;
 		vdev->timeout.d0i3_entry_msg = 100;
+		vdev->timeout.state_dump_msg = 10;
 	} else {
 		vdev->timeout.boot = 1000;
 		vdev->timeout.jsm = 500;

@@ -104,6 +106,7 @@ static void timeouts_init(struct ivpu_device *vdev)
 		else
 			vdev->timeout.autosuspend = 100;
 		vdev->timeout.d0i3_entry_msg = 5;
+		vdev->timeout.state_dump_msg = 10;
 	}
 }
 

index 29b723039a345..13c8a12162e89 100644
@@ -353,6 +353,32 @@ int ivpu_ipc_send_receive(struct ivpu_device *vdev, struct vpu_jsm_msg *req,
 	return ret;
 }
 
+int ivpu_ipc_send_and_wait(struct ivpu_device *vdev, struct vpu_jsm_msg *req,
+			   u32 channel, unsigned long timeout_ms)
+{
+	struct ivpu_ipc_consumer cons;
+	int ret;
+
+	ret = ivpu_rpm_get(vdev);
+	if (ret < 0)
+		return ret;
+
+	ivpu_ipc_consumer_add(vdev, &cons, channel, NULL);
+
+	ret = ivpu_ipc_send(vdev, &cons, req);
+	if (ret) {
+		ivpu_warn_ratelimited(vdev, "IPC send failed: %d\n", ret);
+		goto consumer_del;
+	}
+
+	msleep(timeout_ms);
+
+consumer_del:
+	ivpu_ipc_consumer_del(vdev, &cons);
+	ivpu_rpm_put(vdev);
+	return ret;
+}
+
 static bool
 ivpu_ipc_match_consumer(struct ivpu_device *vdev, struct ivpu_ipc_consumer *cons,
 			struct ivpu_ipc_hdr *ipc_hdr, struct vpu_jsm_msg *jsm_msg)

index fb4de7fb8210e..b4dfb504679ba 100644
@@ -107,5 +107,7 @@ int ivpu_ipc_send_receive_internal(struct ivpu_device *vdev, struct vpu_jsm_msg
 int ivpu_ipc_send_receive(struct ivpu_device *vdev, struct vpu_jsm_msg *req,
 			  enum vpu_ipc_msg_type expected_resp, struct vpu_jsm_msg *resp,
 			  u32 channel, unsigned long timeout_ms);
+int ivpu_ipc_send_and_wait(struct ivpu_device *vdev, struct vpu_jsm_msg *req,
+			   u32 channel, unsigned long timeout_ms);
 
 #endif /* __IVPU_IPC_H__ */

index 88105963c1b28..f7618b605f021 100644
@@ -555,3 +555,11 @@ int ivpu_jsm_dct_disable(struct ivpu_device *vdev)
 	return ivpu_ipc_send_receive_internal(vdev, &req, VPU_JSM_MSG_DCT_DISABLE_DONE, &resp,
 					      VPU_IPC_CHAN_ASYNC_CMD, vdev->timeout.jsm);
 }
+
+int ivpu_jsm_state_dump(struct ivpu_device *vdev)
+{
+	struct vpu_jsm_msg req = { .type = VPU_JSM_MSG_STATE_DUMP };
+
+	return ivpu_ipc_send_and_wait(vdev, &req, VPU_IPC_CHAN_ASYNC_CMD,
+				      vdev->timeout.state_dump_msg);
+}

index e4e42c0ff6e65..9e84d3526a146 100644
@@ -43,4 +43,6 @@ int ivpu_jsm_metric_streamer_info(struct ivpu_device *vdev, u64 metric_group_mas
 				  u64 buffer_size, u32 *sample_size, u64 *info_size);
 int ivpu_jsm_dct_enable(struct ivpu_device *vdev, u32 active_us, u32 inactive_us);
 int ivpu_jsm_dct_disable(struct ivpu_device *vdev);
+int ivpu_jsm_state_dump(struct ivpu_device *vdev);
+
 #endif

index 0110f5ee7d069..848d7468d48ce 100644
@@ -124,6 +124,7 @@ static void ivpu_pm_recovery_work(struct work_struct *work)
 	if (ret)
 		ivpu_err(vdev, "Failed to resume NPU: %d\n", ret);
 
+	ivpu_jsm_state_dump(vdev);
 	ivpu_dev_coredump(vdev);
 
 	atomic_inc(&vdev->pm->reset_counter);

index 848d7468d48ce..fbb61a2c3b19c 100644
@@ -111,41 +111,57 @@ static int ivpu_resume(struct ivpu_device *vdev)
 	return ret;
 }
 
-static void ivpu_pm_recovery_work(struct work_struct *work)
+static void ivpu_pm_reset_begin(struct ivpu_device *vdev)
 {
-	struct ivpu_pm_info *pm = container_of(work, struct ivpu_pm_info, recovery_work);
-	struct ivpu_device *vdev = pm->vdev;
-	char *evt[2] = {"IVPU_PM_EVENT=IVPU_RECOVER", NULL};
-	int ret;
-
-	ivpu_err(vdev, "Recovering the NPU (reset #%d)\n", atomic_read(&vdev->pm->reset_counter));
-
-	ret = pm_runtime_resume_and_get(vdev->drm.dev);
-	if (ret)
-		ivpu_err(vdev, "Failed to resume NPU: %d\n", ret);
-
-	ivpu_jsm_state_dump(vdev);
-	ivpu_dev_coredump(vdev);
+	pm_runtime_disable(vdev->drm.dev);
 
 	atomic_inc(&vdev->pm->reset_counter);
 	atomic_set(&vdev->pm->reset_pending, 1);
 	down_write(&vdev->pm->reset_lock);
+}
+
+static void ivpu_pm_reset_complete(struct ivpu_device *vdev)
+{
+	int ret;
 
-	ivpu_suspend(vdev);
 	ivpu_pm_prepare_cold_boot(vdev);
 	ivpu_jobs_abort_all(vdev);
 	ivpu_ms_cleanup_all(vdev);
 
 	ret = ivpu_resume(vdev);
-	if (ret)
+	if (ret) {
 		ivpu_err(vdev, "Failed to resume NPU: %d\n", ret);
+		pm_runtime_set_suspended(vdev->drm.dev);
+	} else {
+		pm_runtime_set_active(vdev->drm.dev);
+	}
 
 	up_write(&vdev->pm->reset_lock);
 	atomic_set(&vdev->pm->reset_pending, 0);
 
-	kobject_uevent_env(&vdev->drm.dev->kobj, KOBJ_CHANGE, evt);
 	pm_runtime_mark_last_busy(vdev->drm.dev);
-	pm_runtime_put_autosuspend(vdev->drm.dev);
+	pm_runtime_enable(vdev->drm.dev);
+}
+
+static void ivpu_pm_recovery_work(struct work_struct *work)
+{
+	struct ivpu_pm_info *pm = container_of(work, struct ivpu_pm_info, recovery_work);
+	struct ivpu_device *vdev = pm->vdev;
+	char *evt[2] = {"IVPU_PM_EVENT=IVPU_RECOVER", NULL};
+
+	ivpu_err(vdev, "Recovering the NPU (reset #%d)\n", atomic_read(&vdev->pm->reset_counter));
+
+	ivpu_pm_reset_begin(vdev);
+
+	if (!pm_runtime_status_suspended(vdev->drm.dev)) {
+		ivpu_jsm_state_dump(vdev);
+		ivpu_dev_coredump(vdev);
+		ivpu_suspend(vdev);
+	}
+
+	ivpu_pm_reset_complete(vdev);
+
+	kobject_uevent_env(&vdev->drm.dev->kobj, KOBJ_CHANGE, evt);
 }
 
 void ivpu_pm_trigger_recovery(struct ivpu_device *vdev, const char *reason)

@@ -316,16 +332,13 @@ void ivpu_pm_reset_prepare_cb(struct pci_dev *pdev)
 	struct ivpu_device *vdev = pci_get_drvdata(pdev);
 
 	ivpu_dbg(vdev, PM, "Pre-reset..\n");
-	atomic_inc(&vdev->pm->reset_counter);
-	atomic_set(&vdev->pm->reset_pending, 1);
 
-	pm_runtime_get_sync(vdev->drm.dev);
-	down_write(&vdev->pm->reset_lock);
-	ivpu_prepare_for_reset(vdev);
-	ivpu_hw_reset(vdev);
-	ivpu_pm_prepare_cold_boot(vdev);
-	ivpu_jobs_abort_all(vdev);
-	ivpu_ms_cleanup_all(vdev);
+	ivpu_pm_reset_begin(vdev);
+
+	if (!pm_runtime_status_suspended(vdev->drm.dev)) {
+		ivpu_prepare_for_reset(vdev);
+		ivpu_hw_reset(vdev);
+	}
 
 	ivpu_dbg(vdev, PM, "Pre-reset done.\n");
 }

@@ -333,18 +346,12 @@ void ivpu_pm_reset_prepare_cb(struct pci_dev *pdev)
 void ivpu_pm_reset_done_cb(struct pci_dev *pdev)
 {
 	struct ivpu_device *vdev = pci_get_drvdata(pdev);
-	int ret;
 
 	ivpu_dbg(vdev, PM, "Post-reset..\n");
-	ret = ivpu_resume(vdev);
-	if (ret)
-		ivpu_err(vdev, "Failed to set RESUME state: %d\n", ret);
-	up_write(&vdev->pm->reset_lock);
-	atomic_set(&vdev->pm->reset_pending, 0);
-	ivpu_dbg(vdev, PM, "Post-reset done.\n");
 
-	pm_runtime_mark_last_busy(vdev->drm.dev);
-	pm_runtime_put_autosuspend(vdev->drm.dev);
+	ivpu_pm_reset_complete(vdev);
+
+	ivpu_dbg(vdev, PM, "Post-reset done.\n");
 }
 
 void ivpu_pm_init(struct ivpu_device *vdev)