[RESEND v4 1/6] block: free both rq_map and request

[RESEND v4 1/6] block: free both rq_map and request

[Weiping Zhang]

For this error handle, it should free both map and request,
otherwise a memory leak occur.

Allocation:

__blk_mq_alloc_rq_map
	blk_mq_alloc_rq_map
		blk_mq_alloc_rq_map
			tags = blk_mq_init_tags : kzalloc_node:
			tags->rqs = kcalloc_node
			tags->static_rqs = kcalloc_node
	blk_mq_alloc_rqs
		p = alloc_pages_node
		tags->static_rqs[i] = p + offset;

Free:

blk_mq_free_rq_map
	kfree(tags->rqs);
	kfree(tags->static_rqs);
	blk_mq_free_tags
		kfree(tags);

The page allocated in blk_mq_alloc_rqs cannot be released,
so we should use blk_mq_free_map_and_requests here.

blk_mq_free_map_and_requests
	blk_mq_free_rqs
		__free_pages : cleanup for blk_mq_alloc_rqs
	blk_mq_free_rq_map : cleanup for blk_mq_alloc_rq_map

Signed-off-by: Weiping Zhang <zhangweiping@didiglobal.com>
---
 block/blk-mq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index a7785df2c944..f789b3e1b3ab 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2995,7 +2995,7 @@ static int __blk_mq_alloc_rq_maps(struct blk_mq_tag_set *set)
 
 out_unwind:
 	while (--i >= 0)
-		blk_mq_free_rq_map(set->tags[i]);
+		blk_mq_free_map_and_requests(set, i);
 
 	return -ENOMEM;
 }
-- 
2.18.1

[RESEND v4 2/6] block: save previous hardware queue count before udpate

[Weiping Zhang]

blk_mq_realloc_tag_set_tags will update set->nr_hw_queues, so
save old set->nr_hw_queues before call this function.

Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Weiping Zhang <zhangweiping@didiglobal.com>
---
 block/blk-mq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index f789b3e1b3ab..a79afbe60ca6 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -3347,11 +3347,11 @@ static void __blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set,
 		blk_mq_sysfs_unregister(q);
 	}
 
+	prev_nr_hw_queues = set->nr_hw_queues;
 	if (blk_mq_realloc_tag_set_tags(set, set->nr_hw_queues, nr_hw_queues) <
 	    0)
 		goto reregister;
 
-	prev_nr_hw_queues = set->nr_hw_queues;
 	set->nr_hw_queues = nr_hw_queues;
 	blk_mq_update_queue_map(set);
 fallback:
-- 
2.18.1

[RESEND v4 3/6] block: refactor __blk_mq_alloc_rq_maps

[Weiping Zhang]

This patch add a new member nr_allocated_map_rqs to the
struct blk_mq_tag_set to record the number of maps and requests have
been allocated for this tagset.

Now there are two problems when we increase/decrease hardware queue:

For increase, we do not allocate maps and request for the new
allocated hardware queue, it will be fixed in the next patch.

For decrease, when driver decrease hardware queue, set->nr_hw_queues
will be changed firstly in blk_mq_realloc_tag_set_tags or
__blk_mq_update_nr_hw_queues, then blk_mq_realloc_hw_ctxs and
blk_mq_map_swqueue, even blk_mq_free_tag_set have no chance to free
these hardware queue resource, because they iterate hardware queue by
for (i = 0; i < set->nr_hw_queues; i++).

Since request needs lots of memory, it's not easy alloc so many memory
dynamically, espeicially when system is under memory pressure.

This patch allow nr_hw_queues does not equal to the nr_allocated_map_rqs,
to avoid alloc/free memory when change hardware queue count.

Signed-off-by: Weiping Zhang <zhangweiping@didiglobal.com>
---
 block/blk-mq.c         | 26 +++++++++++++++++++-------
 include/linux/blk-mq.h |  1 +
 2 files changed, 20 insertions(+), 7 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index a79afbe60ca6..8393cb50bdc8 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2484,8 +2484,10 @@ static bool __blk_mq_alloc_rq_map(struct blk_mq_tag_set *set, int hctx_idx)
 
 	ret = blk_mq_alloc_rqs(set, set->tags[hctx_idx], hctx_idx,
 				set->queue_depth);
-	if (!ret)
+	if (!ret) {
+		set->nr_allocated_map_rqs++;
 		return true;
+	}
 
 	blk_mq_free_rq_map(set->tags[hctx_idx]);
 	set->tags[hctx_idx] = NULL;
@@ -2499,6 +2501,7 @@ static void blk_mq_free_map_and_requests(struct blk_mq_tag_set *set,
 		blk_mq_free_rqs(set, set->tags[hctx_idx], hctx_idx);
 		blk_mq_free_rq_map(set->tags[hctx_idx]);
 		set->tags[hctx_idx] = NULL;
+		set->nr_allocated_map_rqs--;
 	}
 }
 
@@ -2983,18 +2986,27 @@ void blk_mq_exit_queue(struct request_queue *q)
 	blk_mq_exit_hw_queues(q, set, set->nr_hw_queues);
 }
 
-static int __blk_mq_alloc_rq_maps(struct blk_mq_tag_set *set)
+/*
+ * Only append new map and requests, if new > now, all of these maps and
+ * request will be released when cleanup whole tag set. Because requests
+ * will cost lots memory, if system's memory is under a pressure, it's not
+ * easy to allocate too much memory.
+ */
+static int blk_mq_realloc_map_and_requests(struct blk_mq_tag_set *set, int new)
 {
-	int i;
+	int i, now = set->nr_allocated_map_rqs;
+
+	if (new <= now)
+		return 0;
 
-	for (i = 0; i < set->nr_hw_queues; i++)
+	for (i = now; i < new; i++)
 		if (!__blk_mq_alloc_rq_map(set, i))
 			goto out_unwind;
 
 	return 0;
 
 out_unwind:
-	while (--i >= 0)
+	while (--i >= now)
 		blk_mq_free_map_and_requests(set, i);
 
 	return -ENOMEM;
@@ -3012,7 +3024,7 @@ static int blk_mq_alloc_rq_maps(struct blk_mq_tag_set *set)
 
 	depth = set->queue_depth;
 	do {
-		err = __blk_mq_alloc_rq_maps(set);
+		err = blk_mq_realloc_map_and_requests(set, set->nr_hw_queues);
 		if (!err)
 			break;
 
@@ -3189,7 +3201,7 @@ void blk_mq_free_tag_set(struct blk_mq_tag_set *set)
 {
 	int i, j;
 
-	for (i = 0; i < set->nr_hw_queues; i++)
+	for (i = 0; i < set->nr_allocated_map_rqs; i++)
 		blk_mq_free_map_and_requests(set, i);
 
 	for (j = 0; j < set->nr_maps; j++) {
diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h
index f389d7c724bd..d950435cd3c6 100644
--- a/include/linux/blk-mq.h
+++ b/include/linux/blk-mq.h
@@ -240,6 +240,7 @@ struct blk_mq_tag_set {
 	unsigned int		nr_maps;
 	const struct blk_mq_ops	*ops;
 	unsigned int		nr_hw_queues;
+	unsigned int		nr_allocated_map_rqs;
 	unsigned int		queue_depth;
 	unsigned int		reserved_tags;
 	unsigned int		cmd_size;
-- 
2.18.1

[RESEND v4 4/6] block: alloc map and request for new hardware queue

[Weiping Zhang]

Alloc new map and request for new hardware queue when increse
hardware queue count. Before this patch, it will show a
warning for each new hardware queue, but it's not enough, these
hctx have no maps and reqeust, when a bio was mapped to these
hardware queue, it will trigger kernel panic when get request
from these hctx.

Test environment:
 * A NVMe disk supports 128 io queues
 * 96 cpus in system

A corner case can always trigger this panic, there are 96
io queues allocated for HCTX_TYPE_DEFAULT type, the corresponding kernel
log: nvme nvme0: 96/0/0 default/read/poll queues. Now we set nvme write
queues to 96, then nvme will alloc others(32) queues for read, but
blk_mq_update_nr_hw_queues does not alloc map and request for these new
added io queues. So when process read nvme disk, it will trigger kernel
panic when get request from these hardware context.

Reproduce script:

nr=$(expr `cat /sys/block/nvme0n1/device/queue_count` - 1)
echo $nr > /sys/module/nvme/parameters/write_queues
echo 1 > /sys/block/nvme0n1/device/reset_controller
dd if=/dev/nvme0n1 of=/dev/null bs=4K count=1

[ 8040.805626] ------------[ cut here ]------------
[ 8040.805627] WARNING: CPU: 82 PID: 12921 at block/blk-mq.c:2578 blk_mq_map_swqueue+0x2b6/0x2c0
[ 8040.805627] Modules linked in: nvme nvme_core nf_conntrack_netlink xt_addrtype br_netfilter overlay xt_CHECKSUM xt_MASQUERADE xt_conntrack ipt_REJECT nft_counter nf_nat_tftp nf_conntrack_tftp nft_masq nf_tables_set nft_fib_inet nft_f
ib_ipv4 nft_fib_ipv6 nft_fib nft_reject_inet nf_reject_ipv4 nf_reject_ipv6 nft_reject nft_ct nft_chain_nat nf_nat nf_conntrack tun bridge nf_defrag_ipv6 nf_defrag_ipv4 stp llc ip6_tables ip_tables nft_compat rfkill ip_set nf_tables nfne
tlink sunrpc intel_rapl_msr intel_rapl_common skx_edac nfit libnvdimm x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm irqbypass ipmi_ssif crct10dif_pclmul crc32_pclmul iTCO_wdt iTCO_vendor_support ghash_clmulni_intel intel_
cstate intel_uncore raid0 joydev intel_rapl_perf ipmi_si pcspkr mei_me ioatdma sg ipmi_devintf mei i2c_i801 dca lpc_ich ipmi_msghandler acpi_power_meter acpi_pad xfs libcrc32c sd_mod ast i2c_algo_bit drm_vram_helper drm_ttm_helper ttm d
rm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops
[ 8040.805637]  ahci drm i40e libahci crc32c_intel libata t10_pi wmi dm_mirror dm_region_hash dm_log dm_mod [last unloaded: nvme_core]
[ 8040.805640] CPU: 82 PID: 12921 Comm: kworker/u194:2 Kdump: loaded Tainted: G        W         5.6.0-rc5.78317c+ #2
[ 8040.805640] Hardware name: Inspur SA5212M5/YZMB-00882-104, BIOS 4.0.9 08/27/2019
[ 8040.805641] Workqueue: nvme-reset-wq nvme_reset_work [nvme]
[ 8040.805642] RIP: 0010:blk_mq_map_swqueue+0x2b6/0x2c0
[ 8040.805643] Code: 00 00 00 00 00 41 83 c5 01 44 39 6d 50 77 b8 5b 5d 41 5c 41 5d 41 5e 41 5f c3 48 8b bb 98 00 00 00 89 d6 e8 8c 81 03 00 eb 83 <0f> 0b e9 52 ff ff ff 0f 1f 00 0f 1f 44 00 00 41 57 48 89 f1 41 56
[ 8040.805643] RSP: 0018:ffffba590d2e7d48 EFLAGS: 00010246
[ 8040.805643] RAX: 0000000000000000 RBX: ffff9f013e1ba800 RCX: 000000000000003d
[ 8040.805644] RDX: ffff9f00ffff6000 RSI: 0000000000000003 RDI: ffff9ed200246d90
[ 8040.805644] RBP: ffff9f00f6a79860 R08: 0000000000000000 R09: 000000000000003d
[ 8040.805645] R10: 0000000000000001 R11: ffff9f0138c3d000 R12: ffff9f00fb3a9008
[ 8040.805645] R13: 000000000000007f R14: ffffffff96822660 R15: 000000000000005f
[ 8040.805645] FS:  0000000000000000(0000) GS:ffff9f013fa80000(0000) knlGS:0000000000000000
[ 8040.805646] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 8040.805646] CR2: 00007f7f397fa6f8 CR3: 0000003d8240a002 CR4: 00000000007606e0
[ 8040.805647] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 8040.805647] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 8040.805647] PKRU: 55555554
[ 8040.805647] Call Trace:
[ 8040.805649]  blk_mq_update_nr_hw_queues+0x31b/0x390
[ 8040.805650]  nvme_reset_work+0xb4b/0xeab [nvme]
[ 8040.805651]  process_one_work+0x1a7/0x370
[ 8040.805652]  worker_thread+0x1c9/0x380
[ 8040.805653]  ? max_active_store+0x80/0x80
[ 8040.805655]  kthread+0x112/0x130
[ 8040.805656]  ? __kthread_parkme+0x70/0x70
[ 8040.805657]  ret_from_fork+0x35/0x40
[ 8040.805658] ---[ end trace b5f13b1e73ccb5d3 ]---
[ 8229.365135] BUG: kernel NULL pointer dereference, address: 0000000000000004
[ 8229.365165] #PF: supervisor read access in kernel mode
[ 8229.365178] #PF: error_code(0x0000) - not-present page
[ 8229.365191] PGD 0 P4D 0
[ 8229.365201] Oops: 0000 [#1] SMP PTI
[ 8229.365212] CPU: 77 PID: 13024 Comm: dd Kdump: loaded Tainted: G        W         5.6.0-rc5.78317c+ #2
[ 8229.365232] Hardware name: Inspur SA5212M5/YZMB-00882-104, BIOS 4.0.9 08/27/2019
[ 8229.365253] RIP: 0010:blk_mq_get_tag+0x227/0x250
[ 8229.365265] Code: 44 24 04 44 01 e0 48 8b 74 24 38 65 48 33 34 25 28 00 00 00 75 33 48 83 c4 40 5b 5d 41 5c 41 5d 41 5e c3 48 8d 68 10 4c 89 ef <44> 8b 60 04 48 89 ee e8 dd f9 ff ff 83 f8 ff 75 c8 e9 67 fe ff ff
[ 8229.365304] RSP: 0018:ffffba590e977970 EFLAGS: 00010246
[ 8229.365317] RAX: 0000000000000000 RBX: ffff9f00f6a79860 RCX: ffffba590e977998
[ 8229.365333] RDX: 0000000000000000 RSI: ffff9f012039b140 RDI: ffffba590e977a38
[ 8229.365349] RBP: 0000000000000010 R08: ffffda58ff94e190 R09: ffffda58ff94e198
[ 8229.365365] R10: 0000000000000011 R11: ffff9f00f6a79860 R12: 0000000000000000
[ 8229.365381] R13: ffffba590e977a38 R14: ffff9f012039b140 R15: 0000000000000001
[ 8229.365397] FS:  00007f481c230580(0000) GS:ffff9f013f940000(0000) knlGS:0000000000000000
[ 8229.365415] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 8229.365428] CR2: 0000000000000004 CR3: 0000005f35e26004 CR4: 00000000007606e0
[ 8229.365444] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 8229.365460] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 8229.365476] PKRU: 55555554
[ 8229.365484] Call Trace:
[ 8229.365498]  ? finish_wait+0x80/0x80
[ 8229.365512]  blk_mq_get_request+0xcb/0x3f0
[ 8229.365525]  blk_mq_make_request+0x143/0x5d0
[ 8229.365538]  generic_make_request+0xcf/0x310
[ 8229.365553]  ? scan_shadow_nodes+0x30/0x30
[ 8229.365564]  submit_bio+0x3c/0x150
[ 8229.365576]  mpage_readpages+0x163/0x1a0
[ 8229.365588]  ? blkdev_direct_IO+0x490/0x490
[ 8229.365601]  read_pages+0x6b/0x190
[ 8229.365612]  __do_page_cache_readahead+0x1c1/0x1e0
[ 8229.365626]  ondemand_readahead+0x182/0x2f0
[ 8229.365639]  generic_file_buffered_read+0x590/0xab0
[ 8229.365655]  new_sync_read+0x12a/0x1c0
[ 8229.365666]  vfs_read+0x8a/0x140
[ 8229.365676]  ksys_read+0x59/0xd0
[ 8229.365688]  do_syscall_64+0x55/0x1d0
[ 8229.365700]  entry_SYSCALL_64_after_hwframe+0x44/0xa9

Signed-off-by: Weiping Zhang <zhangweiping@didiglobal.com>
---
 block/blk-mq.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 8393cb50bdc8..1c360b69155c 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -3366,6 +3366,13 @@ static void __blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set,
 
 	set->nr_hw_queues = nr_hw_queues;
 	blk_mq_update_queue_map(set);
+
+	if (blk_mq_realloc_map_and_requests(set, set->nr_hw_queues) < 0) {
+		pr_warn("Updating nr_hw_queues to %d fails, fallback to %d\n",
+			nr_hw_queues, prev_nr_hw_queues);
+		goto reregister;
+	}
+
 fallback:
 	list_for_each_entry(q, &set->tag_list, tag_set_list) {
 		blk_mq_realloc_hw_ctxs(set, q);
-- 
2.18.1

[RESEND v4 5/6] block: rename __blk_mq_alloc_rq_map

[Weiping Zhang]

rename __blk_mq_alloc_rq_map to __blk_mq_alloc_map_and_request,
actually it alloc both map and request, make function name
align with function.

Signed-off-by: Weiping Zhang <zhangweiping@didiglobal.com>
---
 block/blk-mq.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 1c360b69155c..b6c907dbbb30 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2473,7 +2473,7 @@ static void blk_mq_init_cpu_queues(struct request_queue *q,
 	}
 }
 
-static bool __blk_mq_alloc_rq_map(struct blk_mq_tag_set *set, int hctx_idx)
+static bool __blk_mq_alloc_map_and_request(struct blk_mq_tag_set *set, int hctx_idx)
 {
 	int ret = 0;
 
@@ -2527,7 +2527,7 @@ static void blk_mq_map_swqueue(struct request_queue *q)
 		hctx_idx = set->map[HCTX_TYPE_DEFAULT].mq_map[i];
 		/* unmapped hw queue can be remapped after CPU topo changed */
 		if (!set->tags[hctx_idx] &&
-		    !__blk_mq_alloc_rq_map(set, hctx_idx)) {
+		    !__blk_mq_alloc_map_and_request(set, hctx_idx)) {
 			/*
 			 * If tags initialization fail for some hctx,
 			 * that hctx won't be brought online.  In this
@@ -3000,7 +3000,7 @@ static int blk_mq_realloc_map_and_requests(struct blk_mq_tag_set *set, int new)
 		return 0;
 
 	for (i = now; i < new; i++)
-		if (!__blk_mq_alloc_rq_map(set, i))
+		if (!__blk_mq_alloc_map_and_request(set, i))
 			goto out_unwind;
 
 	return 0;
-- 
2.18.1

[RESEND v4 6/6] block: rename blk_mq_alloc_rq_maps

[Weiping Zhang]

rename blk_mq_alloc_rq_maps to blk_mq_alloc_map_and_requests,
this function allocs both map and request, make function name align
with funtion.

Signed-off-by: Weiping Zhang <zhangweiping@didiglobal.com>
---
 block/blk-mq.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index b6c907dbbb30..8ae9e375fe53 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -3017,7 +3017,7 @@ static int blk_mq_realloc_map_and_requests(struct blk_mq_tag_set *set, int new)
  * may reduce the depth asked for, if memory is tight. set->queue_depth
  * will be updated to reflect the allocated depth.
  */
-static int blk_mq_alloc_rq_maps(struct blk_mq_tag_set *set)
+static int blk_mq_alloc_map_and_requests(struct blk_mq_tag_set *set)
 {
 	unsigned int depth;
 	int err;
@@ -3177,7 +3177,7 @@ int blk_mq_alloc_tag_set(struct blk_mq_tag_set *set)
 	if (ret)
 		goto out_free_mq_map;
 
-	ret = blk_mq_alloc_rq_maps(set);
+	ret = blk_mq_alloc_map_and_requests(set);
 	if (ret)
 		goto out_free_mq_map;
 
-- 
2.18.1

Re: [RESEND v4 1/6] block: free both rq_map and request

[Ming Lei]

On Tue, Apr 28, 2020 at 9:29 PM Weiping Zhang
<zhangweiping@didiglobal.com> wrote:
>
> For this error handle, it should free both map and request,
> otherwise a memory leak occur.
>
> Allocation:
>
> __blk_mq_alloc_rq_map
>         blk_mq_alloc_rq_map
>                 blk_mq_alloc_rq_map
>                         tags = blk_mq_init_tags : kzalloc_node:
>                         tags->rqs = kcalloc_node
>                         tags->static_rqs = kcalloc_node
>         blk_mq_alloc_rqs
>                 p = alloc_pages_node
>                 tags->static_rqs[i] = p + offset;
>
> Free:
>
> blk_mq_free_rq_map
>         kfree(tags->rqs);
>         kfree(tags->static_rqs);
>         blk_mq_free_tags
>                 kfree(tags);
>
> The page allocated in blk_mq_alloc_rqs cannot be released,
> so we should use blk_mq_free_map_and_requests here.
>
> blk_mq_free_map_and_requests
>         blk_mq_free_rqs
>                 __free_pages : cleanup for blk_mq_alloc_rqs
>         blk_mq_free_rq_map : cleanup for blk_mq_alloc_rq_map
>
> Signed-off-by: Weiping Zhang <zhangweiping@didiglobal.com>
> ---
>  block/blk-mq.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index a7785df2c944..f789b3e1b3ab 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -2995,7 +2995,7 @@ static int __blk_mq_alloc_rq_maps(struct blk_mq_tag_set *set)
>
>  out_unwind:
>         while (--i >= 0)
> -               blk_mq_free_rq_map(set->tags[i]);
> +               blk_mq_free_map_and_requests(set, i);
>
>         return -ENOMEM;
>  }
> --
> 2.18.1
>

Reviewed-by: Ming Lei <ming.lei@redhat.com>

thanks,
Ming Lei

Re: [RESEND v4 2/6] block: save previous hardware queue count before udpate

[Ming Lei]

On Tue, Apr 28, 2020 at 9:29 PM Weiping Zhang
<zhangweiping@didiglobal.com> wrote:
>
> blk_mq_realloc_tag_set_tags will update set->nr_hw_queues, so
> save old set->nr_hw_queues before call this function.
>
> Reviewed-by: Bart Van Assche <bvanassche@acm.org>
> Signed-off-by: Weiping Zhang <zhangweiping@didiglobal.com>
> ---
>  block/blk-mq.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index f789b3e1b3ab..a79afbe60ca6 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -3347,11 +3347,11 @@ static void __blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set,
>                 blk_mq_sysfs_unregister(q);
>         }
>
> +       prev_nr_hw_queues = set->nr_hw_queues;
>         if (blk_mq_realloc_tag_set_tags(set, set->nr_hw_queues, nr_hw_queues) <
>             0)
>                 goto reregister;
>
> -       prev_nr_hw_queues = set->nr_hw_queues;
>         set->nr_hw_queues = nr_hw_queues;
>         blk_mq_update_queue_map(set);
>  fallback:
> --
> 2.18.1
>

Reviewed-by: Ming Lei <ming.lei@redhat.com>

thanks,
Ming Lei

Re: [RESEND v4 4/6] block: alloc map and request for new hardware queue

[Ming Lei]

On Tue, Apr 28, 2020 at 09:29:59PM +0800, Weiping Zhang wrote:
> Alloc new map and request for new hardware queue when increse
> hardware queue count. Before this patch, it will show a
> warning for each new hardware queue, but it's not enough, these
> hctx have no maps and reqeust, when a bio was mapped to these
> hardware queue, it will trigger kernel panic when get request
> from these hctx.
> 
> Test environment:
>  * A NVMe disk supports 128 io queues
>  * 96 cpus in system
> 
> A corner case can always trigger this panic, there are 96
> io queues allocated for HCTX_TYPE_DEFAULT type, the corresponding kernel
> log: nvme nvme0: 96/0/0 default/read/poll queues. Now we set nvme write
> queues to 96, then nvme will alloc others(32) queues for read, but
> blk_mq_update_nr_hw_queues does not alloc map and request for these new
> added io queues. So when process read nvme disk, it will trigger kernel
> panic when get request from these hardware context.
> 

map and request is supposed to be allocated via __blk_mq_alloc_rq_map() called
from blk_mq_map_swqueue(), so why is such issue triggered?

The reason could be that only queue type of HCTX_TYPE_DEFAULT is handled
in __blk_mq_alloc_rq_map(), and looks all queue types should have been covered
here.

Could you test the following patch and see if the issue can be fixed?
Then we can save one intermediate variable if it helps.


diff --git a/block/blk-mq.c b/block/blk-mq.c
index 12dee4ecd5cc..7e77e27b613e 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2742,19 +2742,6 @@ static void blk_mq_map_swqueue(struct request_queue *q)
 	 * If the cpu isn't present, the cpu is mapped to first hctx.
 	 */
 	for_each_possible_cpu(i) {
-		hctx_idx = set->map[HCTX_TYPE_DEFAULT].mq_map[i];
-		/* unmapped hw queue can be remapped after CPU topo changed */
-		if (!set->tags[hctx_idx] &&
-		    !__blk_mq_alloc_rq_map(set, hctx_idx)) {
-			/*
-			 * If tags initialization fail for some hctx,
-			 * that hctx won't be brought online.  In this
-			 * case, remap the current ctx to hctx[0] which
-			 * is guaranteed to always have tags allocated
-			 */
-			set->map[HCTX_TYPE_DEFAULT].mq_map[i] = 0;
-		}
-
 		ctx = per_cpu_ptr(q->queue_ctx, i);
 		for (j = 0; j < set->nr_maps; j++) {
 			if (!set->map[j].nr_queues) {
@@ -2763,6 +2750,19 @@ static void blk_mq_map_swqueue(struct request_queue *q)
 				continue;
 			}
 
+			/* unmapped hw queue can be remapped after CPU topo changed */
+			hctx_idx = set->map[j].mq_map[i];
+			if (!set->tags[hctx_idx] &&
+			    !__blk_mq_alloc_rq_map(set, hctx_idx)) {
+				/*
+				 * If tags initialization fail for some hctx,
+				 * that hctx won't be brought online.  In this
+				 * case, remap the current ctx to hctx[0] which
+				 * is guaranteed to always have tags allocated
+				 */
+				set->map[j].mq_map[i] = 0;
+			}
+
 			hctx = blk_mq_map_queue_type(q, j, i);
 			ctx->hctxs[j] = hctx;
 			/*


Thanks,
Ming

Re: [RESEND v4 5/6] block: rename __blk_mq_alloc_rq_map

[Ming Lei]

On Tue, Apr 28, 2020 at 09:30:13PM +0800, Weiping Zhang wrote:
> rename __blk_mq_alloc_rq_map to __blk_mq_alloc_map_and_request,
> actually it alloc both map and request, make function name
> align with function.
> 
> Signed-off-by: Weiping Zhang <zhangweiping@didiglobal.com>
> ---
>  block/blk-mq.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index 1c360b69155c..b6c907dbbb30 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -2473,7 +2473,7 @@ static void blk_mq_init_cpu_queues(struct request_queue *q,
>  	}
>  }
>  
> -static bool __blk_mq_alloc_rq_map(struct blk_mq_tag_set *set, int hctx_idx)
> +static bool __blk_mq_alloc_map_and_request(struct blk_mq_tag_set *set, int hctx_idx)
>  {
>  	int ret = 0;
>  
> @@ -2527,7 +2527,7 @@ static void blk_mq_map_swqueue(struct request_queue *q)
>  		hctx_idx = set->map[HCTX_TYPE_DEFAULT].mq_map[i];
>  		/* unmapped hw queue can be remapped after CPU topo changed */
>  		if (!set->tags[hctx_idx] &&
> -		    !__blk_mq_alloc_rq_map(set, hctx_idx)) {
> +		    !__blk_mq_alloc_map_and_request(set, hctx_idx)) {
>  			/*
>  			 * If tags initialization fail for some hctx,
>  			 * that hctx won't be brought online.  In this
> @@ -3000,7 +3000,7 @@ static int blk_mq_realloc_map_and_requests(struct blk_mq_tag_set *set, int new)
>  		return 0;
>  
>  	for (i = now; i < new; i++)
> -		if (!__blk_mq_alloc_rq_map(set, i))
> +		if (!__blk_mq_alloc_map_and_request(set, i))
>  			goto out_unwind;
>  
>  	return 0;
> -- 
> 2.18.1
> 

Reviewed-by: Ming Lei <ming.lei@redhat.com>

-- 
Ming

Re: [RESEND v4 6/6] block: rename blk_mq_alloc_rq_maps

[Ming Lei]

On Tue, Apr 28, 2020 at 09:30:28PM +0800, Weiping Zhang wrote:
> rename blk_mq_alloc_rq_maps to blk_mq_alloc_map_and_requests,
> this function allocs both map and request, make function name align
> with funtion.
> 
> Signed-off-by: Weiping Zhang <zhangweiping@didiglobal.com>
> ---
>  block/blk-mq.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index b6c907dbbb30..8ae9e375fe53 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -3017,7 +3017,7 @@ static int blk_mq_realloc_map_and_requests(struct blk_mq_tag_set *set, int new)
>   * may reduce the depth asked for, if memory is tight. set->queue_depth
>   * will be updated to reflect the allocated depth.
>   */
> -static int blk_mq_alloc_rq_maps(struct blk_mq_tag_set *set)
> +static int blk_mq_alloc_map_and_requests(struct blk_mq_tag_set *set)
>  {
>  	unsigned int depth;
>  	int err;
> @@ -3177,7 +3177,7 @@ int blk_mq_alloc_tag_set(struct blk_mq_tag_set *set)
>  	if (ret)
>  		goto out_free_mq_map;
>  
> -	ret = blk_mq_alloc_rq_maps(set);
> +	ret = blk_mq_alloc_map_and_requests(set);
>  	if (ret)
>  		goto out_free_mq_map;
>  
> -- 
> 2.18.1
> 

Reviewed-by: Ming Lei <ming.lei@redhat.com>

-- 
Ming

Re: [RESEND v4 4/6] block: alloc map and request for new hardware queue

[Weiping Zhang]

On Sun, May 3, 2020 at 11:22 AM Ming Lei <ming.lei@redhat.com> wrote:
>
> On Tue, Apr 28, 2020 at 09:29:59PM +0800, Weiping Zhang wrote:
> > Alloc new map and request for new hardware queue when increse
> > hardware queue count. Before this patch, it will show a
> > warning for each new hardware queue, but it's not enough, these
> > hctx have no maps and reqeust, when a bio was mapped to these
> > hardware queue, it will trigger kernel panic when get request
> > from these hctx.
> >
> > Test environment:
> >  * A NVMe disk supports 128 io queues
> >  * 96 cpus in system
> >
> > A corner case can always trigger this panic, there are 96
> > io queues allocated for HCTX_TYPE_DEFAULT type, the corresponding kernel
> > log: nvme nvme0: 96/0/0 default/read/poll queues. Now we set nvme write
> > queues to 96, then nvme will alloc others(32) queues for read, but
> > blk_mq_update_nr_hw_queues does not alloc map and request for these new
> > added io queues. So when process read nvme disk, it will trigger kernel
> > panic when get request from these hardware context.
> >
>
> map and request is supposed to be allocated via __blk_mq_alloc_rq_map() called
> from blk_mq_map_swqueue(), so why is such issue triggered?
>
> The reason could be that only queue type of HCTX_TYPE_DEFAULT is handled
> in __blk_mq_alloc_rq_map(), and looks all queue types should have been covered
> here.
>
> Could you test the following patch and see if the issue can be fixed?
> Then we can save one intermediate variable if it helps.
>
Hello Ming,

It works well, so the patch-3 can be dropped.
For patch-3, there is no memory leak when decrease hardware queue count,
blk_mq_realloc_hw_ctxs will release the unused rq_map and request by
blk_mq_free_map_and_requests.


>
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index 12dee4ecd5cc..7e77e27b613e 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -2742,19 +2742,6 @@ static void blk_mq_map_swqueue(struct request_queue *q)
>          * If the cpu isn't present, the cpu is mapped to first hctx.
>          */
>         for_each_possible_cpu(i) {
> -               hctx_idx = set->map[HCTX_TYPE_DEFAULT].mq_map[i];
> -               /* unmapped hw queue can be remapped after CPU topo changed */
> -               if (!set->tags[hctx_idx] &&
> -                   !__blk_mq_alloc_rq_map(set, hctx_idx)) {
> -                       /*
> -                        * If tags initialization fail for some hctx,
> -                        * that hctx won't be brought online.  In this
> -                        * case, remap the current ctx to hctx[0] which
> -                        * is guaranteed to always have tags allocated
> -                        */
> -                       set->map[HCTX_TYPE_DEFAULT].mq_map[i] = 0;
> -               }
> -
>                 ctx = per_cpu_ptr(q->queue_ctx, i);
>                 for (j = 0; j < set->nr_maps; j++) {
>                         if (!set->map[j].nr_queues) {
> @@ -2763,6 +2750,19 @@ static void blk_mq_map_swqueue(struct request_queue *q)
>                                 continue;
>                         }
>
> +                       /* unmapped hw queue can be remapped after CPU topo changed */
> +                       hctx_idx = set->map[j].mq_map[i];
> +                       if (!set->tags[hctx_idx] &&
> +                           !__blk_mq_alloc_rq_map(set, hctx_idx)) {
> +                               /*
> +                                * If tags initialization fail for some hctx,
> +                                * that hctx won't be brought online.  In this
> +                                * case, remap the current ctx to hctx[0] which
> +                                * is guaranteed to always have tags allocated
> +                                */
> +                               set->map[j].mq_map[i] = 0;
> +                       }
> +
>                         hctx = blk_mq_map_queue_type(q, j, i);
>                         ctx->hctxs[j] = hctx;
>                         /*
>
>
> Thanks,
> Ming
>