blkdebug: fix "once" rule
PATCH v2 1/1 blkdebug: fix "once" rule
Background:
The blkdebug scripts are currently engineered so that when a debug
event occurs, a prefilter browses a master list of parsed rules for a certain event and adds them to an "active list" of rules to be used for the forthcoming action, provided the events and state numbers match.
Then, once the request is received, the last active rule is used to
inject an error if certain parameters match.
This active list is cleared every time the prefilter injects a new
rule for the first time during a debug event.
The "once" rule currently causes the error injection, if it is
triggered, to only clear the active list. This is insufficient for preventing future injections of the same rule.
Remedy:
This patch /deletes/ the rule from the list that the prefilter
browses, so it is gone for good. In V2, we remove only the rule of interest from the active list instead of allowing the "once" rule to clear the entire list of active rules.
Impact:
This affects iotests 026. Several ENOSPC tests that used "once" can
be seen to have output that shows multiple failure messages. After this patch, the error messages tend to be smaller and less severe, but the injection can still be seen to be working. I have patched the expected output to expect the smaller error messages.
V2:
- Remove only the offending "once" rule from the active list instead of clearing the entire active list.
--- block/blkdebug.c | 9 +++++---- tests/qemu-iotests/026.out | 24 ++++-------------------- 2 files changed, 9 insertions(+), 24 deletions(-)
block/blkdebug.c +5 -4
@@ -473,14 +473,15 @@ static BlockAIOCB *inject_error(BlockDriverState *bs,struct BlkdebugAIOCB *acb;QEMUBH *bh; -if (rule->options.inject.once) { -QSIMPLEQ_INIT (&s->active_rules); - } -if (rule->options.inject.immediately) {return NULL ; } +if (rule->options.inject.once) { +QSIMPLEQ_REMOVE (&s->active_rules, rule, BlkdebugRule, active_next); +remove_rule (rule); + } + acb =qemu_aio_get (&blkdebug_aiocb_info, bs, cb, opaque); acb->ret = -error;
tests/qemu-iotests/026.out +4 -20
@@ -140,19 +140,13 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_update; errno: 5; imm: off; once: on; write -Failed to flush the L2 table cache: Input/output error write failed: Input/output error - -127 leaked clusters were found on the image. -This means waste of disk space, but no harm to data. +No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_update; errno: 5; imm: off; once: on; write -b -Failed to flush the L2 table cache: Input/output error write failed: Input/output error - -127 leaked clusters were found on the image. -This means waste of disk space, but no harm to data. +No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_update; errno: 5; imm: off; once: off; write
@@ -174,19 +168,13 @@ This means waste of disk space, but no harm to data. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_update; errno: 28; imm: off; once: on; write -Failed to flush the L2 table cache: No space left on device write failed: No space left on device - -127 leaked clusters were found on the image. -This means waste of disk space, but no harm to data. +No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_update; errno: 28; imm: off; once: on; write -b -Failed to flush the L2 table cache: No space left on device write failed: No space left on device - -127 leaked clusters were found on the image. -This means waste of disk space, but no harm to data. +No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_update; errno: 28; imm: off; once: off; write
@@ -356,13 +344,11 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_update_part; errno: 5; imm: off; once: on; write -Failed to flush the refcount block cache: Input/output error write failed: Input/output error No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_update_part; errno: 5; imm: off; once: on; write -b -Failed to flush the refcount block cache: Input/output error write failed: Input/output error No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824
@@ -382,13 +368,11 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_update_part; errno: 28; imm: off; once: on; write -Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_update_part; errno: 28; imm: off; once: on; write -b -Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824
Signature · 1 line
1.9.3
Quoted 29 lines · John Snow
Background:
The blkdebug scripts are currently engineered so that when a debug
event occurs, a prefilter browses a master list of parsed rules for a certain event and adds them to an "active list" of rules to be used for the forthcoming action, provided the events and state numbers match.Then, once the request is received, the last active rule is used to
inject an error if certain parameters match.This active list is cleared every time the prefilter injects a new
rule for the first time during a debug event.The "once" rule currently causes the error injection, if it is
triggered, to only clear the active list. This is insufficient for preventing future injections of the same rule.Remedy:
This patch /deletes/ the rule from the list that the prefilter
browses, so it is gone for good. In V2, we remove only the rule of interest from the active list instead of allowing the "once" rule to clear the entire list of active rules.Impact:
This affects iotests 026. Several ENOSPC tests that used "once" can
be seen to have output that shows multiple failure messages. After this patch, the error messages tend to be smaller and less severe, but the injection can still be seen to be working. I have patched the expected output to expect the smaller error messages.
Quoted 3 lines
V2:
- Remove only the offending "once" rule from the active list instead of clearing the entire active list.
Could you move this version information under the three dashes so it won't be included in the commit message?
Quoted 2 lines
Signed-off-by John Snow <[email protected]>---
Here, that is.
Quoted 18 lines
block/blkdebug.c | 9 +++++---- tests/qemu-iotests/026.out | 24 ++++-------------------- 2 files changed, 9 insertions(+), 24 deletions(-)block/blkdebug.c +0
-4@@ -473,14 +473,15 @@ static BlockAIOCB *inject_error(BlockDriverState *bs,struct BlkdebugAIOCB *acb;QEMUBH *bh; -if (rule->options.inject.once) { -QSIMPLEQ_INIT (&s->active_rules); - } -if (rule->options.inject.immediately) {return NULL ;
block/blkdebug.c static BlockAIOCB *inject_error(BlockDriverState *bs, · line 487
Quoted 10 lines
block/blkdebug.c +5
-0} +if (rule->options.inject.once) { +QSIMPLEQ_REMOVE (&s->active_rules, rule, BlkdebugRule, active_next); +remove_rule (rule); + } + acb =qemu_aio_get (&blkdebug_aiocb_info, bs, cb, opaque); acb->ret = -error;
The new code itself looks good, but why did you move it after the immediately block? Because now, if "immediately" is set, "once" will be ignored:
$ ./qemu-img create -f qcow2 test.qcow2 64M; ./qemu-io -c 'aio_write 0 64k' "json:{'driver':'qcow2','file':{'driver':'blkdebug','image':{'driver':'file','filename':'test.qcow2'},'inject-error':[{'event':'write_aio','once':true,'immediately':true}]}}" Formatting 'test.qcow2', fmt=qcow2 size=67108864 encryption=off cluster_size=65536 lazy_refcounts=off aio_write failed: Input/output error
Failed to flush the L2 table cache: Input/output error
Failed to flush the refcount block cache: Input/output error
(I know, I know, I should have noticed in v1...)
Max
Quoted 84 lines · Max Reitz
Quoted 29 lines · John Snow
Background:
The blkdebug scripts are currently engineered so that when a debug
event occurs, a prefilter browses a master list of parsed rules for a certain event and adds them to an "active list" of rules to be used for the forthcoming action, provided the events and state numbers match.Then, once the request is received, the last active rule is used to
inject an error if certain parameters match.This active list is cleared every time the prefilter injects a new
rule for the first time during a debug event.The "once" rule currently causes the error injection, if it is
triggered, to only clear the active list. This is insufficient for preventing future injections of the same rule.Remedy:
This patch /deletes/ the rule from the list that the prefilter
browses, so it is gone for good. In V2, we remove only the rule of interest from the active list instead of allowing the "once" rule to clear the entire list of active rules.Impact:
This affects iotests 026. Several ENOSPC tests that used "once" can
be seen to have output that shows multiple failure messages. After this patch, the error messages tend to be smaller and less severe, but the injection can still be seen to be working. I have patched the expected output to expect the smaller error messages.Quoted 3 lines
V2:
- Remove only the offending "once" rule from the active list instead of clearing the entire active list.Could you move this version information under the three dashes so it won't be included in the commit message?
Quoted 2 lines
Signed-off-by John Snow <[email protected]>---
Here, that is.
Quoted 17 lines
block/blkdebug.c | 9 +++++---- tests/qemu-iotests/026.out | 24 ++++-------------------- 2 files changed, 9 insertions(+), 24 deletions(-)block/blkdebug.c +0
-4@@ -473,14 +473,15 @@ static BlockAIOCB *inject_error(BlockDriverState *bs, struct BlkdebugAIOCB *acb;QEMUBH *bh; -if (rule->options.inject.once) { -QSIMPLEQ_INIT (&s->active_rules); - } -if (rule->options.inject.immediately) {block/blkdebug.c static BlockAIOCB *inject_error(BlockDriverState · line 486
Quoted 10 lines
block/blkdebug.c +5
-0return NULL ; } +if (rule->options.inject.once) { +QSIMPLEQ_REMOVE (&s->active_rules, rule, BlkdebugRule, active_next); +remove_rule (rule); + } + acb =qemu_aio_get (&blkdebug_aiocb_info, bs, cb, opaque); acb->ret = -error;The new code itself looks good, but why did you move it after the immediately block? Because now, if "immediately" is set, "once" will be ignored:
$ ./qemu-img create -f qcow2 test.qcow2 64M; ./qemu-io -c 'aio_write 0 64k' "json:{'driver':'qcow2','file':{'driver':'blkdebug','image':{'driver':'file','filename':'test.qcow2'},'inject-error':[{'event':'write_aio','once':true,'immediately':true}]}}"Formatting 'test.qcow2', fmt=qcow2 size=67108864 encryption=off cluster_size=65536 lazy_refcounts=off
aio_writefailed: Input/output error
Failed to flush the L2 table cache: Input/output error Failed to flush the refcount block cache: Input/output error
Quoted 4 lines
(I know, I know, I should have noticed in v1...)
Max
augh, durr.
Because I didn't want the deletion of the rule to trigger a segfault, but I wasn't thinking about the interaction of both rules here. Duhhh.