Making the usage details a little better. #35
Merged
yehudasa merged 1 commit intoceph:masterfrom Jan 23, 2013
cholcombe973:master
Merged
Making the usage details a little better. #35yehudasa merged 1 commit intoceph:masterfrom cholcombe973:master
yehudasa merged 1 commit intoceph:masterfrom
cholcombe973:master
Conversation
Improved the usage message.
yehudasa
added a commit
that referenced
this pull request
Jan 23, 2013
Making the usage details a little better.
Member
|
lol, or I guess Yehuda is ok with the original text :) |
chamdoo
pushed a commit
to chamdoo/ceph
that referenced
this pull request
Nov 13, 2015
Remove utilities/.DS_Store file.
XinzeChi
added a commit
to XinzeChi/ceph
that referenced
this pull request
Jan 29, 2016
os: fix perf counter for file journal
ivancich
added a commit
to ivancich/ceph-fork
that referenced
this pull request
Sep 13, 2017
a9e777f Merge pull request ceph#35 from joscollin/wip-warning-dmclock 45c63a5 Merge pull request ceph#36 from ceph/wip-fix-functional 40a9fe5 With upgrade to gcc 7.1.1 need to include <functional> in additional header files. a2dd155 src/dmclock_server.h: silence warning from -Wmaybe-uninitialized 0d6c7c6 Merge pull request ceph#24 from bspark8/wip_online_client_info_f a5e8cde Add unit test for update_client_info, dynamic_cli_info_f. 1faf352 Modify each client's QoS parameter applied time from client_rec creation time to each request's tagging creation time. git-subtree-dir: src/dmclock git-subtree-split: a9e777f
joscollin
referenced
this pull request
in joscollin/ceph
Sep 14, 2017
a9e777f Merge pull request #35 from joscollin/wip-warning-dmclock 45c63a5 Merge pull request #36 from ceph/wip-fix-functional 40a9fe5 With upgrade to gcc 7.1.1 need to include <functional> in additional header files. a2dd155 src/dmclock_server.h: silence warning from -Wmaybe-uninitialized 0d6c7c6 Merge pull request #24 from bspark8/wip_online_client_info_f a5e8cde Add unit test for update_client_info, dynamic_cli_info_f. 1faf352 Modify each client's QoS parameter applied time from client_rec creation time to each request's tagging creation time. git-subtree-dir: src/dmclock git-subtree-split: a9e777f
joscollin
referenced
this pull request
in joscollin/ceph
Sep 14, 2017
Closed
rjfd
added a commit
to rjfd/ceph
that referenced
this pull request
Jan 26, 2018
mgr/dashboard_v2/frontend: add support for CI coverage and unit testing
sdpeters
pushed a commit
to sdpeters/ceph
that referenced
this pull request
May 29, 2019
…y existing pool file with the same name should be deleted. Signed-off-by: Yuan Lu <yuan.y.lu@intel.com>
sdpeters
added a commit
to sdpeters/ceph
that referenced
this pull request
May 29, 2019
for Ceph RBD. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Resorting to pruning out tests that fail with RWL. Don't add RWL on images with the MIGRATING feature bit They src and dest may have the same image name. Also don't create image writeback at all if there is no image cache configured. Make deep copy test use different image names for src and dest Fix tests for added async completions, used whether RWL is enabled or not. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> This block guard implementation isn't used anymore Handle zero length requests Tighten up block guard release. Remove assert for what wasn't a bug. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Skip these tests when RWL is enabled for now. It's unclear whether they should work with the cache enabled or not. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Do a flush() before shut_down() to complete all in-flight writes and append a final sync point. Flush to RADOS on ::flush() Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make DiscardCopyup test not simultaneously open the same image > once Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Replaced ineffective pmem_bl copies. TestInternal passes now with RWL enabled. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> TestFUA is incompatible with RWL (multiple open of same image) Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Internal flush may be called before init succeeds Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make RWL perf logging configurable Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Require !RWL for tests that open images > once Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Log image name and ID in key ops. Use async completion to avoid a deadlock seen during testing Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Flush/invalidate image cache during Resize Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Disable more tests that open images > once. DeepCopy tests shoudl work on RWL, but don't. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make invalidate after flush configurable Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> When RWL is enabled, set environment variable to skip actual PMDK persist work Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> All migration tests skipped when RWL is enabled. Migration probably needs to disable ImageCache first. One more DeepCopy test that dosn't work with RWL. Tweak tests to not open images > once Tests can close all images before fixture teardown. Avoids reopens that fil with write-back cache. Allow saved image_ctxs to be closed by name Close the image before re-opening it Skip tests that reopen images when RWL is enabled Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Disable slow PMDK persistence in tests Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Consolidate constructors Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Force init, shutdown, and flush() completions to op_work_queue Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Didn't need a special case for image cache after all Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix apparent copy-and-paste error from KernelDevice.cc Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Revert gratuitous whitespace changes Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Adds IMAGE_CACHE feature bit. Moves image cache init logic to ImageState This is in prep. for initializing ImageCache during exclusive lock acquire. Add types for image cache config Can fetch image cache state during open Record ImageCache state on image close Init & shut down image cache in exclusive lock acquire/release Disable checks for dirty. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Lock tests expect specific feature tests Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Remove deprecated io prio calls Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Include image-cache feature options Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make mock tests work with IMAGE_CACHE feature on Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make more tests work with IMAGE_CACHE feature enabled Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> More adaptation to the feature bit universe Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Reworked ImageRequestWQ<>::flush() to use ::aio_flush() again, but with a FlushSource arg. Avoids use of flush_image(), which I think was getting blocked with the user requests in the DeepCopy test, which was failing. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Flush to journal for internal or user flush. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Undo previous. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Add IMAGE_CACHE cases to RBD unit test Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix apparent copy-and-paste error Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Don't build PMDK or RWL on ARM while PMDK on ARM64 is still experimental Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Skip test_librbd when features includes IMAGE_CACHE, but WITH_RWL is not defined. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make read_only flag const, since that's how it's used Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix message Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Don't init image cache when opening R/O unless there's a dirty one Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Add evn vars to inhibit flush and retire on close for RWL debugging Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Add ImageCache action to build Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Add env vars for testing image cache rbd tool actions There's a regression here, but I'm committing anyway so I can migrate to gitlab now. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix lockdep issue in shutdown when not flushing/retiring. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Use FLUSH_SOURCE_SHUTDOWN to identify shutown flushes. In general, a whoile lot of plumbing to get the ability to suppress RWL flushing on close, so we can test removing and discarding the image-cache with the rbd tool. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix rbd cli test output for new image-cache comand and --no-flush option to bench Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Clean up logging from previous test Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix expected calls of block_writes() which now include the flush source. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Always init exclusive lock when the image cache feature is enabled. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Remove bogus assert Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Checkpoint rbd image-cache command Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Revert "Always init exclusive lock when the image cache feature is enabled." This reverts commit 314c82f. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix message Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Track image-cache option changes Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fixed some things clang didn't like, which may have broken rtags Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Split WITH_PMEM into WITH_PMEM_DEVICE and WITH_PMDK WITH_PMDK causes the PMDK (/NVML) to be built. WITH_PMEM_DEVICE causes the PMEM OSD to be built. WITH_PMEM_DEVICE, or WITH_RWL requires WITH_PMDK Primordial support for WITH_PMDK_PKG enables use of PMDK installed on platform. We'd probably like finer grained control over this (build most of Ceph without PMDK installed, and make PMDK an install dependency for only a few selected things). We can now enable RWL without requiring the PMEM OSD to build and pass tests. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Don't assert if exclusive lock release failed. On failure, these things may not have been done. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Also make etags file ctags -e, into src/TAGS. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Revert "Don't assert if exclusive lock release failed." This reverts commit 184ba6f5e97357380d30f8012f28c34c38e6818c Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Disable almost all deep copy tests when RWL is enabled These used to pass, and made interesting test cases, but it probably doesn't make sense to do this with the cache enabled anyway. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Ignore etags file TAGS Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Adds ReplicatedWriteLogInternal.h Hopefully trims enough off of ReplicatedWriteLog.cc that gitlab will show the diffs in a merge request. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Shift more things from RWL.cc to RWLInternal.h Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Finish moving lines Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> One more try for gitlab-friendly file sizes. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> To fix IO hange problem caused by aio_flush. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Add IO test for RWL It does IO test with RWL, and tries to cover the test case when RWL cache is full. But it is not guranteed. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix the bug 'First IO hangs if RWL can't be initialized'. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Update default and minimum value of rwl_size. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fixes the image-cache show command Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Update RWL full case to use default value. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Add pool name in the name of rwl cache file. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fixed rebase issues. Look closer at whether FLUSH_SOURCE_WRITEBACK should be a no-op in RWL, or an assert. Look closer at discard granularity passed to lower cache. This replaces skip_partial. We were always passing skip_partial = false, so the RWL and the image got the same discard results (regions which read as zero, vs. whatever was there before). Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix expected usage output Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Probably rebase noise Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Repeat compare buffer as necessary to match length of read in compare-and-write Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Disable some more migration tests when IMAGE_CACHE feature is enabled. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> fix image-cache disable feature command Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Start rwl threads after rwl_init. Signed-off-by: Li Xiaoyan <xiaoyan.li@intel.com> change ENOENT to EPERM when open image cache failed; Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> fixed issue ceph#35: If the image metadata says there is no cache, any existing pool file with the same name should be deleted. Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> only image cache state.present=0 and has pool file, then del the pool file with the same name; Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> do this in rwl_int() function; Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> Pass through everything to writeback interface Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> fix ceph#41:IO hang when rwl_size = 10G or 20G and without rate_iops limit, While there's no IO problem when rwl_size=1G Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> add high/low water log entries threshold and trigger it to retire entries when there's no enough log entries; Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> clean indent issue; Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> modify indentation based on Scott's feedback; Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> Fix the issue of flushing rwl cache when showing image cache. Signed-off-by: Xiaoyan Li <xiaoyan.li@intel.com> Fix image-cache show issue. Signed-off-by: Xiaoyan Li <xiaoyan.li@intel.com> image-cache discard-dirty CLI command Signed-off-by: Mahati Chamarthy <mahati.chamarthy@intel.com> pmemobj_close won't fail. Signed-off-by: Xiaoyan Li <xiaoyan.li@intel.com> Rebase repairs. Mostly trivial except async op handling in flush. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make imageFlushRequest::send_request() more like it was before the rebase. All in pursuit of issue ceph#48 after the latest rebase. The aio_comp() operations in comments there cause an assert in init or open now. FLUSH_SOURCE_WRITEBACK requests complete synchonously now, like they did before the rebase. None of this fixes the test assert in issue ceph#48. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Replace namespace declaration that is present in master but not our branch Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Supplied initial values for things not initialized by all constructors I saw compiler warnings about these fields being used uninitialized and thought it might be related to the rebase regression (it probably isn't). It does look like these could be left uninitialized if the ReplicatedWriteLogSpec constructor with no argumets was used. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Bad merge Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix unit test problem. Signed-off-by: Xiaoyan Li <xiaoyan.li@intel.com>
sdpeters
pushed a commit
to sdpeters/ceph
that referenced
this pull request
May 29, 2019
…y existing pool file with the same name should be deleted. Signed-off-by: Yuan Lu <yuan.y.lu@intel.com>
sdpeters
added a commit
to sdpeters/ceph
that referenced
this pull request
May 31, 2019
for Ceph RBD. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Resorting to pruning out tests that fail with RWL. Don't add RWL on images with the MIGRATING feature bit They src and dest may have the same image name. Also don't create image writeback at all if there is no image cache configured. Make deep copy test use different image names for src and dest Fix tests for added async completions, used whether RWL is enabled or not. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> This block guard implementation isn't used anymore Handle zero length requests Tighten up block guard release. Remove assert for what wasn't a bug. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Skip these tests when RWL is enabled for now. It's unclear whether they should work with the cache enabled or not. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Do a flush() before shut_down() to complete all in-flight writes and append a final sync point. Flush to RADOS on ::flush() Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make DiscardCopyup test not simultaneously open the same image > once Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Replaced ineffective pmem_bl copies. TestInternal passes now with RWL enabled. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> TestFUA is incompatible with RWL (multiple open of same image) Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Internal flush may be called before init succeeds Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make RWL perf logging configurable Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Require !RWL for tests that open images > once Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Log image name and ID in key ops. Use async completion to avoid a deadlock seen during testing Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Flush/invalidate image cache during Resize Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Disable more tests that open images > once. DeepCopy tests shoudl work on RWL, but don't. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make invalidate after flush configurable Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> When RWL is enabled, set environment variable to skip actual PMDK persist work Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> All migration tests skipped when RWL is enabled. Migration probably needs to disable ImageCache first. One more DeepCopy test that dosn't work with RWL. Tweak tests to not open images > once Tests can close all images before fixture teardown. Avoids reopens that fil with write-back cache. Allow saved image_ctxs to be closed by name Close the image before re-opening it Skip tests that reopen images when RWL is enabled Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Disable slow PMDK persistence in tests Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Consolidate constructors Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Force init, shutdown, and flush() completions to op_work_queue Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Didn't need a special case for image cache after all Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix apparent copy-and-paste error from KernelDevice.cc Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Revert gratuitous whitespace changes Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Adds IMAGE_CACHE feature bit. Moves image cache init logic to ImageState This is in prep. for initializing ImageCache during exclusive lock acquire. Add types for image cache config Can fetch image cache state during open Record ImageCache state on image close Init & shut down image cache in exclusive lock acquire/release Disable checks for dirty. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Lock tests expect specific feature tests Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Remove deprecated io prio calls Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Include image-cache feature options Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make mock tests work with IMAGE_CACHE feature on Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make more tests work with IMAGE_CACHE feature enabled Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> More adaptation to the feature bit universe Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Reworked ImageRequestWQ<>::flush() to use ::aio_flush() again, but with a FlushSource arg. Avoids use of flush_image(), which I think was getting blocked with the user requests in the DeepCopy test, which was failing. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Flush to journal for internal or user flush. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Undo previous. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Add IMAGE_CACHE cases to RBD unit test Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix apparent copy-and-paste error Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Don't build PMDK or RWL on ARM while PMDK on ARM64 is still experimental Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Skip test_librbd when features includes IMAGE_CACHE, but WITH_RWL is not defined. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make read_only flag const, since that's how it's used Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix message Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Don't init image cache when opening R/O unless there's a dirty one Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Add evn vars to inhibit flush and retire on close for RWL debugging Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Add ImageCache action to build Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Add env vars for testing image cache rbd tool actions There's a regression here, but I'm committing anyway so I can migrate to gitlab now. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix lockdep issue in shutdown when not flushing/retiring. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Use FLUSH_SOURCE_SHUTDOWN to identify shutown flushes. In general, a whoile lot of plumbing to get the ability to suppress RWL flushing on close, so we can test removing and discarding the image-cache with the rbd tool. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix rbd cli test output for new image-cache comand and --no-flush option to bench Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Clean up logging from previous test Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix expected calls of block_writes() which now include the flush source. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Always init exclusive lock when the image cache feature is enabled. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Remove bogus assert Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Checkpoint rbd image-cache command Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Revert "Always init exclusive lock when the image cache feature is enabled." This reverts commit 314c82f. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix message Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Track image-cache option changes Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fixed some things clang didn't like, which may have broken rtags Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Split WITH_PMEM into WITH_PMEM_DEVICE and WITH_PMDK WITH_PMDK causes the PMDK (/NVML) to be built. WITH_PMEM_DEVICE causes the PMEM OSD to be built. WITH_PMEM_DEVICE, or WITH_RWL requires WITH_PMDK Primordial support for WITH_PMDK_PKG enables use of PMDK installed on platform. We'd probably like finer grained control over this (build most of Ceph without PMDK installed, and make PMDK an install dependency for only a few selected things). We can now enable RWL without requiring the PMEM OSD to build and pass tests. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Don't assert if exclusive lock release failed. On failure, these things may not have been done. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Also make etags file ctags -e, into src/TAGS. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Revert "Don't assert if exclusive lock release failed." This reverts commit 184ba6f5e97357380d30f8012f28c34c38e6818c Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Disable almost all deep copy tests when RWL is enabled These used to pass, and made interesting test cases, but it probably doesn't make sense to do this with the cache enabled anyway. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Ignore etags file TAGS Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Adds ReplicatedWriteLogInternal.h Hopefully trims enough off of ReplicatedWriteLog.cc that gitlab will show the diffs in a merge request. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Shift more things from RWL.cc to RWLInternal.h Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Finish moving lines Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> One more try for gitlab-friendly file sizes. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> To fix IO hange problem caused by aio_flush. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Add IO test for RWL It does IO test with RWL, and tries to cover the test case when RWL cache is full. But it is not guranteed. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix the bug 'First IO hangs if RWL can't be initialized'. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Update default and minimum value of rwl_size. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fixes the image-cache show command Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Update RWL full case to use default value. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Add pool name in the name of rwl cache file. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fixed rebase issues. Look closer at whether FLUSH_SOURCE_WRITEBACK should be a no-op in RWL, or an assert. Look closer at discard granularity passed to lower cache. This replaces skip_partial. We were always passing skip_partial = false, so the RWL and the image got the same discard results (regions which read as zero, vs. whatever was there before). Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix expected usage output Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Probably rebase noise Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Repeat compare buffer as necessary to match length of read in compare-and-write Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Disable some more migration tests when IMAGE_CACHE feature is enabled. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> fix image-cache disable feature command Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Start rwl threads after rwl_init. Signed-off-by: Li Xiaoyan <xiaoyan.li@intel.com> change ENOENT to EPERM when open image cache failed; Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> fixed issue ceph#35: If the image metadata says there is no cache, any existing pool file with the same name should be deleted. Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> only image cache state.present=0 and has pool file, then del the pool file with the same name; Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> do this in rwl_int() function; Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> Pass through everything to writeback interface Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> fix ceph#41:IO hang when rwl_size = 10G or 20G and without rate_iops limit, While there's no IO problem when rwl_size=1G Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> add high/low water log entries threshold and trigger it to retire entries when there's no enough log entries; Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> clean indent issue; Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> modify indentation based on Scott's feedback; Signed-off-by: Yuan Lu <yuan.y.lu@intel.com> Fix the issue of flushing rwl cache when showing image cache. Signed-off-by: Xiaoyan Li <xiaoyan.li@intel.com> Fix image-cache show issue. Signed-off-by: Xiaoyan Li <xiaoyan.li@intel.com> image-cache discard-dirty CLI command Signed-off-by: Mahati Chamarthy <mahati.chamarthy@intel.com> pmemobj_close won't fail. Signed-off-by: Xiaoyan Li <xiaoyan.li@intel.com> Rebase repairs. Mostly trivial except async op handling in flush. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Make imageFlushRequest::send_request() more like it was before the rebase. All in pursuit of issue ceph#48 after the latest rebase. The aio_comp() operations in comments there cause an assert in init or open now. FLUSH_SOURCE_WRITEBACK requests complete synchonously now, like they did before the rebase. None of this fixes the test assert in issue ceph#48. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Replace namespace declaration that is present in master but not our branch Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Supplied initial values for things not initialized by all constructors I saw compiler warnings about these fields being used uninitialized and thought it might be related to the rebase regression (it probably isn't). It does look like these could be left uninitialized if the ReplicatedWriteLogSpec constructor with no argumets was used. Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Bad merge Signed-off-by: Scott Peterson <scott.d.peterson@intel.com> Fix unit test problem. Signed-off-by: Xiaoyan Li <xiaoyan.li@intel.com>
galsalomon66
pushed a commit
that referenced
this pull request
Jan 14, 2021
Signed-off-by: Albin Antony <aantony@redhat.com> Co-authored-by: Albin Antony <aantony@localhost.localdomain>
m-ildefons
pushed a commit
to m-ildefons/ceph
that referenced
this pull request
Sep 7, 2022
rgw/sfs: updated CHANGELOG.md
jmolmo
pushed a commit
to jmolmo/ceph
that referenced
this pull request
Dec 19, 2023
mgr/call_home_agent: fix reports frequency
Moisi
pushed a commit
to Moisi/ceph
that referenced
this pull request
Oct 14, 2024
…libstdc++ Merge in STORAGEENG/ceph from bugfix/OBJGEN1-917-osd-crashes-in-ceph-v18.2.1-13 to aka_version_18.2.1 * commit '5ed8f5be3008cd240b41287a697bf361ea4c590a': strip that LDFlags for all removing a link flag when building with static libstdc++
athanatos
pushed a commit
to athanatos/ceph
that referenced
this pull request
Feb 14, 2025
src/dmclock_server.h: silence warning from -Wmaybe-uninitialized
tchaikov
added a commit
to tchaikov/ceph
that referenced
this pull request
Feb 27, 2026
When co_waiter is destroyed, the cancellation slot may still hold a
reference to the op_cancellation callback which captures 'this'. If
the cancellation signal is emitted after co_waiter is destroyed (e.g.,
during co_throttle shutdown), it results in a stack-use-after-scope
error.
Fix by:
- Adding a cancel_slot member to store the cancellation slot
- Adding a destructor that clears the slot before destruction
- Modifying get() to store the slot for later cleanup
This ensures the cancellation callback is removed before the co_waiter
object goes out of scope, preventing use-after-scope errors.
This issue was identified by ASan:
```
==21453==ERROR: AddressSanitizer: stack-use-after-scope on address 0x7a1364f050c8 at pc 0x603d79ff0d51 bp 0x7ffc1edf78c0 sp 0x7ffc1edf78b8
READ of size 1 at 0x7a1364f050c8 thread T0
#0 0x603d79ff0d50 in std::_Optional_base_impl<boost::asio::detail::awaitable_handler<boost::asio::any_io_executor, std::__exception_ptr::exception_ptr>, std::_Optional_base<boost::asio::detail::awaitable_handler<boost::asio::any_io_executor, std::__ex
ception_ptr::exception_ptr>, false, false>>::_M_is_engaged() const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/optional:471:58
#1 0x603d79ff8874 in std::optional<boost::asio::detail::awaitable_handler<boost::asio::any_io_executor, std::__exception_ptr::exception_ptr>>::operator bool() const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/optional:985:22
#2 0x603d79ff9d5f in ceph::async::co_waiter<void, boost::asio::any_io_executor>::cancel() /ceph/src/common/async/co_waiter.h:153:9
#3 0x603d79ff9c32 in ceph::async::co_waiter<void, boost::asio::any_io_executor>::op_cancellation::operator()(boost::asio::cancellation_type) /ceph/src/common/async/co_waiter.h:112:15
#4 0x603d79ff9a6e in boost::asio::detail::cancellation_handler<ceph::async::co_waiter<void, boost::asio::any_io_executor>::op_cancellation>::call(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_signal.hpp:56:5
#5 0x603d79fb9125 in boost::asio::cancellation_signal::emit(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_signal.hpp:99:17
#6 0x603d79fd6c31 in boost::asio::cancellation_state::impl<boost::asio::cancellation_filter<(boost::asio::cancellation_type)1>, boost::asio::cancellation_filter<(boost::asio::cancellation_type)1>>::operator()(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_state.hpp:222:23
#7 0x603d79fd696e in boost::asio::detail::cancellation_handler<boost::asio::cancellation_state::impl<boost::asio::cancellation_filter<(boost::asio::cancellation_type)1>, boost::asio::cancellation_filter<(boost::asio::cancellation_type)1>>>::call(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_signal.hpp:56:5
#8 0x603d79fb9125 in boost::asio::cancellation_signal::emit(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_signal.hpp:99:17
#9 0x603d79fee03a in boost::asio::detail::co_spawn_cancellation_handler<boost::asio::cancellation_slot_binder<ceph::async::detail::co_throttle_impl<boost::asio::any_io_executor>::child_completion, boost::asio::cancellation_slot>, boost::asio::any_io_executor, void>::operator()(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/impl/co_spawn.hpp:296:13
#10 0x603d79fede9e in boost::asio::detail::cancellation_handler<boost::asio::detail::co_spawn_cancellation_handler<boost::asio::cancellation_slot_binder<ceph::async::detail::co_throttle_impl<boost::asio::any_io_executor>::child_completion, boost::asio::cancellation_slot>, boost::asio::any_io_executor, void>>::call(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_signal.hpp:56:5
#11 0x603d79fb9125 in boost::asio::cancellation_signal::emit(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_signal.hpp:99:17
#12 0x603d79fe7135 in ceph::async::detail::co_throttle_impl<boost::asio::any_io_executor>::cancel() /ceph/src/common/async/detail/co_throttle_impl.h:122:17
ceph#13 0x603d79fe701c in ceph::async::co_throttle<boost::asio::any_io_executor>::cancel() /ceph/src/common/async/co_throttle.h:110:11
ceph#14 0x603d79fe27a8 in ceph::async::co_throttle<boost::asio::any_io_executor>::~co_throttle() /ceph/src/common/async/co_throttle.h:76:5
ceph#15 0x603d79f98dce in ceph::async::co_throttle_spawn_shutdown_Test::TestBody()::$_0::operator()() const (.destroy) /ceph/src/test/common/test_async_co_throttle.cc:264:3
ceph#16 0x603d79fe25ec in std::__n4861::coroutine_handle<void>::destroy() const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/coroutine:137:30
ceph#17 0x603d79fe2584 in boost::asio::detail::awaitable_frame_base<boost::asio::any_io_executor>::destroy() /opt/ceph/include/boost/asio/impl/awaitable.hpp:512:11
ceph#18 0x603d79fb79a9 in boost::asio::awaitable<void, boost::asio::any_io_executor>::~awaitable() /opt/ceph/include/boost/asio/awaitable.hpp:77:15
ceph#19 0x603d79f7fb0a in boost::asio::awaitable<boost::asio::detail::awaitable_thread_entry_point, boost::asio::any_io_executor> boost::asio::detail::co_spawn_entry_point<ceph::async::capture(std::optional<std::__exception_ptr::exception_ptr>&)::$_0, boost::asio::any_io_executor, boost::asio::detail::awaitable_as_function<void, boost::asio::any_io_executor>>(boost::asio::awaitable<void, boost::asio::any_io_executor>*, boost::asio::detail::co_spawn_state<ceph::async::capture(std::optional<std::__exception_ptr::exception_ptr>&)::$_0, boost::asio::any_io_executor, boost::asio::detail::awaitable_as_function<void, boost::asio::any_io_executor>, void>) (.destroy) /opt/ceph/include/boost/asio/impl/co_spawn.hpp:205:5
ceph#20 0x603d79fe25ec in std::__n4861::coroutine_handle<void>::destroy() const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/coroutine:137:30
ceph#21 0x603d79fe2584 in boost::asio::detail::awaitable_frame_base<boost::asio::any_io_executor>::destroy() /opt/ceph/include/boost/asio/impl/awaitable.hpp:512:11
ceph#22 0x603d79fd4fc9 in boost::asio::awaitable<boost::asio::detail::awaitable_thread_entry_point, boost::asio::any_io_executor>::~awaitable() /opt/ceph/include/boost/asio/awaitable.hpp:77:15
ceph#23 0x603d79fde3d4 in boost::asio::detail::awaitable_thread<boost::asio::any_io_executor>::~awaitable_thread()::'lambda'()::~() /opt/ceph/include/boost/asio/impl/awaitable.hpp:692:11
ceph#24 0x603d79fdf034 in boost::asio::detail::binder0<boost::asio::detail::awaitable_thread<boost::asio::any_io_executor>::~awaitable_thread()::'lambda'()>::~binder0() /opt/ceph/include/boost/asio/detail/bind_handler.hpp:30:7
ceph#25 0x603d79fe0501 in void boost::asio::detail::executor_function::complete<boost::asio::detail::binder0<boost::asio::detail::awaitable_thread<boost::asio::any_io_executor>::~awaitable_thread()::'lambda'()>, std::allocator<void>>(boost::asio::detail::executor_function::impl_base*, bool) /opt/ceph/include/boost/asio/detail/executor_function.hpp:115:3
ceph#26 0x603d79fdc152 in boost::asio::detail::executor_function::~executor_function() /opt/ceph/include/boost/asio/detail/executor_function.hpp:52:7
ceph#27 0x603d79ffcea8 in boost::asio::detail::executor_op<boost::asio::detail::executor_function, std::allocator<void>, boost::asio::detail::scheduler_operation>::do_complete(void*, boost::asio::detail::scheduler_operation*, boost::system::error_code const&, unsigned long) /opt/ceph/include/boost/asio/detail/executor_op.hpp:73:3
ceph#28 0x603d79fce17c in boost::asio::detail::scheduler_operation::destroy() /opt/ceph/include/boost/asio/detail/scheduler_operation.hpp:45:5
ceph#29 0x603d79fd0380 in boost::asio::detail::scheduler::shutdown() /opt/ceph/include/boost/asio/detail/impl/scheduler.ipp:174:10
ceph#30 0x603d79fd483c in boost::asio::detail::service_registry::shutdown_services() /opt/ceph/include/boost/asio/detail/impl/service_registry.ipp:44:14
ceph#31 0x603d79fd4735 in boost::asio::execution_context::shutdown() /opt/ceph/include/boost/asio/impl/execution_context.ipp:48:22
ceph#32 0x603d79fb8c08 in boost::asio::io_context::~io_context() /opt/ceph/include/boost/asio/impl/io_context.ipp:65:3
ceph#33 0x603d79f4a284 in ceph::async::co_throttle_spawn_shutdown_Test::TestBody() /ceph/src/test/common/test_async_co_throttle.cc:274:1
ceph#34 0x603d7a0fdd8d in void testing::internal::HandleSehExceptionsInMethodIfSupported<testing::Test, void>(testing::Test*, void (testing::Test::*)(), char const*) /ceph/src/googletest/googletest/src/gtest.cc:2653:10
ceph#35 0x603d7a0b49e5 in void testing::internal::HandleExceptionsInMethodIfSupported<testing::Test, void>(testing::Test*, void (testing::Test::*)(), char const*) /ceph/src/googletest/googletest/src/gtest.cc:2689:14
ceph#36 0x603d7a06f0bd in testing::Test::Run() /ceph/src/googletest/googletest/src/gtest.cc:2728:5
```
Fixes: https://tracker.ceph.com/issues/75231
Signed-off-by: Kefu Chai <k.chai@proxmox.com>
tchaikov
added a commit
to tchaikov/ceph
that referenced
this pull request
Feb 27, 2026
When co_waiter is destroyed, the cancellation slot may still hold a
reference to the op_cancellation callback which captures 'this'. If
the cancellation signal is emitted after co_waiter is destroyed (e.g.,
during co_throttle shutdown), it results in a stack-use-after-scope
error.
Fix by:
- Adding a cancel_slot member to store the cancellation slot
- Adding a destructor that clears the slot before destruction
- Modifying get() to store the slot for later cleanup
This ensures the cancellation callback is removed before the co_waiter
object goes out of scope, preventing use-after-scope errors.
This issue was identified by ASan:
```
==21453==ERROR: AddressSanitizer: stack-use-after-scope on address 0x7a1364f050c8 at pc 0x603d79ff0d51 bp 0x7ffc1edf78c0 sp 0x7ffc1edf78b8
READ of size 1 at 0x7a1364f050c8 thread T0
#0 0x603d79ff0d50 in std::_Optional_base_impl<boost::asio::detail::awaitable_handler<boost::asio::any_io_executor, std::__exception_ptr::exception_ptr>, std::_Optional_base<boost::asio::detail::awaitable_handler<boost::asio::any_io_executor, std::__ex
ception_ptr::exception_ptr>, false, false>>::_M_is_engaged() const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/optional:471:58
#1 0x603d79ff8874 in std::optional<boost::asio::detail::awaitable_handler<boost::asio::any_io_executor, std::__exception_ptr::exception_ptr>>::operator bool() const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/optional:985:22
#2 0x603d79ff9d5f in ceph::async::co_waiter<void, boost::asio::any_io_executor>::cancel() /ceph/src/common/async/co_waiter.h:153:9
#3 0x603d79ff9c32 in ceph::async::co_waiter<void, boost::asio::any_io_executor>::op_cancellation::operator()(boost::asio::cancellation_type) /ceph/src/common/async/co_waiter.h:112:15
#4 0x603d79ff9a6e in boost::asio::detail::cancellation_handler<ceph::async::co_waiter<void, boost::asio::any_io_executor>::op_cancellation>::call(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_signal.hpp:56:5
#5 0x603d79fb9125 in boost::asio::cancellation_signal::emit(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_signal.hpp:99:17
#6 0x603d79fd6c31 in boost::asio::cancellation_state::impl<boost::asio::cancellation_filter<(boost::asio::cancellation_type)1>, boost::asio::cancellation_filter<(boost::asio::cancellation_type)1>>::operator()(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_state.hpp:222:23
#7 0x603d79fd696e in boost::asio::detail::cancellation_handler<boost::asio::cancellation_state::impl<boost::asio::cancellation_filter<(boost::asio::cancellation_type)1>, boost::asio::cancellation_filter<(boost::asio::cancellation_type)1>>>::call(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_signal.hpp:56:5
#8 0x603d79fb9125 in boost::asio::cancellation_signal::emit(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_signal.hpp:99:17
#9 0x603d79fee03a in boost::asio::detail::co_spawn_cancellation_handler<boost::asio::cancellation_slot_binder<ceph::async::detail::co_throttle_impl<boost::asio::any_io_executor>::child_completion, boost::asio::cancellation_slot>, boost::asio::any_io_executor, void>::operator()(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/impl/co_spawn.hpp:296:13
#10 0x603d79fede9e in boost::asio::detail::cancellation_handler<boost::asio::detail::co_spawn_cancellation_handler<boost::asio::cancellation_slot_binder<ceph::async::detail::co_throttle_impl<boost::asio::any_io_executor>::child_completion, boost::asio::cancellation_slot>, boost::asio::any_io_executor, void>>::call(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_signal.hpp:56:5
#11 0x603d79fb9125 in boost::asio::cancellation_signal::emit(boost::asio::cancellation_type) /opt/ceph/include/boost/asio/cancellation_signal.hpp:99:17
#12 0x603d79fe7135 in ceph::async::detail::co_throttle_impl<boost::asio::any_io_executor>::cancel() /ceph/src/common/async/detail/co_throttle_impl.h:122:17
ceph#13 0x603d79fe701c in ceph::async::co_throttle<boost::asio::any_io_executor>::cancel() /ceph/src/common/async/co_throttle.h:110:11
ceph#14 0x603d79fe27a8 in ceph::async::co_throttle<boost::asio::any_io_executor>::~co_throttle() /ceph/src/common/async/co_throttle.h:76:5
ceph#15 0x603d79f98dce in ceph::async::co_throttle_spawn_shutdown_Test::TestBody()::$_0::operator()() const (.destroy) /ceph/src/test/common/test_async_co_throttle.cc:264:3
ceph#16 0x603d79fe25ec in std::__n4861::coroutine_handle<void>::destroy() const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/coroutine:137:30
ceph#17 0x603d79fe2584 in boost::asio::detail::awaitable_frame_base<boost::asio::any_io_executor>::destroy() /opt/ceph/include/boost/asio/impl/awaitable.hpp:512:11
ceph#18 0x603d79fb79a9 in boost::asio::awaitable<void, boost::asio::any_io_executor>::~awaitable() /opt/ceph/include/boost/asio/awaitable.hpp:77:15
ceph#19 0x603d79f7fb0a in boost::asio::awaitable<boost::asio::detail::awaitable_thread_entry_point, boost::asio::any_io_executor> boost::asio::detail::co_spawn_entry_point<ceph::async::capture(std::optional<std::__exception_ptr::exception_ptr>&)::$_0, boost::asio::any_io_executor, boost::asio::detail::awaitable_as_function<void, boost::asio::any_io_executor>>(boost::asio::awaitable<void, boost::asio::any_io_executor>*, boost::asio::detail::co_spawn_state<ceph::async::capture(std::optional<std::__exception_ptr::exception_ptr>&)::$_0, boost::asio::any_io_executor, boost::asio::detail::awaitable_as_function<void, boost::asio::any_io_executor>, void>) (.destroy) /opt/ceph/include/boost/asio/impl/co_spawn.hpp:205:5
ceph#20 0x603d79fe25ec in std::__n4861::coroutine_handle<void>::destroy() const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/coroutine:137:30
ceph#21 0x603d79fe2584 in boost::asio::detail::awaitable_frame_base<boost::asio::any_io_executor>::destroy() /opt/ceph/include/boost/asio/impl/awaitable.hpp:512:11
ceph#22 0x603d79fd4fc9 in boost::asio::awaitable<boost::asio::detail::awaitable_thread_entry_point, boost::asio::any_io_executor>::~awaitable() /opt/ceph/include/boost/asio/awaitable.hpp:77:15
ceph#23 0x603d79fde3d4 in boost::asio::detail::awaitable_thread<boost::asio::any_io_executor>::~awaitable_thread()::'lambda'()::~() /opt/ceph/include/boost/asio/impl/awaitable.hpp:692:11
ceph#24 0x603d79fdf034 in boost::asio::detail::binder0<boost::asio::detail::awaitable_thread<boost::asio::any_io_executor>::~awaitable_thread()::'lambda'()>::~binder0() /opt/ceph/include/boost/asio/detail/bind_handler.hpp:30:7
ceph#25 0x603d79fe0501 in void boost::asio::detail::executor_function::complete<boost::asio::detail::binder0<boost::asio::detail::awaitable_thread<boost::asio::any_io_executor>::~awaitable_thread()::'lambda'()>, std::allocator<void>>(boost::asio::detail::executor_function::impl_base*, bool) /opt/ceph/include/boost/asio/detail/executor_function.hpp:115:3
ceph#26 0x603d79fdc152 in boost::asio::detail::executor_function::~executor_function() /opt/ceph/include/boost/asio/detail/executor_function.hpp:52:7
ceph#27 0x603d79ffcea8 in boost::asio::detail::executor_op<boost::asio::detail::executor_function, std::allocator<void>, boost::asio::detail::scheduler_operation>::do_complete(void*, boost::asio::detail::scheduler_operation*, boost::system::error_code const&, unsigned long) /opt/ceph/include/boost/asio/detail/executor_op.hpp:73:3
ceph#28 0x603d79fce17c in boost::asio::detail::scheduler_operation::destroy() /opt/ceph/include/boost/asio/detail/scheduler_operation.hpp:45:5
ceph#29 0x603d79fd0380 in boost::asio::detail::scheduler::shutdown() /opt/ceph/include/boost/asio/detail/impl/scheduler.ipp:174:10
ceph#30 0x603d79fd483c in boost::asio::detail::service_registry::shutdown_services() /opt/ceph/include/boost/asio/detail/impl/service_registry.ipp:44:14
ceph#31 0x603d79fd4735 in boost::asio::execution_context::shutdown() /opt/ceph/include/boost/asio/impl/execution_context.ipp:48:22
ceph#32 0x603d79fb8c08 in boost::asio::io_context::~io_context() /opt/ceph/include/boost/asio/impl/io_context.ipp:65:3
ceph#33 0x603d79f4a284 in ceph::async::co_throttle_spawn_shutdown_Test::TestBody() /ceph/src/test/common/test_async_co_throttle.cc:274:1
ceph#34 0x603d7a0fdd8d in void testing::internal::HandleSehExceptionsInMethodIfSupported<testing::Test, void>(testing::Test*, void (testing::Test::*)(), char const*) /ceph/src/googletest/googletest/src/gtest.cc:2653:10
ceph#35 0x603d7a0b49e5 in void testing::internal::HandleExceptionsInMethodIfSupported<testing::Test, void>(testing::Test*, void (testing::Test::*)(), char const*) /ceph/src/googletest/googletest/src/gtest.cc:2689:14
ceph#36 0x603d7a06f0bd in testing::Test::Run() /ceph/src/googletest/googletest/src/gtest.cc:2728:5
```
`
Signed-off-by: Kefu Chai <k.chai@proxmox.com>
tchaikov
added a commit
to tchaikov/ceph
that referenced
this pull request
Mar 1, 2026
rbd_features_from_string() uses boost::lexical_cast which throws
bad_lexical_cast when the input is not numeric. The exception is caught
and handled as "parse as feature name list instead".
This is normal control flow, but when ASAN's __cxa_throw interceptor is
misconfigured (e.g. with intercept_cxx_exceptions=0 leaving real___cxa_throw
NULL), any exception causes a CHECK failure. Even with a correctly configured
ASAN, throwing exceptions during config initialization adds overhead.
Replace the try/catch pattern with boost::conversion::try_lexical_convert,
which returns false on parse failure instead of throwing. This eliminates
the exception entirely, making the code more efficient and avoiding any
interaction with ASAN's exception interceptor.
The ASan report:
```
AddressSanitizer: CHECK failed: asan_interceptors.cpp:320 "((__interception::real___cxa_throw)) != (0)" (0x0, 0x0) (tid=30529)
#0 0x7d668fac9a09 in CheckUnwind ../../../../src/libsanitizer/asan/asan_rtl.cpp:67
#1 0x7d668faec105 in __sanitizer::CheckFailed(char const*, int, char const*, unsigned long long, unsigned long long) ../../../../src/libsanitizer/sanitizer_common/sanitizer_termination.cpp:86
#2 0x7d668fa4b194 in __interceptor___cxa_throw ../../../../src/libsanitizer/asan/asan_interceptors.cpp:320
#3 0x7d668ae4ec0f in void boost::throw_exception<boost::bad_lexical_cast>(boost::bad_lexical_cast const&) /opt/ceph/include/boost/throw_exception.hpp:165
#4 0x7d668c1e1e0b in void boost::conversion::detail::throw_bad_cast<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, unsigned long>() /opt/ceph/include/boost/lexical_cast/bad_lexical_cast.hpp:93
#5 0x7d668c1e0e05 in unsigned long boost::lexical_cast<unsigned long, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) /opt/ceph/include/boost/lexical_cast.hpp:43
#6 0x7d668c1df609 in librbd::rbd_features_from_string(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::ostream*) /ceph/src/librbd/Features.cc:67
#7 0x7d668b255a35 in get_rbd_options()::$_2::operator()(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const src/common/options/rbd_options.cc:44
#8 0x7d668b255806 in int std::__invoke_impl<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>(std::__invoke_other, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/invoke.h:61
#9 0x7d668b255754 in std::enable_if<is_invocable_r_v<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>, int>::type std::__invoke_r<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>(get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/invoke.h:114
#10 0x7d668b25563c in std::_Function_handler<int (std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*), get_rbd_options()::$_2>::_M_invoke(std::_Any_data const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/std_function.h:290
#11 0x7d668af924f1 in std::function<int (std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*)>::operator()(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/std_function.h:591
#12 0x7d668af8c495 in Option::pre_validate(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const /ceph/src/common/options.cc:94
ceph#13 0x7d668aef9ec3 in md_config_t::md_config_t(ConfigValues&, ConfigTracker const&, bool) /ceph/src/common/config.cc:208
ceph#14 0x7d668ae0040d in ceph::common::ConfigProxy::ConfigProxy(bool) /ceph/src/common/config_proxy.h:70
ceph#15 0x7d668adf3f5d in ceph::common::CephContext::CephContext(unsigned int, ceph::common::CephContext::create_options const&) /ceph/src/common/ceph_context.cc:730
ceph#16 0x7d668adf3785 in ceph::common::CephContext::CephContext(unsigned int, code_environment_t, int) /ceph/src/common/ceph_context.cc:724
ceph#17 0x7d668aef5ee6 in common_preinit(CephInitParameters const&, code_environment_t, int) /ceph/src/common/common_init.cc:40
ceph#18 0x7d668d4a47af in rados_create_cct(char const*, CephInitParameters*) /ceph/src/librados/librados_c.cc:120
ceph#19 0x7d668d4a49db in _rados_create2 /ceph/src/librados/librados_c.cc:168
ceph#20 0x7d668d94dc4b in __pyx_pf_5rados_5Rados_2__setup /ceph/build/src/pybind/rados/rados_processed.c:13219
ceph#21 0x7d668d94dc4b in __pyx_pw_5rados_5Rados_3__setup /ceph/build/src/pybind/rados/rados_processed.c:12703
ceph#22 0x7d668d94a347 in __Pyx_CyFunction_CallAsMethod /ceph/build/src/pybind/rados/rados_processed.c:93157
ceph#23 0x58ac5086d0ba in _PyObject_MakeTpCall (/usr/bin/python3.10+0x1810ba)
ceph#24 0x58ac508843da (/usr/bin/python3.10+0x1983da)
ceph#25 0x58ac50885076 in PyVectorcall_Call (/usr/bin/python3.10+0x199076)
ceph#26 0x7d668d94cdd7 in __Pyx_PyObject_Call /ceph/build/src/pybind/rados/rados_processed.c:90994
ceph#27 0x7d668d94cdd7 in __pyx_pf_5rados_5Rados___init__ /ceph/build/src/pybind/rados/rados_processed.c:12474
ceph#28 0x7d668d94cdd7 in __pyx_pw_5rados_5Rados_1__init__ /ceph/build/src/pybind/rados/rados_processed.c:12443
ceph#29 0x58ac5086d43a (/usr/bin/python3.10+0x18143a)
ceph#30 0x58ac50884d3a in PyObject_Call (/usr/bin/python3.10+0x198d3a)
ceph#31 0x58ac508637de in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1777de)
ceph#32 0x58ac5087702b in _PyFunction_Vectorcall (/usr/bin/python3.10+0x18b02b)
ceph#33 0x58ac508615fe in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1755fe)
ceph#34 0x58ac5087702b in _PyFunction_Vectorcall (/usr/bin/python3.10+0x18b02b)
ceph#35 0x58ac508615fe in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1755fe)
```
Signed-off-by: Kefu Chai <k.chai@proxmox.com>
tchaikov
added a commit
to tchaikov/ceph
that referenced
this pull request
Mar 1, 2026
rbd_features_from_string() uses boost::lexical_cast which throws
bad_lexical_cast when the input is not numeric. The exception is caught
and handled as "parse as feature name list instead".
This is normal control flow, but when ASAN's __cxa_throw interceptor is
misconfigured (e.g. with intercept_cxx_exceptions=0 leaving real___cxa_throw
NULL), any exception causes a CHECK failure. Even with a correctly configured
ASAN, throwing exceptions during config initialization adds overhead.
Replace the try/catch pattern with boost::conversion::try_lexical_convert,
which returns false on parse failure instead of throwing. This eliminates
the exception entirely, making the code more efficient and avoiding any
interaction with ASAN's exception interceptor.
The ASan report:
```
AddressSanitizer: CHECK failed: asan_interceptors.cpp:320 "((__interception::real___cxa_throw)) != (0)" (0x0, 0x0) (tid=30529)
#0 0x7d668fac9a09 in CheckUnwind ../../../../src/libsanitizer/asan/asan_rtl.cpp:67
#1 0x7d668faec105 in __sanitizer::CheckFailed(char const*, int, char const*, unsigned long long, unsigned long long) ../../../../src/libsanitizer/sanitizer_common/sanitizer_termination.cpp:86
#2 0x7d668fa4b194 in __interceptor___cxa_throw ../../../../src/libsanitizer/asan/asan_interceptors.cpp:320
#3 0x7d668ae4ec0f in void boost::throw_exception<boost::bad_lexical_cast>(boost::bad_lexical_cast const&) /opt/ceph/include/boost/throw_exception.hpp:165
#4 0x7d668c1e1e0b in void boost::conversion::detail::throw_bad_cast<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, unsigned long>() /opt/ceph/include/boost/lexical_cast/bad_lexical_cast.hpp:93
#5 0x7d668c1e0e05 in unsigned long boost::lexical_cast<unsigned long, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) /opt/ceph/include/boost/lexical_cast.hpp:43
#6 0x7d668c1df609 in librbd::rbd_features_from_string(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::ostream*) /ceph/src/librbd/Features.cc:67
#7 0x7d668b255a35 in get_rbd_options()::$_2::operator()(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const src/common/options/rbd_options.cc:44
#8 0x7d668b255806 in int std::__invoke_impl<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>(std::__invoke_other, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/invoke.h:61
#9 0x7d668b255754 in std::enable_if<is_invocable_r_v<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>, int>::type std::__invoke_r<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>(get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/invoke.h:114
#10 0x7d668b25563c in std::_Function_handler<int (std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*), get_rbd_options()::$_2>::_M_invoke(std::_Any_data const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/std_function.h:290
#11 0x7d668af924f1 in std::function<int (std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*)>::operator()(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/std_function.h:591
#12 0x7d668af8c495 in Option::pre_validate(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const /ceph/src/common/options.cc:94
ceph#13 0x7d668aef9ec3 in md_config_t::md_config_t(ConfigValues&, ConfigTracker const&, bool) /ceph/src/common/config.cc:208
ceph#14 0x7d668ae0040d in ceph::common::ConfigProxy::ConfigProxy(bool) /ceph/src/common/config_proxy.h:70
ceph#15 0x7d668adf3f5d in ceph::common::CephContext::CephContext(unsigned int, ceph::common::CephContext::create_options const&) /ceph/src/common/ceph_context.cc:730
ceph#16 0x7d668adf3785 in ceph::common::CephContext::CephContext(unsigned int, code_environment_t, int) /ceph/src/common/ceph_context.cc:724
ceph#17 0x7d668aef5ee6 in common_preinit(CephInitParameters const&, code_environment_t, int) /ceph/src/common/common_init.cc:40
ceph#18 0x7d668d4a47af in rados_create_cct(char const*, CephInitParameters*) /ceph/src/librados/librados_c.cc:120
ceph#19 0x7d668d4a49db in _rados_create2 /ceph/src/librados/librados_c.cc:168
ceph#20 0x7d668d94dc4b in __pyx_pf_5rados_5Rados_2__setup /ceph/build/src/pybind/rados/rados_processed.c:13219
ceph#21 0x7d668d94dc4b in __pyx_pw_5rados_5Rados_3__setup /ceph/build/src/pybind/rados/rados_processed.c:12703
ceph#22 0x7d668d94a347 in __Pyx_CyFunction_CallAsMethod /ceph/build/src/pybind/rados/rados_processed.c:93157
ceph#23 0x58ac5086d0ba in _PyObject_MakeTpCall (/usr/bin/python3.10+0x1810ba)
ceph#24 0x58ac508843da (/usr/bin/python3.10+0x1983da)
ceph#25 0x58ac50885076 in PyVectorcall_Call (/usr/bin/python3.10+0x199076)
ceph#26 0x7d668d94cdd7 in __Pyx_PyObject_Call /ceph/build/src/pybind/rados/rados_processed.c:90994
ceph#27 0x7d668d94cdd7 in __pyx_pf_5rados_5Rados___init__ /ceph/build/src/pybind/rados/rados_processed.c:12474
ceph#28 0x7d668d94cdd7 in __pyx_pw_5rados_5Rados_1__init__ /ceph/build/src/pybind/rados/rados_processed.c:12443
ceph#29 0x58ac5086d43a (/usr/bin/python3.10+0x18143a)
ceph#30 0x58ac50884d3a in PyObject_Call (/usr/bin/python3.10+0x198d3a)
ceph#31 0x58ac508637de in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1777de)
ceph#32 0x58ac5087702b in _PyFunction_Vectorcall (/usr/bin/python3.10+0x18b02b)
ceph#33 0x58ac508615fe in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1755fe)
ceph#34 0x58ac5087702b in _PyFunction_Vectorcall (/usr/bin/python3.10+0x18b02b)
ceph#35 0x58ac508615fe in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1755fe)
```
Signed-off-by: Kefu Chai <k.chai@proxmox.com>
tchaikov
added a commit
to tchaikov/ceph
that referenced
this pull request
Mar 1, 2026
rbd_features_from_string() uses boost::lexical_cast which throws
bad_lexical_cast when the input is not numeric. The exception is caught
and handled as "parse as feature name list instead".
This is normal control flow, but when ASAN's __cxa_throw interceptor is
misconfigured (e.g. with intercept_cxx_exceptions=0 leaving real___cxa_throw
NULL), any exception causes a CHECK failure. Even with a correctly configured
ASAN, throwing exceptions during config initialization adds overhead.
Replace the try/catch pattern with boost::conversion::try_lexical_convert,
which returns false on parse failure instead of throwing. This eliminates
the exception entirely, making the code more efficient and avoiding any
interaction with ASAN's exception interceptor.
The ASan report:
```
AddressSanitizer: CHECK failed: asan_interceptors.cpp:320 "((__interception::real___cxa_throw)) != (0)" (0x0, 0x0) (tid=30529)
#0 0x7d668fac9a09 in CheckUnwind ../../../../src/libsanitizer/asan/asan_rtl.cpp:67
#1 0x7d668faec105 in __sanitizer::CheckFailed(char const*, int, char const*, unsigned long long, unsigned long long) ../../../../src/libsanitizer/sanitizer_common/sanitizer_termination.cpp:86
#2 0x7d668fa4b194 in __interceptor___cxa_throw ../../../../src/libsanitizer/asan/asan_interceptors.cpp:320
#3 0x7d668ae4ec0f in void boost::throw_exception<boost::bad_lexical_cast>(boost::bad_lexical_cast const&) /opt/ceph/include/boost/throw_exception.hpp:165
#4 0x7d668c1e1e0b in void boost::conversion::detail::throw_bad_cast<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, unsigned long>() /opt/ceph/include/boost/lexical_cast/bad_lexical_cast.hpp:93
#5 0x7d668c1e0e05 in unsigned long boost::lexical_cast<unsigned long, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) /opt/ceph/include/boost/lexical_cast.hpp:43
#6 0x7d668c1df609 in librbd::rbd_features_from_string(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::ostream*) /ceph/src/librbd/Features.cc:67
#7 0x7d668b255a35 in get_rbd_options()::$_2::operator()(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const src/common/options/rbd_options.cc:44
#8 0x7d668b255806 in int std::__invoke_impl<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>(std::__invoke_other, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/invoke.h:61
#9 0x7d668b255754 in std::enable_if<is_invocable_r_v<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>, int>::type std::__invoke_r<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>(get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/invoke.h:114
#10 0x7d668b25563c in std::_Function_handler<int (std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*), get_rbd_options()::$_2>::_M_invoke(std::_Any_data const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/std_function.h:290
#11 0x7d668af924f1 in std::function<int (std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*)>::operator()(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/std_function.h:591
#12 0x7d668af8c495 in Option::pre_validate(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const /ceph/src/common/options.cc:94
ceph#13 0x7d668aef9ec3 in md_config_t::md_config_t(ConfigValues&, ConfigTracker const&, bool) /ceph/src/common/config.cc:208
ceph#14 0x7d668ae0040d in ceph::common::ConfigProxy::ConfigProxy(bool) /ceph/src/common/config_proxy.h:70
ceph#15 0x7d668adf3f5d in ceph::common::CephContext::CephContext(unsigned int, ceph::common::CephContext::create_options const&) /ceph/src/common/ceph_context.cc:730
ceph#16 0x7d668adf3785 in ceph::common::CephContext::CephContext(unsigned int, code_environment_t, int) /ceph/src/common/ceph_context.cc:724
ceph#17 0x7d668aef5ee6 in common_preinit(CephInitParameters const&, code_environment_t, int) /ceph/src/common/common_init.cc:40
ceph#18 0x7d668d4a47af in rados_create_cct(char const*, CephInitParameters*) /ceph/src/librados/librados_c.cc:120
ceph#19 0x7d668d4a49db in _rados_create2 /ceph/src/librados/librados_c.cc:168
ceph#20 0x7d668d94dc4b in __pyx_pf_5rados_5Rados_2__setup /ceph/build/src/pybind/rados/rados_processed.c:13219
ceph#21 0x7d668d94dc4b in __pyx_pw_5rados_5Rados_3__setup /ceph/build/src/pybind/rados/rados_processed.c:12703
ceph#22 0x7d668d94a347 in __Pyx_CyFunction_CallAsMethod /ceph/build/src/pybind/rados/rados_processed.c:93157
ceph#23 0x58ac5086d0ba in _PyObject_MakeTpCall (/usr/bin/python3.10+0x1810ba)
ceph#24 0x58ac508843da (/usr/bin/python3.10+0x1983da)
ceph#25 0x58ac50885076 in PyVectorcall_Call (/usr/bin/python3.10+0x199076)
ceph#26 0x7d668d94cdd7 in __Pyx_PyObject_Call /ceph/build/src/pybind/rados/rados_processed.c:90994
ceph#27 0x7d668d94cdd7 in __pyx_pf_5rados_5Rados___init__ /ceph/build/src/pybind/rados/rados_processed.c:12474
ceph#28 0x7d668d94cdd7 in __pyx_pw_5rados_5Rados_1__init__ /ceph/build/src/pybind/rados/rados_processed.c:12443
ceph#29 0x58ac5086d43a (/usr/bin/python3.10+0x18143a)
ceph#30 0x58ac50884d3a in PyObject_Call (/usr/bin/python3.10+0x198d3a)
ceph#31 0x58ac508637de in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1777de)
ceph#32 0x58ac5087702b in _PyFunction_Vectorcall (/usr/bin/python3.10+0x18b02b)
ceph#33 0x58ac508615fe in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1755fe)
ceph#34 0x58ac5087702b in _PyFunction_Vectorcall (/usr/bin/python3.10+0x18b02b)
ceph#35 0x58ac508615fe in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1755fe)
```
Signed-off-by: Kefu Chai <k.chai@proxmox.com>
tchaikov
added a commit
to tchaikov/ceph
that referenced
this pull request
Mar 1, 2026
rbd_features_from_string() uses boost::lexical_cast which throws
bad_lexical_cast when the input is not numeric. The exception is caught
and handled as "parse as feature name list instead".
This is normal control flow, but when ASAN's __cxa_throw interceptor is
misconfigured (e.g. with intercept_cxx_exceptions=0 leaving real___cxa_throw
NULL), any exception causes a CHECK failure. Even with a correctly configured
ASAN, throwing exceptions during config initialization adds overhead.
Replace the try/catch pattern with boost::conversion::try_lexical_convert,
which returns false on parse failure instead of throwing. This eliminates
the exception entirely, making the code more efficient and avoiding any
interaction with ASAN's exception interceptor.
The ASan report:
```
AddressSanitizer: CHECK failed: asan_interceptors.cpp:320 "((__interception::real___cxa_throw)) != (0)" (0x0, 0x0) (tid=30529)
#0 0x7d668fac9a09 in CheckUnwind ../../../../src/libsanitizer/asan/asan_rtl.cpp:67
#1 0x7d668faec105 in __sanitizer::CheckFailed(char const*, int, char const*, unsigned long long, unsigned long long) ../../../../src/libsanitizer/sanitizer_common/sanitizer_termination.cpp:86
#2 0x7d668fa4b194 in __interceptor___cxa_throw ../../../../src/libsanitizer/asan/asan_interceptors.cpp:320
#3 0x7d668ae4ec0f in void boost::throw_exception<boost::bad_lexical_cast>(boost::bad_lexical_cast const&) /opt/ceph/include/boost/throw_exception.hpp:165
#4 0x7d668c1e1e0b in void boost::conversion::detail::throw_bad_cast<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, unsigned long>() /opt/ceph/include/boost/lexical_cast/bad_lexical_cast.hpp:93
#5 0x7d668c1e0e05 in unsigned long boost::lexical_cast<unsigned long, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) /opt/ceph/include/boost/lexical_cast.hpp:43
#6 0x7d668c1df609 in librbd::rbd_features_from_string(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::ostream*) /ceph/src/librbd/Features.cc:67
#7 0x7d668b255a35 in get_rbd_options()::$_2::operator()(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const src/common/options/rbd_options.cc:44
#8 0x7d668b255806 in int std::__invoke_impl<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>(std::__invoke_other, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/invoke.h:61
#9 0x7d668b255754 in std::enable_if<is_invocable_r_v<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>, int>::type std::__invoke_r<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>(get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/invoke.h:114
#10 0x7d668b25563c in std::_Function_handler<int (std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*), get_rbd_options()::$_2>::_M_invoke(std::_Any_data const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/std_function.h:290
#11 0x7d668af924f1 in std::function<int (std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*)>::operator()(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/std_function.h:591
#12 0x7d668af8c495 in Option::pre_validate(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const /ceph/src/common/options.cc:94
ceph#13 0x7d668aef9ec3 in md_config_t::md_config_t(ConfigValues&, ConfigTracker const&, bool) /ceph/src/common/config.cc:208
ceph#14 0x7d668ae0040d in ceph::common::ConfigProxy::ConfigProxy(bool) /ceph/src/common/config_proxy.h:70
ceph#15 0x7d668adf3f5d in ceph::common::CephContext::CephContext(unsigned int, ceph::common::CephContext::create_options const&) /ceph/src/common/ceph_context.cc:730
ceph#16 0x7d668adf3785 in ceph::common::CephContext::CephContext(unsigned int, code_environment_t, int) /ceph/src/common/ceph_context.cc:724
ceph#17 0x7d668aef5ee6 in common_preinit(CephInitParameters const&, code_environment_t, int) /ceph/src/common/common_init.cc:40
ceph#18 0x7d668d4a47af in rados_create_cct(char const*, CephInitParameters*) /ceph/src/librados/librados_c.cc:120
ceph#19 0x7d668d4a49db in _rados_create2 /ceph/src/librados/librados_c.cc:168
ceph#20 0x7d668d94dc4b in __pyx_pf_5rados_5Rados_2__setup /ceph/build/src/pybind/rados/rados_processed.c:13219
ceph#21 0x7d668d94dc4b in __pyx_pw_5rados_5Rados_3__setup /ceph/build/src/pybind/rados/rados_processed.c:12703
ceph#22 0x7d668d94a347 in __Pyx_CyFunction_CallAsMethod /ceph/build/src/pybind/rados/rados_processed.c:93157
ceph#23 0x58ac5086d0ba in _PyObject_MakeTpCall (/usr/bin/python3.10+0x1810ba)
ceph#24 0x58ac508843da (/usr/bin/python3.10+0x1983da)
ceph#25 0x58ac50885076 in PyVectorcall_Call (/usr/bin/python3.10+0x199076)
ceph#26 0x7d668d94cdd7 in __Pyx_PyObject_Call /ceph/build/src/pybind/rados/rados_processed.c:90994
ceph#27 0x7d668d94cdd7 in __pyx_pf_5rados_5Rados___init__ /ceph/build/src/pybind/rados/rados_processed.c:12474
ceph#28 0x7d668d94cdd7 in __pyx_pw_5rados_5Rados_1__init__ /ceph/build/src/pybind/rados/rados_processed.c:12443
ceph#29 0x58ac5086d43a (/usr/bin/python3.10+0x18143a)
ceph#30 0x58ac50884d3a in PyObject_Call (/usr/bin/python3.10+0x198d3a)
ceph#31 0x58ac508637de in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1777de)
ceph#32 0x58ac5087702b in _PyFunction_Vectorcall (/usr/bin/python3.10+0x18b02b)
ceph#33 0x58ac508615fe in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1755fe)
ceph#34 0x58ac5087702b in _PyFunction_Vectorcall (/usr/bin/python3.10+0x18b02b)
ceph#35 0x58ac508615fe in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1755fe)
```
Signed-off-by: Kefu Chai <k.chai@proxmox.com>
tchaikov
added a commit
to tchaikov/ceph
that referenced
this pull request
Mar 18, 2026
rbd_features_from_string() uses boost::lexical_cast which throws
bad_lexical_cast when the input is not numeric. The exception is caught
and handled as "parse as feature name list instead".
This is normal control flow, but when ASAN's __cxa_throw interceptor is
misconfigured (e.g. with intercept_cxx_exceptions=0 leaving real___cxa_throw
NULL), any exception causes a CHECK failure. Even with a correctly configured
ASAN, throwing exceptions during config initialization adds overhead.
Replace the try/catch pattern with boost::conversion::try_lexical_convert,
which returns false on parse failure instead of throwing. This eliminates
the exception entirely, making the code more efficient and avoiding any
interaction with ASAN's exception interceptor.
The ASan report:
```
AddressSanitizer: CHECK failed: asan_interceptors.cpp:320 "((__interception::real___cxa_throw)) != (0)" (0x0, 0x0) (tid=30529)
#0 0x7d668fac9a09 in CheckUnwind ../../../../src/libsanitizer/asan/asan_rtl.cpp:67
#1 0x7d668faec105 in __sanitizer::CheckFailed(char const*, int, char const*, unsigned long long, unsigned long long) ../../../../src/libsanitizer/sanitizer_common/sanitizer_termination.cpp:86
#2 0x7d668fa4b194 in __interceptor___cxa_throw ../../../../src/libsanitizer/asan/asan_interceptors.cpp:320
#3 0x7d668ae4ec0f in void boost::throw_exception<boost::bad_lexical_cast>(boost::bad_lexical_cast const&) /opt/ceph/include/boost/throw_exception.hpp:165
#4 0x7d668c1e1e0b in void boost::conversion::detail::throw_bad_cast<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, unsigned long>() /opt/ceph/include/boost/lexical_cast/bad_lexical_cast.hpp:93
#5 0x7d668c1e0e05 in unsigned long boost::lexical_cast<unsigned long, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) /opt/ceph/include/boost/lexical_cast.hpp:43
#6 0x7d668c1df609 in librbd::rbd_features_from_string(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::ostream*) /ceph/src/librbd/Features.cc:67
#7 0x7d668b255a35 in get_rbd_options()::$_2::operator()(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const src/common/options/rbd_options.cc:44
#8 0x7d668b255806 in int std::__invoke_impl<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>(std::__invoke_other, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/invoke.h:61
#9 0x7d668b255754 in std::enable_if<is_invocable_r_v<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>, int>::type std::__invoke_r<int, get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*>(get_rbd_options()::$_2&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/invoke.h:114
#10 0x7d668b25563c in std::_Function_handler<int (std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*), get_rbd_options()::$_2>::_M_invoke(std::_Any_data const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*&&) /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/std_function.h:290
#11 0x7d668af924f1 in std::function<int (std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*)>::operator()(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/std_function.h:591
#12 0x7d668af8c495 in Option::pre_validate(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >*) const /ceph/src/common/options.cc:94
ceph#13 0x7d668aef9ec3 in md_config_t::md_config_t(ConfigValues&, ConfigTracker const&, bool) /ceph/src/common/config.cc:208
ceph#14 0x7d668ae0040d in ceph::common::ConfigProxy::ConfigProxy(bool) /ceph/src/common/config_proxy.h:70
ceph#15 0x7d668adf3f5d in ceph::common::CephContext::CephContext(unsigned int, ceph::common::CephContext::create_options const&) /ceph/src/common/ceph_context.cc:730
ceph#16 0x7d668adf3785 in ceph::common::CephContext::CephContext(unsigned int, code_environment_t, int) /ceph/src/common/ceph_context.cc:724
ceph#17 0x7d668aef5ee6 in common_preinit(CephInitParameters const&, code_environment_t, int) /ceph/src/common/common_init.cc:40
ceph#18 0x7d668d4a47af in rados_create_cct(char const*, CephInitParameters*) /ceph/src/librados/librados_c.cc:120
ceph#19 0x7d668d4a49db in _rados_create2 /ceph/src/librados/librados_c.cc:168
ceph#20 0x7d668d94dc4b in __pyx_pf_5rados_5Rados_2__setup /ceph/build/src/pybind/rados/rados_processed.c:13219
ceph#21 0x7d668d94dc4b in __pyx_pw_5rados_5Rados_3__setup /ceph/build/src/pybind/rados/rados_processed.c:12703
ceph#22 0x7d668d94a347 in __Pyx_CyFunction_CallAsMethod /ceph/build/src/pybind/rados/rados_processed.c:93157
ceph#23 0x58ac5086d0ba in _PyObject_MakeTpCall (/usr/bin/python3.10+0x1810ba)
ceph#24 0x58ac508843da (/usr/bin/python3.10+0x1983da)
ceph#25 0x58ac50885076 in PyVectorcall_Call (/usr/bin/python3.10+0x199076)
ceph#26 0x7d668d94cdd7 in __Pyx_PyObject_Call /ceph/build/src/pybind/rados/rados_processed.c:90994
ceph#27 0x7d668d94cdd7 in __pyx_pf_5rados_5Rados___init__ /ceph/build/src/pybind/rados/rados_processed.c:12474
ceph#28 0x7d668d94cdd7 in __pyx_pw_5rados_5Rados_1__init__ /ceph/build/src/pybind/rados/rados_processed.c:12443
ceph#29 0x58ac5086d43a (/usr/bin/python3.10+0x18143a)
ceph#30 0x58ac50884d3a in PyObject_Call (/usr/bin/python3.10+0x198d3a)
ceph#31 0x58ac508637de in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1777de)
ceph#32 0x58ac5087702b in _PyFunction_Vectorcall (/usr/bin/python3.10+0x18b02b)
ceph#33 0x58ac508615fe in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1755fe)
ceph#34 0x58ac5087702b in _PyFunction_Vectorcall (/usr/bin/python3.10+0x18b02b)
ceph#35 0x58ac508615fe in _PyEval_EvalFrameDefault (/usr/bin/python3.10+0x1755fe)
```
Signed-off-by: Kefu Chai <k.chai@proxmox.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Improved the usage message.