kou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2067874304
It seems that `--with-tls` is related:
https://github.com/apache/arrow/pull/41279#issuecomment-2067874205
--
This is an automated message from the Apache Git Service.
To respond to the
pitrou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2056957706
It's more a case of DLL finalization coming before whatever unit of work is
still being executed in some worker thread, AFAIU. We had a similar problem
with S3 (but resulting in worse
felipecrv commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2056949564
Looks like yet another case of a library polluting the thread-local storage
area :)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log
pitrou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-205584
> I think that `--with-tls` isn't the cause of this.
Well, we can wait for an update conda-forge package to make sure that's the
case.
--
This is an automated message from the
kou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2053787467
I've also investigated this more.
I think that `--with-tls` isn't the cause of this.
I could reproduce this by the following program that only uses `libxml2` and
`std::thread`:
kou commented on code in PR #41163:
URL: https://github.com/apache/arrow/pull/41163#discussion_r1563245965
##
cpp/src/arrow/util/config.h.cmake:
##
@@ -57,6 +57,7 @@
#cmakedefine ARROW_GCS
#cmakedefine ARROW_HDFS
#cmakedefine ARROW_S3
+#cmakedefine ARROW_TEST_MEMCHECK
pitrou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2051897809
I've reported https://github.com/conda-forge/libxml2-feedstock/issues/117
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub
pitrou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2051880913
Ok, that's because `--with-tls` doesn't get passed here:
https://github.com/conda-forge/libxml2-feedstock/blob/main/recipe/build.sh
--
This is an automated message from the Apache
pitrou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2051877190
Judging by the disassembly of `xmlGetThreadLocalStorage` on conda-forge's
libxml2, it seems that `USE_TLS` doesn't get enabled in that package.
--
This is an automated message from the
felipecrv commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2051875729
And the Azure SDK depends on it:
```
_deps/azure_sdk-build/sdk/storage/azure-storage-common/azure-storage-common-cppConfig.cmake
35: find_dependency(LibXml2)
```
--
pitrou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2051867304
> Where does `libxml2` come in our case? Is it vendored by the Azure SDK?
Something else?
Answering myself: it's version 2.12.6 from conda-forge.
--
This is an automated message
pitrou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2051834340
Note that `xmlNewGlobalState` seems to be some kind of legacy function
("backwards compatibility
of libxml2 to pre-thread-safe behaviour"):
pitrou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2051819043
> But I suspect it's something in the Azure SDK (exception-safety failure?)
because `GetFileInfoWithSelectorFromContainer` doesn't do complicated memory
management.
That may
felipecrv commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2051811767
> Ok... This seems to assume that Valgrind is buggy and doesn't handle a
calloc() call correctly, while our Azure FS implementation (and/or the Azure
SDK) is spotless and cannot
pitrou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2051322908
Ok... This seems to assume that Valgrind is buggy and doesn't handle a
calloc() call correctly, while our Azure FS implementation (and/or the Azure
SDK) is spotless and cannot contain any
kou commented on code in PR #41163:
URL: https://github.com/apache/arrow/pull/41163#discussion_r1562240202
##
cpp/src/arrow/util/config.h.cmake:
##
@@ -57,6 +57,7 @@
#cmakedefine ARROW_GCS
#cmakedefine ARROW_HDFS
#cmakedefine ARROW_S3
+#cmakedefine ARROW_TEST_MEMCHECK
kou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2051302206
My understanding is
https://github.com/apache/arrow/pull/40567/files#diff-7d2cbf9c6abf1ee980b8cf5a79e222543c17b3d93b12c4c0c9fa123b1bb200b6R398-R411
.
--
This is an automated message from
pitrou commented on code in PR #41163:
URL: https://github.com/apache/arrow/pull/41163#discussion_r1562232462
##
cpp/src/arrow/util/config.h.cmake:
##
@@ -57,6 +57,7 @@
#cmakedefine ARROW_GCS
#cmakedefine ARROW_HDFS
#cmakedefine ARROW_S3
+#cmakedefine ARROW_TEST_MEMCHECK
kou commented on code in PR #41163:
URL: https://github.com/apache/arrow/pull/41163#discussion_r1562230914
##
cpp/src/arrow/util/config.h.cmake:
##
@@ -57,6 +57,7 @@
#cmakedefine ARROW_GCS
#cmakedefine ARROW_HDFS
#cmakedefine ARROW_S3
+#cmakedefine ARROW_TEST_MEMCHECK
pitrou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2051262752
@kou Did we try to understand _why_ a memory leak was reported? We shouldn't
just silence the errors that annoy us.
cc @felipecrv
--
This is an automated message from the Apache
pitrou commented on code in PR #41163:
URL: https://github.com/apache/arrow/pull/41163#discussion_r1562204543
##
cpp/src/arrow/util/config.h.cmake:
##
@@ -57,6 +57,7 @@
#cmakedefine ARROW_GCS
#cmakedefine ARROW_HDFS
#cmakedefine ARROW_S3
+#cmakedefine ARROW_TEST_MEMCHECK
conbench-apache-arrow[bot] commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2051216955
After merging your PR, Conbench analyzed the 4 benchmarking runs that have
been run so far on merge-commit 6e8ac430816b84d295ddf89709c4b42e558b2f7e.
There were
kou merged PR #41163:
URL: https://github.com/apache/arrow/pull/41163
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail:
kou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2050860758
+1
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe,
github-actions[bot] commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2050826523
Revision: 5f4255fc3f37b6352527847befcfd848fe8a7d05
Submitted crossbow builds: [ursacomputing/crossbow @
kou commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2050824817
@github-actions crossbow submit test-conda-cpp-valgrind
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above
github-actions[bot] commented on PR #41163:
URL: https://github.com/apache/arrow/pull/41163#issuecomment-2050824889
:warning: GitHub issue #41004 **has been automatically assigned in GitHub**
to PR creator.
--
This is an automated message from the Apache Git Service.
To respond to the
27 matches
Mail list logo
