Bug#1063951: fixed in barectf 3.1.2-2
On 2024-03-24 18:28, Timo Röhling wrote: Control: reopen -1 Control: notfixed -1 3.1.2-2 On Fri, 15 Mar 2024 21:44:05 + Debian FTP Masters wrote: * [2375fb4] Disable pytest8 deprecation warnings (Closes: #1063951) - Also (Closes: #1066742) built with pytest8 Unfortunately, this was not a permanent solution, because the features which PytestRemovedIn8Warning has been warning about, have been actually removed now (pytest 8.1). Cheers Timo They seem to have renamed 'PytestRemovedIn8Warning' to 'PytestRemovedIn9Warning' and we can't ignore a non-existent exception type without generating an error. When I remove it from the ignore list and just keep the 2 DeprecationWarning types, I can successfully run the test suite with pytest 8.1.1. I'll go for this until upstream adds support for pytest8. Michael
Bug#1043037: Reasonable fork of MLMMJ available now
On Sat, 23 Dec 2023 18:10:34 -0500 Chris Knadle wrote: retitle 1043037 Switch upstream source to fork after release of MLMMJ 1.4 summary 1043037 MLMMJ upstream is dead since 2017, but users of MLMMJ have created a usable fork with new features; it's time to switch to it thanks New location for ongoing fork MLMMJ releases (current release is 1.4.3): https://codeberg.org/mlmmj/mlmmj/releases Hi, I've prepared an updated mlmmj 1.4.4 package in a personal repo [1], I've also been working with upstream to fix some issues I've encountered while updating the packaging code. Please have a look and feel free to take only the pieces you like, I don't run an mlmmj server yet so other than running the test suite it's quite untested. There are some test failures on Salsa-CI which I'm unable to reproduce locally at the moment. I would like to eventually migrate an existing mailman2 system to mlmmj after trixie is release, hence this effort. Cheers, Michael [1] https://salsa.debian.org/mjeanson/mlmmj
Bug#1051075: babeltrace2: dpkg-gensymbols: error: some symbols or patterns disappeared in the symbols file
On 2023-09-02 06:53, zhangdandan wrote: Source: babeltrace2 Version: 2.0.5-1 Severity: wishlist Tags: sid ftbfs User: debian-de...@lists.debian.org Usertags: loongarch64 Dear maintainers, When compiling the package babeltrace2 for loong64 in the Debian Package Auto-Building environment [1], The error message is as follows. dpkg-gensymbols: error: some symbols or patterns disappeared in the symbols file: see diff output below dpkg-gensymbols: warning: debian/libbabeltrace2-0/DEBIAN/symbols doesn't match completely debian/libbabeltrace2-0.symbols --- debian/libbabeltrace2-0.symbols (libbabeltrace2-0_2.0.5-1_loong64) +++ dpkg-gensymbolsee2M2Y 2023-09-01 15:10:42.916098851 + @@ -193,14 +193,14 @@ __bt_get_end_section_component_class_descriptors@Base 2.0.0 __bt_get_end_section_plugin_descriptor_attributes@Base 2.0.0 __bt_get_end_section_plugin_descriptors@Base 2.0.0 - __start___bt_plugin_component_class_descriptor_attributes@Base 2.0.0 - __start___bt_plugin_component_class_descriptors@Base 2.0.0 - __start___bt_plugin_descriptor_attributes@Base 2.0.0 - __start___bt_plugin_descriptors@Base 2.0.0 - __stop___bt_plugin_component_class_descriptor_attributes@Base 2.0.0 - __stop___bt_plugin_component_class_descriptors@Base 2.0.0 - __stop___bt_plugin_descriptor_attributes@Base 2.0.0 - __stop___bt_plugin_descriptors@Base 2.0.0 +#MISSING: 2.0.5-1# __start___bt_plugin_component_class_descriptor_attributes@Base 2.0.0 +#MISSING: 2.0.5-1# __start___bt_plugin_component_class_descriptors@Base 2.0.0 +#MISSING: 2.0.5-1# __start___bt_plugin_descriptor_attributes@Base 2.0.0 +#MISSING: 2.0.5-1# __start___bt_plugin_descriptors@Base 2.0.0 +#MISSING: 2.0.5-1# __stop___bt_plugin_component_class_descriptor_attributes@Base 2.0.0 +#MISSING: 2.0.5-1# __stop___bt_plugin_component_class_descriptors@Base 2.0.0 +#MISSING: 2.0.5-1# __stop___bt_plugin_descriptor_attributes@Base 2.0.0 +#MISSING: 2.0.5-1# __stop___bt_plugin_descriptors@Base 2.0.0 bt_clock_class_borrow_user_attributes@Base 2.0.0 bt_clock_class_borrow_user_attributes_const@Base 2.0.0 bt_clock_class_create@Base 2.0.0 dh_makeshlibs: error: failing due to earlier errors The full build log is available from the link [2]. Please consider updating the symbols file. If you have any questions, you can contact me at any time. [1]:https://buildd.debian.org/status/package.php?p=babeltrace2=sid [2]:https://buildd.debian.org/status/fetch.php?pkg=babeltrace2=loong64=2.0.5-1=1693581057=0 thanks, Dandan Zhang Hi, I'd like to understand first why those symbols are not present when building on loongarch, do you have an idea? Thanks, Michael
Bug#1054088: sanoid: let dh_installsystemd choose the location of units
On 2023-10-16 15:38, Helmut Grohne wrote: Source: sanoid Version: 2.2.0-1 Tags: patch User: helm...@debian.org Usertags: dep17m2 We want to move all aliased files from / to /usr to finalize the /usr-merge transition via DEP17. sanoid is affected, because it installs two systemd units. One of its units is installed via dh_installsystemd and will move to /usr when the placement in dh_installsystemd changes. The other unit is hard coded to /lib/systemd/system via debian/sanoid.install. Rather than moving it to /usr, I recommend installing it using dh_installsystemd as well, because it'll revert back to the old location when uploading to bookworm-backports, which will comply with the file move moratorium that still is in effect for bookworm. I'm attaching a patch for your convenience. Helmut Thanks for the patch, it will be included in the next upload.
Bug#1053376: RM: lttnganalyses -- ROM; obsolete, dead upstream
Package: ftp.debian.org Severity: normal User: ftp.debian@packages.debian.org Usertags: remove X-Debbugs-Cc: lttnganaly...@packages.debian.org Control: affects -1 + src:lttnganalyses Hi, Please remove lttnganalyses, it's been unmaintained for a while and broken with recent python versions. Thanks!
Bug#1052595: liburcu: Please add support for loong64
On 2023-09-24 22:51, JiaLing Zhang wrote: Source: liburcu Version: 0.14.0-1 Severity: serious Tags: ftbfs loongarch64 Hello! the liburcu build failed in buildd , the upstream have support loongarch. Please update the source. thanks! JiaLing [1]: https://github.com/urcu/userspace-rcu/commit/dc46a9c324ae94d89da41ea9a3f97503115df88e Hi, It seems like atomic operations on RISC-V with current GCC are buggy [1] [2], we will have to wait for the release of GCC 13.3 before we can add support for loong64 to liburcu. Thanks, Michael [1] https://github.com/urcu/userspace-rcu/commit/46980605309e922d14f646c6705d184cb674c0f7 [2] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104831
Bug#1053358: Breaks if user-defined zfs properties contain whitespace
Hi, Can you open a bug upstream? I would prefer to go with a fix approved by upstream. Also I'm not sure I understand under which circumstances the content of '$value' could be controlled by an 'adversary'? Can you explain shortly what would be an exploitation scenario you envision? Thanks, Michael
Bug#1035464: bullseye-pu: package lttng-modules/2.12.5-1+deb11u1
On 2023-09-29 05:16, Andreas Beckmann wrote: On Sat, 23 Sep 2023 21:53:12 +0100 "Adam D. Barratt" wrote: On Wed, 2023-05-03 at 11:34 -0400, Michael Jeanson wrote: > Fix the dkms build of lttng-modules against the current bullseye > kernel 5.10.0-22. > Please go ahead; sorry for the delay. Hi Michael, I've uploaded your proposed changes to DELAYED/1 to ensure it gets into the next point release (scheduled for Oct 7). Andreas Thanks for taking care of this. Michael
Bug#1036106: unblock: lttng-modules/2.13.9-1
Package: release.debian.org Severity: normal User: release.debian@packages.debian.org Usertags: unblock X-Debbugs-Cc: lttng-modu...@packages.debian.org Control: affects -1 + src:lttng-modules Please unblock package lttng-modules The 2.13.9 release currently in testing contains fixes to build against the current 5.10.0-22 kernel in stable, keeping the current 2.13.8 release might result in a failed upgrade from bulllseye to bookworm. [ Reason ] The latest 5.10.0 kernel in stable results in a build failure with 2.13.8. [ Impact ] Potential failed upgrade to bookworm. [ Tests ] Manually tested with 5.10.0-22 and 6.1.0-5. [ Risks ] Low, this is a not widely used package and the changes between 2.13.8 and 2.13.9 are very targeted. [ Checklist ] [x] all changes are documented in the d/changelog [x] I reviewed all changes and I approve them [x] attach debdiff against the package in testing unblock lttng-modules/2.13.9-1 diff -Nru lttng-modules-2.13.8/ChangeLog lttng-modules-2.13.9/ChangeLog --- lttng-modules-2.13.8/ChangeLog 2023-01-13 16:08:06.0 -0500 +++ lttng-modules-2.13.9/ChangeLog 2023-03-03 10:39:24.0 -0500 @@ -1,3 +1,9 @@ +2023-03-03 (Canadian Bacon Day) LTTng modules 2.13.9 + * fix: jbd2: use the correct print format (v5.4.229) + * fix: jbd2 upper bound for v5.10.163 + * fix: jbd2: use the correct print format (v5.10.163) + * fix: btrfs: move accessor helpers into accessors.h (v6.2) + 2023-01-13 (National Sticker Day) LTTng modules 2.13.8 * fix: jbd2: use the correct print format * Fix: in_x32_syscall was introduced in v4.7.0 diff -Nru lttng-modules-2.13.8/debian/changelog lttng-modules-2.13.9/debian/changelog --- lttng-modules-2.13.8/debian/changelog 2023-01-16 11:47:18.0 -0500 +++ lttng-modules-2.13.9/debian/changelog 2023-03-07 14:12:32.0 -0500 @@ -1,3 +1,10 @@ +lttng-modules (2.13.9-1) unstable; urgency=medium + + * [2f1b62b] New upstream version 2.13.9 +- Bugfix release, adds support for v6.2 and multiple stable kernels. + + -- Michael Jeanson Tue, 07 Mar 2023 14:12:32 -0500 + lttng-modules (2.13.8-1) unstable; urgency=medium * [813bc03] New upstream version 2.13.8 diff -Nru lttng-modules-2.13.8/include/instrumentation/events/btrfs.h lttng-modules-2.13.9/include/instrumentation/events/btrfs.h --- lttng-modules-2.13.8/include/instrumentation/events/btrfs.h 2023-01-13 16:08:06.0 -0500 +++ lttng-modules-2.13.9/include/instrumentation/events/btrfs.h 2023-03-03 10:39:24.0 -0500 @@ -9,6 +9,10 @@ #include #include +#if (LTTNG_LINUX_VERSION_CODE >= LTTNG_KERNEL_VERSION(6,2,0)) +#include <../fs/btrfs/accessors.h> +#endif + #ifndef _TRACE_BTRFS_DEF_ #define _TRACE_BTRFS_DEF_ struct btrfs_root; diff -Nru lttng-modules-2.13.8/include/instrumentation/events/jbd2.h lttng-modules-2.13.9/include/instrumentation/events/jbd2.h --- lttng-modules-2.13.8/include/instrumentation/events/jbd2.h 2023-01-13 16:08:06.0 -0500 +++ lttng-modules-2.13.9/include/instrumentation/events/jbd2.h 2023-03-03 10:39:24.0 -0500 @@ -28,6 +28,8 @@ ) #if (LTTNG_LINUX_VERSION_CODE >= LTTNG_KERNEL_VERSION(6,2,0) \ + || LTTNG_KERNEL_RANGE(5,4,229, 5,5,0) \ + || LTTNG_KERNEL_RANGE(5,10,163, 5,11,0) \ || LTTNG_KERNEL_RANGE(5,15,87, 5,16,0) \ || LTTNG_KERNEL_RANGE(6,0,18, 6,1,0) \ || LTTNG_KERNEL_RANGE(6,1,4, 6,2,0)) @@ -96,6 +98,8 @@ #endif #if (LTTNG_LINUX_VERSION_CODE >= LTTNG_KERNEL_VERSION(6,2,0) \ + || LTTNG_KERNEL_RANGE(5,4,229, 5,5,0) \ + || LTTNG_KERNEL_RANGE(5,10,163, 5,11,0) \ || LTTNG_KERNEL_RANGE(5,15,87, 5,16,0) \ || LTTNG_KERNEL_RANGE(6,0,18, 6,1,0) \ || LTTNG_KERNEL_RANGE(6,1,4, 6,2,0)) @@ -138,6 +142,8 @@ ) #if (LTTNG_LINUX_VERSION_CODE >= LTTNG_KERNEL_VERSION(6,2,0) \ + || LTTNG_KERNEL_RANGE(5,4,229, 5,5,0) \ + || LTTNG_KERNEL_RANGE(5,10,163, 5,11,0) \ || LTTNG_KERNEL_RANGE(5,15,87, 5,16,0) \ || LTTNG_KERNEL_RANGE(6,0,18, 6,1,0) \ || LTTNG_KERNEL_RANGE(6,1,4, 6,2,0)) diff -Nru lttng-modules-2.13.8/include/lttng/tracer.h lttng-modules-2.13.9/include/lttng/tracer.h --- lttng-modules-2.13.8/include/lttng/tracer.h 2023-01-13 16:08:06.0 -0500 +++ lttng-modules-2.13.9/include/lttng/tracer.h 2023-03-03 10:39:24.0 -0500 @@ -28,7 +28,7 @@ #define LTTNG_MODULES_MAJOR_VERSION 2 #define LTTNG_MODULES_MINOR_VERSION 13 -#define LTTNG_MODULES_PATCHLEVEL_VERSION 8 +#define LTTNG_MODULES_PATCHLEVEL_VERSION 9 #define LTTNG_MODULES_EXTRAVERSION "" #define LTTNG_VERSION_NAME "Nordicité"
Bug#1035464: bullseye-pu: package lttng-modules/2.12.5-1+deb11u1
Package: release.debian.org Severity: normal Tags: bullseye User: release.debian@packages.debian.org Usertags: pu X-Debbugs-Cc: lttng-modu...@packages.debian.org Control: affects -1 + src:lttng-modules [ Reason ] Fix the dkms build of lttng-modules against the current bullseye kernel 5.10.0-22. [ Impact ] I'ts currently impossible to use the lttng kernel tracer with the latest bullseye kernel. [ Tests ] Tested manually on a bullseye virtual machine. [ Risks ] Minimal, won't be more broken than it actually is. [ Checklist ] [*] *all* changes are documented in the d/changelog [*] I reviewed all changes and I approve them [*] attach debdiff against the package in (old)stable [*] the issue is verified as fixed in unstable [ Changes ] Backport the minimum list of upstream patches to support building on 5.10.0-22. diff -Nru lttng-modules-2.12.5/debian/changelog lttng-modules-2.12.5/debian/changelog --- lttng-modules-2.12.5/debian/changelog 2021-02-17 17:12:39.0 -0500 +++ lttng-modules-2.12.5/debian/changelog 2023-05-03 11:13:07.0 -0400 @@ -1,3 +1,20 @@ +lttng-modules (2.12.5-1+deb11u1) bullseye; urgency=medium + + * Fix build on linux 5.10.0-22 (Closes: #1035364) + + [ Michael Jeanson ] + * [a952a3a] Adjust gbp.conf for bullseye stable update + + [ Povilas Kanapickas ] + * [ab16ac0] Add patch to fix build on Linux 5.10.137..5.11 + * [25013d7] Add patch to fix build on Linux 5.10.119..5.11 + + [ Michael Jeanson ] + * [90a214b] dkms: conditionnaly include lttng-probe-random.ko + * [be2eaa4] Add patch to fix build on Linux 5.10.163..5.11 + + -- Michael Jeanson Wed, 03 May 2023 11:13:07 -0400 + lttng-modules (2.12.5-1) unstable; urgency=medium * [8e0b514] New upstream version 2.12.5 diff -Nru lttng-modules-2.12.5/debian/gbp.conf lttng-modules-2.12.5/debian/gbp.conf --- lttng-modules-2.12.5/debian/gbp.conf2021-02-17 17:12:39.0 -0500 +++ lttng-modules-2.12.5/debian/gbp.conf2023-05-01 15:01:42.0 -0400 @@ -1,3 +1,3 @@ [DEFAULT] -upstream-branch=upstream/latest -debian-branch=debian/sid +upstream-branch=upstream/stable-2.12.5 +debian-branch=debian/bullseye diff -Nru lttng-modules-2.12.5/debian/lttng-modules-dkms.dkms.in lttng-modules-2.12.5/debian/lttng-modules-dkms.dkms.in --- lttng-modules-2.12.5/debian/lttng-modules-dkms.dkms.in 2021-02-17 17:12:39.0 -0500 +++ lttng-modules-2.12.5/debian/lttng-modules-dkms.dkms.in 2023-05-01 15:07:32.0 -0400 @@ -169,10 +169,12 @@ DEST_MODULE_LOCATION[$i]="/extra/probes" i=$((i+1)) -BUILT_MODULE_NAME[$i]="lttng-probe-random" -BUILT_MODULE_LOCATION[$i]="probes/" -DEST_MODULE_LOCATION[$i]="/extra/probes" -i=$((i+1)) +if [ -f "$kernel_source_dir/include/trace/events/random.h" ]; then +BUILT_MODULE_NAME[$i]="lttng-probe-random" +BUILT_MODULE_LOCATION[$i]="probes/" +DEST_MODULE_LOCATION[$i]="/extra/probes" +i=$((i+1)) +fi BUILT_MODULE_NAME[$i]="lttng-probe-rcu" BUILT_MODULE_LOCATION[$i]="probes/" diff -Nru lttng-modules-2.12.5/debian/patches/fix-adjust-range-v5.10.137-in-block-probe.patch lttng-modules-2.12.5/debian/patches/fix-adjust-range-v5.10.137-in-block-probe.patch --- lttng-modules-2.12.5/debian/patches/fix-adjust-range-v5.10.137-in-block-probe.patch 1969-12-31 19:00:00.0 -0500 +++ lttng-modules-2.12.5/debian/patches/fix-adjust-range-v5.10.137-in-block-probe.patch 2023-05-01 15:04:13.0 -0400 @@ -0,0 +1,92 @@ +From bee932ee7580bfa50e58cf4bb1e1bf98a0a80b15 Mon Sep 17 00:00:00 2001 +From: Michael Jeanson +Date: Mon, 22 Aug 2022 14:16:27 -0400 +Subject: [PATCH] fix: adjust range v5.10.137 in block probe + +See upstream commit, backported in v5.10.137 : + +commit 1cb3032406423b25aa984854b4d78e0100d292dd +Author: Christoph Hellwig +Date: Thu Dec 3 17:21:39 2020 +0100 + +block: remove the request_queue to argument request based +tracepoints + +[ Upstream commit a54895fa057c67700270777f7661d8d3c7fda88a ] + +The request_queue can trivially be derived from the request. + +Change-Id: I01f96a437641421faf993b4b031171c372bd0374 +Signed-off-by: Michael Jeanson +Signed-off-by: Mathieu Desnoyers +--- + instrumentation/events/lttng-module/block.h | 18 -- + 1 file changed, 12 insertions(+), 6 deletions(-) + +diff --git a/instrumentation/events/lttng-module/block.h b/instrumentation/events/lttng-module/block.h +index 4212b048..9eb77e7a 100644 +--- a/instrumentation/events/lttng-module/block.h b/instrumentation/events/lttng-module/block.h +@@ -266,7 +266,8 @@ LTTNG_TRACEPOINT_EVENT_INSTANCE(block_rq_with_error, block_rq_abort, + ) + #endif + +-#if (LTTNG_LINUX_VERSION_CODE >= LTTNG_KERNEL_VERSION(5,11,0)) ++#if (LTTNG_LINUX_VERSION_CODE >= LTTNG_KERNEL_VERSION(5,11,0) \ ++ || LTTNG_KERNEL_RANGE(5,10,137, 5,11,0)) + /** + * block_rq_requeue - place block IO request
Bug#995268: RFA: popt -- lib for parsing cmdline parameters - development files
Package: wnpp Severity: normal Control: affects -1 src:popt I request an adopter for the popt package, I don't have time to properly maintain it. The package is in pretty good shape but there is some patches and bugs to triage and address. The package description is: Popt was heavily influenced by the getopt() and getopt_long() functions, but it allows more powerful argument expansion. It can parse arbitrary argv[] style arrays and automatically set variables based on command line arguments. It also allows command line arguments to be aliased via configuration files and includes utility functions for parsing arbitrary strings into argv[] arrays using shell-like rules. . This package contains the popt static library and header file.
Bug#982270: linux: [armhf] several imx6 systems unable to detect ethernet
This looks to be fixed in '5.10.38-1', at least on the single Wandboard Quad (rev C1) I tested, networking works at boot. I'll update the kernel on the boards that are CI workers and report back in a couple days if the network is stable.
Bug#976920: liburcu: FTBFS on ppc64el: dh_auto_test: error: make -j160 check VERBOSE=1 regtest returned exit code 2
On Wed, Dec 9, 2020 at 4:09 AM Lucas Nussbaum wrote: > > Source: liburcu > Version: 0.12.1-1 > Severity: serious > Justification: FTBFS on ppc64el > Tags: bullseye sid ftbfs > Usertags: ftbfs-20201209 ftbfs-bullseye ftbfs-ppc64el > > Hi, > > During a rebuild of all packages in sid, your package failed to build > on ppc64el. At the same time, it did not fail on amd64. > > I'm marking this bug as severity:serious since your package currently has > ppc64el binary packages in unstable (so this is a regression). > That is a lot of CPUs, I'll check with the upstream author why there is a 128 threads limit in the test suite.
Bug#973706: buster-pu: package lttng-modules/2.10.8-1
Package: release.debian.org Severity: normal Tags: buster User: release.debian@packages.debian.org Usertags: pu Hi, The attached diff fixes a build failure of the dkms modules on the 4.19.0-11 buster linux kernel. This was reported at: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=972321 I'd appreciate if you consider allowing this upload to buster. Thanks, Michael diff -Nru lttng-modules-2.10.8/debian/changelog lttng-modules-2.10.8/debian/changelog --- lttng-modules-2.10.8/debian/changelog 2018-11-02 16:13:20.0 -0400 +++ lttng-modules-2.10.8/debian/changelog 2020-11-03 11:46:36.0 -0500 @@ -1,3 +1,10 @@ +lttng-modules (2.10.8-1+deb10u1) buster; urgency=medium + + * [5c8aed8] Update debian/gbp.conf for buster + * [16882db] Fix build on >= 4.19.0-10 kernels (Closes: #972321) + + -- Michael Jeanson Tue, 03 Nov 2020 11:46:36 -0500 + lttng-modules (2.10.8-1) unstable; urgency=medium * [7037820] New upstream version 2.10.8 diff -Nru lttng-modules-2.10.8/debian/gbp.conf lttng-modules-2.10.8/debian/gbp.conf --- lttng-modules-2.10.8/debian/gbp.conf2018-07-04 11:21:39.0 -0400 +++ lttng-modules-2.10.8/debian/gbp.conf2020-11-03 11:30:17.0 -0500 @@ -1,3 +1,3 @@ [DEFAULT] -upstream-branch=upstream/latest -debian-branch=debian/sid +upstream-branch=upstream/2.10.8 +debian-branch=debian/buster diff -Nru lttng-modules-2.10.8/debian/patches/0001-fix-writeback-Fix-sync-livelock-due-to-b_dirty_time-.patch lttng-modules-2.10.8/debian/patches/0001-fix-writeback-Fix-sync-livelock-due-to-b_dirty_time-.patch --- lttng-modules-2.10.8/debian/patches/0001-fix-writeback-Fix-sync-livelock-due-to-b_dirty_time-.patch 1969-12-31 19:00:00.0 -0500 +++ lttng-modules-2.10.8/debian/patches/0001-fix-writeback-Fix-sync-livelock-due-to-b_dirty_time-.patch 2020-11-03 11:30:17.0 -0500 @@ -0,0 +1,122 @@ +From 8e6be22dd4bf96b08447d075ce2fe3916f47764b Mon Sep 17 00:00:00 2001 +From: Michael Jeanson +Date: Mon, 31 Aug 2020 14:16:01 -0400 +Subject: [PATCH] fix: writeback: Fix sync livelock due to b_dirty_time + processing (v5.9) + +See upstream commit: + + commit f9cae926f35e8230330f28c7b743ad088611a8de + Author: Jan Kara + Date: Fri May 29 16:08:58 2020 +0200 + +writeback: Fix sync livelock due to b_dirty_time processing + +When we are processing writeback for sync(2), move_expired_inodes() +didn't set any inode expiry value (older_than_this). This can result in +writeback never completing if there's steady stream of inodes added to +b_dirty_time list as writeback rechecks dirty lists after each writeback +round whether there's more work to be done. Fix the problem by using +sync(2) start time is inode expiry value when processing b_dirty_time +list similarly as for ordinarily dirtied inodes. This requires some +refactoring of older_than_this handling which simplifies the code +noticeably as a bonus. + +Change-Id: I99e505965d278565ae512a30842cdce888e0c84a +Signed-off-by: Michael Jeanson +Signed-off-by: Mathieu Desnoyers +--- + .../events/lttng-module/writeback.h | 46 +-- + 1 file changed, 33 insertions(+), 13 deletions(-) + +diff --git a/instrumentation/events/lttng-module/writeback.h b/instrumentation/events/lttng-module/writeback.h +index c472b33..353d58a 100644 +--- a/instrumentation/events/lttng-module/writeback.h b/instrumentation/events/lttng-module/writeback.h +@@ -371,34 +371,55 @@ LTTNG_TRACEPOINT_EVENT_WBC_INSTANCE(wbc_balance_dirty_wait, writeback_wbc_balanc + #endif + LTTNG_TRACEPOINT_EVENT_WBC_INSTANCE(wbc_writepage, writeback_wbc_writepage) + +-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,1,0)) ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(5,9,0) || \ ++ LTTNG_KERNEL_RANGE(5,8,6, 5,9,0) || \ ++ LTTNG_KERNEL_RANGE(5,4,62, 5,5,0) || \ ++ LTTNG_KERNEL_RANGE(4,19,143, 4,20,0) || \ ++ LTTNG_KERNEL_RANGE(4,14,196, 4,15,0) || \ ++ LTTNG_KERNEL_RANGE(4,9,235, 4,10,0) || \ ++ LTTNG_KERNEL_RANGE(4,4,235, 4,5,0) || \ ++ LTTNG_UBUNTU_KERNEL_RANGE(4,15,18,119, 4,16,0,0)) ++LTTNG_TRACEPOINT_EVENT(writeback_queue_io, ++ TP_PROTO(struct bdi_writeback *wb, ++ struct wb_writeback_work *work, ++ unsigned long dirtied_before, ++ int moved), ++ TP_ARGS(wb, work, dirtied_before, moved), ++ TP_FIELDS( ++ ctf_array_text(char, name, dev_name(wb->bdi->dev), 32) ++ ctf_integer(unsigned long, older, dirtied_before) ++ ctf_integer(int, moved, moved) ++ ) ++) ++#elif (LINUX_VERSION_CODE >= KERNEL_VERSION(3,2,0)) + LTTNG_TRACEPOINT_EVENT(writeback_queue_io, + TP_PROTO(struct bdi_writeback *wb, +-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,2,0)) +struct wb_writeback_work *work, +-#else +- unsigned long *older_than_this, +-#endif +int moved), +-#if (LINUX_VERSION_CODE &g
Bug#972321: lttng-modules-dkms fails to build since 10.6
On Wed, Oct 21, 2020 at 5:06 AM wrote: > > Dear Maintainer, > > Do you have any idea on this issue? > > Best, > > Fukui Hi, The instrumentation of a writeback probe was modified in newer 4.19 kernels, a patch [1] to lttng-modules will be required to fix the build. I'll push a stable update but it might take a while to land in buster-updates. In the meantime, you can apply the patch locally or build the modules from the latest 2.10 release tarball which already includes the patch. Michael [1] https://github.com/lttng/lttng-modules/commit/7318a1e88c8097733f5a05529069e4e03c142cae
Bug#968020: RM: lttv -- ROM; Dead upstream
Package: ftp.debian.org Severity: normal Hi ftp-master, Please remove src:lttv, it is unmaintained upstream, the last release was in 2013. It's usefulness is very limited as it can't reliably open traces produced with any currently maintained LTTng version. Thanks, Michael
Bug#949765: ITP: babeltrace2 -- A trace manipulation toolkit
Package: wnpp Severity: wishlist Owner: Michael Jeanson * Package name: babeltrace2 Version : 2.0.0 Upstream Author : Jérémie Galarneau * URL : https://www.babeltrace.org/ * License : MIT Programming Lang: C Description : A trace manipulation toolkit The Babeltrace 2 project offers a library with a C API, Python 3 bindings, and a command-line tool which makes it easy for mere mortals to view, convert, transform, and analyze traces. Babeltrace 2 is also the reference parser implementation of the Common Trace Format (CTF), a versatile trace format followed by various tracers and tools such as LTTng and barectf. This package is required because the API and the basename of the library has changed since babeltrace1, current dependencies like gdb and ceph will need to be ported to this new API and both package will have to be co-installable for a while.
Bug#948257: depmod: ERROR: ../libkmod/libkmod.c:515 lookup_builtin_file() could not open builtin file '/var/tmp/mkinitramfs_N1a1Mk/lib/modules/5.4.0-1-amd64/modules.builtin.bin'
On 2020-01-06 7:46 p.m., Benjamin Poirier wrote: > I'm not sure if it's related but I saw almost the same error on last > upgrade (but for 5.4.0-2): > > depmod: ERROR: ../libkmod/libkmod.c:515 lookup_builtin_file() could not open > builtin file > '/var/tmp/mkinitramfs_J2sneW/lib/modules/5.4.0-2-amd64/modules.builtin.bin' > > and I now get: > > root@vsid:~# lttng list --kernel > Error: Unable to list kernel events: Kernel tracer not available > root@vsid:~# journalctl -u lttng-sessiond.service > [...] > Jan 07 09:33:20 vsid lttng-sessiond[403]: Error: Failed to load kmod library > resources > Jan 07 09:33:20 vsid lttng-sessiond[403]: Warning: No kernel tracer available > > lttng-modules-dkms is installed. I can load the modules manually but I > still get the same error. > Hi, If you had just installed the lttng-modules-dkms and lttng-tools packages, it's possible that the lttng-sessiond deamon was started before the kernel modules were built and so it couldn't load them. Simply restarting the sessiond should fix this. Cheers, Michael
Bug#941814: libpopt: leaks memory for leftover arguments
found #941814 Regressions were reported in gdisk and svox, I reverted the patch until more testing can be done. gdisk/1.0.4-3: https://ci.debian.net/data/autopkgtest/testing/amd64/g/gdisk/3216686/log.gz svox/1.0+git20130326-9: https://ci.debian.net/data/autopkgtest/testing/amd64/s/svox/3216687/log.gz
Bug#933635: popt: The upstream domain is unreachable
On 2019-08-01 4:45 a.m., Paride Legovini wrote: > Source: popt > Version: 1.16-12 > Severity: normal > > The upstream domain listed in d/copyright and d/control: > > http://rpm5.org/files/popt/ > > is not functional anymore. The domain itself it still registered, but no > service appears to be active there. Arch Linux and possibly other > distributions are using Debian as their upstream, see: > > https://git.archlinux32.org/archlinux32/packages/src/branch/master/core/popt/PKGBUILD > > Paride > Until a new upstream emerges, there is not much we can do. Hopefully that will happen sooner than later. Thanks for the report, Michael
Bug#931147: gdb depends on newer libbabeltrace
On 2019-08-24 6:12 a.m., Andreas Beckmann wrote: > Followup-For: Bug #931147 > > Hi, > > I've just opened a PU request for babeltrace and a binNMU request > for gdb etc. to get this fixed in buster, too. > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=935583 > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=935588 > > Andreas > Thanks for doing this, I appreciate the help. Michael
Bug#927010: python3-babeltrace: Event instances all refer to a constantly-overwritten data - unexpected behaviours and segfaults
On 2019-04-13 10:23 a.m., ydir...@free.fr wrote: > Package: python3-babeltrace > Version: 1.5.6-2 > Severity: important > > While investigating why I was getting a segfault while processing a CTF trace > (in which I > do a first pass extracting some info, before a second pass plotting the > data... > while keeping a ref to various events to avoid making copies of everything in > the first pass), I finally understand that even though we get individual > python handles > for all events, they get clobbered by further event parsing. Hi, This is a known limitation of the babeltrace 1 python bindings, you'll have to either deep copy the events you want to do further processing on or iterate multiple times on the trace collection. This limitation will be fixed in the bt2 bindings and you'll be able to keep references to events. Regards, Michael
Bug#894824: popt FTCBFS: api-sanity-checker dependency unsatisfiable
On 2018-12-28 7:24 a.m., Niels Thykier wrote: > Hi Michael, > > Could I perhaps convince you to do an upload with this patch in > time for buster? It will enable us to remove one hack/work around > in the "bootstrap.sh" script currently in use for bootstrapping > new architectures (as well as improve our cross-building stats for > buster). Also, you mention that you have applied it in git but I > cannot see it in salsa (admittedly, I only checked the master > branch). Have you perhaps forgotten to push it? > > If you would like some additional changes/improvements to popt > (albeit unrelated to this bug), then I can think of: > > * The override for dh_auto_build might be redundant now that > debhelper passes --disable-silent-rules (as of > debhelper/9.20140613) > > * The popt looks like a package that might be able to build with > "Rules-Requires-Root: no" set in the Source in d/control. - If that > fails in sid because the upstream build system relies on root, > please set "Rules-Requires-Root: binary-targets" explicitly. This > will enable us to flip the default and make "(fake)root" opt-in > rather than opt-out in the future. > > Thanks, ~Niels Sorry for the lack of reply, I've just uploaded a package with the fix included as well as addressing your other comments. Michael
Bug#910577: gdb-multiarch: libbabeltrace1 dependency not tight enough
On 2018-12-11 2:54 p.m., Adrian Bunk wrote: > Most important is to fix the root cause: > > $ cat /var/lib/dpkg/info/libbabeltrace1\:amd64.shlibs > libbabeltrace-ctf-metadata 1 libbabeltrace1 > libbabeltrace-ctf-text 1 libbabeltrace1 > libbabeltrace-ctf 1 libbabeltrace1 > ... > $ > > Reverse dependencies could then be fixed for buster by binNMU. Ok, so if I understand correctly. Either, "libbabeltrace-ctf1" should be added to the dependencies of the 3 ctf shared objects in the shlibs file. That way packages that link with any of theses libs will have an explicit dependency on "libbabeltrace-ctf1", which will work with both old and new babeltrace packages. Or the dependency on "libbabeltrace1" should be versioned to explicitly pull a package that includes the ctf sub-library. Do you have an opinion on which of these solutions is preferable. Thanks, Michael
Bug#910577: gdb-multiarch: libbabeltrace1 dependency not tight enough
On 2018-11-26 3:51 p.m., Adrian Bunk wrote: > Control: reassign -1 libbabeltrace1 1.5.3-2 > Control: severity -1 serious > Control: affects -1 gdb-multiarch > > On Mon, Oct 08, 2018 at 11:35:10AM +0200, Matthias Urlichs wrote: >> Package: gdb-multiarch >> Version: 8.1-4 >> Severity: important >> >> gdb-multiarch requires libbabeltrace-ctf1. >> >> --\ Versions of libbabeltrace-ctf1 (3) >> >> p1.5.1-1 >> i1.5.6-1 >> i A libbabeltrace1 1.5.6-1 >> >> Apparently the build process picks up the fact that libbabeltrace-1.5.6-1 >> contains the -ctf sub-library, but then doesn't depend on >=1.5.6. >> Thus 1.5.1 ends up being installed, which doesn't contain the plugin. >> ... > > Reassigning to libbabeltrace1 where this bug should be fixed. > > cu > Adrian > Hi, I'm quite unsure how to fix this properly. I understand the problem but every solution I come up with involves modifying already released packages. I'm open to any suggestion. Michael
Bug#878816: stretch-pu: package lttng-modules/2.9.0-1+deb9u1
Hi, Here is an updated debdiff with a fix for an additional bug, both are failure to build the dkms modules on recent kernels. I've also addressed the metadata problems on #864404 to properly report the fixed version in unstable. Changes: lttng-modules (2.9.0-1+deb9u1) stable; urgency=medium * [ee40323] Fix build on linux-rt 4.9 kernels. (Closes: #864404) * [b20f74a] Fix build on >= 4.9.0-3 kernels (Closes: #889901) Cheers, Michael diff -Nru lttng-modules-2.9.0/debian/changelog lttng-modules-2.9.0/debian/changelog --- lttng-modules-2.9.0/debian/changelog2016-11-29 18:13:55.0 -0500 +++ lttng-modules-2.9.0/debian/changelog2018-08-10 12:47:14.0 -0400 @@ -1,3 +1,11 @@ +lttng-modules (2.9.0-1+deb9u1) stable; urgency=medium + + * [c3d8eab] Stretch gbp branch config + * [ee40323] Fix build on linux-rt 4.9 kernels. (Closes: #864404) + * [b20f74a] Fix build on >= 4.9.0-3 kernels (Closes: #889901) + + -- Michael Jeanson Fri, 10 Aug 2018 12:47:14 -0400 + lttng-modules (2.9.0-1) unstable; urgency=medium * [500ac85] New upstream version 2.9.0 diff -Nru lttng-modules-2.9.0/debian/gbp.conf lttng-modules-2.9.0/debian/gbp.conf --- lttng-modules-2.9.0/debian/gbp.conf 1969-12-31 19:00:00.0 -0500 +++ lttng-modules-2.9.0/debian/gbp.conf 2018-08-10 12:10:26.0 -0400 @@ -0,0 +1,3 @@ +[DEFAULT] +upstream-branch=upstream/2.9.0 +debian-branch=debian/stretch diff -Nru lttng-modules-2.9.0/debian/patches/fix-debian-kernel-version-parsing.patch lttng-modules-2.9.0/debian/patches/fix-debian-kernel-version-parsing.patch --- lttng-modules-2.9.0/debian/patches/fix-debian-kernel-version-parsing.patch 1969-12-31 19:00:00.0 -0500 +++ lttng-modules-2.9.0/debian/patches/fix-debian-kernel-version-parsing.patch 2018-08-10 12:38:51.0 -0400 @@ -0,0 +1,66 @@ +From 671137d5cda4a7415b34246bd8842650a7c6ddd9 Mon Sep 17 00:00:00 2001 +From: Michael Jeanson +Date: Tue, 9 Jan 2018 15:43:19 -0500 +Subject: [PATCH] Fix: debian kernel version parsing + +The debian version script only worked for ckt kernels and that was fine +until now because we only had checks for those versions in the code. + +ckt (Canonical Kernel Team) kernels were used for a while during the jessie +cycle, their versionning is a bit different. They track the upstream vanilla +stable updates but they don't update the minor version number and instead add +an additionnal -cktX. They were all 3.16.7-cktX and after a while the version +switched back to upstream style at 3.16.36. + +Knowing that, we can compare regular debian and ckt kernel versions +using this scheme : + + MAJOR.PATCHLEVEL.SUBLEVEL.CKT.DEBABI.DEBPATCH + +And setting CKT to zero for non-ckt kernels. + +Signed-off-by: Michael Jeanson +Signed-off-by: Mathieu Desnoyers +--- + abi-debian-version.sh | 18 ++ + 1 file changed, 10 insertions(+), 8 deletions(-) + +diff --git a/abi-debian-version.sh b/abi-debian-version.sh +index 310d6a8..36da212 100755 +--- a/abi-debian-version.sh b/abi-debian-version.sh +@@ -12,24 +12,26 @@ fi + + # Assuming KPATH is the target kernel headers directory + DEB_PACKAGE_VERSION=$(sed -rn 's/^#define LINUX_PACKAGE_ID " Debian (.*)"/\1/p' ${KPATH}/include/generated/package.h) ++ + # Ignore backports part + DEB_PACKAGE_VERSION=$(echo ${DEB_PACKAGE_VERSION} | sed -r 's/~(bpo|deb).*//') ++ ++# ckt (Canonical Kernel Team) kernels were used for a while during the jessie ++# cycle, their versionning is a bit different. They track the upstream vanilla ++# stable updates but they don't update the minor version number and instead add ++# an additionnal -cktX. They were all 3.16.7-cktX and after a while the version ++# switched back to upstream style at 3.16.36. ++ + # Get -ckt update number, if present + KERNEL_CKT_UPDATE=$(echo ${DEB_PACKAGE_VERSION} | sed -rn 's/^[0-9]+\.[0-9]+\.[0-9]+-ckt([0-9]+).*/\1/p') +- +-# Only care about the rest if it is a -ckt kernel, making sure we do not +-# clash with older Debian kernels (e.g. Debian 3.2.65-1+deb7u2). +-if [ -z "${KERNEL_CKT_UPDATE}" ]; then +- echo 0 +- exit 0 +-fi ++test -n "${KERNEL_CKT_UPDATE}" || KERNEL_CKT_UPDATE=0 + + # Get package revision + DEB_PACKAGE_REVISION=$(echo ${DEB_PACKAGE_VERSION} | sed -r 's/.*-([^-]+)$/\1/') + # Get non-sec update number + DEB_PACKAGE_REVISION_BASE=$(echo ${DEB_PACKAGE_REVISION} | sed -r 's/^([0-9]+).*/\1/') + # Get security update number, if present +-DEB_PACKAGE_REVISION_SECURITY=$(echo ${DEB_PACKAGE_REVISION} | sed -rn 's/.*\+(squeeze|deb[0-9])+u([0-9]+)$/\1/p') ++DEB_PACKAGE_REVISION_SECURITY=$(echo ${DEB_PACKAGE_REVISION} | sed -rn 's/.*\+(squeeze|deb[0-9]+)+u([0-9]+)$/\2/p') + test -n "${DEB_PACKAGE_REVISION_SECURITY}" || DEB_PACKAGE_REVISION_SECURITY=0 + # Combine all update numbers into one + DEB_API_VERSION=$((KERNEL_CKT_UPDATE * 1 + DEB_PACKAGE_REVISION_BASE * 100 + DEB_PACKAGE_REVISION_SECURITY)) diff -Nru lttng-modules-2.9
Bug#897483: ust: FTBFS: LTTngUst.c:20:10: fatal error: org_lttng_ust_LTTngUst.h: No such file or directory
On 2018-05-02 04:07 PM, Lucas Nussbaum wrote: > Source: ust > Version: 2.10.1-2 > Severity: serious > Tags: buster sid > User: debian...@lists.debian.org > Usertags: qa-ftbfs-20180502 qa-ftbfs > Justification: FTBFS on amd64 > > Hi, > > During a rebuild of all packages in sid, your package failed to build on > amd64. > > Relevant part (hopefully): >> /bin/bash ../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -I.. >> -I../include/lttng -I../include -I../include >> -I/usr/lib/jvm/default-java/include >> -I/usr/lib/jvm/default-java/include/linux -Wdate-time -D_FORTIFY_SOURCE=2 >> -Wall -g -O2 -fdebug-prefix-map=/<>=. -fstack-protector-strong >> -Wformat -Werror=format-security -c -o LTTngUst.lo LTTngUst.c >> libtool: compile: gcc -DHAVE_CONFIG_H -I. -I.. -I../include/lttng >> -I../include -I../include -I/usr/lib/jvm/default-java/include >> -I/usr/lib/jvm/default-java/include/linux -Wdate-time -D_FORTIFY_SOURCE=2 >> -Wall -g -O2 -fdebug-prefix-map=/<>=. -fstack-protector-strong >> -Wformat -Werror=format-security -c LTTngUst.c -fPIC -DPIC -o >> .libs/LTTngUst.o >> LTTngUst.c:20:10: fatal error: org_lttng_ust_LTTngUst.h: No such file or >> directory >> #include "org_lttng_ust_LTTngUst.h" >> ^~ >> compilation terminated. >> make[4]: *** [Makefile:501: LTTngUst.lo] Error 1 > > The full build log is available from: >http://aws-logs.debian.net/2018/05/02/ust_2.10.1-2_unstable.log > > A list of current common problems and possible solutions is available at > http://wiki.debian.org/qa.debian.org/FTBFS . You're welcome to contribute! > > About the archive rebuild: The rebuild was done on EC2 VM instances from > Amazon Web Services, using a clean, minimal and up-to-date chroot. Every > failed build was retried once to eliminate random failures. > Hi, That's surprising, I did a source upload a week or so ago that built fine on the buildds. Looking at both build logs the difference seems to be that the default jdk changed from 9 to 10. I'll try to reproduce the failure on my system and fix it. Thanks for the heads up, Michael
Bug#894824: popt FTCBFS: api-sanity-checker dependency unsatisfiable
On 2018-04-04 11:24, Helmut Grohne wrote: > Source: popt > Version: 1.16-11 > Tags: patch > User: helm...@debian.org > Usertags: rebootstrap > > popt fails to cross build from source, because its build dependency on > api-sanity-checker is unsatisfiable. In general, Architecture: all > packages can never satisfy cross Build-Depends unless marked Multi-Arch: > foreign. In this case however, api-sanity-checker is entirely > unnecessary, because it is only used for testing and we cannot test > during cross builds anyway. Thus annotating the dependency with > is sufficient to get satisfiable cross Build-Depends again. I > used reproducible builds to verify that this change does not affect the > resulting binary packages. Please consider applying the attached patch. > > Helmut > Hi, Thanks for the patch, I've applied it in git, I'll wait on some other changes to upload unless this is blocking something on your side? Regards, Michael
Bug#894600: libpopt0: package priority change?
On 2018-04-02 02:34, Daniel Vacek wrote: > Package: libpopt0 > Version: 1.16-11 > Severity: normal > > Hi, I noticed package priority changed from 'important' to 'optional' without > mentioning this in changelog. I was wondering if this is intended or by a > mistake? > > --nX > Hi, The priority change should of been in the changelog, my mistake. This was done on the recommendation of lintian that libraries should not have a priority higher than optional because they will always be pulled in by dependencies. Regards, Michael
Bug#889901: lttng-modules-dkms-2.9.0-1 fails to build with latest kernel (4.9.0-5-amd64) in debian 9.3
On 2018-02-08 10:27, srikanth krishnakar wrote: > > Do we have a fix or workaround available for this failure ? > This is fixed in the latest upstream stable-2.9 update: http://lttng.org/files/lttng-modules/lttng-modules-2.9.8.tar.bz2 You might want to build from source, it will probably take a while for the updated package to land in stretch. Cheers, Michael
Bug#885556: multipath-udeb: depends on a non-udeb package: liburcu6
On 2018-01-14 22:02, Cyril Brulebois wrote: > clone 885556 -1 reassign -1 src:liburcu 0.10.0-2 retitle -1 please > provide a udeb severity -1 important block 885556 by -1 thanks > > > Hi, > > Cyril Brulebois(2017-12-28): >> [Please keep debian-boot@ in copy of your replies.] >> >> Hi, >> >> Your udeb package depends on a non-udeb package (liburcu6), >> making it uninstallable on all architectures. > > So, looking at the source, it seems liburcu really isn't an option, > as it's hardcoded in a few makefiles, etc. I've drafted a patch to > add a udeb to src:liburcu, which you'll find attached. > > Let's see if I got the BTS dance right. :) > > > Cheers, Hi, I've uploaded 0.10.0-3 with the included patch and some other minor packaging fixes, it's sitting in the NEW queue because of the added udeb. Regards, Michael
Bug#882434: stretch-pu: package ust/2.9.0-2+deb9u1
Package: release.debian.org Severity: normal Tags: stretch User: release.debian@packages.debian.org Usertags: pu Hi, The attached diff fixes a bug that makes the python3-lttngust package completely broken unless the corresponding liblttng-ust-dev is also installed. The original python code load the library using ctypes without specifying a soname. This fix was reported and merged upstream. Fixed in unstable: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=882366 Regards, Michael -- System Information: Debian Release: 9.1 APT prefers stable APT policy: (500, 'stable') Architecture: i386 (i686) Kernel: Linux 4.9.0-4-686-pae (SMP w/2 CPU cores) Locale: LANG=en_CA.UTF-8, LC_CTYPE=en_CA.UTF-8 (charmap=UTF-8), LANGUAGE=en_CA:en (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) diff -Nru ust-2.9.0/debian/changelog ust-2.9.0/debian/changelog --- ust-2.9.0/debian/changelog 2017-03-08 12:04:25.0 -0500 +++ ust-2.9.0/debian/changelog 2017-11-22 14:45:44.0 -0500 @@ -1,3 +1,10 @@ +ust (2.9.0-2+deb9u1) stable; urgency=medium + + * [5ffa17d] Set gbp branch config + * [8e770e4] Fix python3-lttngust load un-versioned library (Closes: #882366) + + -- Michael Jeanson <mjean...@ubuntu.com> Wed, 22 Nov 2017 14:45:44 -0500 + ust (2.9.0-2) unstable; urgency=medium * [b8d4e77] Add missing liblttng-ust-fd.so.* (Closes: #857166) diff -Nru ust-2.9.0/debian/gbp.conf ust-2.9.0/debian/gbp.conf --- ust-2.9.0/debian/gbp.conf 1969-12-31 19:00:00.0 -0500 +++ ust-2.9.0/debian/gbp.conf 2017-11-22 14:44:31.0 -0500 @@ -0,0 +1,3 @@ +[DEFAULT] +upstream-branch=upstream/2.9.0 +debian-branch=debian/stretch diff -Nru ust-2.9.0/debian/patches/fix-specify-soname-in-python-lttngust-loadlibrary.patch ust-2.9.0/debian/patches/fix-specify-soname-in-python-lttngust-loadlibrary.patch --- ust-2.9.0/debian/patches/fix-specify-soname-in-python-lttngust-loadlibrary.patch 1969-12-31 19:00:00.0 -0500 +++ ust-2.9.0/debian/patches/fix-specify-soname-in-python-lttngust-loadlibrary.patch 2017-11-22 14:45:15.0 -0500 @@ -0,0 +1,30 @@ +From 00ee1adfe1e34d43494227781f6662b0a21b7c4b Mon Sep 17 00:00:00 2001 +From: Michael Jeanson <mjean...@efficios.com> +Date: Tue, 21 Nov 2017 11:11:15 -0500 +Subject: [PATCH] Fix: specify SONAME in python-lttngust LoadLibrary + +When loading the python agent library with ctypes in the python +bindings, specify the SONAME. This will make sure we load the proper +library in the event of a SONAME bump and the bindings will work without +having to install the "dev" package which in most distros contains the +non-versionned ".so". + +Signed-off-by: Michael Jeanson <mjean...@efficios.com> +Signed-off-by: Mathieu Desnoyers <mathieu.desnoy...@efficios.com> +--- + python-lttngust/lttngust/loghandler.py | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +diff --git a/python-lttngust/lttngust/loghandler.py b/python-lttngust/lttngust/loghandler.py +index e82cf5c5..6f144cac 100644 +--- a/python-lttngust/lttngust/loghandler.py b/python-lttngust/lttngust/loghandler.py +@@ -22,7 +22,7 @@ + + + class _Handler(logging.Handler): +-_LIB_NAME = 'liblttng-ust-python-agent.so' ++_LIB_NAME = 'liblttng-ust-python-agent.so.0' + + def __init__(self): + super(self.__class__, self).__init__(level=logging.NOTSET) diff -Nru ust-2.9.0/debian/patches/series ust-2.9.0/debian/patches/series --- ust-2.9.0/debian/patches/series 2016-11-29 18:21:51.0 -0500 +++ ust-2.9.0/debian/patches/series 2017-11-22 14:45:15.0 -0500 @@ -1,3 +1,4 @@ fix-incompatible-java-bytecode-format.patch use-python3.patch javah-doesnt-generate-class-files.patch +fix-specify-soname-in-python-lttngust-loadlibrary.patch
Bug#882366: python3-lttngust tries to load unversionned agent library
Package: python3-lttngust Version: 2.10.0-3 Severity: important The python3-lttngust package contains bindings that use ctypes to interact with a dedicated library provided by the liblttng-ust-python-agent0. The python code loads the library by name but it targets the un-versioned ".so" which is part of the dev package. This means that when installed without the dev package the bindings won't work. -- System Information: Debian Release: buster/sid APT prefers testing APT policy: (500, 'testing') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.13.0-1-amd64 (SMP w/2 CPU cores) Locale: LANG=en_CA.UTF-8, LC_CTYPE=en_CA.UTF-8 (charmap=UTF-8), LANGUAGE=en_CA:en (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages python3-lttngust depends on: ii liblttng-ust-python-agent0 2.10.0-3 ii python3 3.6.3-2 python3-lttngust recommends no packages. python3-lttngust suggests no packages. -- no debconf information
Bug#878816: stretch-pu: package lttng-modules/2.9.0-1+deb9u1
On 19/11/17 09:50, Julien Cristau wrote: > On Mon, Oct 30, 2017 at 11:31:39 -0400, Michael Jeanson wrote: > >> So, I'll tag bug #864404 as stretch and add the fix to my next unstable >> upload without a close statement. >> > Both of those sound like the wrong thing to do. Fair enough, but I'm more interested by what would be the right thing to do?
Bug#878816: stretch-pu: package lttng-modules/2.9.0-1+deb9u1
On 2017-10-29 14:11, Adam D. Barratt wrote: > Control: tags -1 + moreinfo > > On Mon, 2017-10-16 at 16:22 -0400, Michael Jeanson wrote: >> The attached diff fixes a build failure of the dkms modules on the >> linux-rt flavor of the debian linux kernel. This was reported at: >> >> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=864404 > > The metadata for that bug suggests that it affects the package in > unstable and is not yet fixed there - is that correct? If so, please > get the fix applied in unstable first; otherwise, please fix the > metadata to include an appropriate "fixed" version. > > Regards, > > Adam > There is currently no linux-image-rt in testing/unstable and this fix is only relevant for linux rt 4.9, so I didn't think it was necessary. But I guess it could be useful if someone want to backport lttng 2.10 to stretch with the 4.9 rt kernel. So, I'll tag bug #864404 as stretch and add the fix to my next unstable upload without a close statement. Regards, Michael
Bug#879088: babeltrace: package merge makes gdb, linux-perf etc. uninstallable
On 2017-10-19 04:18, Adrian Bunk wrote: > Control: retitle -1 Please make libbabeltrace-ctf{1,-dev} provides versioned > > On Thu, Oct 19, 2017 at 08:28:46AM +0100, Simon McVittie wrote: >> ... >> Please do one of these as soon as possible: >> ... > > None of these is required. > > Making the libbabeltrace-ctf{1,-dev} provides versioned is all that is > required to fix the versioned (build) dependencies. > > cu > Adrian > Sorry for that mess, I will upload a fixed package with the versioned provides. Thanks for pointing me in the right direction. Cheers, Michael
Bug#878816: stretch-pu: package lttng-modules/2.9.0-1+deb9u1
Package: release.debian.org Severity: normal Tags: stretch User: release.debian@packages.debian.org Usertags: pu Hi, The attached diff fixes a build failure of the dkms modules on the linux-rt flavor of the debian linux kernel. This was reported at: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=864404 I'd appreciate if you consider allowing this upload to stretch. Thanks, Michael -- System Information: Debian Release: 9.1 APT prefers stable APT policy: (500, 'stable') Architecture: i386 (i686) Kernel: Linux 4.9.0-4-686-pae (SMP w/2 CPU cores) Locale: LANG=en_CA.UTF-8, LC_CTYPE=en_CA.UTF-8 (charmap=UTF-8), LANGUAGE=en_CA:en (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) diff --git a/debian/changelog b/debian/changelog index f458a42..0dfc818 100644 --- a/debian/changelog +++ b/debian/changelog @@ -1,3 +1,10 @@ +lttng-modules (2.9.0-1+deb9u1) stable; urgency=medium + + * [c3d8eab] Stretch gbp branch config + * [ee40323] Fix build on linux-rt 4.9 kernels. (Closes: #864404) + + -- Michael Jeanson <mjean...@ubuntu.com> Mon, 16 Oct 2017 15:25:06 -0400 + lttng-modules (2.9.0-1) unstable; urgency=medium * [500ac85] New upstream version 2.9.0 diff --git a/debian/gbp.conf b/debian/gbp.conf new file mode 100644 index 000..b174468 --- /dev/null +++ b/debian/gbp.conf @@ -0,0 +1,3 @@ +[DEFAULT] +upstream-branch=upstream/2.9.0 +debian-branch=debian/stretch diff --git a/debian/patches/fix-linux-rt-4.9-sched.patch b/debian/patches/fix-linux-rt-4.9-sched.patch new file mode 100644 index 000..63b5303 --- /dev/null +++ b/debian/patches/fix-linux-rt-4.9-sched.patch @@ -0,0 +1,31 @@ +Fix build of sched instrumentation on linux-rt >= 4.9.27-rt8 +--- a/instrumentation/events/lttng-module/sched.h b/instrumentation/events/lttng-module/sched.h +@@ -540,7 +540,26 @@ + ) + #endif + +-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,37)) ++#if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,12,0) || \ ++ (LTTNG_KERNEL_RANGE(4,9,27, 4,10,0) && defined(CONFIG_PREEMPT_RT_FULL))) ++/* ++ * Tracepoint for showing priority inheritance modifying a tasks ++ * priority. ++ */ ++LTTNG_TRACEPOINT_EVENT(sched_pi_setprio, ++ ++ TP_PROTO(struct task_struct *tsk, struct task_struct *pi_task), ++ ++ TP_ARGS(tsk, pi_task), ++ ++ TP_FIELDS( ++ ctf_array_text(char, comm, tsk->comm, TASK_COMM_LEN) ++ ctf_integer(pid_t, tid, tsk->pid) ++ ctf_integer(int, oldprio, tsk->prio - MAX_RT_PRIO) ++ ctf_integer(int, newprio, pi_task ? pi_task->prio - MAX_RT_PRIO : tsk->prio - MAX_RT_PRIO) ++ ) ++) ++#elif (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,37)) + /* + * Tracepoint for showing priority inheritance modifying a tasks + * priority. diff --git a/debian/patches/series b/debian/patches/series new file mode 100644 index 000..ada1602 --- /dev/null +++ b/debian/patches/series @@ -0,0 +1 @@ +fix-linux-rt-4.9-sched.patch
Bug#876148: ust: FTBFS w/GCJ (on hppa)
On 2017-09-20 21:52, Aaron M. Ucko wrote: > Michael Jeanson <mjean...@efficios.com> writes: > >> The ust java bindings won't build with gcj, however they are optional >> and not required for a normal use of ust. I'd like to just disable them >> on platforms that don't have an openjdk port, I added a 'nojava' [1] >> build profile [2] to the packaging code that could be used for that but >> I haven't found a way to have it auto-enabled for hppa builds. > > Sounds good! AFAICT, it *technically* works to add a conditional > > export DEB_BUILD_PROFILES += nojava That's what I initially tried but at the end of the build, dh_install would still try to grab the files for the java packages which should of been disabled. > > near the top of debian/rules. However, that's kind of a hack, so I > suppose it would be better to specify !hppa alongside in > debian/control[1] and have debian/rules consult a private variable > conditionalized on both the architecture and the profile: > > ifeq ($(filter nojava,$(DEB_BUILD_PROFILES))$(filter > hppa,$(DEB_HOST_ARCH)),) > java_ok = 1 > endif > > ifeq ($(java_ok),1) > # ... > endif > > Thanks! The problem with this approach is that the java packages are arch 'all' so we can't exclude an architecture, plus I'm pretty sure negations are not valid in the 'Architecture' field. > > [1] [!hppa] in Build-Depends, Architecture: !hppa > for the relevant binary packages. >
Bug#876148: ust: FTBFS w/GCJ (on hppa)
On 2017-09-18 21:54, Aaron M. Ucko wrote: > Source: ust > Version: 2.9.1-1 > Severity: important > Tags: upstream > Justification: fails to build from source > User: debian-h...@lists.debian.org > > Builds of ust with GCJ (to which default-jdk necessarily still boils > down on hppa, admittedly not a release architecture) have been failing: > > CLASSPATH=.:./.${CLASSPATH:+":$CLASSPATH"} gcj -C -d . -source 1.6 -target > 1.6 org/lttng/ust/LTTngUst.java > gcj: error: 1.6: No such file or directory > gcj: error: 1.6: No such file or directory > gcj: error: unrecognized command line option '-source'; did you mean > '-fsource='? > gcj: error: unrecognized command line option '-target'; did you mean > '-ftarget='? > Makefile:510: recipe for target 'classnoinst.stamp' failed > make[4]: *** [classnoinst.stamp] Error 1 > > Could you please take a look? If building with GCJ is infeasible, > please specifically require OpenJDK so that autobuilders for GCJ-only > architectures (i.e., hppa) don't bother trying to cover ust. > > Thanks! > Hi, The ust java bindings won't build with gcj, however they are optional and not required for a normal use of ust. I'd like to just disable them on platforms that don't have an openjdk port, I added a 'nojava' [1] build profile [2] to the packaging code that could be used for that but I haven't found a way to have it auto-enabled for hppa builds. Michael [1] https://github.com/debian-lttng/lttng-ust/commit/b72deebefe21f4b8f884502cc72e18752bb66ba6 [2] https://wiki.debian.org/BuildProfileSpec
Bug#875837: fuji: Missing fuji(8) man page
Package: fuji Version: 1.0.2-1+b2 Severity: normal Tags: patch Dear Maintainer, The usr/sbin/fuji binary has no corresponding man page, I wrote one and attached it to this message. Feel free to add more information to it. Cheers, Michael -- System Information: Debian Release: buster/sid APT prefers testing APT policy: (500, 'testing') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.12.0-1-amd64 (SMP w/2 CPU cores) Locale: LANG=en_CA.UTF-8, LC_CTYPE=en_CA.UTF-8 (charmap=UTF-8), LANGUAGE=en_CA:en (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages fuji depends on: ii init-system-helpers 1.49 ii libc62.24-17 fuji recommends no packages. fuji suggests no packages. -- no debconf information .TH "FUJI" 8 "2017-09-14" "Fuji MQTT Gateway" "fuji" .SH NAME fuji \- Fuji Message Queue Telemetry Transport Gateway .SH SYNOPSIS .PP \fBfuji\fR [ \fB\-c\fR \fIconfig\fR ] [ \fB\-d\fR ] [ \fB\-v\fR ] .SH DESCRIPTION Fuji is an MQTT gateway daemon written in Go\&. It is designed to be run as a standalone daemon process\&. .PP It will read sensor data from serial ports and send them to an MQTT broker\&. .SH OPTIONS .TP \-c \fIconfig\fR Uses the directives in the file \fIconfig\fR on startup\&. When omitted the default \fI/etc/fuji\-gw/config.toml\fR is used\&. .TP \-d Run fuji in debug mode\&. .TP \-v Print the version of fuji, and then exit\&. .SH AUTHOR This manual page was written by Michael Jeanson <mjean...@ubuntu.com> for the Debian project (but may be used by others). .SH COPYRIGHT Copyright \(co 2015\-2016 Shiguredo Inc. <f...@shiguredo.jp> Licensed under the Apache License, Version 2.0, see <\fBfile:///usr/share/common\-licenses/Apache\-2.0\fR>\&.
Bug#867465: ceph-base: purging ceph-base deletes /etc/ceph/rbdmap owned by ceph-common
Hi, Here is a simple fix for this bug, both ceph-common and ceph-base delete 'etc/ceph' on purge in their postrm script but only ceph-common owns files in this directory. Just remove the cleanup code from the ceph-base postrm script. Cheers, Michael >From bb8312996cf030e0846f458009558b42145a6061 Mon Sep 17 00:00:00 2001 From: Michael Jeanson <mjean...@efficios.com> Date: Thu, 14 Sep 2017 15:28:51 -0400 Subject: [PATCH] Don't clean dirs on postrm in ceph-base (Closes: #867465) Cleanup of the /etc/ceph directory on purge is already handled by ceph-common which owns files in the directory. --- debian/ceph-base.postrm | 2 -- 1 file changed, 2 deletions(-) diff --git a/debian/ceph-base.postrm b/debian/ceph-base.postrm index 48e4853c0..31272029c 100644 --- a/debian/ceph-base.postrm +++ b/debian/ceph-base.postrm @@ -24,8 +24,6 @@ case "$1" in ;; purge) - rm -rf /var/log/ceph - rm -rf /etc/ceph ;; upgrade|failed-upgrade|abort-install|abort-upgrade|disappear) -- 2.14.1
Bug#864404: Info received ()
On 2017-06-08 05:37, Gavin Lambert wrote: > So, I've checked upstream, and it looks like this is *partially* fixed > there, by these two commits: > > https://bugs.lttng.org/projects/lttng-modules/repository/revisions/673e9a0300912e68d516cf06e0553c11b28cddbf/diff/instrumentation/events/lttng-module/sched.h > > > https://bugs.lttng.org/projects/lttng-modules/repository/revisions/262ab18947d8568ef6357d5bfd84d11537cb4b9f/diff/instrumentation/events/lttng-module/sched.h > > > Presumably these changes are in the version in sid, although I haven't > verified that. > > However by themselves I don't think this will solve the problem for > stretch, due to the missing localversion-rt file causing the check for > LTTNG_RT_KERNEL_RANGE to fail even though the 4.9-rt kernel is intended > to pass. > > Maybe this is actually a bug in the kernel package now? Or should this > have a different way of detecting the Debian -rt kernel source? > Hi, I introduced the rt kernel detection code in lttng-modules based on the 'localversion-rt' file because I couldn’t find another way to identify those kernels, unfortunately this file is present in the rt kernel source tree but is not installed in the kernel headers package. You could ask the maintainers of the linux-rt package if they would be willing to add it to their headers package. Ideally the rt kernel developers would add rt specific version defines in the kernel headers like most other custom kernels do. Michael
Bug#860154: liburcu: Please add architecture support for m68k
On 2017-04-12 05:01, John Paul Adrian Glaubitz wrote: > Hi! > > liburcu is currently missing architecture support for m68k > and therefore fails to build from source [1]. > > The attached patch adds minimal support for m68k, analogue > to what was done for sh4. > > Thanks, > Adrian Hi, m68k support is something I have looked into before but the lack of available hardware to do basic testing on has prevented me from making any progress. If it's something you can arrange I'd be interested. I'm very hesitant on merging this patch without having at least run the full regression test suite on real hardware. Regards, Michael
Bug#854778: liburcu 32 bit and 64 bit versions cannot install simultaneously
On 2017-02-28 09:20, Matthew Wilcox wrote: > I think you may need to rebuild them differently from this in order to > have it work. > > I just installed zlib1g-dev:i386 and :amd64 to have an example to look > at. Each of these packages contains the same files under /usr/share and > /usr/include, and dpkg happily installed them both without special flags > on my part. > > Try “apt-get source liburcu-dev”, and edit the debian/control file’s > Multi-Arch line for liburcu-dev. Then rebuild it and see if it behaves > differently when you try to install them both at the same time. I've just pushed 0.9.3-2 to experimental with multiarch enabled on the dev packages. I had to move the includes to the platform specific directory since they differ depending on the plaform. I've tested multiarch installs and I've rebuilt packages that depend on liburcu successfully, I would appreciate if you could test the packages and report back here. They should show up in the archive soon. Cheers, Michael
Bug#854778: liburcu 32 bit and 64 bit versions cannot install simultaneously
On 2017-02-10 04:12, Rehas Sachdeva wrote: > Package: liburcu-dev > Version: 0.9.1-3 > > I followed the instructions here, > https://wiki.debian.org/Multiarch/Implementation > to add i386 architecture. > > I am able to build 64-bit binaries but I am not able to force 32-bit > build using -m32 flag, as mentioned here http://liburcu.org/. Below is > the error message: > > /usr/bin/ld: skipping incompatible //usr/local/lib/liburcu.so when > searching for -lurcu > /usr/bin/ld: skipping incompatible //usr/local/lib/liburcu.a when > searching for -lurcu > /usr/bin/ld: cannot find -lurcu > collect2: error: ld returned 1 exit status > > I am using GNU/Linux 4.4.0-62-generic, on x86_64 machine. > > Thanks. Hi, I'm not sure exactly what you are trying to achieve but the files you refer to in "/usr/local" are not part of the liburcu debian package and must be artifacts of a local build and install. I'll need more details on what you are trying to do and what steps you took to get there. Regards, Michael
Bug#841886: Please upgrade babeltrace to the 1.5 series for Stretch
I've just pushed 1.5.0~rc1-2 with the proper include files in libbabeltrace-ctf-dev. I'll keep this bug open until 1.5.0 final is uploaded. Cheers, Michael On 2016-10-24 15:53, Sebastian Andrzej Siewior wrote: > So I tried to build the perf ctf I noticed that I need > > diff --git a/debian/libbabeltrace-ctf-dev.install > b/debian/libbabeltrace-ctf-dev.install > --- a/debian/libbabeltrace-ctf-dev.install > +++ b/debian/libbabeltrace-ctf-dev.install > @@ -1,4 +1,5 @@ > usr/include/babeltrace/ctf/* > +usr/include/babeltrace/ctf-ir/* > usr/lib/*/libbabeltrace-ctf.a > usr/lib/*/libbabeltrace-ctf.so > usr/lib/*/libbabeltrace-ctf-text.a > > to get it compiled & linked. The resulting binary was able to covert a > stream so it looks good. >
Bug#841886: Please upgrade babeltrace to the 1.5 series for Stretch
After discussing it with Jérémie the rc1 was officially released, I've just pushed it to Debian, should be available in unstable tomorrow. Cheers, Michael On 2016-10-24 11:06, Michael Jeanson wrote: > Hi, > > I'm not too keen on packaging something that is not even in RC yet, I've > taken a look at the state of this branch and it currently only adds > symbols to libabeltrace1 which by my understanding would not require a > transition. So I guess we don't have to worry about the November fifth > freeze. > > That being said, I would really like to get this in Debian as soon as > possible, I'll ask the maintainer what is the ETA for the release. > > Michael > > On 2016-10-24 03:21, Sebastian Andrzej Siewior wrote: >> Package: babeltrace >> Version: 1.4.0-3 >> Severity: wishlist >> >> Could you please bump babeltrace to the 1.5 series which is prepared in >> https://github.com/jgalar/babeltrace/tree/stable-1.5-staging >> >> This is upstream kind answer to what is needed to get the `perf' command >> (from the linux-perf package) linked against libbabeltrace and so >> support a perf to CTF conversion. >> >> If you need any support on packaging etc. please let me know, I am glad >> to help to get this done before the transition freeze on 5th November. >> >> Sebastian >> >
Bug#841886: Please upgrade babeltrace to the 1.5 series for Stretch
Hi, I'm not too keen on packaging something that is not even in RC yet, I've taken a look at the state of this branch and it currently only adds symbols to libabeltrace1 which by my understanding would not require a transition. So I guess we don't have to worry about the November fifth freeze. That being said, I would really like to get this in Debian as soon as possible, I'll ask the maintainer what is the ETA for the release. Michael On 2016-10-24 03:21, Sebastian Andrzej Siewior wrote: > Package: babeltrace > Version: 1.4.0-3 > Severity: wishlist > > Could you please bump babeltrace to the 1.5 series which is prepared in > https://github.com/jgalar/babeltrace/tree/stable-1.5-staging > > This is upstream kind answer to what is needed to get the `perf' command > (from the linux-perf package) linked against libbabeltrace and so > support a perf to CTF conversion. > > If you need any support on packaging etc. please let me know, I am glad > to help to get this done before the transition freeze on 5th November. > > Sebastian >
Bug#839684: RM: lttngtop -- ROM; broken in unstable, abandoned upstream
Package: ftp.debian.org Severity: normal Hi, lttngtop has been broken for a while in unstable because of updates to he lttng stack, upstream has officialy abandoned this project and no fixes are expected. Thanks, Michael Jeanson
Bug#821795: lttng-modules-dkms: module FTBFS in jessie against Linux 3.16: error: conflicting types for 'trace_mm_page_alloc_extfrag'
- On Jul 12, 2016, at 5:17 PM, Jon Bernard jbern...@debian.org wrote: > * Michael Jeanson <mjean...@ubuntu.com> wrote: >> Hi Jon, >> >> Here is an updated debdiff, it works on a fully updated jessie. Should >> we track these updates in a branch of the collab-maint repo? > > Excellent, thanks. This patch works and I've put it into a dedicated > stable branch 'jessie' so this will be easier to track. > > In looking at the stable update procedure, I need to request permission > to submit an update and I wonder: would it be more valuable to update > our Jessie version (2.5.1) to the latest supported stable version > upstream (2.5.6) and rebase these patches on that version? Give that > upstream supports a stable branch (which I think is unusual, but very > awesome), this looks like a compelling approach. > > Thoughts? If minor version bumps are allowed in Debian, then it would make a lot more sense to update to the latest 2.5.x bugfix release. Michael
Bug#821795: lttng-modules-dkms: module FTBFS in jessie against Linux 3.16: error: conflicting types for 'trace_mm_page_alloc_extfrag'
Hi Jon, Here is an updated debdiff, it works on a fully updated jessie. Should we track these updates in a branch of the collab-maint repo? Michael diff -Nru lttng-modules-2.5.1/debian/changelog lttng-modules-2.5.1/debian/changelog --- lttng-modules-2.5.1/debian/changelog2014-10-24 11:11:35.0 -0400 +++ lttng-modules-2.5.1/debian/changelog2016-07-04 04:49:26.0 -0400 @@ -1,3 +1,9 @@ +lttng-modules (2.5.1-2~test2) unstable; urgency=medium + + * Add patches + + -- Michael Jeanson <mjean...@ubuntu.com> Mon, 04 Jul 2016 17:24:58 -0400 + lttng-modules (2.5.1-1) unstable; urgency=medium * [106d0a8] New upstream version 2.5.1 diff -Nru lttng-modules-2.5.1/debian/lttng-modules-dkms.dkms.in lttng-modules-2.5.1/debian/lttng-modules-dkms.dkms.in --- lttng-modules-2.5.1/debian/lttng-modules-dkms.dkms.in 2014-10-24 11:11:35.0 -0400 +++ lttng-modules-2.5.1/debian/lttng-modules-dkms.dkms.in 2016-07-04 04:47:00.0 -0400 @@ -10,7 +10,7 @@ KCONFIG=true fi -MAKE[$i]="make -C $kernel_source_dir M=$dkms_tree/$PACKAGE_NAME/$PACKAGE_VERSION/build modules" +MAKE[$i]="make -C $kernel_source_dir M=$dkms_tree/$PACKAGE_NAME/$PACKAGE_VERSION/build modules V=1" BUILT_MODULE_NAME[$i]="lttng-lib-ring-buffer" BUILT_MODULE_LOCATION[$i]="lib/" @@ -158,13 +158,6 @@ DEST_MODULE_LOCATION[$i]="/extra/probes" i=$((i+1)) -if [ "$KCONFIG" = "true" ] && [ -n "$CONFIG_REGMAP" ]; then -BUILT_MODULE_NAME[$i]="lttng-probe-regmap" -BUILT_MODULE_LOCATION[$i]="probes/" -DEST_MODULE_LOCATION[$i]="/extra/probes" -i=$((i+1)) -fi - if [ "$KCONFIG" = "true" ] && [ -n "$CONFIG_REGULATOR" ]; then BUILT_MODULE_NAME[$i]="lttng-probe-regulator" BUILT_MODULE_LOCATION[$i]="probes/" diff -Nru lttng-modules-2.5.1/debian/patches/0001-Fix-discover-Debian-API.patch lttng-modules-2.5.1/debian/patches/0001-Fix-discover-Debian-API.patch --- lttng-modules-2.5.1/debian/patches/0001-Fix-discover-Debian-API.patch 1969-12-31 19:00:00.0 -0500 +++ lttng-modules-2.5.1/debian/patches/0001-Fix-discover-Debian-API.patch 2016-07-04 04:41:24.0 -0400 @@ -0,0 +1,257 @@ +From 0aadd8424c7c21259f80da4c97a2b9fef92bd2e9 Mon Sep 17 00:00:00 2001 +From: Mathieu Desnoyers <mathieu.desnoy...@efficios.com> +Date: Mon, 27 Apr 2015 10:57:34 -0400 +Subject: [PATCH 1/2] Fix: discover Debian API + +Signed-off-by: Mathieu Desnoyers <mathieu.desnoy...@efficios.com> +--- + Makefile | 12 ++-- + Makefile.ABI.workarounds | 11 + abi-debian-version.sh | 30 ++ + instrumentation/events/lttng-module/kmem.h | 91 +- + lib/Makefile | 4 ++ + lttng-kernel-version.h | 14 + + probes/Makefile| 4 ++ + 7 files changed, 161 insertions(+), 5 deletions(-) + create mode 100644 Makefile.ABI.workarounds + create mode 100755 abi-debian-version.sh + +--- a/Makefile b/Makefile +@@ -5,6 +5,8 @@ + ifneq ($(KERNELRELEASE),) + ifneq ($(CONFIG_TRACEPOINTS),) + ++KERNELDIR=${LTTNG_KERNELDIR} ++ + lttng_check_linux_version = $(shell pwd)/include/linux/version.h + lttng_check_generated_linux_version = $(shell pwd)/include/generated/uapi/linux/version.h + +@@ -19,6 +21,8 @@ + endif + endif + ++include $(KBUILD_EXTMOD)/Makefile.ABI.workarounds ++ + obj-m += lttng-ring-buffer-client-discard.o + obj-m += lttng-ring-buffer-client-overwrite.o + obj-m += lttng-ring-buffer-metadata-client.o +@@ -67,14 +71,14 @@ + CFLAGS = $(EXTCFLAGS) + + default: +- $(MAKE) -C $(KERNELDIR) M=$(PWD) modules ++ LTTNG_KERNELDIR=$(KERNELDIR) $(MAKE) -C $(KERNELDIR) M=$(PWD) modules + + modules_install: +- $(MAKE) -C $(KERNELDIR) M=$(PWD) modules_install ++ LTTNG_KERNELDIR=$(KERNELDIR) $(MAKE) -C $(KERNELDIR) M=$(PWD) modules_install + + clean: +- $(MAKE) -C $(KERNELDIR) M=$(PWD) clean ++ LTTNG_KERNELDIR=$(KERNELDIR) $(MAKE) -C $(KERNELDIR) M=$(PWD) clean + + %.i: %.c +- $(MAKE) -C $(KERNELDIR) M=$(PWD) $@ ++ LTTNG_KERNELDIR=$(KERNELDIR) $(MAKE) -C $(KERNELDIR) M=$(PWD) $@ + endif # KERNELRELEASE +--- /dev/null b/Makefile.ABI.workarounds +@@ -0,0 +1,11 @@ ++# Work-around for distro-specific public modules ABI breakages. ++# Some distributions break the public module instrumentation ABI ++# compared to upstream stable kernels without providing other mean than ++# the kernel EXTRAVERSION to figure it out. Translate this information ++# into a define visible from the C preprocessor. ++ ++DEB_API_VERSION=$(shell $(KBUILD_EXTMOD)/abi-debian-version.sh $(CURDIR)) ++ ++ifneq ($(DEB_API_VERSION), 0) ++ccflags-y += -DDEBIAN_API_VERSION=$(DEB_API_VERSION) ++endif +--- /dev/null
Bug#821795: lttng-modules-dkms: module FTBFS in jessie against Linux 3.16: error: conflicting types for 'trace_mm_page_alloc_extfrag'
- On Jul 1, 2016, at 10:29 PM, Jon Bernard jbern...@debian.org wrote: > Hey Michael, > > I just attempted a build with your diff and I'm seeing the attached > errors. It looks like the debian-specific ifdefs are not quite right, > does any of this look familiar? I'll look at it closer over the holiday > if you don't get to it sooner. Looks like the jessie kernel was updated since I made this debdiff, I'll look into making a new patch, it should be pretty straightforward. Michael
Bug#821795: lttng-modules-dkms: module FTBFS in jessie against Linux 3.16: error: conflicting types for 'trace_mm_page_alloc_extfrag'
Hi, Here is a debdiff against the current package fixing the build failure, I'll ask Jon to make a new upload. Michael lttng-modules-2.5.1-fix-build.debdiff Description: Binary data
Bug#819654: ITP: barectf -- A code generator to write native CTF binary streams
Package: wnpp Severity: wishlist Owner: Michael Jeanson <mjean...@ubuntu.com> * Package name: barectf Version : 2.1.4 Upstream Author : Philippe Proulx <ppro...@efficios.com> * URL : http://barectf.org/ * License : MIT Programming Lang: Python Description : A C99 code generator to write native CTF binary streams barectf is a command-line utility which generates C99 code that is able to write native Common Trace Format (CTF) binary streams. The target audience of barectf is developers who need to trace bare metal systems (without an operating system). The code produced by barectf is pure C99 and can be lightweight enough to fit on a tiny microcontroller.
Bug#805559: FTBFS: Depends: libbabeltrace-dev (< 1.3) but 1.3.1-1 is to be installed
On Thu, 19 Nov 2015 16:46:19 + Iain Lane <la...@debian.org> wrote: > On Thu, Nov 19, 2015 at 11:25:58AM -0500, Michael Jeanson wrote: > > […] > > I've pushed the fix to collab-maint but I don't have upload rights > > yet, would you mind uploading it? > > I'd love to, but it still fails. :( That's what happens when I try to go too fast, it's now fixed and tested. > > sbuild-build-depends-lttngtop-dummy : Depends: libbabeltrace-dev (< > 1.3) but 1.3.1-1 is to be installed > > > […] > > By the way, I'm looking for a second advocate for my DM application, > > I've been co-maintaining the lttng related packages with Jon > > Bernard for more than a year now but haven't worked with any other > > DD yet. > > I'd have to work with you a bit more (just drive-bying here really) > before being happy to do that - but you only strictly need one > advocate, even if they encourage more. Yeah I'd like to have at least 2 sponsors, it just seems more appropriate. > > Cheers, >
Bug#805559: FTBFS: Depends: libbabeltrace-dev (< 1.3) but 1.3.1-1 is to be installed
On Thu, 19 Nov 2015 13:56:37 + Iain Lanewrote: > Package: lttngtop > Version: 0.3-2 > Severity: serious > Tags: patch > User: ubuntu-de...@lists.ubuntu.com > Usertags: origin-ubuntu xenial ubuntu-patch > > Hi, > > lttngtop fails to build because it has an upper bound on a versioned > dependency which is unsatisfiable in unstable now. > > Some packages could not be installed. This may mean that you have > requested an impossible situation or if you are using the unstable > distribution that some required packages have not yet been created > or been moved out of Incoming. > The following information may help to resolve the situation: > > The following packages have unmet dependencies: >sbuild-build-depends-lttngtop-dummy : Depends: libbabeltrace-dev > (< 1.3) but 1.3.1-1 is to be installed Depends: libbabeltrace-ctf-dev > (< 1.3) but 1.3.1-1 is to be installed E: Unable to correct problems, > you have held broken packages. apt-get failed. > Package installation failed > > I tried dropping the bounds (attached patch) and it builds for me. > > Are they necessary? If so, could you fix the package to work with the > new babeltrace? > > Cheers, > Hi, I've pushed the fix to collab-maint but I don't have upload rights yet, would you mind uploading it? For the record, the post 1.2 master branch of babeltrace is incompatible with lttngtop and was supposed to be the next release. However, as the release date kept sliding, it was decided to apply some patches on the 1.2 branch and release it as 1.3 and I forgot to update this package. By the way, I'm looking for a second advocate for my DM application, I've been co-maintaining the lttng related packages with Jon Bernard for more than a year now but haven't worked with any other DD yet. Thanks, Michael
Bug#798244: lttnganalyses ftbfs, missing build dependencies
On 15-09-07 05:19 AM, Matthias Klose wrote: > Package: src:lttnganalyses > Version: 0.3.0-1 > Severity: serious > Tags: sid stretch > > [...] > fakeroot debian/rules binary > dh binary --with python3 --buildsystem=pybuild >dh_testroot -O--buildsystem=pybuild >dh_prep -O--buildsystem=pybuild >dh_auto_install -O--buildsystem=pybuild > I: pybuild base:170: /usr/bin/python3.5 setup.py install --root > /«PKGBUILDDIR»/debian/python3-lttnganalyses > lttnganalysescli needs the babeltrace package. > See https://www.efficios.com/babeltrace for more info. > > E: pybuild pybuild:256: install: plugin distutils failed with: exit code=1: > /usr/bin/python3.5 setup.py install --root > /«PKGBUILDDIR»/debian/python3-lttnganalyses > dh_auto_install: pybuild --install -i python{version} -p 3.5 3.4 --dir . > --dest-dir /«PKGBUILDDIR»/debian/python3-lttnganalyses returned exit code 13 > make: *** [binary] Error 13 > dpkg-buildpackage: error: fakeroot debian/rules binary gave error exit status > 2 > debian/rules:7: recipe for target 'binary' failed > This seems to be related to the transition to python 3.5, lttnganalyses depends on python3-babeltrace which contains a native extension that wasn't built against python 3.5. I'm not sure what's the proper way to fix this. Michael
Bug#793393: ITP: lttnganalyses -- LTTng 2.0 trace analysis tools
Package: wnpp Severity: wishlist Owner: Michael Jeanson mjean...@ubuntu.com * Package name: lttnganalyses Version : 0.3.0 Upstream Author : Julien Desfossez jdesfos...@efficios.com * URL : https://github.com/lttng/lttng-analyses * License : MIT Programming Lang: Python Description : LTTng 2.0 trace analysis tools The LTTng project aims at providing highly efficient tracing tools for Linux. Its tracers help tracking down performance issues and debugging problems involving multiple concurrent processes and threads. Tracing across multiple systems is also possible. . This package contains various tools to analyse LTTng kernel traces and extract monitoring data and metrics. As opposed to other diagnostic or monitoring solutions, this approach is designed to allow users to record their system's activity with a low overhead, wait for a problem to occur and then diagnose its cause offline. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#776924: liburcu: Enable hppa architecture (with patch, using gcc atomic builtins)
On 03/02/2015 11:58, Helge Deller wrote: Package: liburcu Version: 0.8.6-1 Severity: normal Tags: patch The attached patch enables liburucu to sucessfully build on the hppa architecture. The patch is pretty trivial, since hppa now supports the builtin atomic functions from gcc. Can you please apply it for the next upload or apply it to upstream? Attached are two patches: One for the debian package, and one for upstream version. I did signed-off both. Thanks, Helge Hi, I've talked to upstream and they don't have any hppa hardware on hand. They would like to have a look at the output log of make regtest and make bench before applying the patch. If you have access to such hardware would you be able to run those tests and send the output? Thanks, Michael -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#741971: ITP: lttngtop -- A top-like interface to read and browse LTTng traces.
-BEGIN PGP SIGNED MESSAGE- Hash: SHA512 Hi, On 14-03-18 01:34 PM, Holger Levsen wrote: mabye add a word or two about what LTTng is? The LTTng project is a set of highly efficient kernel and userspace tracing tools for Linux. They can be used to debug live systems, analyze performance and monitor with minimal impact on production systems. Most of the LTTng tools are currently packaged in debian by Jon Bernard which showed interest in sponsoring the upload of lttngtop. Regards, Michael -BEGIN PGP SIGNATURE- Version: GnuPG v1 Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQIcBAEBCgAGBQJTKLxbAAoJEIZWH0UhgPz+3xgP/jGe3LoFB+JEuCoN6Ys/9/zu WtvgShyW1JkbBz9xP057DdKcAJneqexRQT4xrE8+DsKLYzgSiqy4wDYDX7MxN4ul KvspS3ch8+BQL0CTGwqrAGjQsCvgUJfWp8pfObJ+sKhu4iSWzxzA35oS+Uq7Sf2+ Dd0bYVhYdWwAs1CJ31O1E4n+pLyNf+476jI8uYjnoumh1YClMRKhra7LTQFp9u+S XyCsS8/3U/vTeztMw/c1nCb5ayx2CI+gZLsACVvS5giskOoEcKGPWCEDTRzdk+6N lZ/YxBy25G4wy9SAqQ+9HP2XIOa0SX2kh9HW+JwVtpgf0kWEJ3agTj+9UV2Gkuto qmvmCMAuIaHVJK4cY5nW+mlV6Qp1znjtK3ZtHtac54Wl5/+3rw8z7fdK02nDzuAo jRcaxbjjWTTJErkUK9SziqPH02+CFJ4kkmnF8KVxVY0SjaTW6UWPxL84BUslisFK 4tWaWkc94E6gVKeiKy9yvPiazXVw7crhDKzYlD1IHqZDtp2M7aZLw/s4nZz39k8r hOifZLVjtArtfhO75PnbARvPUr4pIhYh+BhGeNj4JFLqJYBZbeLmzjuSYkbut5oj pKD+md6Qd8LI7U7fY0UsPiMDK+TKAiYQeXpWCidXZEWe4xcWYyfG1iIxF1mywxOW NH+wafb8BSFPgEqlvB/w =FhZ+ -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#741971: ITP: lttngtop -- A top-like interface to read and browse LTTng traces.
Package: wnpp Severity: wishlist Owner: Michael Jeanson mjean...@ubuntu.com * Package name: lttngtop Version : 0.2 Upstream Author : Julien Desfossez julien.desfos...@efficios.com * URL : http://www.lttng.org/ * License : GPL2 Programming Lang: C Description : A top-like interface to read and browse LTTng traces. Lttngtop is an ncurses interface for reading and browsing traces recorded by the LTTng tracer and displaying various statistics. As of now, the cpu usage, per file/process I/O bandwidth and perf counters are displayed. This version currently only supports offline traces, but a live version is in alpha and will be available for testing soon. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#630634: RFP: ding-libs -- ding-libs is a set of helpful libraries used by SSSD and other projects.
Package: wnpp Severity: wishlist * Package name: ding-libs Version : 0.1.2 Upstream Author : Dmitri Pal d...@redhat.com * URL : https://fedorahosted.org/sssd/ * License : LGPL Programming Lang: C Description : ding-libs is a set of helpful libraries used by SSSD and other projects. ding-libs provides utility functions to manipulate filesystem pathnames, a hash table which dynamically resizes to achieve optimal storage and access time properties, a data type to collect data in a hierarchical structure for easy iteration and serialization, a dynamically growing, reference-counted array, and a library to process configuration files in initialization format (INI) into a library collection data structure. It is required to upgrade the sssd package to 1.5.x -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org