Tom Hromatka [Wed, 9 Feb 2022 19:27:43 +0000 (12:27 -0700)]
api.c: Refactor cgroup_create_cgroup()
Non-functional change to refactor cgroup_create_cgroup(). Move
the contents of this "for" loop to its own function:
for (k = 0; k < cgroup->index; k++)
This will be used as a building block for the next commit to allow
the creation of cgroups with no controllers attached.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
In cgroup_init(), the first line/header of /proc/cgroups helps in
interpreting the values in the remaining of the file. We are interested
in the rest of the file and a temporary buffer of size FILENAME_MAX is
allocated/deallocated after reading the header. The length of the header
is only 56 characters in length, excluding the newline. This patch
reduces it to LL_MAX a.k.a 100 characters.
LL_MAX is used instead of introducing another new length macro.
Signed-off-by: Kamalesh Babulal <kamalesh.babulal@oracle.com> Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Tom Hromatka [Wed, 15 Dec 2021 21:27:26 +0000 (21:27 +0000)]
github: Add a code coverage action
Add a code coverage action to the Github workflow. As part
of this code coverage action, ignore the C file that is
auto-generated from the Cython code. Also ignore a couple Python
header files that were confusing lcov.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Wed, 15 Dec 2021 21:03:28 +0000 (21:03 +0000)]
github: Skip test 38 (cpuset) on VMs hosted by Github Actions
Skip test #38 - 038-cgxget-cpuset_settings.py - on servers hosted
by Github Actions. Test #38 sets/clears the cpuset exclusive bit
and this is not allowed on Github's servers.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Add several unmappable cpuset settings. As the name implies, these
settings have no equivalent in cgroup v2:
cpuset.mem_exclusive
cpuset.mem_hardwall
cpuset.memory_migrate
cpuset.memory_pressure
cpuset.mem_pressure_enabled
cpuset.memory_spread_page
cpuset.memory_spread_slab
cpuset.sched_load_balance
cpuset.sched_relax_domain_level
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Mon, 31 Jan 2022 15:21:00 +0000 (08:21 -0700)]
cgxset: Add a flag to ignore settings that can't be converted
Add a flag, "-i", to cgxset to ignore settings that can't be
converted from cgroup v1 to cgroup v2 or vice versa.
Some cgroup v1 settings, e.g. cpuset.sched_relax_domain_level,
don't have a cgroup v2 equivalent and likely never will. This
flag allows users to ignore errors from these unconvertable
settings.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Fri, 28 Jan 2022 21:22:53 +0000 (14:22 -0700)]
cgxget: Add a flag to ignore settings that can't be converted
Add a flag, "-i", to cgxget to ignore settings that can't be
converted from cgroup v1 to cgroup v2 or vice versa.
Some cgroup v1 settings, e.g. cpuset.sched_relax_domain_level,
don't have a cgroup v2 equivalent and likely never will. This
flag allows users to ignore errors from these unconvertable
settings.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Add an unmappable mapping for the cpu.stat setting. While
the filename, cpu.stat, maps perfectly from v1 <-> v2, the
contents do not. At this time, let's set this as unmappable
and revisit it when we've thought through the best way to
map the contents.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Wed, 15 Dec 2021 21:13:29 +0000 (21:13 +0000)]
abstraction: Add support for unmappable settings
Add a function to explicitly handle settings that cannot
be mapped from v1 to v2 or vice versa. This function returns
ECGNOVERSIONCONVERT which cgxget, cgxset, etc. can then
monitor and if requested by the user, ignore.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Add custom functions to map cpu.cfs_quota_us <-> cpu.max and
cpu.cfs_period_us <-> cpu.max. These mappings can be 1->N or
N->1 and utilize custom values for limits, "-1" in cgroup v1
and "max" in cgroup v2.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Wed, 15 Dec 2021 20:39:52 +0000 (20:39 +0000)]
wrapper: Add function to remove name/value pair from a controller
Add a function to remove a name/value settings pair from
a cgroup_controller instance. As part of this change, I did a small
refactor on cgroup_free_controllers() to make the code more reusable
throughout libcgroup.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Wed, 15 Dec 2021 20:39:22 +0000 (20:39 +0000)]
api.c: Add const qualifier to parameters in cgroup_read_value_begin()
Make two string parameters passed into cgroup_read_value_begin()
fully constant, i.e. const char * const. These parameters,
controller and name, are only used as read-only variables
within cgroup_read_value_begin() and are not intended to be
modified.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Wed, 15 Dec 2021 20:37:37 +0000 (20:37 +0000)]
api.c/wrapper/abstraction: Save previous setting name in control_value
When there's an N->1 mapping, the abstraction layer needs
to know which "1" setting to map back to. Add a field to
struct control_value that saves off the original setting,
so that the abstraction layer can know how to map back from
an N->1 setting.
A good example of this is cpu.max/cpu.cfs_quota_us. The
abstraction layer will map cpu.cfs_quota_us to cpu.max.
During the return conversion, the abstraction layer needs
to know whether to map cpu.max to cfs_quota_us or
cfs_period_us.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Wed, 15 Dec 2021 21:11:28 +0000 (21:11 +0000)]
abstraction: Add name conversion function
Add a function to the abstraction layer that can convert from
one cgroup setting name to another setting name. The values are
not adjusted by this function. For example, this can be used for
conversions like cpuset.effective_cpus to cpuset.cpus.effective.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Fri, 28 Jan 2022 17:22:49 +0000 (10:22 -0700)]
cgxset: Introduce cgxset
Add a new libcgroup tool - cgxset. cgxset is based upon cgset,
but it supports converting from cgroup version to another.
For example, a request like the following will work on a system
running the cgroup v1 cpu controller or the cgroup v2 cpu
controller.
$ cgxset -2 -r cpu.weight=42 MyCgroup
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Fri, 28 Jan 2022 17:22:11 +0000 (10:22 -0700)]
cgxget: Introduce cgxget
Add a new libcgroup tool - cgxget. cgxget is based upon cgget,
but it supports converting from cgroup version to another.
For example, a request like the following will work on a system
running the cgroup v1 cpu controller or the cgroup v2 cpu controller.
$ cgxget -2 -r cpu.weight MyCgroup
The return value to the user will match the cgroup version requested.
If the user requests a v1 setting, the return value will be a cgroup
v1 setting. If the user requests a v2 setting, the return value will
be a cgroup v2 setting.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Wed, 15 Dec 2021 20:18:54 +0000 (20:18 +0000)]
abstraction: Add function to convert from one cgroup version to another
Add a function - cgroup_convert_cgroup() - that can convert from
one cgroup version to another. This function will be the building
block for the libcgroup abstraction layer.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Wed, 15 Dec 2021 20:14:11 +0000 (20:14 +0000)]
abstraction: Add integer conversion function
Add a function to the abstraction layer that can convert from
one integer setting to another. For example, this can be used
for simple conversions like cpu.shares to cpu.weight and vice
versa.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Wed, 15 Dec 2021 19:37:36 +0000 (19:37 +0000)]
api.c: Remove static from cgroup_copy_controller_values()
Remove static from cgroup_copy_controller_values() in api.c
so that other files within libcgroup can invoke it. It will
be used by the abstraction layer code in a subsequent commit.
Also, declare *dst as a const pointer and *src as a const *
const pointer. Subsequent abstraction layer checkins make
heavy use of const and without these changes, warnings (which
continuous integration converts to errors) are thrown.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Tom Hromatka [Wed, 15 Dec 2021 19:30:30 +0000 (19:30 +0000)]
wrapper.c: Store the version in struct cgroup_controller
Store the cgroup version in struct cgroup_controller. Each
controller can be mounted separately, and thus the cgroup
version is applicable at the controller level.
Note that this commit requires cgroup_init() to be called by
cgconfigparser. Storing the version in the cgroup_controller
struct requires the cg_mount_table[] to be populated.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com> Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Kamalesh Babulal [Thu, 27 Jan 2022 19:33:52 +0000 (12:33 -0700)]
api.c: add support for cgroup.threads
cgroup_build_tasks_procs_path(), that builds the proc path for moving
the tid, doesn't check for the cgroup type and appends cgroup.procs for
both cgroup v1/v2. With cgroup v2, this results in building a wrong
proc path if the cgroup type is threaded. This patch adds support to
check for the cgroup v2 type and build the proc path according to the
below table:
cgroup.type | proc file | Description |
----------------+----------------+------------------------------------+
domain | cgroup.procs | regular cgroup, all threads gets |
| | moved to the cgroup. |
domain threaded | cgroup.procs | root of threaded subtree, all |
| | threads gets moved to the cgroup. |
threaded | cgroup.threads | cgroup under threaded subtree, |
| | single threads with matching tid, |
| | gets moved to the cgroup. |
domain invalid | N/A | invalid state of a cgroup under the|
| | threaded subtree. |
Closes: https://github.com/libcgroup/libcgroup/issues/64 Signed-off-by: Kamalesh Babulal <kamalesh.babulal@oracle.com> Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Kamalesh Babulal [Tue, 18 Jan 2022 15:22:13 +0000 (08:22 -0700)]
api.c: cache cgrules with CGFLAG_USECACHE flag
pam_cgroup.so plugin uses /etc/cgrules.conf to assign processes
to the requested cgroup. This works well with active
cgrulesengd but in the instances where cgrulesengd is disabled,
the rules are not read and cached by default. Fix this is by
reloading and caching the rules when called with CGFLAG_USECACHE
flag.
Fixes: https://github.com/libcgroup/libcgroup/issues/84 Reported-by: Kent Overstreet <kent.overstreet@gmail.com> Signed-off-by: Kamalesh Babulal <kamalesh.babulal@oracle.com> Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
TJH: Minor commit comment changes
Ingo van Lil [Thu, 16 Dec 2021 20:32:23 +0000 (13:32 -0700)]
Initialize cgroup version for named V1 hierarchy
Commit 3aac9889 added the 'version' field in cg_mount_table_s and
initializes it in cgroup_process_v1_mnt(). The same initialization is
required in a second place dealing with named hierarchies.
Signed-off-by: Ingo van Lil <inguin@gmx.de> Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Tom Hromatka [Tue, 7 Dec 2021 16:36:51 +0000 (09:36 -0700)]
github: Force a fresh clone on all self-hosted runners
The cgroup v2 job is being run on a self-hosted runner that
re-uses the workspace. Delete the local workspace on all
self-hosted runners prior to cloning the repo.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Tom Hromatka [Tue, 7 Dec 2021 16:14:26 +0000 (09:14 -0700)]
tests: Update submodule to the latest commit on libcgroup-tests main branch
Closes: https://github.com/libcgroup/libcgroup/issues/71 Reported-by: Ross Burton <ross.burton@arm.com> Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Ross Burton [Thu, 2 Dec 2021 10:51:28 +0000 (10:51 +0000)]
automake: idiomatically mark test libraries as convenience libraries
Instead of installing then deleting, mark the libraries which are only
used at build time and should not be installed as noinst, so they don't
get installed at all.
Signed-off-by: Ross Burton <ross.burton@arm.com> Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Tom Hromatka [Fri, 12 Nov 2021 13:55:20 +0000 (06:55 -0700)]
automake: Don't install testing libraries
libcgset and libcgfortesting are libraries that are used by the
unit test framework and are not intended to be installed when
`make install` is invoked. Delete them from the install path.
Closes: https://github.com/libcgroup/libcgroup/issues/62 Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com> Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Tom Hromatka [Fri, 12 Nov 2021 13:51:36 +0000 (06:51 -0700)]
automake: Link libcgset with libcgroup
When linking with the -Wl,--no-undefined flag enabled, libcgset
fails due to undefined symbols. Fix this by enabling this flag
and linking with libcgroup.la
Closes: https://github.com/libcgroup/libcgroup/issues/66 Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Kamalesh Babulal [Thu, 21 Oct 2021 21:18:28 +0000 (15:18 -0600)]
github: re-arrange workflow run commands
The current workflow executes the functional tests from the make check
called at the top-level source directory. Though this runs the tests
case, it also has an unintended side effect of re-running gunit tests
for every functional test combination in the workflow. Fix it, by
re-arranging the function test run commands, to change into
tests/ftests directory before executing make check.
Signed-off-by: Kamalesh Babulal <kamalesh.babulal@oracle.com> Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Ross Burton [Fri, 10 Sep 2021 17:24:47 +0000 (11:24 -0600)]
configure.ac: check for the library containing fts_open
The musl C library doesn't provide fts.h[1], so libcgroup doesn't
compile with musl. However, there is a standalone implementation of fts
for musl users[2] which can be used.
Use AC_SEARCH_LIBS to search for fts_open, which will check if it is part
of libc, and if not look in libfts, then set LIBS if needed.
Signed-off-by: Ross Burton <ross.burton@arm.com>
[TJH: minor formatting changes to match existing configure.ac] Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Tom Hromatka [Wed, 21 Jul 2021 15:40:45 +0000 (09:40 -0600)]
README.md: Display status of main branch in build status badge
The master branch was renamed to main, but the build status
badge wasn't updated as part of this change. Display the
build status of the main branch in README.md
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Tom Hromatka [Tue, 6 Jul 2021 15:16:09 +0000 (09:16 -0600)]
github: Delete the tests folder before bootstrap
On self-hosted github machines, the working directory for
libcgroup is reused from one run to the next. This is fine
for the root of the directory as the proper git clean/reset
commands are issues. But this is problematic for the git
submodules like the tests/ folder because clean/reset is nott
run there.
Remove the tests/ directory entirely and rely on ./bootstrap.sh
to recreate and populate it.
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Replace __BEGIN_DECLS and __END_DECLS with extern "C".
The macros __BEGIN_DECLS and __END_DECLS are a GNU-ism found in
glibc and uClibc, but not musl. We replace them by the more general
extern "C" { ... } block exposed only if we have __cplusplus.
Forward ported from a patch by Anthony G. Basile <blueness@gentoo.org>.
Signed-off-by: Alex Xu (Hello71) <alex_y_xu@yahoo.ca> Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Tomasz Kłoczko [Wed, 12 May 2021 20:35:40 +0000 (20:35 +0000)]
automake: replace INCLUDES by CPPFLAGS
automake shows warnings like below:
src/Makefile.am:16: warning: 'INCLUDES' is the old name for 'AM_CPPFLAGS' (or '*_CPPFLAGS')
src/Makefile.am: installing 'build-aux/depcomp'
src/bindings/Makefile.am:2: warning: 'INCLUDES' is the old name for 'AM_CPPFLAGS' (or '*_CPPFLAGS')
src/daemon/Makefile.am:3: warning: 'INCLUDES' is the old name for 'AM_CPPFLAGS' (or '*_CPPFLAGS')
src/pam/Makefile.am:1: warning: 'INCLUDES' is the old name for 'AM_CPPFLAGS' (or '*_CPPFLAGS')
src/tools/Makefile.am:3: warning: 'INCLUDES' is the old name for 'AM_CPPFLAGS' (or '*_CPPFLAGS')
Signed-off-by: Tomasz Kłoczko <kloczek@github.com> Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Tom Hromatka [Tue, 27 Apr 2021 17:08:38 +0000 (17:08 +0000)]
lscgroup: Fix strncpy() truncation warning
Fix the following strncpy() string truncation warning:
In function ‘strncat’,
inlined from ‘print_cgroup’ at lscgroup.c:149:4:
/usr/include/x86_64-linux-gnu/bits/string_fortified.h:136:10: warning: ‘__builtin___strncat_chk’ output
may be truncated copying between 0 and 4095 bytes from a string of length 4095 [-Wstringop-truncation]
136 | return __builtin___strncat_chk (__dest, __src, __len, __bos (__dest));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Tom Hromatka [Tue, 27 Apr 2021 17:04:21 +0000 (17:04 +0000)]
lssubsys: Fix strncpy() truncation warning
Fix the following strncpy() string truncation warning:
In function ‘strncpy’,
inlined from ‘print_all_controllers_in_hierarchy’ at lssubsys.c:131:4,
inlined from ‘cgroup_list_all_controllers’ at lssubsys.c:225:9,
inlined from ‘main’ at lssubsys.c:293:8:
/usr/include/x86_64-linux-gnu/bits/string_fortified.h:106:10: warning: ‘__builtin_strncpy’ output may be truncated copying 4095 bytes from a string of length 4095 [-Wstringop-truncation]
106 | return __builtin___strncpy_chk (__dest, __src, __len, __bos (__dest));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
Tom Hromatka [Tue, 27 Apr 2021 17:03:18 +0000 (17:03 +0000)]
lssubsys: Fix strncpy() truncation warning
Fix the following strncpy() string truncation warning:
In function ‘strncpy’,
inlined from ‘print_all_controllers_in_hierarchy’ at lssubsys.c:127:4,
inlined from ‘cgroup_list_all_controllers’ at lssubsys.c:223:9,
inlined from ‘main’ at lssubsys.c:291:8:
/usr/include/x86_64-linux-gnu/bits/string_fortified.h:106:10: warning: ‘__builtin_strncpy’ output may be truncated copying 4095 bytes from a string of length 4095 [-Wstringop-truncation]
106 | return __builtin___strncpy_chk (__dest, __src, __len, __bos (__dest));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Signed-off-by: Tom Hromatka <tom.hromatka@oracle.com>
projects / thirdparty / libcgroup.git / log
