Changes in Makefile

2cbb20b0 - tracing: Disable branch profiling in noinstr code

Fri, 21 Mar 2025 19:53:32 +0000

tracing: Disable branch profiling in noinstr codeCONFIG_TRACE_BRANCH_PROFILING inserts a call to ftrace_likely_update()for each use of likely() or unlikely(). That breaks noinstr rules ifthe affected function is annotated as noinstr.Disable branch profiling for files with noinstr functions. In additionto some individual files, this also includes the entire arch/x86subtree, as well as the kernel/entry, drivers/cpuidle, and drivers/idledirectories, all of which are noinstr-heavy.Due to the nature of how sched binaries are built by combining multiple.c files into one, branch profiling is disabled more broadly across thesched code than would otherwise be needed.This fixes many warnings like the following: vmlinux.o: warning: objtool: do_syscall_64+0x40: call to ftrace_likely_update() leaves .noinstr.text section vmlinux.o: warning: objtool: __rdgsbase_inactive+0x33: call to ftrace_likely_update() leaves .noinstr.text section vmlinux.o: warning: objtool: handle_bug.isra.0+0x198: call to ftrace_likely_update() leaves .noinstr.text section ...Reported-by: Ingo Molnar Suggested-by: Steven Rostedt Signed-off-by: Josh Poimboeuf Signed-off-by: Ingo Molnar Acked-by: Thomas Gleixner Cc: Linus Torvalds Link: https://lore.kernel.org/r/fb94fc9303d48a5ed370498f54500cc4c338eb6d.1742586676.git.jpoimboe@kernel.org List of files: /linux-6.15/lib/Makefile

c104c160 - Kunit to check the longest symbol length

Sun, 02 Mar 2025 22:15:18 +0000

Kunit to check the longest symbol lengthThe longest length of a symbol (KSYM_NAME_LEN) was increased to 512in the reference [1]. This patch adds kunit test suite to check the longestsymbol length. These tests verify that the longest symbol length definedis supported.This test can also help other efforts for longer symbol length,like [2].The test suite defines one symbol with the longest possible length.The first test verify that functions with names of the createdsymbol, can be called or not.The second test, verify that the symbols are created (ornot) in the kernel symbol table.[1] https://lore.kernel.org/lkml/20220802015052.10452-6-ojeda@kernel.org/[2] https://lore.kernel.org/lkml/20240605032120.3179157-1-song@kernel.org/Link: https://lore.kernel.org/r/20250302221518.76874-1-sergio.collado@gmail.comTested-by: Martin Rodriguez Reboredo Reviewed-by: Shuah Khan Reviewed-by: Rae Moar Signed-off-by: Sergio González Collado Link: https://github.com/Rust-for-Linux/linux/issues/504Reviewed-by: Rae Moar Acked-by: David Gow Signed-off-by: Shuah Khan List of files: /linux-6.15/lib/Makefile

97c1f302 - scanf: convert self-test to KUnit

Fri, 07 Mar 2025 11:27:36 +0000

scanf: convert self-test to KUnitConvert the scanf() self-test to a KUnit test.In the interest of keeping the patch reasonably-sized this doesn'trefactor the tests into proper parameterized tests - it's all one bigtest case.Reviewed-by: David Gow Reviewed-by: Petr Mladek Tested-by: Petr Mladek Signed-off-by: Tamir Duberstein Link: https://lore.kernel.org/r/20250307-scanf-kunit-convert-v9-3-b98820fa39ff@gmail.comSigned-off-by: Kees Cook List of files: /linux-6.15/lib/Makefile

7a79e7da - printf: convert self-test to KUnit

Fri, 07 Mar 2025 22:08:56 +0000

printf: convert self-test to KUnitConvert the printf() self-test to a KUnit test.In the interest of keeping the patch reasonably-sized this doesn'trefactor the tests into proper parameterized tests - it's all one bigtest case.Signed-off-by: Tamir Duberstein Reviewed-by: Petr Mladek Tested-by: Petr Mladek Link: https://lore.kernel.org/r/20250307-printf-kunit-convert-v6-1-4d85c361c241@gmail.comSigned-off-by: Kees Cook List of files: /linux-6.15/lib/Makefile

df7fcbef - vdso: Add generic time data storage

Tue, 04 Feb 2025 12:05:37 +0000

vdso: Add generic time data storageHistorically each architecture defined their own way to store the vDSOdata page. Add a generic mechanism to provide storage for that page.Furthermore this generic storage will be extended to also provideuniform storage for *non*-time-related data, like the random state orarchitecture-specific data. These will have their own pages and datastructures, so rename 'vdso_data' into 'vdso_time_data' to make thatsplit clear from the name.Also introduce a new consistent naming scheme for the symbols related tothe vDSO, which makes it clear if the symbol is accessible fromuserspace or kernel space and the type of data behind the symbol.The generic fault handler contains an optimization to prefault the vvarpage when the timens page is accessed. This was lifted from s390 and x86.Co-developed-by: Nam Cao Signed-off-by: Nam Cao Signed-off-by: Thomas Weißschuh Signed-off-by: Thomas Gleixner Link: https://lore.kernel.org/all/20250204-vdso-store-rng-v3-5-13a4669dfc8c@linutronix.de List of files: /linux-6.15/lib/Makefile

b341f6fd - blackhole_dev: convert self-test to KUnit

Sat, 08 Feb 2025 19:26:43 +0000

blackhole_dev: convert self-test to KUnitConvert this very simple smoke test to a KUnit test.Add a missing `htons` call that was spotted[0] by kernel test robot after initial conversion to KUnit.Link: https://lore.kernel.org/oe-kbuild-all/202502090223.qCYMBjWT-lkp@intel.com/ [0]Signed-off-by: Tamir Duberstein Link: https://patch.msgid.link/20250208-blackholedev-kunit-convert-v2-1-182db9bd56ec@gmail.comSigned-off-by: Jakub Kicinski List of files: /linux-6.15/lib/Makefile

db6fe4d6 - lib: Move KUnit tests into tests/ subdirectory

Mon, 02 Dec 2024 07:55:40 +0000

lib: Move KUnit tests into tests/ subdirectoryFollowing from the recent KUnit file naming discussion[1], move allKUnit tests in lib/ into lib/tests/.Link: https://lore.kernel.org/lkml/20240720165441.it.320-kees@kernel.org/ [1]Acked-by: Steven Rostedt (Google) Acked-by: Jakub Kicinski Acked-by: Masami Hiramatsu (Google) Reviewed-by: David Gow Acked-by: Vlastimil Babka Reviewed-by: Rae Moar Link: https://lore.kernel.org/r/20241202075545.3648096-4-davidgow@google.comSigned-off-by: Kees Cook List of files: /linux-6.15/lib/Makefile

feb541bf - lib/crc64-rocksoft: stop wrapping the crypto API

Thu, 30 Jan 2025 03:51:20 +0000

lib/crc64-rocksoft: stop wrapping the crypto APIFollowing what was done for the CRC32 and CRC-T10DIF library functions,get rid of the pointless use of the crypto API and makecrc64_rocksoft_update() call into the library directly. This is fasterand simpler.Remove crc64_rocksoft() (the version of the function that did not take a'crc' argument) since it is unused.Reviewed-by: Ard Biesheuvel Reviewed-by: "Martin K. Petersen" Acked-by: Keith Busch Link: https://lore.kernel.org/r/20250130035130.180676-2-ebiggers@kernel.orgSigned-off-by: Eric Biggers List of files: /linux-6.15/lib/Makefile

05033905 - revert "xarray: port tests to kunit"

Fri, 31 Jan 2025 00:09:20 +0000

revert "xarray: port tests to kunit"Revert c7bb5cf9fc4e ("xarray: port tests to kunit"). It broke the buildwhen compiing the xarray userspace test harness code.Reported-by: Sidhartha Kumar Closes: https://lkml.kernel.org/r/07cf896e-adf8-414f-a629-a808fc26014a@oracle.comCc: David Gow Cc: Matthew Wilcox Cc: Tamir Duberstein Cc: "Liam R. Howlett" Cc: Geert Uytterhoeven Cc: Lorenzo Stoakes Signed-off-by: Andrew Morton List of files: /linux-6.15/lib/Makefile

c7bb5cf9 - xarray: port tests to kunit

Thu, 05 Dec 2024 15:11:26 +0000

xarray: port tests to kunitMinimally rewrite the XArray unit tests to use kunit. This integratesnicely with existing kunit tools which produce nicer human-readable outputcompared to the existing machinery.Running the xarray tests before this change requires an obscureinvocation```tools/testing/kunit/kunit.py run --arch arm64 --make_options LLVM=1 \ --kconfig_add CONFIG_TEST_XARRAY=y --raw_output=all nothing```which on failure produces```BUG at check_reserve:513...XArray: 6782340 of 6782364 tests passed```and exits 0.Running the xarray tests after this change requires a simpler invocation```tools/testing/kunit/kunit.py run --arch arm64 --make_options LLVM=1 \ xarray```which on failure produces (colors omitted)```[09:50:53] ====================== check_reserve ======================[09:50:53] [FAILED] param-0[09:50:53] # check_reserve: EXPECTATION FAILED at lib/test_xarray.c:536[09:50:53] xa_erase(xa, 12345678) != NULL...[09:50:53] # module: test_xarray[09:50:53] # xarray: pass:26 fail:3 skip:0 total:29[09:50:53] # Totals: pass:28 fail:3 skip:0 total:31[09:50:53] ===================== [FAILED] xarray ======================```and exits 1.Use of richer kunit assertions is intentionally omitted to reduce thescope of the change.[akpm@linux-foundation.org: fix cocci warning] Link: https://lore.kernel.org/oe-kbuild-all/202412081700.YXB3vBbg-lkp@intel.com/Link: https://lkml.kernel.org/r/20241205-xarray-kunit-port-v1-1-ee44bc7aa201@gmail.comSigned-off-by: Tamir Duberstein Cc: Bill Wendling Cc: Christophe Leroy Cc: Geert Uytterhoeven Cc: Justin Stitt Cc: Madhavan Srinivasan Cc: Michael Ellerman Cc: Nathan Chancellor Cc: Naveen N Rao Cc: Nicholas Piggin Cc: Nick Desaulniers Signed-off-by: Andrew Morton List of files: /linux-6.15/lib/Makefile

87fe0a13 - lib/crc32test: delete obsolete crc32test.c

Mon, 02 Dec 2024 01:20:54 +0000

lib/crc32test: delete obsolete crc32test.cDelete crc32test.c, since it has been superseded by crc_kunit.c.Reviewed-by: Ard Biesheuvel Reviewed-by: Martin K. Petersen Acked-by: Geert Uytterhoeven # m68kCc: Vinicius Peixoto Link: https://lore.kernel.org/r/20241202012056.209768-11-ebiggers@kernel.orgSigned-off-by: Eric Biggers List of files: /linux-6.15/lib/Makefile

c14e8536 - lib/crc16_kunit: delete obsolete crc16_kunit.c

Mon, 02 Dec 2024 01:20:53 +0000

lib/crc16_kunit: delete obsolete crc16_kunit.cThis new test showed up in v6.13-rc1. Delete it since it is beingsuperseded by crc_kunit.c, which is more comprehensive (tests multipleCRC variants without duplicating code, includes a benchmark, etc.).Cc: Vinicius Peixoto Link: https://lore.kernel.org/r/20241202012056.209768-10-ebiggers@kernel.orgSigned-off-by: Eric Biggers List of files: /linux-6.15/lib/Makefile

e47d9b1a - lib/crc_kunit.c: add KUnit test suite for CRC library functions

Mon, 02 Dec 2024 01:20:52 +0000

lib/crc_kunit.c: add KUnit test suite for CRC library functionsAdd a KUnit test suite for the crc16, crc_t10dif, crc32_le, crc32_be,crc32c, and crc64_be library functions. It avoids code duplication bysharing most logic among all CRC variants. The test suite includes:- Differential fuzz test of each CRC function against a simple bit-at-a-time reference implementation.- Test for CRC combination, when implemented by a CRC variant.- Optional benchmark of each CRC function with various data lengths.This is intended as a replacement for crc32test and crc16_kunit, as wellas a new test for CRC variants which didn't previously have a test.Reviewed-by: Ard Biesheuvel Reviewed-by: Martin K. Petersen Cc: Vinicius Peixoto Link: https://lore.kernel.org/r/20241202012056.209768-9-ebiggers@kernel.orgSigned-off-by: Eric Biggers List of files: /linux-6.15/lib/Makefile

38a9a512 - lib/crc32: make crc32c() go directly to lib

Mon, 02 Dec 2024 01:08:40 +0000

lib/crc32: make crc32c() go directly to libNow that the lower level __crc32c_le() library function is optimized foreach architecture, make crc32c() just call that instead of taking aninefficient and error-prone detour through the shash API.Note: a future cleanup should make crc32c_le() be the actual libraryfunction instead of __crc32c_le(). That will require updating callersof __crc32c_le() to use crc32c_le() instead, and updating callers ofcrc32c_le() that expect a 'const void *' arg to expect 'const u8 *'instead. Similarly, a future cleanup should remove LIBCRC32C by makingeveryone who is selecting it just select CRC32 directly instead.Reviewed-by: Ard Biesheuvel Link: https://lore.kernel.org/r/20241202010844.144356-16-ebiggers@kernel.orgSigned-off-by: Eric Biggers List of files: /linux-6.15/lib/Makefile

11131415 - lib: util_macros_kunit: add kunit test for util_macros.h

Tue, 05 Nov 2024 14:54:06 +0000

lib: util_macros_kunit: add kunit test for util_macros.hA bug was found in the find_closest() (find_closest_descending() is alsoaffected after some testing), where for certain values with smallprogressions of 1, 2 & 3, the rounding (done by averaging 2 values) causesan incorrect index to be returned.The bug is described in more detail in the commit which fixes the bug. This commit adds a kunit test to validate that the fix works correctly.This kunit test adds some of the arrays (from the driver-sphere) that seemto produce issues with the 'find_closest()' macro. Specifically the onefrom ad7606 driver (with which the bug was found) and from the ina2xxdrivers, which shows the quirk with 'find_closest()' with elements in aarray that have an interval of 3.For the find_closest_descending() tests, the same arrays are used as forthe find_closest(), but in reverse; the idea is that'find_closest_descending()' should return the sames indices as'find_closest()' but in reverse.For testing both macros, there are 4 special arrays created, one fortesting find_closest{_descending}() for arrays of progressions 1, 2, 3 and4. The idea is to show that (for progressions of 1, 2 & 3) the fix worksas expected. When removing the fix, the issues should start to show up.Then an extra array of negative and positive values is added. There arecurrently no such arrays within drivers, but one could expect that thesemacros behave correctly even for such arrays.To run this kunit: ./tools/testing/kunit/kunit.py run "*util_macros*"Link: https://lkml.kernel.org/r/20241105145406.554365-2-aardelean@baylibre.comSigned-off-by: Alexandru Ardelean Cc: Bartosz Golaszewski Cc: Greg Kroah-Hartman Signed-off-by: Andrew Morton List of files: /linux-6.15/lib/Makefile

92a8b224 - lib/min_heap: introduce non-inline versions of min heap API functions

Sun, 20 Oct 2024 04:01:51 +0000

lib/min_heap: introduce non-inline versions of min heap API functionsPatch series "Enhance min heap API with non-inline functions andoptimizations", v2.Add non-inline versions of the min heap API functions in lib/min_heap.cand updates all users outside of kernel/events/core.c to use thesenon-inline versions. To mitigate the performance impact of indirectfunction calls caused by the non-inline versions of the swap and comparefunctions, a builtin swap has been introduced that swaps elements based ontheir size. Additionally, it micro-optimizes the efficiency of the minheap by pre-scaling the counter, following the same approach as inlib/sort.c. Documentation for the min heap API has also been added to thecore-api section.This patch (of 10):All current min heap API functions are marked with '__always_inline'. However, as the number of users increases, inlining these functionseverywhere leads to a increase in kernel size.In performance-critical paths, such as when perf events are enabled andmin heap functions are called on every context switch, it is important toretain the inline versions for optimal performance. To balance this, theoriginal inline functions are kept, and additional non-inline versions ofthe functions have been added in lib/min_heap.c.Link: https://lkml.kernel.org/r/20241020040200.939973-1-visitorckw@gmail.comLink: https://lore.kernel.org/20240522161048.8d8bbc7b153b4ecd92c50666@linux-foundation.orgLink: https://lkml.kernel.org/r/20241020040200.939973-2-visitorckw@gmail.comSigned-off-by: Kuan-Wei Chiu Suggested-by: Andrew Morton Cc: Adrian Hunter Cc: Arnaldo Carvalho de Melo Cc: Ching-Chun (Jim) Huang Cc: Coly Li Cc: Ian Rogers Cc: Ingo Molnar Cc: Jiri Olsa Cc: Jonathan Corbet Cc: Kent Overstreet Cc: Kuan-Wei Chiu Cc: "Liang, Kan" Cc: Mark Rutland Cc: Matthew Sakai Cc: Matthew Wilcox (Oracle) Cc: Namhyung Kim Cc: Peter Zijlstra Signed-off-by: Andrew Morton List of files: /linux-6.15/lib/Makefile

bf9850f6 - lib/Makefile: make union-find compilation conditional on CONFIG_CPUSETS

Fri, 11 Oct 2024 14:12:14 +0000

lib/Makefile: make union-find compilation conditional on CONFIG_CPUSETSCurrently, cpuset is the only user of the union-find implementation. Compiling union-find in all configurations unnecessarily increases thecode size when building the kernel without cgroup support. Modify thebuild system to compile union-find only when CONFIG_CPUSETS is enabled.Link: https://lore.kernel.org/lkml/1ccd6411-5002-4574-bb8e-3e64bba6a757@redhat.com/Link: https://lkml.kernel.org/r/20241011141214.87096-1-visitorckw@gmail.comSigned-off-by: Kuan-Wei Chiu Suggested-by: Waiman Long Acked-by: Waiman Long Acked-by: Tejun Heo Reviewed-by: Christoph Hellwig Cc: Ching-Chun (Jim) Huang Cc: Johannes Weiner Cc: Michal Koutný Cc: Xavier Cc: Zefan Li Signed-off-by: Andrew Morton List of files: /linux-6.15/lib/Makefile

5d042707 - lib/crc16_kunit.c: add KUnit tests for crc16

Sat, 12 Oct 2024 07:43:49 +0000

lib/crc16_kunit.c: add KUnit tests for crc16Add Kunit tests for the kernel's implementation of the standard CRC-16algorithm (). The test data consists of 100randomly-generated test cases, validated against a naive CRC-16implementation.This test follows roughly the same logic as lib/crc32test.c, but withoutthe performance measurements.Link: https://lkml.kernel.org/r/20241012-crc16-kunit-v3-1-0ca75cb58ca9@lkcamp.devSigned-off-by: Vinicius Peixoto Co-developed-by: Enzo Bertoloti Signed-off-by: Enzo Bertoloti Co-developed-by: Fabricio Gasperin Signed-off-by: Fabricio Gasperin Suggested-by: David Laight Cc: Brendan Higgins Cc: David Gow Cc: Rae Moar Signed-off-by: Andrew Morton List of files: /linux-6.15/lib/Makefile

b660d0a2 - New implementation for IO memcpy and IO memset

Mon, 28 Oct 2024 13:42:24 +0000

New implementation for IO memcpy and IO memsetThe IO memcpy and IO memset functions in asm-generic/io.h simply callmemcpy and memset. This can lead to alignment problems or faults onarchitectures that do not define their own version and fall back tothese defaults.This patch introduces new implementations for IO memcpy and IO memset,that use read{l,q} accessor functions, align accesses to machine wordsize, and resort to byte accesses when the target memory is not aligned.For new architectures and existing ones that were using the oldfallbacks these functions are save to use, because IO memory constraintsare taken into account. Moreover, architectures with similarimplementations can now use these new versions, not needing to implementtheir own.Reviewed-by: Yann Sionneau Signed-off-by: Julian Vetter Signed-off-by: Arnd Bergmann List of files: /linux-6.15/lib/Makefile

84b4a51f - selftests: add new kallsyms selftests

Mon, 21 Oct 2024 19:11:44 +0000

selftests: add new kallsyms selftestsWe lack find_symbol() selftests, so add one. This let's us stress testimprovements easily on find_symbol() or optimizations. It also inherentlyallows us to test the limits of kallsyms on Linux today.We test a pathalogical use case for kallsyms by introducing moduleswhich are automatically written for us with a larger number of symbols.We have 4 kallsyms test modules:A: has KALLSYSMS_NUMSYMS exported symbolsB: uses one of A's symbolsC: adds KALLSYMS_SCALE_FACTOR * KALLSYSMS_NUMSYMS exportedD: adds 2 * the symbols than CBy using anything much larger than KALLSYSMS_NUMSYMS as 10,000 andKALLSYMS_SCALE_FACTOR of 8 we segfault today. So we're capped ataround 160000 symbols somehow today. We can inpsect that issue atour leasure later, but for now the real value to this test is thatthis will easily allow us to test improvements on find_symbol().We want to enable this test on allyesmodconfig builds so we can'tuse this combination, so instead just use a safe value for now andbe informative on the Kconfig symbol documentation about where ourthresholds are for testers. We default then to KALLSYSMS_NUMSYMS ofjust 100 and KALLSYMS_SCALE_FACTOR of 8.On x86_64 we can use perf, for other architectures we just use 'time'and allow for customizations. For example a future enhancements couldbe done for parisc to check for unaligned accesses which triggers aspecial special exception handler assembler code inside the kernel.The negative impact on performance is so large on parisc that itkeeps track of its accesses on /proc/cpuinfo as UAH:IRQ: CPU0 CPU13: 1332 0 SuperIO ttyS07: 1270013 0 SuperIO pata_ns8741564: 320023012 320021431 CPU timer65: 17080507 20624423 CPU IPIUAH: 10948640 58104 Unaligned access handler trapsWhile at it, this tidies up lib/ test modules to allow us to havea new directory for them. The amount of test modules under lib/is insane.This should also hopefully showcase how to start doing basicself module writing code, which may be more useful for more complexcases later in the future.Signed-off-by: Luis Chamberlain List of files: /linux-6.15/lib/Makefile