| Message ID | 20250131225444.1144932-1-petr.vorel@gmail.com |
|---|---|
| State | New |
| Headers | show |
| Series | [1/1] ltp: Update to 20250130 | expand |
Hi Petr, Thanks for the patch. On Fri, 2025-01-31 at 23:54 +0100, Petr Vorel wrote: > Removed 0001-sched_attr-Do-not-define-for-glibc-2.41.patch > (accepted upstream) > > Signed-off-by: Petr Vorel <petr.vorel@gmail.com> > --- > Hi all, > > could please anybody test this patch? (I'm an LTP upstream developer > with very little Yocto / openembedded knowledge). > > Would it be possible also to test if 0001-Add-__clear_cache- > declaration-for-clang.patch > is still needed? As Cyril noted at [1], it sounds like clang bug > (__clear_cache() is a pre-defined function and as such the signature > should be part of the compiler). > > @Tim any change your kirk recipe is merged upstream? > > Kind regards, > Petr > > [1] https://lore.kernel.org/ltp/Zmg9WZ1TitB7D_RC@yuki/ > [2] https://git.yoctoproject.org/poky-contrib/log/?h=timo/kirk I ran it through our automated testing and whilst the main tests ran ok, it failed on musl: https://autobuilder.yoctoproject.org/valkyrie/#/builders/6/builds/932/steps/11/logs/stdio https://autobuilder.yoctoproject.org/valkyrie/#/builders/3/builds/948/steps/11/logs/stdio Cheers, Richard
Hi Richard, ... > I ran it through our automated testing and whilst the main tests ran > ok, it failed on musl: > https://autobuilder.yoctoproject.org/valkyrie/#/builders/6/builds/932/steps/11/logs/stdio > https://autobuilder.yoctoproject.org/valkyrie/#/builders/3/builds/948/steps/11/logs/stdio Thank you for running the tests for me. Although we have musl (via Alpine) in LTP CI, this error slipped in: | CC testcases/kernel/syscalls/listen/listen01 | In file included from listmount01.c:22: | listmount.h: In function 'listmount': | listmount.h:18:16: error: variable 'req' has initializer but incomplete type | 18 | struct mnt_id_req req = { | | ^~~~~~~~~~ | listmount.h:19:18: error: 'struct mnt_id_req' has no member named 'size' | 19 | .size = MNT_ID_REQ_SIZE_VER0, | | ^~~~ Fortunately fix looks quite obvious (missing autotools guard). I'll send Cc to LTP + v2 for yocto. I'll try to force myself to run the yocto build myself to not bother you with testing for me. Kind regards, Petr > Cheers, > Richard
diff --git a/meta/recipes-extended/ltp/ltp/0001-sched_attr-Do-not-define-for-glibc-2.41.patch b/meta/recipes-extended/ltp/ltp/0001-sched_attr-Do-not-define-for-glibc-2.41.patch deleted file mode 100644 index 58d665c7dd..0000000000 --- a/meta/recipes-extended/ltp/ltp/0001-sched_attr-Do-not-define-for-glibc-2.41.patch +++ /dev/null @@ -1,42 +0,0 @@ -From 38ac9ed7a4f7e2dcedb3c13d807ad0575435bab4 Mon Sep 17 00:00:00 2001 -From: Khem Raj <raj.khem@gmail.com> -Date: Wed, 30 Oct 2024 14:37:12 -0700 -Subject: [PATCH] sched_attr: Do not define for glibc >= 2.41 - -glibc 2.41+ has added [1] definitions for sched_setattr and sched_getattr functions -and struct sched_attr. Therefore, it needs to be checked for here as well before -defining sched_attr - -Define sched_attr conditionally on SCHED_ATTR_SIZE_VER0 - -Fixes builds with glibc/trunk - -[1] https://sourceware.org/git/?p=glibc.git;a=commitdiff;h=21571ca0d70302909cf72707b2a7736cf12190a0;hp=298bc488fdc047da37482f4003023cb9adef78f8 - -Upstream-Status: Submitted [https://lists.linux.it/pipermail/ltp/2024-October/040678.html] -Signed-off-by: Khem Raj <raj.khem@gmail.com> ---- - include/lapi/sched.h | 3 +++ - 1 file changed, 3 insertions(+) - -diff --git a/include/lapi/sched.h b/include/lapi/sched.h -index 26fdb628d..e1c7ff011 100644 ---- a/include/lapi/sched.h -+++ b/include/lapi/sched.h -@@ -15,6 +15,8 @@ - #include "lapi/syscalls.h" - #include "lapi/sched.h" - -+/* sched_attr is not defined in glibc < 2.41 */ -+#ifndef SCHED_ATTR_SIZE_VER0 - struct sched_attr { - uint32_t size; - -@@ -44,6 +46,7 @@ static inline int sched_getattr(pid_t pid, struct sched_attr *attr, - { - return syscall(__NR_sched_getattr, pid, attr, size, flags); - } -+#endif - - #ifndef HAVE_CLONE3 - struct clone_args { diff --git a/meta/recipes-extended/ltp/ltp_20240930.bb b/meta/recipes-extended/ltp/ltp_20250130.bb similarity index 98% rename from meta/recipes-extended/ltp/ltp_20240930.bb rename to meta/recipes-extended/ltp/ltp_20250130.bb index 3fe8772ea7..908a037959 100644 --- a/meta/recipes-extended/ltp/ltp_20240930.bb +++ b/meta/recipes-extended/ltp/ltp_20250130.bb @@ -24,12 +24,11 @@ TUNE_CCARGS:remove:x86-64 = "-mfpmath=sse" CFLAGS:append:powerpc64 = " -D__SANE_USERSPACE_TYPES__" CFLAGS:append:mipsarchn64 = " -D__SANE_USERSPACE_TYPES__" -SRCREV = "60f81419c43f94ec182827ef0b9eb5baeb303419" +SRCREV = "a7c31dff7edc089a32e990765e12952cc4d7666a" SRC_URI = "git://github.com/linux-test-project/ltp.git;branch=master;protocol=https \ file://0001-Remove-OOM-tests-from-runtest-mm.patch \ file://0001-Add-__clear_cache-declaration-for-clang.patch \ - file://0001-sched_attr-Do-not-define-for-glibc-2.41.patch \ " S = "${WORKDIR}/git"
Removed 0001-sched_attr-Do-not-define-for-glibc-2.41.patch (accepted upstream) Signed-off-by: Petr Vorel <petr.vorel@gmail.com> --- Hi all, could please anybody test this patch? (I'm an LTP upstream developer with very little Yocto / openembedded knowledge). Would it be possible also to test if 0001-Add-__clear_cache-declaration-for-clang.patch is still needed? As Cyril noted at [1], it sounds like clang bug (__clear_cache() is a pre-defined function and as such the signature should be part of the compiler). @Tim any change your kirk recipe is merged upstream? Kind regards, Petr [1] https://lore.kernel.org/ltp/Zmg9WZ1TitB7D_RC@yuki/ [2] https://git.yoctoproject.org/poky-contrib/log/?h=timo/kirk ...ed_attr-Do-not-define-for-glibc-2.41.patch | 42 ------------------- .../ltp/{ltp_20240930.bb => ltp_20250130.bb} | 3 +- 2 files changed, 1 insertion(+), 44 deletions(-) delete mode 100644 meta/recipes-extended/ltp/ltp/0001-sched_attr-Do-not-define-for-glibc-2.41.patch rename meta/recipes-extended/ltp/{ltp_20240930.bb => ltp_20250130.bb} (98%)