On 2021-03-21 10:02 p.m., Yi Fan Yu wrote:
3.17.0.rc2 will be very close to the 3.17.0 release
It seems that 3.17 is out now:
https://www.valgrind.org/downloads/current.html
still need to refresh
drd-musl-fix.patch
This post-release commit, that makes the Makefile.a aware of
the musl.supp suppression filter might help musl ptest resutls:
dde556d51 (HEAD -> master, origin/master, origin/HEAD) Add musl.supp to
SUPP_FILES
and musl.supp is included in the valgring git repo and likely the
tarball so we can drop it from the recipe is the content is the same.
TESTING RESULTS:
qemu x86-64:
FAIL: drd/tests/swapcontext
FAIL: gdbserver_tests/hgtls
=== Test Summary ===
TOTAL: 736
PASSED: 693
FAILED: 2
SKIPPED: 41
Compared to:
=== Test Summary ===
TOTAL: 408
PASSED: 390
FAILED: 0
SKIPPED: 18
DURATION: 738
https://autobuilder.yocto.io/pub/non-release/20210321-1/testresults/qemuarm64-ptest/
swap-context.vgtest has been disabled
a workaround for gdbserver_tests/hgtls has been implemented
+ install ${S}/none/tests/tls.c ${D}${PTEST_PATH}/gdbserver_tests/tls.c
OTHER CHANGES:
libdir is now libexecdir
Signed-off-by: Yi Fan Yu <yifan...@windriver.com>
---
...Disable-drd-tests-swapcontext.vgtest.patch | 49 +++++
...gtest-hanging-on-newer-glibc-and-or-.patch | 200 ------------------
.../valgrind/0001-drd-Port-to-Fedora-33.patch | 48 -----
.../valgrind/valgrind/0001-drd-musl-fix.patch | 12 +-
...eak_cmsg.stderr.exp-adjust-tmp-paths.patch | 2 +-
...ch-drd_pthread_intercepts.c-for-musl.patch | 39 ++++
.../valgrind/valgrind/fixed-perl-path.patch | 20 +-
.../valgrind/valgrind/musl.supp | 46 ++++
.../valgrind/valgrind/run-ptest | 3 +-
.../valgrind/valgrind/s390x_vec_op_t.patch | 24 ++-
...{valgrind_3.16.1.bb => valgrind_3.17.0.bb} | 23 +-
11 files changed, 185 insertions(+), 281 deletions(-)
create mode 100644
meta/recipes-devtools/valgrind/valgrind/0001-Disable-drd-tests-swapcontext.vgtest.patch
delete mode 100644
meta/recipes-devtools/valgrind/valgrind/0001-Fix-nlcontrolc.vgtest-hanging-on-newer-glibc-and-or-.patch
delete mode 100644
meta/recipes-devtools/valgrind/valgrind/0001-drd-Port-to-Fedora-33.patch
create mode 100644
meta/recipes-devtools/valgrind/valgrind/0001-patch-drd_pthread_intercepts.c-for-musl.patch
create mode 100644 meta/recipes-devtools/valgrind/valgrind/musl.supp
rename meta/recipes-devtools/valgrind/{valgrind_3.16.1.bb =>
valgrind_3.17.0.bb} (93%)
diff --git
a/meta/recipes-devtools/valgrind/valgrind/0001-Disable-drd-tests-swapcontext.vgtest.patch
b/meta/recipes-devtools/valgrind/valgrind/0001-Disable-drd-tests-swapcontext.vgtest.patch
new file mode 100644
index 0000000000..9e78a9cb61
--- /dev/null
+++
b/meta/recipes-devtools/valgrind/valgrind/0001-Disable-drd-tests-swapcontext.vgtest.patch
@@ -0,0 +1,49 @@
+From e3b587b1cfa7dde1484c88b2af67c49a1c6166e1 Mon Sep 17 00:00:00 2001
+From: Yi Fan Yu <yifan...@windriver.com>
+Date: Fri, 19 Mar 2021 16:35:26 -0400
+Subject: [PATCH] Disable drd/tests/swapcontext.vgtest
+
+Test fails on both x86-64 and arm64.
+
+Unbuidable on musl with switchcontext.
typo: Unbuildable
Are you going to work on this for a while or just report it upstream?
What happens if you run it?
+
+Signed-off-by: Yi Fan Yu <yifan...@windriver.com>
+---
+ drd/tests/Makefile.am | 10 ----------
+ 1 file changed, 10 deletions(-)
+
+diff --git a/drd/tests/Makefile.am b/drd/tests/Makefile.am
+index 79e076246..5df76150e 100644
+--- a/drd/tests/Makefile.am
++++ b/drd/tests/Makefile.am
+@@ -267,8 +267,6 @@ EXTRA_DIST = \
+ sem_open_traced.vgtest \
+ sem_wait.stderr.exp \
+ sem_wait.vgtest \
+- swapcontext.stderr.exp \
+- swapcontext.vgtest \
+ sigalrm.stderr.exp \
+ sigalrm.vgtest \
+ sigaltstack.stderr.exp \
+@@ -408,7 +406,6 @@ check_PROGRAMS = \
+ sem_as_mutex \
+ sem_open \
+ sigalrm \
+- swapcontext \
If swapcontext builds for all our targets/libc combos,
I prefer to build it, install it but then move it aside in the run-ptest
script so that there's less of a barrier for someone to try to fix it.
I guess the downside to that approach is that there's no patch for us
to notice so we should probably track it with a YP bug so do as you
please I guess. :-)
+ threaded-fork \
+ trylock \
+ unit_bitmap \
+@@ -588,10 +585,3 @@ std_thread2_LDFLAGS = -Wl,--no-as-needed
+ endif
+
+ sem_wait_SOURCES = sem_wait.cpp
+-
+-swapcontext_SOURCES = swapcontext.c
+-swapcontext_CFLAGS = $(AM_CFLAGS) -g
+-if VGCONF_OS_IS_SOLARIS
+-swapcontext_CFLAGS += -D__EXTENSIONS__
+-endif
+-
+--
+2.29.2
+
diff --git
a/meta/recipes-devtools/valgrind/valgrind/0001-Fix-nlcontrolc.vgtest-hanging-on-newer-glibc-and-or-.patch
b/meta/recipes-devtools/valgrind/valgrind/0001-Fix-nlcontrolc.vgtest-hanging-on-newer-glibc-and-or-.patch
deleted file mode 100644
index 98cbcd132c..0000000000
---
a/meta/recipes-devtools/valgrind/valgrind/0001-Fix-nlcontrolc.vgtest-hanging-on-newer-glibc-and-or-.patch
+++ /dev/null
@@ -1,200 +0,0 @@
-From 83c24e31df6932a6d4fced179050c6d8d8c6f3b5 Mon Sep 17 00:00:00 2001
-From: Philippe Waroquiers <philippe.waroqui...@skynet.be>
-Date: Sun, 7 Mar 2021 22:29:27 +0100
-Subject: [PATCH] Fix nlcontrolc.vgtest hanging on newer glibc and/or arm64
-
-This test verifies that GDB can interrupt a process with all threads
-blocked in a long select syscall.
-The test used to terminate by having GDB modifying the select argument.
-However, modifying the select argument works only for specific arch
-and/or specific versions of glibc.
-The test then blocks on other architectures/glibc versions.
-
-The previous version of the test was:
- * first launching sleepers so as to have all threads blocked in long select
- * interrupting these threads
- * changing the select time arg so that the threads burn cpu
- * and then change variables to have the program exit.
-
-The new version does:
- * first launches sleepers so that all threads are burning cpu.
- * interrupting these threads
- * change the local variables of sleepers so that the threads will
- block in a long select syscall
- * interrupt these threads
- * kill the program.
-
-With this new version, we still check the behaviour of gdb+vgdbserver
-for both burning and sleep threads, but without having the termination
-depending on modifying select syscall argument.
-
-Tested on debian amd64 and on ubuntu arm64 (to check the test does not hang
-on an arm64 platform).
-
-Upstream-Status: Backport
-
-From commit on master:
-c79180a3afcf65902e578646c3b716cc749db406
-
-Signed-off-by: Yi Fan Yu <yifan...@windriver.com>
----
- gdbserver_tests/nlcontrolc.stderr.exp | 4 +-
- gdbserver_tests/nlcontrolc.stdinB.gdb | 57 +++++++++++++++-----------
- gdbserver_tests/nlcontrolc.stdoutB.exp | 25 ++++++-----
- gdbserver_tests/nlcontrolc.vgtest | 12 +++---
- 4 files changed, 56 insertions(+), 42 deletions(-)
-
-diff --git a/gdbserver_tests/nlcontrolc.stderr.exp
b/gdbserver_tests/nlcontrolc.stderr.exp
-index ac75bb3da..b63a9a988 100644
---- a/gdbserver_tests/nlcontrolc.stderr.exp
-+++ b/gdbserver_tests/nlcontrolc.stderr.exp
-@@ -3,9 +3,9 @@ Nulgrind, the minimal Valgrind tool
- (action at startup) vgdb me ...
-
-
--loops/sleep_ms/burn/threads_spec/affinity: 1000000000 1000000000 1000000000
BSBSBSBS 1
-+loops/sleep_ms/burn/threads_spec/affinity: 1000000000 0 100000 BSBSBSBS 1
- Brussels ready to sleep and/or burn
- London ready to sleep and/or burn
- Petaouchnok ready to sleep and/or burn
- main ready to sleep and/or burn
--
-+Gdb request to kill this process
-diff --git a/gdbserver_tests/nlcontrolc.stdinB.gdb
b/gdbserver_tests/nlcontrolc.stdinB.gdb
-index 667ece18d..ea4fcd530 100644
---- a/gdbserver_tests/nlcontrolc.stdinB.gdb
-+++ b/gdbserver_tests/nlcontrolc.stdinB.gdb
-@@ -9,32 +9,43 @@ shell ./simulate_control_c
--vgdb-prefix=./vgdb-prefix-nlcontrolc 1 grep main nl
- #
- continue
- #
--# Here, all tasks should be blocked in a loooonnnng select, all in WaitSys
--info threads
--# We will unblock them by changing their timeout argument
--# To avoid going into the frame where the timeval arg is,
--# it has been defined as global variables, as the nr
--# of calls on the stack differs between 32bits and 64bits,
--# and/or between OS.
--# ensure select finishes in a few milliseconds max:
--p t[0].tv_sec = 0
--p t[1].tv_sec = 0
--p t[2].tv_sec = 0
--p t[3].tv_sec = 0
--#
--# We will change the burning parameters in a few seconds
-+# Threads are burning cpu now
-+# We would like to fully test info threads here, but which thread are Runnable
-+# or Yielding is unpredictable. With a recent enough gdb, check the nr of
-+# threads by state using pipe commands and grep/wc.
-+init-if-undefined $_gdb_major = 0
-+init-if-undefined $_gdb_minor = 0
-+if $_gdb_major >= 9
-+ | info threads | grep VgTs_Runnable | wc -l
-+ | info threads | grep VgTs_Yielding | wc -l
-+else
-+ echo 1\n
-+ echo 3\n
-+end
-+# We change the variables so that all the threads are blocked in a syscall
-+p burn = 0
-+p sleepms = 1000000
-+#
-+#
- shell ./simulate_control_c --vgdb-prefix=./vgdb-prefix-nlcontrolc 1 grep
changed nlcontrolc.stdoutB.out
- #
--echo changed burning parameters\n
-+echo changed burning parameters to sleeping parameters\n
- continue
-+# Here, all tasks should be blocked in a loooonnnng select, all in WaitSys
-+info threads
-+# We reset the sleepms to 0. The threads should still be blocked in the
syscall
-+p sleepms = 0
-+shell ./simulate_control_c --vgdb-prefix=./vgdb-prefix-nlcontrolc 1 grep
reset nlcontrolc.stdoutB.out
- #
--# Threads are burning cpu now
--# We would like to test info threads here, but which thread are Runnable or
Yielding
--# is unpredictable.
--# info threads
--p burn = 0
--p loops = 0
--p report_finished = 0
-+echo reset to sleeping parameters\n
- continue
--# and the process should stop very quickly now
-+# threads should still be blocked in a loooonnnng select, all in WaitSys
-+info threads
-+if $_gdb_major >= 9
-+ | info threads | grep VgTs_WaitSys | wc -l
-+else
-+ echo 4\n
-+end
-+# Make the process die.
-+kill
- quit
-diff --git a/gdbserver_tests/nlcontrolc.stdoutB.exp
b/gdbserver_tests/nlcontrolc.stdoutB.exp
-index e8a5ff8ba..2e8dc8498 100644
---- a/gdbserver_tests/nlcontrolc.stdoutB.exp
-+++ b/gdbserver_tests/nlcontrolc.stdoutB.exp
-@@ -1,18 +1,21 @@
- Continuing.
- Program received signal SIGTRAP, Trace/breakpoint trap.
-+do_burn () at sleepers.c:41
-+41 for (i = 0; i < burn; i++) loopnr++;
-+ > > > > > >1
-+3
-+$1 = 0
-+$2 = 1000000
-+changed burning parameters to sleeping parameters
-+Continuing.
-+Program received signal SIGTRAP, Trace/breakpoint trap.
- 0x........ in syscall ...
- * 1 Thread .... (tid 1 VgTs_WaitSys) 0x........ in syscall ...
--$1 = 0
--$2 = 0
- $3 = 0
--$4 = 0
--changed burning parameters
-+reset to sleeping parameters
- Continuing.
- Program received signal SIGTRAP, Trace/breakpoint trap.
--do_burn () at sleepers.c:41
--41 for (i = 0; i < burn; i++) loopnr++;
--$5 = 0
--$6 = 0
--$7 = 0
--Continuing.
--Program exited normally.
-+0x........ in syscall ...
-+* 1 Thread .... (tid 1 VgTs_WaitSys) 0x........ in syscall ...
-+ > > > >4
-+Kill the program being debugged? (y or n) [answered Y; input not from
terminal]
-diff --git a/gdbserver_tests/nlcontrolc.vgtest
b/gdbserver_tests/nlcontrolc.vgtest
-index bb5308403..09edfcaba 100644
---- a/gdbserver_tests/nlcontrolc.vgtest
-+++ b/gdbserver_tests/nlcontrolc.vgtest
-@@ -4,16 +4,16 @@
- # and modify some variables
- # the user can control-c an process with all threads in Running/Yielding
- # and modify some variables
--# sleepers is started with argument so that it will compute during ages.
--# The variable modifications means it will exit in a reasonable time.
--# This test is disabled on Solaris because modifying select/poll/ppoll timeout
--# has no effect if a thread is already blocked in that syscall.
-+# sleepers is started so that it burns CPU.
-+# We then interrupt the process.
-+# We modify variables so that instead of burning cpu, sleepers blocks
-+# all threads in a select syscall.
- prog: sleepers
--args: 1000000000 1000000000 1000000000 BSBSBSBS 1
-+args: 1000000000 0 100000 BSBSBSBS 1
- vgopts: --tool=none --vgdb=yes --vgdb-error=0
--vgdb-prefix=./vgdb-prefix-nlcontrolc
- stderr_filter: filter_stderr
- # Bug 338633 nlcontrol hangs on arm64 currently.
--prereq: test -e gdb -a -f vgdb.invoker && ! ../tests/arch_test arm64 && !
../tests/os_test solaris
-+prereq: test -e gdb -a -f vgdb.invoker && ! ../tests/os_test solaris
- progB: gdb
- argsB: --quiet -l 60 --nx ./sleepers
- stdinB: nlcontrolc.stdinB.gdb
---
-2.29.2
-
diff --git
a/meta/recipes-devtools/valgrind/valgrind/0001-drd-Port-to-Fedora-33.patch
b/meta/recipes-devtools/valgrind/valgrind/0001-drd-Port-to-Fedora-33.patch
deleted file mode 100644
index 37f6ea667d..0000000000
--- a/meta/recipes-devtools/valgrind/valgrind/0001-drd-Port-to-Fedora-33.patch
+++ /dev/null
@@ -1,48 +0,0 @@
-From 15330adf7c2471fbaa6a0818db07078d81dbff97 Mon Sep 17 00:00:00 2001
-From: Bart Van Assche <bvanass...@acm.org>
-Date: Sat, 19 Sep 2020 08:08:59 -0700
-Subject: [PATCH] drd: Port to Fedora 33
-
-Apparently on Fedora 33 the POSIX thread functions exist in both libc and
-libpthread. Hence this patch that intercepts the pthread functions in
-libc. See also https://bugs.kde.org/show_bug.cgi?id=426144 .
-
-Signed-off-by: Bart Van Assche <bvanass...@acm.org>
-
-This patch was imported from the valgrind sourceware server
-(https://sourceware.org/git/?p=valgrind.git;a=commit;h=15330adf7c2471fbaa6a0818db07078d81dbff97)
-It was modified to remove the changes to the valgrind NEWS file,
-as these are difficult to maintain and don't impact the valgrind
-code itself.
-
-Upstream-Status: Backport
-
-Signed-off-by: Stacy Gaikovaia <stacy.gaikov...@windriver.com>
----
- drd/drd_pthread_intercepts.c | 9 +++++++++
- 1 file changed, 10 insertions(+)
-
-diff --git a/drd/drd_pthread_intercepts.c b/drd/drd_pthread_intercepts.c
-index 58c45aaec..c2882e5ab 100644
---- a/drd/drd_pthread_intercepts.c
-+++ b/drd/drd_pthread_intercepts.c
-@@ -174,7 +174,16 @@ static int never_true;
- ret_ty VG_WRAP_FUNCTION_ZZ(VG_Z_LIBC_SONAME,zf) argl_decl \
- { return implf argl; }
- #else
-+/*
-+ * On Linux, intercept both the libc and the libpthread functions. At
-+ * least glibc 2.32.9000 (Fedora 34) has an implementation of all pthread
-+ * functions in both libc and libpthread. Older glibc versions only have an
-+ * implementation of the pthread functions in libpthread.
-+ */
- #define PTH_FUNC(ret_ty, zf, implf, argl_decl, argl) \
-+ ret_ty VG_WRAP_FUNCTION_ZZ(VG_Z_LIBC_SONAME,zf) argl_decl; \
-+ ret_ty VG_WRAP_FUNCTION_ZZ(VG_Z_LIBC_SONAME,zf) argl_decl \
-+ { return implf argl; } \
- ret_ty VG_WRAP_FUNCTION_ZZ(VG_Z_LIBPTHREAD_SONAME,zf) argl_decl; \
- ret_ty VG_WRAP_FUNCTION_ZZ(VG_Z_LIBPTHREAD_SONAME,zf) argl_decl \
- { return implf argl; }
---
-2.25.1
-
diff --git a/meta/recipes-devtools/valgrind/valgrind/0001-drd-musl-fix.patch
b/meta/recipes-devtools/valgrind/valgrind/0001-drd-musl-fix.patch
index e96bf3c611..a3268714b8 100644
--- a/meta/recipes-devtools/valgrind/valgrind/0001-drd-musl-fix.patch
+++ b/meta/recipes-devtools/valgrind/valgrind/0001-drd-musl-fix.patch
@@ -1,13 +1,5 @@
-The changes in 0001-drd-Port-to-Fedora-33.patch break builds on musl. These
-need a __GLIBC__ guard to ensure musl builds continue to work.
-
-Upstream-Status: Pending
-Signed-off-by: Richard Purdie <richard.pur...@linuxfoundation.org>
-
-Index: valgrind-3.16.1/drd/drd_pthread_intercepts.c
-===================================================================
---- valgrind-3.16.1.orig/drd/drd_pthread_intercepts.c
-+++ valgrind-3.16.1/drd/drd_pthread_intercepts.c
+--- a/drd/drd_pthread_intercepts.c
++++ b/drd/drd_pthread_intercepts.c
@@ -180,6 +180,7 @@ static int never_true;
* functions in both libc and libpthread. Older glibc versions only have an
* implementation of the pthread functions in libpthread.
diff --git
a/meta/recipes-devtools/valgrind/valgrind/0001-none-tests-fdleak_cmsg.stderr.exp-adjust-tmp-paths.patch
b/meta/recipes-devtools/valgrind/valgrind/0001-none-tests-fdleak_cmsg.stderr.exp-adjust-tmp-paths.patch
index d641998d72..e421ac6c80 100644
---
a/meta/recipes-devtools/valgrind/valgrind/0001-none-tests-fdleak_cmsg.stderr.exp-adjust-tmp-paths.patch
+++
b/meta/recipes-devtools/valgrind/valgrind/0001-none-tests-fdleak_cmsg.stderr.exp-adjust-tmp-paths.patch
@@ -19,7 +19,7 @@ index 6e5a797b9..0ede1ab18 100644
@@ -25,10 +25,10 @@ Open file descriptor ...: /dev/null
- FILE DESCRIPTORS: 6 open at exit.
+ FILE DESCRIPTORS: 6 open (3 std) at exit.
-Open file descriptor ...: /tmp/data2
+Open file descriptor ...: ...
...
diff --git
a/meta/recipes-devtools/valgrind/valgrind/0001-patch-drd_pthread_intercepts.c-for-musl.patch
b/meta/recipes-devtools/valgrind/valgrind/0001-patch-drd_pthread_intercepts.c-for-musl.patch
new file mode 100644
index 0000000000..8af1be8ead
--- /dev/null
+++
b/meta/recipes-devtools/valgrind/valgrind/0001-patch-drd_pthread_intercepts.c-for-musl.patch
@@ -0,0 +1,39 @@
+From db26c19477894c9473959a2e292c77fd5b5edeb4 Mon Sep 17 00:00:00 2001
+From: Yi Fan Yu <yifan...@windriver.com>
+Date: Fri, 19 Mar 2021 14:11:02 -0700
+Subject: [PATCH] patch drd_pthread_intercepts.c for musl
+
+Signed-off-by: Yi Fan Yu <yifan...@windriver.com>
+---
+ drd/drd_pthread_intercepts.c | 8 ++++++++
+ 1 file changed, 8 insertions(+)
+
+diff --git a/drd/drd_pthread_intercepts.c b/drd/drd_pthread_intercepts.c
+index 28c0de78b..82c612a7e 100644
+--- a/drd/drd_pthread_intercepts.c
++++ b/drd/drd_pthread_intercepts.c
+@@ -169,6 +169,7 @@ static int never_true;
+ return pth_func_result; \
+ }
+ #elif defined(VG_WRAP_THREAD_FUNCTION_LIBC_ONLY)
++#ifdef __GLIBC__
+ #define PTH_FUNC(ret_ty, zf, implf, argl_decl, argl) \
+ ret_ty VG_WRAP_FUNCTION_ZZ(VG_Z_LIBC_SONAME,zf) argl_decl; \
+ ret_ty VG_WRAP_FUNCTION_ZZ(VG_Z_LIBC_SONAME,zf) argl_decl \
+@@ -182,6 +183,13 @@ static int never_true;
+ ret_ty VG_WRAP_FUNCTION_ZZ(VG_Z_LIBPTHREAD_SONAME,zf) argl_decl \
+ { return implf argl; }
+ #else
++#define PTH_FUNC(ret_ty, zf, implf, argl_decl, argl) \
++ ret_ty VG_WRAP_FUNCTION_ZZ(VG_Z_LIBPTHREAD_SONAME,zf) argl_decl; \
++ ret_ty VG_WRAP_FUNCTION_ZZ(VG_Z_LIBPTHREAD_SONAME,zf) argl_decl \
++ { return implf argl; }
++#endif
++#define
Is this line needed?
++#else
+ # error "Unknown platform/thread wrapping"
+ #endif
+
+--
+2.17.1
+
diff --git a/meta/recipes-devtools/valgrind/valgrind/fixed-perl-path.patch
b/meta/recipes-devtools/valgrind/valgrind/fixed-perl-path.patch
index b431d3356e..db6867f625 100644
--- a/meta/recipes-devtools/valgrind/valgrind/fixed-perl-path.patch
+++ b/meta/recipes-devtools/valgrind/valgrind/fixed-perl-path.patch
@@ -1,4 +1,4 @@
-From 67e7a690107efb16d6d5aebfe420b64a552accdf Mon Sep 17 00:00:00 2001
+From d85cc45e0ddeda68adf594dead715964cb32d0e7 Mon Sep 17 00:00:00 2001
From: Qing He <qing...@intel.com>
Date: Tue, 31 Aug 2010 22:51:58 +0800
Subject: [PATCH] valgrind: fix perl scripts
@@ -22,10 +22,11 @@ Signed-off-by: Maxin B. John <maxin.j...@intel.com>
cachegrind/cg_diff.in | 2 +-
massif/ms_print.in | 2 +-
perf/vg_perf.in | 2 +-
- 4 files changed, 4 insertions(+), 4 deletions(-)
+ tests/vg_regtest.in | 2 +-
+ 5 files changed, 5 insertions(+), 5 deletions(-)
diff --git a/cachegrind/cg_annotate.in b/cachegrind/cg_annotate.in
-index 69365e8..978265d 100644
+index fea114b..5e814fd 100644
--- a/cachegrind/cg_annotate.in
+++ b/cachegrind/cg_annotate.in
@@ -1,4 +1,4 @@
@@ -35,7 +36,7 @@ index 69365e8..978265d 100644
##--------------------------------------------------------------------##
##--- Cachegrind's annotator. cg_annotate.in ---##
diff --git a/cachegrind/cg_diff.in b/cachegrind/cg_diff.in
-index 395460b..05873cc 100755
+index 9d9258e..d0f0ec7 100755
--- a/cachegrind/cg_diff.in
+++ b/cachegrind/cg_diff.in
@@ -1,4 +1,4 @@
@@ -45,7 +46,7 @@ index 395460b..05873cc 100755
##--------------------------------------------------------------------##
##--- Cachegrind's differencer. cg_diff.in ---##
diff --git a/massif/ms_print.in b/massif/ms_print.in
-index e6fae89..3b85b40 100755
+index a206ce4..df1bc31 100755
--- a/massif/ms_print.in
+++ b/massif/ms_print.in
@@ -1,4 +1,4 @@
@@ -55,7 +56,7 @@ index e6fae89..3b85b40 100755
##--------------------------------------------------------------------##
##--- Massif's results printer ms_print.in ---##
diff --git a/perf/vg_perf.in b/perf/vg_perf.in
-index 7a80cb0..28f6156 100644
+index 90ee1d2..c585096 100644
--- a/perf/vg_perf.in
+++ b/perf/vg_perf.in
@@ -1,4 +1,4 @@
@@ -65,7 +66,7 @@ index 7a80cb0..28f6156 100644
##--- Valgrind performance testing script vg_perf ---##
##--------------------------------------------------------------------##
diff --git a/tests/vg_regtest.in b/tests/vg_regtest.in
-index cb05b52..032e947 100755
+index 0fe6341..1b45eb7 100755
--- a/tests/vg_regtest.in
+++ b/tests/vg_regtest.in
@@ -1,4 +1,4 @@
@@ -74,5 +75,6 @@ index cb05b52..032e947 100755
##--------------------------------------------------------------------##
##--- Valgrind regression testing script vg_regtest ---##
##--------------------------------------------------------------------##
----
-2.4.0
+--
+2.29.2
+
diff --git a/meta/recipes-devtools/valgrind/valgrind/musl.supp
b/meta/recipes-devtools/valgrind/valgrind/musl.supp
Part of the git repo and likely the tarball as mentioned above.
new file mode 100644
index 0000000000..864172a242
--- /dev/null
+++ b/meta/recipes-devtools/valgrind/valgrind/musl.supp
@@ -0,0 +1,46 @@
+# Suppressions for musl libc
+# See: https://www.openwall.com/lists/musl/2017/06/15/4
+
+{
+ musl-dynlink-false-positive1
+ Memcheck:Leak
+ fun:calloc
+ fun:load_direct_deps
+ fun:load_deps
+ fun:load_deps
+ fun:__dls3
+ fun:__dls2b
+ fun:__dls2
+}
+
+{
+ musl-dynlink-false-positive2
+ Memcheck:Leak
+ fun:calloc
+ fun:load_direct_deps
+ fun:load_deps
+ fun:load_deps
+ fun:__dls3
+ fun:__dls2
+}
+
+{
+ musl-dynlink-false-positive3
+ Memcheck:Leak
+ fun:calloc
+ fun:load_library
+ fun:load_preload
+ fun:__dls3
+ fun:__dls2b
+ fun:__dls2
+}
+
+{
+ musl-dynlink-false-positive4
+ Memcheck:Leak
+ fun:calloc
+ fun:load_library
+ fun:load_preload
+ fun:__dls3
+ fun:__dls2
+}
diff --git a/meta/recipes-devtools/valgrind/valgrind/run-ptest
b/meta/recipes-devtools/valgrind/valgrind/run-ptest
index e8a1a668fc..3e0205fe6e 100755
--- a/meta/recipes-devtools/valgrind/valgrind/run-ptest
+++ b/meta/recipes-devtools/valgrind/valgrind/run-ptest
@@ -7,6 +7,7 @@
# Randy MacLeod <randy.macl...@windriver.com>
###############################################################
VALGRIND_LIB=@libdir@/valgrind
+VALGRIND_LIBEXECDIR=@libexecdir@/valgrind
VALGRIND_BIN=@bindir@/valgrind
LOG="${VALGRIND_LIB}/ptest/valgrind_ptest_$(date +%Y%m%d-%H%M%S).log"
@@ -33,7 +34,7 @@ fi
cd ${VALGRIND_LIB}/ptest && ./tests/vg_regtest \
--valgrind=${VALGRIND_BIN} \
- --valgrind-lib=${VALGRIND_LIB} \
+ --valgrind-lib=${VALGRIND_LIBEXECDIR} \
--yocto-ptest \
gdbserver_tests ${TOOLS} ${EXP_TOOLS} \
2>&1|tee ${LOG}
diff --git a/meta/recipes-devtools/valgrind/valgrind/s390x_vec_op_t.patch
b/meta/recipes-devtools/valgrind/valgrind/s390x_vec_op_t.patch
index eea671da0a..bde1241dc7 100644
--- a/meta/recipes-devtools/valgrind/valgrind/s390x_vec_op_t.patch
+++ b/meta/recipes-devtools/valgrind/valgrind/s390x_vec_op_t.patch
@@ -1,3 +1,8 @@
+From 8154d38bac5cdb3675cfdaf562ab9da01988b263 Mon Sep 17 00:00:00 2001
+From: Yi Fan Yu <yifan...@windriver.com>
+Date: Wed, 17 Mar 2021 12:16:31 -0400
+Subject: [PATCH 19/20] remove s390x_vec_op_t
+
s390x_vec_op_t is not needed anywhere, only elements of enum are accessed
removing it ensures that valgrind can be built with -fno-common option
@@ -6,14 +11,25 @@ ld: ../../VEX/libvex-amd64-linux.a(libvex_amd64_linux_a-guest_s390_helpers.o):/u
Upstream-Status: Pending
Signed-off-by: Khem Raj <raj.k...@gmail.com>
+
+Signed-off-by: Yi Fan Yu <yifan...@windriver.com>
+---
+ VEX/priv/guest_s390_defs.h | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/VEX/priv/guest_s390_defs.h b/VEX/priv/guest_s390_defs.h
+index 9054290..dab8002 100644
--- a/VEX/priv/guest_s390_defs.h
+++ b/VEX/priv/guest_s390_defs.h
-@@ -286,7 +286,7 @@ enum {
- S390_VEC_OP_VFCHE = 18,
- S390_VEC_OP_VFTCI = 19,
- S390_VEC_OP_LAST = 20 // supposed to be the last element in enum
+@@ -284,7 +284,7 @@ typedef enum {
+ S390_VEC_OP_VBPERM,
+ S390_VEC_OP_VMSL,
+ S390_VEC_OP_LAST // supposed to be the last element in enum
-} s390x_vec_op_t;
+};
/* Arguments of s390x_dirtyhelper_vec_op(...) which are packed into one
ULong variable.
+--
+2.29.2
+
diff --git a/meta/recipes-devtools/valgrind/valgrind_3.16.1.bb
b/meta/recipes-devtools/valgrind/valgrind_3.17.0.bb
similarity index 93%
rename from meta/recipes-devtools/valgrind/valgrind_3.16.1.bb
rename to meta/recipes-devtools/valgrind/valgrind_3.17.0.bb
index 9dbd9f74c4..1c1ca74262 100644
--- a/meta/recipes-devtools/valgrind/valgrind_3.16.1.bb
+++ b/meta/recipes-devtools/valgrind/valgrind_3.17.0.bb
@@ -12,7 +12,7 @@ DEPENDS = " \
${@bb.utils.contains('DISTRO_FEATURES', 'ptest', 'boost', '', d)} \
"
-SRC_URI = "https://sourceware.org/pub/valgrind/valgrind-${PV}.tar.bz2 \
+SRC_URI = "https://sourceware.org/pub/valgrind/valgrind-${PV}.RC2.tar.bz2 \
file://fixed-perl-path.patch \
file://Added-support-for-PPC-instructions-mfatbu-mfatbl.patch \
file://run-ptest \
@@ -42,13 +42,15 @@ SRC_URI =
"https://sourceware.org/pub/valgrind/valgrind-${PV}.tar.bz2 \
file://s390x_vec_op_t.patch \
file://0001-none-tests-fdleak_cmsg.stderr.exp-adjust-tmp-paths.patch \
file://0001-memcheck-tests-Fix-timerfd-syscall-test.patch \
- file://0001-drd-Port-to-Fedora-33.patch \
- file://0001-drd-musl-fix.patch \
- file://0001-helgrind-Intercept-libc-functions.patch \
-
file://0001-Fix-nlcontrolc.vgtest-hanging-on-newer-glibc-and-or-.patch \
+ file://musl.supp;subdir=${S} \
+ file://0001-Disable-drd-tests-swapcontext.vgtest.patch \
"
-SRC_URI[md5sum] = "d1b153f1ab17cf1f311705e7a83ef589"
-SRC_URI[sha256sum] =
"c91f3a2f7b02db0f3bc99479861656154d241d2fdb265614ba918cc6720a33ca"
+
+S = "${WORKDIR}/valgrind-${PV}.RC2"
+
+SRC_URI[sha256sum] =
"3f27d2176ef9f316bca2ef89abb7cf9da2d99c69bcc4cd0fc86b9debbf55ede5"
+SRC_URI[md5] = "5dcf7c42635e19b074714c53f3a57580"
+
UPSTREAM_CHECK_REGEX = "valgrind-(?P<pver>\d+(\.\d+)+)\.tar"
COMPATIBLE_HOST = '(i.86|x86_64|arm|aarch64|mips|powerpc|powerpc64).*-linux'
@@ -100,6 +102,7 @@ CACHED_CONFIGUREVARS += "ac_cv_path_PERL='/usr/bin/env
perl'"
# to -O2 but uses -O0 for some specific test apps etc. Passing our own flags
# (via CFLAGS) means we interfere with that. Only pass DEBUG_FLAGS to it
# which fixes build path issue in DWARF.
+# this might be key to our issue
Can you explain what you mean by the comment above?
SELECTED_OPTIMIZATION = "${DEBUG_FLAGS}"
do_configure_prepend () {
@@ -134,7 +137,7 @@ RRECOMMENDS_${PN} += "${TCLIBC}-dbg"
RDEPENDS_${PN}-ptest += " bash coreutils file \
gdb libgomp \
- perl \
+ perl perl-modules \
We usually like to keep things as lean as we can even for ptests.
I believe that if you add perl-modules then you could drop the list on
the next two lines but I don't think that's what we want for the recipe.
What's your plan, was that a quick hack to get this out and reviewed?
perl-module-file-basename perl-module-file-glob perl-module-getopt-long \
perl-module-overloading \
procps sed ${PN}-dbg ${PN}-src"
@@ -223,6 +226,7 @@ do_install_ptest() {
# handle multilib
sed -i s:@libdir@:${libdir}:g ${D}${PTEST_PATH}/run-ptest
+ sed -i s:@libexecdir@:${libexecdir}:g ${D}${PTEST_PATH}/run-ptest
sed -i s:@bindir@:${bindir}:g ${D}${PTEST_PATH}/run-ptest
# This test fails on the host as well, using both 3.15 and git master (as of Jan 24 2020)
@@ -233,6 +237,9 @@ do_install_ptest() {
# via dwarfsrcfiles either, so it needs to be installed manually.
mkdir -p ${D}/usr/src/debug/${PN}/${EXTENDPE}${PV}-${PR}/${BP}/none/tests/
install ${S}/none/tests/tls.c
${D}/usr/src/debug/${PN}/${EXTENDPE}${PV}-${PR}/${BP}/none/tests/
+
+ # workaround to pass gdbserver_tests/hgtls (make gdb find tls.c)
+ install ${S}/none/tests/tls.c ${D}${PTEST_PATH}/gdbserver_tests/tls.c
If it's a work-around, should we open a bug to track the proper fix or
do you think this will suffice for the release?
Good work. Please reply in line where it makes sense
and then send a 3.17.0!
../Randy
}
# avoid stripping some generated binaries otherwise some of the tests will fail
--
# Randy MacLeod
# Wind River Linux
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#149781):
https://lists.openembedded.org/g/openembedded-core/message/149781
Mute This Topic: https://lists.openembedded.org/mt/81515521/21656
Group Owner: openembedded-core+ow...@lists.openembedded.org
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-
