Message ID | 20231019165050.2134920-1-adhemerval.zanella@linaro.org |
---|---|
State | Accepted |
Commit | 83d13972f23546758b600ba940e0d53248dd0339 |
Headers | show |
Series | malloc: Fix tst-tcfree3 build csky-linux-gnuabiv2 with fortify source | expand |
LGTM but would it make sense to have -fno-builtin-malloc for *all* the
malloc tests?
Reviewed-by: DJ Delorie <dj@redhat.com>
On 19/10/23 14:59, DJ Delorie wrote: > > LGTM but would it make sense to have -fno-builtin-malloc for *all* the > malloc tests? > > Reviewed-by: DJ Delorie <dj@redhat.com> > I think it would be valuable, this is what I did in azanella/clang [1] to avoid clang optimize ways most of the calls. [1] https://sourceware.org/git/?p=glibc.git;a=blobdiff;f=malloc/Makefile;h=0e47b8446f59fab9d05364543aff8d988f640a3e;hp=c1db3347d88e112fcdfd78c90564c6d5d95abb33;hb=ca27004e1096ab2942dc770acccf835e52bc9d74;hpb=9fd866d9a7ac9b55b6a796df207aa3cd0899e451
diff --git a/malloc/Makefile b/malloc/Makefile index c1db3347d8..f6cdf7bf0b 100644 --- a/malloc/Makefile +++ b/malloc/Makefile @@ -298,6 +298,8 @@ CPPFLAGS-malloc.c += -DUSE_TCACHE=1 # Uncomment this for test releases. For public releases it is too expensive. #CPPFLAGS-malloc.o += -DMALLOC_DEBUG=1 +CFLAGS-tst-tcfree3.c += -fno-builtin-malloc -fno-builtin-free + sLIBdir := $(shell echo $(slibdir) | sed 's,lib\(\|64\)$$,\\\\$$LIB,') $(objpfx)mtrace: mtrace.pl diff --git a/malloc/tst-tcfree3.c b/malloc/tst-tcfree3.c index ce6d958f8a..30886b6f49 100644 --- a/malloc/tst-tcfree3.c +++ b/malloc/tst-tcfree3.c @@ -19,9 +19,6 @@ #include <malloc.h> #include <string.h> -/* Prevent GCC from optimizing away any malloc/free pairs. */ -#pragma GCC optimize ("O0") - static int do_test (void) {