This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH] Build gdb.opt/inline-*.exp tests at -O0, rely on __attribute__((always_inline)) (was: Re: [PATCH v3 24/34] Push thread->control.command_interp to the struct thread_fsm)
- From: Thomas Preudhomme <thomas dot preudhomme at foss dot arm dot com>
- To: Pedro Alves <palves at redhat dot com>
- Cc: gdb-patches at sourceware dot org
- Date: Fri, 01 Jul 2016 16:24:38 +0100
- Subject: Re: [PATCH] Build gdb.opt/inline-*.exp tests at -O0, rely on __attribute__((always_inline)) (was: Re: [PATCH v3 24/34] Push thread->control.command_interp to the struct thread_fsm)
- Authentication-results: sourceware.org; auth=none
- References: <1462538104-19109-25-git-send-email-palves at redhat dot com> <8587009 dot ikB25uYBLk at e108577-lin> <20144b4c-11ee-fc84-e3ad-b9992f14ce15 at redhat dot com>
On Friday 01 July 2016 13:05:35 Pedro Alves wrote:
> On 07/01/2016 12:02 PM, Thomas Preudhomme wrote:
> > The new tests added by this patch fail for arm-none-eabi targets because
> > -O2 leads to instructions to be reordered widely. In this case, the
> > instruction that follows the first one for line 64 is related to line 70
> > so the test is failing. Shouldn't this test be compiled with -Og and
> > probably also -finline- small-functions -findirect-inlining
> > -fpartial-inlining which relates to inlining and are included in -O2.
>
> Or even just plain -O0. See the commit log below. WDYT?
I'm not very familiar with GDB but the description looks great indeed. Thanks!
Best regards,
Thomas
>
> --------------
> Subject: [PATCH] Build gdb.opt/inline-*.exp tests at -O0, rely on
> __attribute__((always_inline))
>
> A test recently added to gdb.opt/inline-cmds.exp fails for
> arm-none-eabi targets because -O2 leads to instructions to be
> reordered widely.
>
> I guess it might have made sense years ago to enable optimization in
> these tests, but I fail to see the need for that nowadays.
>
> Using -O0 while relying on __attribute__((always_inline)), which is
> already used in the tests [1] [2], avoids this sort of trouble, while
> still exercising the inlining-related use cases that are the focus of
> these tests.
>
> I think that nowadays we can safely assume that all compilers we care
> about support __attribute__((always_inline)) or similar.
>
> [1] - Except one spot that missed it.
>
> [2] - Note that the .exp files make sure the frames that should have
> been inlined are indeed inlined, with "info frame".
>
> gdb/testsuite/ChangeLog:
> 2016-07-01 Pedro Alves <palves@redhat.com>
>
> * gdb.opt/inline-break.exp: Remove optimize=-O2.
> * gdb.opt/inline-bt.exp: Likewise.
> * gdb.opt/inline-cmds.exp: Remove optimize=-O2 and add
> additional_flags=-Winline.
> * gdb.opt/inline-locals.exp: Likewise.
> * gdb.opt/inline-markers.c (ATTR): Define.
> (inlined_fn): Use it.
> ---
> gdb/testsuite/gdb.opt/inline-break.exp | 2 +-
> gdb/testsuite/gdb.opt/inline-bt.exp | 2 +-
> gdb/testsuite/gdb.opt/inline-cmds.exp | 2 +-
> gdb/testsuite/gdb.opt/inline-locals.exp | 2 +-
> gdb/testsuite/gdb.opt/inline-markers.c | 8 +++++++-
> 5 files changed, 11 insertions(+), 5 deletions(-)
>
> diff --git a/gdb/testsuite/gdb.opt/inline-break.exp
> b/gdb/testsuite/gdb.opt/inline-break.exp index b2aa22e..ac56b04 100644
> --- a/gdb/testsuite/gdb.opt/inline-break.exp
> +++ b/gdb/testsuite/gdb.opt/inline-break.exp
> @@ -20,7 +20,7 @@
> standard_testfile
>
> if { [prepare_for_testing $testfile.exp $testfile $srcfile \
> - {debug optimize=-O2 additional_flags=-Winline}] } {
> + {debug additional_flags=-Winline}] } {
> return -1
> }
>
> diff --git a/gdb/testsuite/gdb.opt/inline-bt.exp
> b/gdb/testsuite/gdb.opt/inline-bt.exp index 63d76e2..13c6993 100644
> --- a/gdb/testsuite/gdb.opt/inline-bt.exp
> +++ b/gdb/testsuite/gdb.opt/inline-bt.exp
> @@ -17,7 +17,7 @@ standard_testfile .c inline-markers.c
>
> if {[prepare_for_testing $testfile.exp $testfile \
> [list $srcfile $srcfile2] \
> - {debug optimize=-O2 additional_flags=-Winline}]} {
> + {debug additional_flags=-Winline}]} {
> return -1
> }
>
> diff --git a/gdb/testsuite/gdb.opt/inline-cmds.exp
> b/gdb/testsuite/gdb.opt/inline-cmds.exp index 684f4dd..6c84848 100644
> --- a/gdb/testsuite/gdb.opt/inline-cmds.exp
> +++ b/gdb/testsuite/gdb.opt/inline-cmds.exp
> @@ -19,7 +19,7 @@ set MIFLAGS "-i=mi"
> standard_testfile .c inline-markers.c
>
> if {[prepare_for_testing $testfile.exp $testfile \
> - [list $srcfile $srcfile2] {debug optimize=-O2}]} {
> + [list $srcfile $srcfile2] {debug additional_flags=-Winline}]} {
> return -1
> }
>
> diff --git a/gdb/testsuite/gdb.opt/inline-locals.exp
> b/gdb/testsuite/gdb.opt/inline-locals.exp index df2253a..36f7ed2 100644
> --- a/gdb/testsuite/gdb.opt/inline-locals.exp
> +++ b/gdb/testsuite/gdb.opt/inline-locals.exp
> @@ -16,7 +16,7 @@
> standard_testfile .c inline-markers.c
>
> if {[prepare_for_testing $testfile.exp $testfile \
> - [list $srcfile $srcfile2] {debug optimize=-O2}]} {
> + [list $srcfile $srcfile2] {debug additional_flags=-Winline}]} {
> return -1
> }
>
> diff --git a/gdb/testsuite/gdb.opt/inline-markers.c
> b/gdb/testsuite/gdb.opt/inline-markers.c index cf92e79..41f8a38 100644
> --- a/gdb/testsuite/gdb.opt/inline-markers.c
> +++ b/gdb/testsuite/gdb.opt/inline-markers.c
> @@ -13,6 +13,12 @@
> You should have received a copy of the GNU General Public License
> along with this program. If not, see <http://www.gnu.org/licenses/>.
> */
>
> +#ifdef __GNUC__
> +# define ATTR __attribute__((always_inline))
> +#else
> +# define ATTR
> +#endif
> +
> extern int x, y;
> extern volatile int z;
>
> @@ -26,7 +32,7 @@ void marker(void)
> x += y - z; /* set breakpoint 2 here */
> }
>
> -inline void inlined_fn(void)
> +inline ATTR void inlined_fn(void)
> {
> x += y + z;
> }