Re: [PATCH] Fix gdb.base/step-indirect-call-thunk.exp

[Bruno Larsen <blarsen@redhat.com>]

On 6/15/22 13:21, Carl Love via Gdb-patches wrote:
> GDB maintainers:
> 
> The gdb regression test gdb.base/step-indirect-call-thunk.exp currently
> does not run on X86 due to a compile error related to incompatible gcc
> command line argument.  Secondly, the gcc command line arguments that
> are used are specific to Intel thus generating an unsupported command
> line error when compiled on other architectures.
> 
> This patch fixes the command line arguments so the test will compile on
> X86.  It also adds a check so the test will only run on X86.
> 
> Please let me know if this patch is acceptable for mainline.
> 
>                             Carl Love
> 


Hi Carl!

Thanks for looking at this. The code part of the patch looks good, but I'd suggest a bit
of a change to the commit message.


> 
> --------------------------------------------------------------
> Fix gdb.base/step-indirect-call-thunk.exp
> 
> This test fails on Intel X86-64 with the error:
> 
> Executing on host: gcc  -fno-stack-protector  -fdiagnostics-color=never
> -mindirect-branch=thunk -mfunction-return=thunk -c -g
> -o /.../gdb/testsuite/outputs/gdb.base/step-indirect-call-thunk/step-indirect-call-thunk0.o
> /.../gdb/testsuite/gdb.base/step-indirect-call-thunk.c
> (timeout = 300) builtin_spawn -ignore SIGHUP gcc -fno-stack-protector
> -fdiagnostics-color=never -mindirect-branch=thunk -mfunction-return=thunk -c
> -g -o /.../gdb/testsuite/outputs/gdb.base/step-indirect-call-thunk/step-indirect-call-thunk0.o
> /.../binutils-gdb-current/gdb/testsuite/gdb.base/step-indirect-call-thunk.c
> /.../gdb/testsuite/gdb.base/step-indirect-call-thunk.c:
>   In function 'inc': /.../gdb/testsuite/gdb.base/step-indirect-call-thunk.c:
> 22:1: error: '-mindirect-branch' and '-fcf-protection' are not compatible
>     22 | {                /* inc.1 */
> 
> As stated in the error message the default "-fcf-protection" and
> "-mindirect-branch' are in compatible.  The fcf-protection argument needs to be
> "-fcf-protection=none" for the test to compile on Intel.

This problem doesn't happen on my machine, which is using gcc-8.5.0; I think the default
-fcf-protection value was changed somewhere between gcc-8.5.0 and the one you're using.
I'd mention something along these lines:

	"Due to changes in the default value of -fcf-protection on newer gccs, the test ... can fail."

Also, remember that the commit message should assume the title of the commit was not read,
so please use the full test name instead of "This test".


> 
> The test also fails on PowerPC as the "-mindirect-branch' is an Intel specific
> GCC command line argument.  A check for X86 is added so the test will only run
> on X86 platforms.

I'd remove the specific mention to PowerPC and just say that -mindirect-branch is x86
specific.

With these, I'd give an OK to this patch, but I can't approve it for pushing.

Cheers!
Bruno Larsen
> 
> The patch has been tested and verified on Power 10 and Intel X86-64 systems with
> no regressions.

> ---
>   gdb/testsuite/gdb.base/step-indirect-call-thunk.exp | 6 +++++-
>   1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/gdb/testsuite/gdb.base/step-indirect-call-thunk.exp b/gdb/testsuite/gdb.base/step-indirect-call-thunk.exp
> index 761e1d9a280..7c1b53c99be 100644
> --- a/gdb/testsuite/gdb.base/step-indirect-call-thunk.exp
> +++ b/gdb/testsuite/gdb.base/step-indirect-call-thunk.exp
> @@ -15,7 +15,11 @@
>   
>   standard_testfile
>   
> -set cflags "-mindirect-branch=thunk -mfunction-return=thunk"
> +if { ![istarget "x86*"] } {
> +    return
> +}
> +
> +set cflags "-mindirect-branch=thunk -mfunction-return=thunk -fcf-protection=none"
>   if { [prepare_for_testing "failed to prepare" $testfile $srcfile \
>           [list debug "additional_flags=$cflags"]] } {
>       return -1