On Wed, Nov 19, 2025 at 10:24 PM Jim Tsung-Chun Lin(林宗俊) <[email protected]>
wrote:

> So there is not enough history on these 2. They were added to the repo as
> -O9 back in 1997 when it looks like the testsuite was merged in a whole to
> the cvs. I suspect these testcases were failing at -O3 back then. So
> changing it to -O0 seems wrong and unwanted. Can you explain why you want
> to change them from -O9? GCC treats -O[4-9] the same as -O3 without a
> warning so there is no failure with GCC. Are you trying to compile the
> testsuite with say clang?
> I’m compiling these two tests with Clang, and Clang currently does not
> translate -O9 to -O3 under LTO.
> As a result, the linker fails because it does not recognize -O9. The Clang
> driver should probably handle -O9
> by treating it as -O3 and passing -O3 to the linker.
>
> That said, I don’t think these tests are meant to cover the -O9 option. I
> agree that we should use -O3 for both tests.
>

This is a GCC testsuite so changing it is not really a good idea;
especially since these testcases are almost 30 years old now. Maybe you
should have an extra wrapper when running this testsuite to translate
-O[4-9] into -O3.

Thanks,
Andrew Pinski



>
> ------------------------------
> *寄件者:* Andrew Pinski <[email protected]>
> *寄件日期:* 2025年11月20日 13:52
> *收件者:* Jim Tsung-Chun Lin(林宗俊) <[email protected]>
> *副本:* [email protected] <[email protected]>
> *主旨:* Re: [PATCH 1/1] Fix typo in eh49.C and eh52.C: O9 -> O0
>
> [EXTERNAL MAIL]
>
> On Wed, Nov 19, 2025 at 9:28 PM Jim Lin <[email protected]> wrote:
> >
> > I believe there is no reason to use -O9 for these two testcases. So I
> > assumed it was a typo. The number 9 is next to 0 on the keyboard.
>
> So there is not enough history on these 2. They were added to the repo
> as -O9 back in 1997 when it looks like the testsuite was merged in a
> whole to the cvs.
> I suspect these testcases were failing at -O3 back then. So changing
> it to -O0 seems wrong and unwanted.
> Can you explain why you want to change them from -O9? GCC treats
> -O[4-9] the same as -O3 without a warning so there is no failure with
> GCC. Are you trying to compile the testsuite with say clang?
>
> Thanks,
> Andrew
>
>
> > ---
> >  gcc/testsuite/g++.old-deja/g++.mike/eh49.C | 2 +-
> >  gcc/testsuite/g++.old-deja/g++.mike/eh52.C | 2 +-
> >  2 files changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/gcc/testsuite/g++.old-deja/g++.mike/eh49.C
> b/gcc/testsuite/g++.old-deja/g++.mike/eh49.C
> > index 1380ca02d82..ac6e52baa20 100644
> > --- a/gcc/testsuite/g++.old-deja/g++.mike/eh49.C
> > +++ b/gcc/testsuite/g++.old-deja/g++.mike/eh49.C
> > @@ -1,5 +1,5 @@
> >  // { dg-do run { xfail sparc64-*-elf arm-*-pe } }
> > -// { dg-options "-fexceptions -O9" }
> > +// { dg-options "-fexceptions -O0" }
> >
> >  void main1() {
> >    throw 1;
> > diff --git a/gcc/testsuite/g++.old-deja/g++.mike/eh52.C
> b/gcc/testsuite/g++.old-deja/g++.mike/eh52.C
> > index 18b477c098b..7f44dba265a 100644
> > --- a/gcc/testsuite/g++.old-deja/g++.mike/eh52.C
> > +++ b/gcc/testsuite/g++.old-deja/g++.mike/eh52.C
> > @@ -1,5 +1,5 @@
> >  // { dg-do run  }
> > -// { dg-options "-fexceptions -O9" }
> > +// { dg-options "-fexceptions -O0" }
> >
> >  int promote_mode (int mode, int *punsignedp)
> >  {
> > --
> > 2.43.5
> >
>

Reply via email to