On Thu, 14 Aug 2014, Jakub Jelinek wrote:
> Hi!
>
> Apparently using numbers without dots in 4.9/5 Regression has the
> disadvantage that some bugs make it to the query even when they should not,
> like
> [4.8 Regression] ICE in set_address_disp, at rtlanal.c:5537
> or
> avx512f-ceil-sfix-vec-2.c
Hi!
Apparently using numbers without dots in 4.9/5 Regression has the
disadvantage that some bugs make it to the query even when they should not,
like
[4.8 Regression] ICE in set_address_disp, at rtlanal.c:5537
or
avx512f-ceil-sfix-vec-2.c and avx512f-floor-sfix-vec-2.c FAIL on Solaris9/x86
Matchi
On Wed, 13 Aug 2014, Jakub Jelinek wrote:
> On Wed, Aug 13, 2014 at 02:21:29PM +0200, Richard Biener wrote:
> >
> > The release management team has aggreed to go forward with the
> > change to bump the major with each further release.
> >
> > The following aims to document the details of the ver
On Wed, Aug 13, 2014 at 02:21:29PM +0200, Richard Biener wrote:
>
> The release management team has aggreed to go forward with the
> change to bump the major with each further release.
>
> The following aims to document the details of the versioning scheme we
> intend to use for GCC 5 and up.
>
The release management team has aggreed to go forward with the
change to bump the major with each further release.
The following aims to document the details of the versioning scheme we
intend to use for GCC 5 and up.
Summary in non-html: Development of GCC 5 will happen as
GCC 5.0.0 (experiment
