This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: 3.4 regressions: are 2.95 regressions still actual
On Jan 24, 2004, at 17:01, Marc Espie wrote:
In article
<Pine.LNX.4.44.0401120921120.21333-100000@acapulco.ices.utexas.edu>
you write:
The whole point of this is: moving milestones, or closing PRs as not
important is a _political_ decision, not a technical one. I don't
want one
of our bugzilla people to just go around and do so at will. For this
job,
we have the RM, which is a political position. We can assume that he
got
his position by common consensus, so he's got a political mandate to
do
so. None of us bugmasters has this.
Shudder.
Okay, tell me that's not true. I misunderstood, right ?
Yes, closing as will not fix for a certain release like 3.3.3 because
it is not being worked
on getting fixed for that release but is already fixed for 3.4.
Did I get this right ?
No you did not, see above.
There are PRs getting *closed* that correspond to *actual* problems
that
*do not get fixed*, but get classified as *not important*.
No the bugs which are being closed are bugs which are fixed already on
the head (or a future
major release).
I can understand marking a bug as low priority, or postponed until
after
the next release, but to close a PR as `not important' ???
But all of those bug being closed are already fixed for the next major
release.
We have only closed bugs where it is fixed or the problem can no longer
be a problem.
For an example when we remove support for dwarf-1 should we keep open
the bugs dealing
with dwarf-1 for every even though it is removed from GCC's sources or
a target which is
no longer in GCC's sources?
Do you even think the person who reported that bug in the first place
will even bother to file a bug-report the next time ???
Well some cases like -fssa (which has been removed for 3.4) where it is
mentioned in the
documentation as experimental and known to be bugy, should we keep the
bug open even though
no one is working on -fssa (SSA for RTL) at all.
Thanks,
Andrew Pinski
one of the bug masters