This is the mail archive of the
newlib@sourceware.org
mailing list for the newlib project.
Re: [PATCH] Update newlib so that it passes libc++'s tests
- From: JF Bastien <jfb at chromium dot org>
- To: Craig Howland <howland at lgsinnovations dot com>
- Cc: "newlib at sourceware dot org" <newlib at sourceware dot org>, Jeff Johnston <jjohnstn at redhat dot com>
- Date: Fri, 20 Dec 2013 14:31:46 -0500
- Subject: Re: [PATCH] Update newlib so that it passes libc++'s tests
- Authentication-results: sourceware.org; auth=none
- References: <CABdywOcnSpU=r5NGDDzhea4gxALh8LRL4A9vRY31wFjLhtF5zA at mail dot gmail dot com> <Pine dot LNX dot 4 dot 64 dot 1312170153500 dot 29265 at digraph dot polyomino dot org dot uk> <CABdywOdE5rkMLOyK3NrM+k+OrpRLkbjPp5oddqDz9vv5wKkd0A at mail dot gmail dot com> <52B220A8 dot 4010901 at LGSInnovations dot com> <CABdywOcPo-_RSjBiWeZNeMaB5Aih23OT7u3qXL896NfcqEu6bg at mail dot gmail dot com> <156240854 dot 42718969 dot 1387564988624 dot JavaMail dot root at redhat dot com> <52B498C3 dot 8000108 at LGSInnovations dot com>
>> FWIW, I'm willing to wait if you guys are close to consensus on patch.
>> Otherwise, my intention
>> was to make the snapshot today.
>>
>> My opinion is, if you can make it as good or better than it was and you
>> can achieve the libc++
>> criteria, then fix the exceptions later. It is a minor issue to add a
>> macro in sys/config.h for
>> any platforms that don't have their compiler setting __WCHAR_MIN__ /
>> __WCHAR_MAX__ and don't want
>> the logic below.
Either is fine with me, let me know what to do.
> No, I was referring to the breakage of making the assumption when you don't
> know. (I was assuming the fix that Joseph pointed out.) The idea was to
> apply the patch now only to wchar.h--an improvement over what is there now,
> but still with the issue of a possibly-incorrect fallback value--but to not
> make the addition to stdint.h. At a future time, do the config mess and
> then add to stdint.h. This is just a thought to avoid introducing an issue
> to stdint.h.
> If this does not achieve the goal of fixing the libc++ test, then I tend to
> agree with Jeff as to it's being an improvement worth doing, even if it
> could use a (complicated) cleanup.
I see. So if I modified my patch to only change the WCHAR_* value for
wchar.h and not stdint.h then you think we'd be good to go?
I also think a partial fix that's incrementally more correct is better
than none at all, but I also want to newlib folks to agree that my
change is indeed good!