Re: [PATCH 0/4] bitpos expansion summary reloaded

[Jan Kratochvil <jan dot kratochvil at redhat dot com>]

Hi Siddhesh,

is anywhere FSF GDB GIT commit hash which all this work is for?


On Thu, 27 Sep 2012 15:30:53 +0200, Siddhesh Poyarekar wrote:
> I have also updated my repository of splint output reports here:
> 
> http://git.siddhesh.in/cgit.cgi/splint.git/

BTW I cannot "git clone" it.  Possibly using downloads of each file from this
web interface I see no logs or timestamps and there are too many files, most of
which are probably no longer worth checking.

So far I do not see how to review the patch as it is not annotated which
change is for which valid splint warning.  Comment from other maintainers is
welcome, annotating will take several more weeks of work.  Without the
annotation the only risk is that some changes are accidentally needless.
As "annotation" I call the /^x/ line below:
       if (info->onstack)
        {
-         int n = info->length;
+         LONGEST n = info->length;
xFIXED(Expand n): (xtensa-tdep.c:1886):  VARINIT(n):     (LONGEST to int)        [info->length]
          CORE_ADDR offset = sp + info->u.offset;


But it is also known that the changeset is not absolutely minimal, in some
cases the type was extended as it is very unclear the LONGEST type is in fact
not needed there and int would be enough.  I do not have such example now but
IIRC I just left some such extensions without comment as valid ones.



Thanks,
Jan