This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] Re: Bootstrap broken on powerpc-apple-darwin

Andrew Pinski <pinskia@physics.uc.edu> writes:

> The follow patch from Ian gets me further then a failure in libjava
> which is PR 14580.
> 
> I applied the follow patch also because it worked for me and RTH oked
> it in IRC.

Hmmm, I do think it should be

      ret = mmap_result != (void *) MAP_FAILED;

      /* Sanity check for broken MAP_FIXED.  */
      if (ret && mmap_result != addr)
	abort ();

rather than this:

      ret = mmap_result == (void *) MAP_FAILED;

      /* Sanity check for broken MAP_FIXED.  */
      if (!ret && mmap_result != addr)
	abort ();

The function is supposed to return 1 if the data is fully loaded.  If
the mmap succeeds, then the data is fully loaded.

Returning 0 will cause gt_pch_restore() to read in the data, which is
not necessary.

Ian
Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]