Does ggc-page.c require MMAP_DEV_ZERO and MMAP_ANON?

[H. J. Lu]

On Mon, Jun 23, 2003 at 08:11:50AM -0700, H. J. Lu wrote:
> I notice a strange gcc bug where I got bootstrap comparison failure
> when MMAP_DEV_ZERO and MMAP_ANON weren't available due to an ia64
> kernel bug. I got
> 
> Gcc from gcc-3_3-rhl-branch in cvs on Sun Jun 22 16:04:11 UTC 2003
> failed to bootstrap under Linux/ia64 on Itanium 2 running RHAS 2.1.
> The last known working one was checked out on Wed Jun 18 15:04:48 UTC
> 2003. The same gcc bootstrapped successfully on Itanium 1 running
> RHAS 2.1. The error message is
> 
> Bootstrap comparison failure!
> java/decl.o differs
> make[3]: *** [compare] Error 1
> make[3]: Leaving directory `/export/build/gnu/gcc-3.3-redhat/build-ia64-
> linux/gcc'
> make[2]: *** [bootstrap] Error 2
> make[2]: Leaving directory `/export/build/gnu/gcc-3.3-redhat/build-ia64-linux'
> Command exited with non-zero status 2
> 
> t turned out that it is a gcc bug exposed by a kernel bug. When I
> configured gcc, I got
> 
> checking for working mmap from /dev/zero... no
> checking for working mmap with MAP_ANON(YMOUS)... no
> 
> As the result, I had
> 
> /* #undef HAVE_MMAP_DEV_ZERO */
> /* #undef HAVE_MMAP_ANON */
> #define HAVE_MMAP_FILE 1
> 
> in auto-host.h. Apparently, ggc-page.c doesn't work too well with
> this configuration.
> 

It turns out that it may be the same bug as

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=3163

I reopened the bug and requested the fix to be moved to 3.3 branch.


H.J.