Dear Sirs,
I still cannot compile with gcc-3.0 from 08.04.
Yesterday I tried -ac5 (same problem reported
earlier) and using egcs-2.91.66 for _only_
the peoblematic files (sys.c exec.c binfmt_elf.c
and two others which I dont remember) but the
kernel could not boot.
Now I get:
gcc -D__KERNEL__ -I/glc/build/linux/include -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -fno-strict-aliasing -pipe -mpreferred-stack-boundary=2 -march=i686 -DEXPORT_SYMTAB -c signal.c
gcc -D__KERNEL__ -I/glc/build/linux/include -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -fno-strict-aliasing -pipe -mpreferred-stack-boundary=2 -march=i686 -DEXPORT_SYMTAB -c sys.c
sys.c: In function `sys_gethostname':
/glc/build/linux/include/asm/rwsem-xadd.h:153: inconsistent operand constraints in an `asm'
make[2]: *** [sys.o] Error 1
make[2]: Leaving directory `/glc/build/linux/kernel'
make[1]: *** [first_rule] Error 2
make[1]: Leaving directory `/glc/build/linux/kernel'
make: *** [_dir_kernel] Error 2
ecce:/usr/src/linux #
And ver_linux reports:
ecce:/usr/src/linux # sh scripts/ver_linux
If some fields are empty or look unusual you may have an old version.
Compare to the current minimal requirements in Documentation/Changes.
Linux ecce.homeip.net 2.4.3-ac3 #3 Sun Apr 8 22:06:09 /etc/localtime 2001 i686 unknown
Gnu C 3.0
Gnu make 3.77
binutils 2.10.0.33
util-linux 2.11a
mount 2.11a
modutils 2.4.5
e2fsprogs 1.19
reiserfsprogs 3.x.0j
PPP 2.4.1b2
Linux C Library x 1 root root 1248080 Apr 8 21:14 /lib/libc.so.6
Dynamic linker (ldd) 2.2
Procps 1.2.11
Net-tools 1.59
Kbd 0.99
Sh-utils 1.16
Modules Loaded
ecce:/usr/src/linux #
LIBC6 is originally from SuSE 6.2 but I updated with the one from SuSE 7.1
manually (I think I have both here?? will recompile soon glibc-2.2.2)
If any of you could help me, thanks in advance.
The problem _did_ _not_ exist with same gcc-3 and 2.4.3-ac3.
It doesn't vanish when I get the full patches, not the interdiffs.
-mirabilos
On Sat, Apr 14, 2001 at 09:09:00PM +0000, Thorsten Glaser Geuer wrote:
> Dear Sirs,
> I still cannot compile with gcc-3.0 from 08.04.
Yes ? Who said gcc-3.0 is suitable compiler ?
No doubt it some day will be the default compiler, but not yet.
For that matter, what "gcc-3.0" ??
Looking at http://gcc.gnu.org/ I see *NO* gcc-3.0 being
released, only CVS tag of 3.0 branch which aims for stability
and release.
> --
/Matti Aarnio
On Sun, Apr 15, 2001 at 01:03:35AM +0300, Matti Aarnio wrote:
> On Sat, Apr 14, 2001 at 09:09:00PM +0000, Thorsten Glaser Geuer wrote:
> > Dear Sirs,
> > I still cannot compile with gcc-3.0 from 08.04.
>
> Yes ? Who said gcc-3.0 is suitable compiler ?
>
> No doubt it some day will be the default compiler, but not yet.
Sorry. Who said it should not be tested? How else it could get
'default compiler'? If the gcc-3.0 would start giving errors
on some old code then it could be gcc bug. But this rwsem code
is couple of days old. It is good to let it through stricter
error checking, I guess. This rwsem is very in-flux code. eg.
2.4.4-pre2 did not compile. ac[56] with um-arch do not compile.
If nothing else convinces, then AFAIK both Linus and Alan
expressed interest in seeing reports of using newer gcc's.
Remember, gcc-3.0 is in freeze since Feb 12. (Ofcourse they
also suggested against using random CVS snapshot as default
compiler in distrubution).
--
marko
On Sun, 15 Apr 2001, Marko Kreen wrote:
> Sorry. Who said it should not be tested? How else it could get
> 'default compiler'? If the gcc-3.0 would start giving errors
> on some old code then it could be gcc bug. But this rwsem code
> is couple of days old. It is good to let it through stricter
> error checking, I guess. This rwsem is very in-flux code. eg.
> 2.4.4-pre2 did not compile. ac[56] with um-arch do not compile.
For what its worth, I got the same error on 2.4.3-ac5, using gcc 2.91.66.
I did seem messages fly by in on the list about a few in -ac5, but
haven't gone back to dig them out.
--
-- John E. Jasen ([email protected])
-- In theory, theory and practise are the same. In practise, they aren't.