Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Problem compiling xorg-server{-snap} on recent -CURRENT

0 views
Skip to first unread message

Jonathan Fosburgh

unread,
May 20, 2007, 11:54:44 PM5/20/07
to
I am trying to upgrade my Xorg installation to 7.2. I believe everything has
upgraded successfully except for xorg-server. While trying to compile either
xorg-server port, the compilation process hangs in hw/xfree86/scanpci at
xf86ScanPci.c. The compiler just sits there, rapidly chewing up about 800MB
of memory and doesn't proceed. I am running -CURRENT from Saturday morning
CDT (approx 10AM), so this is post symver and post gcc-4.2. I would imagine
this combination of software wasn't thoroughly tested. My system is an
Athlon64 3200+ (running i386). Uname output is:

FreeBSD asgard.fosburgh.org 7.0-CURRENT FreeBSD 7.0-CURRENT #3: Sat May 19
12:16:39 CDT 2007 root@:/usr/obj/usr/src/sys/vmbsd i386

This is a very recently built system and quite clean in terms of what software
is installed, prior to upgrading to 7.2 I only had about 350 packages
installed, and the bulk of that was kde and support programs and libraries.

Has anyone managed to compile xorg-server on a very recent -CURRENT?
_______________________________________________
freebsd...@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-curre...@freebsd.org"

--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-...@muc.de

Wes Morgan

unread,
May 21, 2007, 1:13:44 AM5/21/07
to
On Sun, 20 May 2007, Jonathan Fosburgh wrote:

> I am trying to upgrade my Xorg installation to 7.2. I believe everything has
> upgraded successfully except for xorg-server. While trying to compile either
> xorg-server port, the compilation process hangs in hw/xfree86/scanpci at
> xf86ScanPci.c. The compiler just sits there, rapidly chewing up about 800MB
> of memory and doesn't proceed. I am running -CURRENT from Saturday morning
> CDT (approx 10AM), so this is post symver and post gcc-4.2. I would imagine
> this combination of software wasn't thoroughly tested. My system is an
> Athlon64 3200+ (running i386). Uname output is:
>
> FreeBSD asgard.fosburgh.org 7.0-CURRENT FreeBSD 7.0-CURRENT #3: Sat May 19
> 12:16:39 CDT 2007 root@:/usr/obj/usr/src/sys/vmbsd i386
>
> This is a very recently built system and quite clean in terms of what software
> is installed, prior to upgrading to 7.2 I only had about 350 packages
> installed, and the bulk of that was kde and support programs and libraries.
>
> Has anyone managed to compile xorg-server on a very recent -CURRENT?

Yes, just give it time. I thought mine had hung on this file as well, but
it eventually compiled it.


--
This .signature sanitized for your protection

Michiel Boland

unread,
May 21, 2007, 2:26:15 AM5/21/07
to
My build of xorg-server died. The box ran out of swap space. I have 512M
RAM + 1G swap. Someone please tell me this is a glitch in the new gcc. I
don't want to add ram just to be able to compile a simple program. :)

cc -DHAVE_CONFIG_H -I. -I. -I../../../include -I../../../include
-I../../../include -I../../../include -I../../../include
-I../../../include -I../../../include -I../../../include
-I../../../hw/xfree86 -I../../../hw/xfree86/include
-I../../../hw/xfree86/common -I../../../hw/xfree86/os-support
-I../../../hw/xfree86/os-support/bus -I../../../os -DHAVE_XORG_CONFIG_H
-Wall -Wpointer-arith -Wstrict-prototypes -Wmissing-prototypes
-Wmissing-declarations -Wnested-externs -fno-strict-aliasing -D_BSD_SOURCE
-DHAS_FCHOWN -DHAS_STICKY_DIR_BIT -D_THREAD_SAFE -I/usr/local/include
-I/usr/local/include/freetype2 -I../../../include -I../../../include
-I../../../Xext -I../../../composite -I../../../damageext
-I../../../xfixes -I../../../Xi -I../../../mi -I../../../miext/shadow
-I../../../miext/damage -I../../../render -I../../../randr -I../../../fb
-O2 -fno-strict-aliasing -pipe -MT xf86ScanPci.lo -MD -MP -MF
.deps/xf86ScanPci.Tpo -c xf86ScanPci.c -fPIC -DPIC -o .libs/xf86ScanPci.o
cc: Internal error: Killed: 9 (program cc1)
Please submit a full bug report.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.
*** Error code 1

FreeBSD aja.boland.org 7.0-CURRENT FreeBSD 7.0-CURRENT #0: Sun May 20
19:26:27 CEST 2007 ro...@aja.boland.org:/usr/obj/usr/src/sys/AJA amd64

$ gcc -v
Using built-in specs.
Target: amd64-undermydesk-freebsd
Configured with: FreeBSD/amd64 system compiler
Thread model: posix
gcc version 4.2.0 20070514 [FreeBSD]

Kris Kennaway

unread,
May 21, 2007, 4:26:57 AM5/21/07
to
On Sun, May 20, 2007 at 10:54:44PM -0500, Jonathan Fosburgh wrote:
> I am trying to upgrade my Xorg installation to 7.2. I believe everything has
> upgraded successfully except for xorg-server. While trying to compile either
> xorg-server port, the compilation process hangs in hw/xfree86/scanpci at
> xf86ScanPci.c. The compiler just sits there, rapidly chewing up about 800MB
> of memory and doesn't proceed. I am running -CURRENT from Saturday morning
> CDT (approx 10AM), so this is post symver and post gcc-4.2. I would imagine
> this combination of software wasn't thoroughly tested. My system is an
> Athlon64 3200+ (running i386). Uname output is:
>
> FreeBSD asgard.fosburgh.org 7.0-CURRENT FreeBSD 7.0-CURRENT #3: Sat May 19
> 12:16:39 CDT 2007 root@:/usr/obj/usr/src/sys/vmbsd i386
>
> This is a very recently built system and quite clean in terms of what software
> is installed, prior to upgrading to 7.2 I only had about 350 packages
> installed, and the bulk of that was kde and support programs and libraries.
>
> Has anyone managed to compile xorg-server on a very recent -CURRENT?

It worked the last time I tried. That was with gcc 4.2 but without
symver. I cannot confirm how much memory it may have required though.

Kris

Dag-Erling Smørgrav

unread,
May 21, 2007, 11:12:42 AM5/21/07
to
Michiel Boland <mic...@boland.org> writes:
> My build of xorg-server died. The box ran out of swap space. I have
> 512M RAM + 1G swap. Someone please tell me this is a glitch in the new
> gcc. I don't want to add ram just to be able to compile a simple
> program. :)

The quick fix is to build at a lower optimization level. Advanced
optimizations can be very memory-consuming, especially when compiling
unusually large source files, or source files which contain unusually
large functions.

DES
--
Dag-Erling Smørgrav - d...@des.no

Michiel Boland

unread,
May 21, 2007, 2:28:38 PM5/21/07
to
>> My build of xorg-server died. The box ran out of swap space. I have
>> 512M RAM + 1G swap. Someone please tell me this is a glitch in the new
>> gcc. I don't want to add ram just to be able to compile a simple
>> program. :)
>
> The quick fix is to build at a lower optimization level. Advanced
> optimizations can be very memory-consuming, especially when compiling
> unusually large source files, or source files which contain unusually
> large functions.

Ok, that appears to do the trick. Compile without any optimization at all.
But I wonder: what is the point of a huge object file with 10000 or so
symbols, of which I most likely will use only one or two? I thought the
whole point of the new xorg was the modularity. (I'm afraid I'm getting a
bit off-topic now, so I will quickly stop my ranting here. :)

Alex Dupre

unread,
May 22, 2007, 2:08:44 AM5/22/07
to
Dag-Erling Smørgrav wrote:
>> My build of xorg-server died. The box ran out of swap space. I have
>> 512M RAM + 1G swap. Someone please tell me this is a glitch in the new
>> gcc. I don't want to add ram just to be able to compile a simple
>> program. :)
>
> The quick fix is to build at a lower optimization level. Advanced
> optimizations can be very memory-consuming, especially when compiling
> unusually large source files, or source files which contain unusually
> large functions.

I confirm the issue and the fix with latest -CURRENT (gcc 4.2 and symbol
versioning). I have less RAM and swap than Michiel, and 800MB of total
memory are not enough to build xf86scanpci.c with default CFLAGS (and
even with less aggressive -O). I had to remove completely the
optimization flag, but only for this single source file, all the rest
can be compiled with -O2.

--
Alex Dupre

Peter Johnson

unread,
May 26, 2007, 7:45:36 PM5/26/07
to
Indeed this is a known bug with gcc 4.2.0:

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30052

There's a patch there but I don't know if it's been committed yet to
gcc.

--
Peter Johnson

Garrett Cooper

unread,
May 26, 2007, 9:36:23 PM5/26/07
to
Peter Johnson wrote:
> Indeed this is a known bug with gcc 4.2.0:
>
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30052
>
> There's a patch there but I don't know if it's been committed yet to
> gcc.

It says that it will be committed to 4.2.1, so technically it hasn't
been done yet I think..

-Garrett

0 new messages