Problem while updating to Kernel 2.2.14
Hey Guys, maybe somebody can help me. After the security announce of SuSE I updated the kernel sources for compiling to kernel 2.2.14. I only updated the package "lx_suse" because it was the only one which was installed before. While compiling I get following error message: --- snip --- time.c: In function 'do_gettimeofday' time.c: 259: fixed or forbidden register 0 (ax) was spilled for class AREG This may be due to a compiler bug or to impossible asm statements or clauses cpp: output pipe has been closed make[1]: *** [time.o] Error 1 make: leaving directory ... make: *** [_dir_arch/i386/kernel] Error 2 --- snap --- The kernel was not finished. I found two differnt files "time.c": one in /usr/src/linux/kernel and one in /usr/src/linux/arch/i386/kernel Do you know what's wrong ? Bye Holger
Hey Guys,
maybe somebody can help me.
After the security announce of SuSE I updated the kernel sources for compiling to kernel 2.2.14. I only updated the package "lx_suse" because it was the only one which was installed before.
While compiling I get following error message:
--- snip ---
time.c: In function 'do_gettimeofday' time.c: 259: fixed or forbidden register 0 (ax) was spilled for class AREG This may be due to a compiler bug or to impossible asm statements or clauses cpp: output pipe has been closed make[1]: *** [time.o] Error 1 make: leaving directory ... make: *** [_dir_arch/i386/kernel] Error 2
--- snap ---
The kernel was not finished.
I found two differnt files "time.c": one in /usr/src/linux/kernel and one in /usr/src/linux/arch/i386/kernel
Do you know what's wrong ?
On Fri, 19 May 2000 Holger-Schmidt@arcor.net wrote: the problematic file is /usr/src/linux/arch/i386/kernel/time.c which triggers a bug in egcs-1.1.2 and can only be compiled with gcc-2.95.2 or gcc-2.7.2.3. AFAIR removing the inline in linux/include/linux/time.h in the definition of normalize_timeval should work for egcs-1.1.2.
Bye Holger
-- with kind regards (mit freundlichem Grinsen), Ruediger Oertel (ro@suse.de) ---------------------------------------------------------- does "DONT PANIC" give a hint ?
I am also experiencing exact the same problem as Holger when trying to compile a pentium/SCSI-optimized kernel. I updated the package lx_suse and k.deflt on SuSe 6.3. This is quite annoying. Why does SuSE distribute kernel updates that don't even build at runtime? Anyhow, could somebody please tell me which line exactly to remove in linux/include/linux/time.h since I'm not too familiar with the C language? Thanks a lot. Best regards, Philippe Wiede pw@megapublic.com Ruediger Oertel wrote:
On Fri, 19 May 2000 Holger-Schmidt@arcor.net wrote:
.... After the security announce of SuSE I updated the kernel sources for compiling to kernel 2.2.14. I only updated the package "lx_suse" because it was the only one which was installed before.
While compiling I get following error message:
--- snip ---
time.c: In function 'do_gettimeofday' time.c: 259: fixed or forbidden register 0 (ax) was spilled for class AREG This may be due to a compiler bug or to impossible asm statements or clauses cpp: output pipe has been closed make[1]: *** [time.o] Error 1 make: leaving directory ... make: *** [_dir_arch/i386/kernel] Error 2
--- snap ---
The kernel was not finished.
I found two differnt files "time.c": one in /usr/src/linux/kernel and one in /usr/src/linux/arch/i386/kernel
Do you know what's wrong ?
the problematic file is /usr/src/linux/arch/i386/kernel/time.c which triggers a bug in egcs-1.1.2 and can only be compiled with gcc-2.95.2 or gcc-2.7.2.3.
AFAIR removing the inline in linux/include/linux/time.h in the definition of normalize_timeval should work for egcs-1.1.2. -- with kind regards (mit freundlichem Grinsen), Ruediger Oertel (ro@suse.de)
On Fri, 19 May 2000, Megapublic Intranet wrote:
I am also experiencing exact the same problem as Holger when trying to compile a pentium/SCSI-optimized kernel. I updated the package lx_suse and k.deflt on SuSe 6.3. This is quite annoying. Why does SuSE distribute kernel updates that don't even build at runtime? It builds - on 6.4 ... as stated this is a bug in the 6.3 compiler.
Anyhow, could somebody please tell me which line exactly to remove in linux/include/linux/time.h since I'm not too familiar with the C language? Thanks a lot. I've attached a patch that should work.
-- with kind regards (mit freundlichem Grinsen), Ruediger Oertel (ro@suse.de) ---------------------------------------------------------- does "DONT PANIC" give a hint ?
Dear All On Fri, 19 May 2000, you wrote:
On Fri, 19 May 2000 Holger-Schmidt@arcor.net wrote:
After the security announce of SuSE I updated the kernel sources for compiling to kernel 2.2.14. I only updated the package "lx_suse" because it was the only one which was installed before.
I'm sorry that I sent this to this list but I thought that my own problem might raise one or two security issues that I'm not too sure about. Last night I downloaded....... ftp://ftp.suse.com/pub/suse/i386/update/6.4/kernel/k_deflt.rpm be6ee213f0cafd4dac5c51a2a8d100f0 ftp://ftp.suse.com/pub/suse/i386/update/6.4/kernel/k_eide.rpm b900eb9f47c94df5cc15721e5f96a58e ftp://ftp.suse.com/pub/suse/i386/update/6.4/kernel/k_i386.rpm 37deca6ee856c3242a13c2a24f32fc7f ftp://ftp.suse.com/pub/suse/i386/update/6.4/d1/lx_suse-2.2.14.SuSE-24.i386.rpm and installed them into my own PC which I'm using to send this mail. After installation and reboot I got the following error message..... ran out of input data --System halted This was after the word LILO appeared and the system tried to load from there. How do I get out of this ?? This morning I started to sort out my machine by starting the rescue system. I found that the option "boot installed system" allowed me to get into my system and I am now using it to send this mail to you. Perhaps I can just remove the k_deflt.rpm k_eide.rpm and the k_i386.rpm. That will solve the problem ???? Thank you -- Richard Sheffield UK
Richard wrote:
After installation and reboot I got the following error message.....
ran out of input data
--System halted ... sort out my machine by starting the rescue system. I found that the option "boot installed system" allowed me to get into my system and I am now using it to send this mail to you. Perhaps I can just remove the k_deflt.rpm k_eide.rpm and the k_i386.rpm. That will solve the problem ????
The problem is likely caused by a erroneous lilo installation. After updating the packages you should have startet Yast and select the kernel to boot. After that, LILO would have been run automatically which was neccessary, because LILO cannot go on with a new kernel without beeing run again. So, if you can get into your system, you might just type (as root) yast --mask kernel --autoexit which brings you to the kernel-menue automatically. Select the kernel to boot. Afterwards Yast will ask you if it should run LILO. Answer yes, and the problem should be gone... In general, the "ran out of input data" often disappears by just running lilo. Best, Frank -- Dipl.-Inform. Frank Steiner mailto:fst@informatik.uni-kiel.de Lehrstuhl f. Programmiersprachen mailto:fsteiner@web.de CAU Kiel, Olshausenstraße 40 Phone: +49 431 880-7265, Fax: -7613 D-24098 Kiel, Germany http://www.informatik.uni-kiel.de/~fst/ *** Windows 2000 - Die drei grossen G - Gesehen. Gelacht! Gelöscht :) ***
Last night I downloaded....... and installed them into my own PC which I'm using to send this mail. After installation and reboot I got the following error message..... ran out of input data --System halted you have to run lilo after installation. lilo updates the boot record to load the new kernel. if you don't run lilo there is no kernel to boot for
On Fri, 19 May 2000, Richard wrote: lilo. just boot the installed system with the CD and type lilo that's it! regards Markus Gaugusch
On Fri, 19 May 2000, Richard wrote:
--System halted
This was after the word LILO appeared and the system tried to load from there. How do I get out of this ?? This morning I started to sort out my machine by starting the rescue system. I found that the option "boot installed system" allowed me to get into my system and I am now using it to send this mail to you. Perhaps I can just remove the k_deflt.rpm k_eide.rpm and the k_i386.rpm. That will solve the problem ????
Perhaps you just need to rerun lilo. regards mw -- Markus Werner mwr@hud.de Hönigsberg & Düvel Datentechnik GmbH BZO Haus 20, 38840 Wolfsburg Tel: +49 173 927 361 5 http://www.hud.de
On a suse 6.4 machine messages got rotated at midnight but no new messages file was created. the lasttime syslogd reload did it, why do i need that. the files are on a raid5 md. -- --- Engelbert Gruber --- SSG Fintl,Gruber,Lassnig A6140 Telfs Untermarkt 9 Tel. ++43-5262-64727 ---
participants (8)
-
engelbert.gruber＠ssg.co.at -
Frank Steiner -
Holger-Schmidt＠arcor.net -
Markus Gaugusch -
Markus Werner -
Megapublic Intranet -
Richard -
Ruediger Oertel