2.6.9 won't compile
Dear all,
I downloaded version 2.6.9 of the kernel from kernel.org
I went throught menuconfig, importing the configuration of my current 2.6.15-28-686.
I then tried to compile it but the compilation exits with an error (below).
I also tried to do make mrproper, got rid of the .config file, made menuconfig again without importing anything, and when compiling it failed again.
The reason for me to use this kernel is because I have a VIA VT82C 693A/598/686 chipset and I suffer from perf problems (which google says only happen since 2.6.10).
Might I request to be personally CC'ed in the answers/comments posted in response to this post ?
Many thanks in advance for your comments and help!
With kind regards,
Pedro Mullor
CHK include/linux/version.h
make[1]: `arch/i386/kernel/asm-offsets.s' is up to date.
CHK include/linux/compile.h
CC arch/i386/kernel/process.oarch/i386/kernel/process.c: In function ‘show_regs’:arch/i386/kernel/process.c:252: warning: pointer targets in passing argument 2 of ‘show_trace’ differ in signedness {standard input}: Assembler messages:
{standard input}:813: Error: suffix or operands invalid for mov' {standard input}:814: Error: suffix or operands invalid for mov' {standard input}:1040: Error: suffix or operands invalid for mov' {standard input}:1041: Error: suffix or operands invalid for mov' {standard input}:1121: Error: suffix or operands invalid for mov' {standard input}:1122: Error: suffix or operands invalid for mov' {standard input}:1226: Error: suffix or operands invalid for mov' {standard input}:1227: Error: suffix or operands invalid for mov' {standard input}:1290: Error: suffix or operands invalid for mov' {standard input}:1302: Error: suffix or operands invalid for mov' make[1]: *** [arch/i386/kernel/process.o] Error 1
make: *** [arch/i386/kernel] Error 2
Quoted 13 lines · Pedro Mullor Jiménez
Dear all,
I downloaded version 2.6.9 of the kernel from kernel.org
I went throught menuconfig, importing the configuration of my current 2.6.15-28-686.I then tried to compile it but the compilation exits with an error (below).
I also tried to do make mrproper, got rid of the .config file, made menuconfig again without importing anything, and when compiling it failed again.
Quoted 7 lines
The reason for me to use this kernel is because I have a VIA VT82C 693A/598/686 chipset and I suffer from perf problems (which google says only happen since 2.6.10).
Might I request to be personally CC'ed in the answers/comments posted in response to this post ?
I don't think 2.6.9 was gcc-4.x compatible, what version of gcc are you using ?
regards,
Benoit
Quoted 9 lines · Pedro Mullor Jiménez
Dear all,
I downloaded version 2.6.9 of the kernel from kernel.org
I went throught menuconfig, importing the configuration of my current 2.6.15-28-686.I then tried to compile it but the compilation exits with an error (below).
Quoted 7 lines
I also tried to do make mrproper, got rid of the .config file, made menuconfig again without importing anything, and when compiling it failed again.
The reason for me to use this kernel is because I have a VIA VT82C 693A/598/686 chipset and I suffer from perf problems (which google says only happen since 2.6.10).
Can you give a pointer or description of this problem?
The preferred solution for you would not be to run an ancient kernel with tons of known security holes but to get this problem fixed.
Quoted 21 lines
Might I request to be personally CC'ed in the answers/comments posted in response to this post ?
Many thanks in advance for your comments and help!
With kind regards,
Pedro MullorCHK
include/linux/version.h
make[1]: `arch/i386/kernel/asm-offsets.s' is up to date.
CHKinclude/linux/compile.h
CCarch/i386/kernel/process.oarch/i386/kernel/process.c: In function ‘show_regs’:arch/i386/kernel/process.c:252: warning: pointer targets in passing argument 2 of ‘show_trace’ differ in signedness {standard input}: Assembler messages:
{standard input}:813: Error: suffix or operands invalid formov' {standard input}:814: Error: suffix or operands invalid formov' {standard input}:1040: Error: suffix or operands invalid formov' {standard input}:1041: Error: suffix or operands invalid formov'
Quoted 8 lines
{standard input}:1121: Error: suffix or operands invalid for
mov' {standard input}:1122: Error: suffix or operands invalid formov' {standard input}:1226: Error: suffix or operands invalid formov' {standard input}:1227: Error: suffix or operands invalid formov' {standard input}:1290: Error: suffix or operands invalid formov' {standard input}:1302: Error: suffix or operands invalid formov' make[1]: *** [arch/i386/kernel/process.o] Error 1
make: *** [arch/i386/kernel] Error 2
Older kernels use an assembler syntax no longer supported by recent binutils.
A patch is at [1].
cu
Adrian
[1] http://www.kernel.org/pub/linux/devel/binutils/linux-2.6-seg-5.patch
Signature · 4 lines
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon SeedQuoted 9 lines
[Benoit Boissinot]
I don't think 2.6.9 was gcc-4.x compatible, what version of gcc are you using ?[Robert Hancock]
Your compiler is probably too new to be able to compile 2.6.9. You'd likely be better off trying the latest kernels and making a detailed report of the problems you're having with it..
Quoted 5 lines
[Adrian Bunk]
Can you give a pointer or description of this problem?The preferred solution for you would not be to run an ancient kernel with tons of known security holes but to get this problem fixed.
Dear Benoît, Robert and Adrian, dear all,
first of all, thanks for your time and answers.
I'm on Ubuntu 6.06 LTS. The gcc version used is 4.0.3.
As you pointed out, this may well be the reason why the compilation fails. I'm sorry, but it's been a couple of years since I compiled my last kernel so I might be not as used to as in the time :o)
As in regards to the problem I'm having, I recall having read an article on which someone gave his oppinion about why the latest crop of kernels with the new stable/unstable/patch organization were so ABC while the previous system was so XYZ.
One of the issues argumented was one about old systems with Via chipsets (among them the VT82CXXX) that worked fine on 2.6.9 but that were fine no more since 2.6.10 (performance degradation issue after heavy/ sustained IO loads).
I am truly sorry of not being able to find the exact article again on google, but then I might have read it on slashdot or on another site.
*The/My Issue* is that, after a sustained load of hard-drive IO operations, the machine I have gets "clogged".
The symptoms are :
- when doing FTP (my machine being the server) I begin at 1.4 MB/s. This speed gradually decrements, and after 10 mins I'm doing 500 or 400 KB/s (I'm aware that the actual speed of a FTP transfer should be the mean speed during the transmission of a 10 MB file, but this falls a bit farther IMHO).
- when I leave Azureus (Bittorrent client) running for 10+ mins, *and I have not been using my pc in the meantime* (no mouse, no keyboard usage), the (usb) mouse is "lagged", "dizzy", dragging, the pointer reacts slowly to the movement of the mouse.
- the same happens after I've been doing an FTP transfer for 10 mins+
- also when on Azureus, the sound system plays several, usually one-off (like Gaim "online" sounds), sounds endlessly (or repeats the playback for a min of 3-5 times)
When this happens, I have tried to go on console mode, stop all possible jobs & services, but I end with only the really minimal services + the kernel processes and the system still behaves the same (I tried reproducing a mp3 file on the console and the sound system is still not behaving correctly (I try to stop the program with Ctrl-C but it won't respond and then I have to kill it).
I am sorry I'm not used to issues like this one so I'm not sure where else to look. I wanted to try 2.6.9 to see if the system behaved better/correctly, because I only have had this issue with the two last Linux flavours I installed (SUSE 10 and Ubuntu 6.06), when on Debian (the version before 3.0) it was fine.
Then again, this is not really the scientific method of isolating the variables in the equation and the machine is now 5 years old, so I wonder if there is not something else that I could have overlooked, but so far this is the only issue I've got and, while being a veteran of several Debian Potato + X11 installations, I'm frankly out of options/ideas here ;o)
At the same time, it still gets the job done beautifully so I don't really have a reason to change it.
If you've read this until here, I really thank you for your time. Any ideas will be greately appreciated. If I can help with anything, please let me know. If it's to no use, then I think I'll still survive :o)
Have a nice weekend!
Pedro Mullor
P.S. Please find the specs of my machine :
2x Intel PIII 933, 512 RAM
Asus Cuv4x-d (Via VT82CXXX, AGP : VIA Apollo Pro 133) Creative Labs SB Audigy (rev 03)
USB logitech Optical Mouse
Logitech PS2 Keyboard
hda: ST340824A, ATA DISK drive
hdb: QUANTUM FIREBALLlct10 10, ATA DISK drive
hdc: LITE-ON DVDRW SHW-1635S, ATAPI CD/DVD-ROM drive
hdd: LG DVD-ROM DRD-8160B, ATAPI CD/DVD-ROM drive
eth0: RealTek RTL8139
GeForce2 MX/MX 400] (rev b2)