Subject: GCC code generation...? - msg#00002

Previous Message by Date: click to view message preview

Re: KA53 asc/clock bug

On Sun, Jul 29, 2007 at 11:05:22AM +0200, Hans Rosenfeld wrote: > In asc_vsbus_attach() the SCSI id for the controller is read from some > clock chip NVRAM. This causes a segv in kernel mode for the KA53 which > seems to lack the clock chip. OpenBSD uses a hardcoded SCSI id 6 for all > systems but KA46, but since the existing code works with KA48 and KA49 > too I made the hardcoded value a special case for the KA53. I just noticed that the patch doesn't work as intended for kernel configurations which exclude KA46, KA48 and KA49. The clk_page is still referenced, but since it is not defined anywhere the kernel cannot be linked. I made a new patch that references clk_page only if VAX46, VAX48 or VAX49 are defined. -- %SYSTEM-F-ANARCHISM, The operating system has been overthrown asc_vsbus.diff Description: Text document

Next Message by Date: click to view message preview

System hung...

Running 4.99.27 (cross-built kernel from another architecture) kernel, with 4.99.20 userspace. I'm trying to build the whole system from source, but partway through the build, the whole systems hangs. I have the machine in this state right now, and can try whatever people would suggest. I was running build.sh -x build kernel=GENERIC in one terminal window (connected through ssh). The output was directed to a file, and I was running a tail -f on that file. Another window was running top. That's it. And NIS is running on the machine, and /usr/src is mounted through nfs. The build log file is also written at /usr/src. As I said, the system is hung. The last update in top looks like this: -----------------top-------------- load averages: 1.12, 1.09, 1.08 up 0 days, 6:10 18:30:47 44 processes: 1 runnable, 42 sleeping, 1 on processor CPU states: 78.3% user, 0.0% nice, 16.7% system, 0.0% interrupt, 4.9% idle Memory: 50M Act, 33M Inact, 3336K Wired, 5180K Exec, 57M File, 19M Free Swap: 128M Total, 128M Free PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND 24376 root 59 0 3936K 4800K RUN 0:05 91.09% 23.63% ld 10524 bqt 28 0 288K 1044K CPU 1:53 0.00% 0.00% top ------------------------------------ And the tail -f got to this: -----------------tail-------------- cc -DHAVE_CONFIG_H -I. -I/usr/src/tools/binutils/../../gnu/dist/binutils/ld -I. -D_GNU_SOURCE -I. -I/usr/src/tools/binutils/../../gnu/dist/binutils/ld -I../bfd -I/usr/src/tools/binutils/../../gnu/dist/binutils/ld/../bfd -I/usr/src/tools/binutils/../../gnu/dist/binutils/ld/../include -I/usr/src/tools/binutils/../../gnu/dist/binutils/ld/../intl -I../intl -O -DLOCALEDIR="\"/usr/src/obj/tooldir.NetBSD-4.99.27-vax/share/locale\"" -W -Wall -Wstrict-prototypes -Wmissing-prototypes -O -c evaxnbsd.c /bin/sh ./libtool --mode=link cc -W -Wall -Wstrict-prototypes -Wmissing-prototypes -O -o ld-new ldgram.o ldlex.o lexsup.o ldlang.o mri.o ldctor.o ldmain.o ldwrite.o ldexp.o ldemul.o ldver.o ldmisc.o ldfile.o ldcref.o eelf32vax.o evaxnbsd.o ../bfd/libbfd.la ../libiberty/libiberty.a mkdir .libs cc -W -Wall -Wstrict-prototypes -Wmissing-prototypes -O -o ld-new ldgram.o ldlex.o lexsup.o ldlang.o mri.o ldctor.o ldmain.o ldwrite.o ldexp.o ldemul.o ldver.o ldmisc.o ldfile.o ldcref.o eelf32vax.o evaxnbsd.o ../bfd/.libs/libbfd.a ../libiberty/libiberty.a NOT REBUILDING /usr/src/tools/binutils/../../gnu/dist/binutils/ld/ldver.texi NOT REBUILDING /usr/src/tools/binutils/../../gnu/dist/binutils/ld/ld.1 -------------------------------------- No response from either terminal, nor the console. Breaking into ddb on the console tells me this: -----------------console-------------- Starting cron. Wed Aug 8 12:21:05 CEST 2007 NetBSD/vax (Gnat.BQTnet.SE) (console) login: Stopped in pid 1523.1 (nbmake) at netbsd:cpu_Debugger+0x15: mfpr $ 18, r1 db> bt Not a panic, no active process, ignored. db> ps PID PPID PGRP UID S FLAGS LWPS COMMAND WAIT 22946 391 391 12 2 0x4100 1 pickup >1523 11736 430 0 2 0x4000 1 nbmake 11736 22926 430 0 2 0x4000 1 sh wait 22926 27093 430 0 2 0x4000 1 nbmake wait 27093 25337 430 0 2 0x4000 1 sh wait 25337 160 430 0 2 0x4000 1 nbmake wait 160 3806 430 0 2 0x4000 1 sh wait 10524 14662 10524 2026 2 0x4000 1 top 14662 23760 14662 2026 2 0x4000 1 tcsh pause 23760 11781 11781 2026 2 0x100 1 sshd select 11781 292 11781 0 2 0x4101 1 sshd netio 3806 430 430 0 2 0x4000 1 nbmake wait 437 421 437 0 2 0x4000 1 tail kqread 430 421 430 0 2 0x4000 1 sh wait 421 316 421 0 2 0x4000 1 tcsh pause 316 335 316 2026 2 0x4000 1 tcsh pause 335 362 362 2026 2 0x100 1 sshd select 362 292 362 0 2 0x4101 1 sshd netio 396 391 391 12 2 0x4100 1 qmgr 413 1 413 0 2 0x4000 1 getty 407 1 407 0 2 0 1 cron 409 1 409 0 2 0 1 inetd kqread 391 1 391 0 2 0x4100 1 master 292 1 292 0 2 0 1 sshd select 283 1 283 0 2 0 1 rwhod 269 1 269 0 2 0 1 ntpd 169 1 169 0 2 0 1 ypbind 168 1 168 0 2 0 1 rpcbind 102 1 102 0 2 0 1 syslogd 1 0 1 0 2 0x4001 1 init wait 0 -1 0 0 2 0x20002 13 system * db> ------------------------------------------ Any one who have any thoughts? It's been sitting still like this for atleast four hours now, so I don't think it will wake up... Johnny -- Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt@xxxxxxxxxx || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol

Previous Message by Thread: click to view message preview

Re: KA53 asc/clock bug

On Sun, Jul 29, 2007 at 11:05:22AM +0200, Hans Rosenfeld wrote: > In asc_vsbus_attach() the SCSI id for the controller is read from some > clock chip NVRAM. This causes a segv in kernel mode for the KA53 which > seems to lack the clock chip. OpenBSD uses a hardcoded SCSI id 6 for all > systems but KA46, but since the existing code works with KA48 and KA49 > too I made the hardcoded value a special case for the KA53. I just noticed that the patch doesn't work as intended for kernel configurations which exclude KA46, KA48 and KA49. The clk_page is still referenced, but since it is not defined anywhere the kernel cannot be linked. I made a new patch that references clk_page only if VAX46, VAX48 or VAX49 are defined. -- %SYSTEM-F-ANARCHISM, The operating system has been overthrown asc_vsbus.diff Description: Text document

Next Message by Thread: click to view message preview

System hung...

Running 4.99.27 (cross-built kernel from another architecture) kernel, with 4.99.20 userspace. I'm trying to build the whole system from source, but partway through the build, the whole systems hangs. I have the machine in this state right now, and can try whatever people would suggest. I was running build.sh -x build kernel=GENERIC in one terminal window (connected through ssh). The output was directed to a file, and I was running a tail -f on that file. Another window was running top. That's it. And NIS is running on the machine, and /usr/src is mounted through nfs. The build log file is also written at /usr/src. As I said, the system is hung. The last update in top looks like this: -----------------top-------------- load averages: 1.12, 1.09, 1.08 up 0 days, 6:10 18:30:47 44 processes: 1 runnable, 42 sleeping, 1 on processor CPU states: 78.3% user, 0.0% nice, 16.7% system, 0.0% interrupt, 4.9% idle Memory: 50M Act, 33M Inact, 3336K Wired, 5180K Exec, 57M File, 19M Free Swap: 128M Total, 128M Free PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND 24376 root 59 0 3936K 4800K RUN 0:05 91.09% 23.63% ld 10524 bqt 28 0 288K 1044K CPU 1:53 0.00% 0.00% top ------------------------------------ And the tail -f got to this: -----------------tail-------------- cc -DHAVE_CONFIG_H -I. -I/usr/src/tools/binutils/../../gnu/dist/binutils/ld -I. -D_GNU_SOURCE -I. -I/usr/src/tools/binutils/../../gnu/dist/binutils/ld -I../bfd -I/usr/src/tools/binutils/../../gnu/dist/binutils/ld/../bfd -I/usr/src/tools/binutils/../../gnu/dist/binutils/ld/../include -I/usr/src/tools/binutils/../../gnu/dist/binutils/ld/../intl -I../intl -O -DLOCALEDIR="\"/usr/src/obj/tooldir.NetBSD-4.99.27-vax/share/locale\"" -W -Wall -Wstrict-prototypes -Wmissing-prototypes -O -c evaxnbsd.c /bin/sh ./libtool --mode=link cc -W -Wall -Wstrict-prototypes -Wmissing-prototypes -O -o ld-new ldgram.o ldlex.o lexsup.o ldlang.o mri.o ldctor.o ldmain.o ldwrite.o ldexp.o ldemul.o ldver.o ldmisc.o ldfile.o ldcref.o eelf32vax.o evaxnbsd.o ../bfd/libbfd.la ../libiberty/libiberty.a mkdir .libs cc -W -Wall -Wstrict-prototypes -Wmissing-prototypes -O -o ld-new ldgram.o ldlex.o lexsup.o ldlang.o mri.o ldctor.o ldmain.o ldwrite.o ldexp.o ldemul.o ldver.o ldmisc.o ldfile.o ldcref.o eelf32vax.o evaxnbsd.o ../bfd/.libs/libbfd.a ../libiberty/libiberty.a NOT REBUILDING /usr/src/tools/binutils/../../gnu/dist/binutils/ld/ldver.texi NOT REBUILDING /usr/src/tools/binutils/../../gnu/dist/binutils/ld/ld.1 -------------------------------------- No response from either terminal, nor the console. Breaking into ddb on the console tells me this: -----------------console-------------- Starting cron. Wed Aug 8 12:21:05 CEST 2007 NetBSD/vax (Gnat.BQTnet.SE) (console) login: Stopped in pid 1523.1 (nbmake) at netbsd:cpu_Debugger+0x15: mfpr $ 18, r1 db> bt Not a panic, no active process, ignored. db> ps PID PPID PGRP UID S FLAGS LWPS COMMAND WAIT 22946 391 391 12 2 0x4100 1 pickup >1523 11736 430 0 2 0x4000 1 nbmake 11736 22926 430 0 2 0x4000 1 sh wait 22926 27093 430 0 2 0x4000 1 nbmake wait 27093 25337 430 0 2 0x4000 1 sh wait 25337 160 430 0 2 0x4000 1 nbmake wait 160 3806 430 0 2 0x4000 1 sh wait 10524 14662 10524 2026 2 0x4000 1 top 14662 23760 14662 2026 2 0x4000 1 tcsh pause 23760 11781 11781 2026 2 0x100 1 sshd select 11781 292 11781 0 2 0x4101 1 sshd netio 3806 430 430 0 2 0x4000 1 nbmake wait 437 421 437 0 2 0x4000 1 tail kqread 430 421 430 0 2 0x4000 1 sh wait 421 316 421 0 2 0x4000 1 tcsh pause 316 335 316 2026 2 0x4000 1 tcsh pause 335 362 362 2026 2 0x100 1 sshd select 362 292 362 0 2 0x4101 1 sshd netio 396 391 391 12 2 0x4100 1 qmgr 413 1 413 0 2 0x4000 1 getty 407 1 407 0 2 0 1 cron 409 1 409 0 2 0 1 inetd kqread 391 1 391 0 2 0x4100 1 master 292 1 292 0 2 0 1 sshd select 283 1 283 0 2 0 1 rwhod 269 1 269 0 2 0 1 ntpd 169 1 169 0 2 0 1 ypbind 168 1 168 0 2 0 1 rpcbind 102 1 102 0 2 0 1 syslogd 1 0 1 0 2 0x4001 1 init wait 0 -1 0 0 2 0x20002 13 system * db> ------------------------------------------ Any one who have any thoughts? It's been sitting still like this for atleast four hours now, so I don't think it will wake up... Johnny -- Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt@xxxxxxxxxx || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol