2.6.5-rc2-mm2 still does not boot but it progress : seems to be console font related

2.6.5-rc2-mm2 still does not boot but it progress : seems to be console font related

[Eric Valette]

[-- Attachment #1: Type: text/plain, Size: 1176 bytes --]

Andrew,

I have compiled a completely clean, unpatched (I mean except of course 
rc2-mm2) and I can still not manage to finish booting. However, this 
time, I get a little bit further AND system seems to hang exactly at the 
same place each time (which was not the case with rc2-mm1). In fact I 
managed to have the same behavior after removing the initramfs patches 
from rc2-mm1 and fixing some other things using bk snapshots diffs (SCSI 
st driver).

It hangs when executing :
/etc/init.d/console-screen.sh after displaying:

Setting up general console font..

Attached is a shell trace of a working session on
2.6.5-rc1-mm2. It basically does (twice wonder why!) a loop of :


/usr/bin/consolechars --tty=/dev/vc/[X] -f lat0-16

I also traced the set of system calls /usr/bin/consolechars does via ptrace.

Of course, the same shell script, same binaries and same settings works 
for 2.6.5-rc1-mm2


Does this rings a bell or suggest a culprit patche?

-- 
    __
   /  `                   	Eric Valette
  /--   __  o _.          	6 rue Paul Le Flem
(___, / (_(_(__         	35740 Pace

Tel: +33 (0)2 99 85 26 76	Fax: +33 (0)2 99 85 26 76
E-mail: eric.valette@free.fr




[-- Attachment #2: consolecharTrace --]
[-- Type: text/plain, Size: 15082 bytes --]

execve("/usr/bin/consolechars", ["/usr/bin/consolechars", "--tty=/dev/vc/1", "-f", "lat0-16"], [/* 64 vars */]) = 0
uname({sys="Linux", node="tri-yann", ...}) = 0
brk(0)                                  = 0x804e000
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40017000
access("/etc/ld.so.nohwcap", F_OK)      = -1 ENOENT (No such file or directory)
open("/etc/ld.so.preload", O_RDONLY)    = -1 ENOENT (No such file or directory)
open("/etc/ld.so.cache", O_RDONLY)      = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=83303, ...}) = 0
old_mmap(NULL, 83303, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40018000
close(3)                                = 0
access("/etc/ld.so.nohwcap", F_OK)      = -1 ENOENT (No such file or directory)
open("/lib/libconsole.so.0", O_RDONLY)  = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\360Z\0"..., 512) = 512
fstat64(3, {st_mode=S_IFREG|0644, st_size=60156, ...}) = 0
old_mmap(NULL, 59128, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x4002d000
old_mmap(0x40039000, 12288, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0xc000) = 0x40039000
close(3)                                = 0
access("/etc/ld.so.nohwcap", F_OK)      = -1 ENOENT (No such file or directory)
open("/lib/libcfont.so.0", O_RDONLY)    = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\320\r\0"..., 512) = 512
fstat64(3, {st_mode=S_IFREG|0644, st_size=11784, ...}) = 0
old_mmap(NULL, 14844, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x4003c000
old_mmap(0x4003f000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x2000) = 0x4003f000
close(3)                                = 0
access("/etc/ld.so.nohwcap", F_OK)      = -1 ENOENT (No such file or directory)
open("/lib/libctutils.so.0", O_RDONLY)  = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0@\26\0\000"..., 512) = 512
fstat64(3, {st_mode=S_IFREG|0644, st_size=18288, ...}) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40040000
old_mmap(NULL, 17456, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40041000
old_mmap(0x40045000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x4000) = 0x40045000
close(3)                                = 0
access("/etc/ld.so.nohwcap", F_OK)      = -1 ENOENT (No such file or directory)
open("/lib/tls/libc.so.6", O_RDONLY)    = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\240X\1"..., 512) = 512
fstat64(3, {st_mode=S_IFREG|0644, st_size=1271388, ...}) = 0
old_mmap(NULL, 1281772, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40046000
old_mmap(0x40174000, 36864, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x12d000) = 0x40174000
old_mmap(0x4017d000, 7916, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x4017d000
close(3)                                = 0
set_thread_area({entry_number:-1 -> 6, base_addr:0x40040b40, limit:1048575, seg_32bit:1, contents:0, read_exec_only:0, limit_in_pages:1, seg_not_present:0, useable:1}) = 0
munmap(0x40018000, 83303)               = 0
open("/usr/lib/locale/locale-archive", O_RDONLY|O_LARGEFILE) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=2264496, ...}) = 0
mmap2(NULL, 2097152, PROT_READ, MAP_PRIVATE, 3, 0) = 0x4017f000
close(3)                                = 0
brk(0)                                  = 0x804e000
brk(0x806f000)                          = 0x806f000
brk(0)                                  = 0x806f000
open("/dev/vc/1", O_RDWR)               = 3
ioctl(3, 0x4b33, 0xbfffe9af)            = 0
rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
rt_sigaction(SIGCHLD, NULL, {SIG_DFL}, 8) = 0
rt_sigaction(SIGCHLD, {0x400427f0, [], 0}, {SIG_DFL}, 8) = 0
stat64("lat0-16", 0xbfffe11c)           = -1 ENOENT (No such file or directory)
stat64("lat0-16.gz", 0xbfffe11c)        = -1 ENOENT (No such file or directory)
stat64("lat0-16.lzo", 0xbfffe11c)       = -1 ENOENT (No such file or directory)
stat64("lat0-16.Z", 0xbfffe11c)         = -1 ENOENT (No such file or directory)
stat64("lat0-16.bz2", 0xbfffe11c)       = -1 ENOENT (No such file or directory)
stat64("lat0-16.psf", 0xbfffe11c)       = -1 ENOENT (No such file or directory)
stat64("lat0-16.psf.gz", 0xbfffe11c)    = -1 ENOENT (No such file or directory)
stat64("lat0-16.psf.lzo", 0xbfffe11c)   = -1 ENOENT (No such file or directory)
stat64("lat0-16.psf.Z", 0xbfffe11c)     = -1 ENOENT (No such file or directory)
stat64("lat0-16.psf.bz2", 0xbfffe11c)   = -1 ENOENT (No such file or directory)
stat64("lat0-16.cp", 0xbfffe11c)        = -1 ENOENT (No such file or directory)
stat64("lat0-16.cp.gz", 0xbfffe11c)     = -1 ENOENT (No such file or directory)
stat64("lat0-16.cp.lzo", 0xbfffe11c)    = -1 ENOENT (No such file or directory)
stat64("lat0-16.cp.Z", 0xbfffe11c)      = -1 ENOENT (No such file or directory)
stat64("lat0-16.cp.bz2", 0xbfffe11c)    = -1 ENOENT (No such file or directory)
stat64("lat0-16.fnt", 0xbfffe11c)       = -1 ENOENT (No such file or directory)
stat64("lat0-16.fnt.gz", 0xbfffe11c)    = -1 ENOENT (No such file or directory)
stat64("lat0-16.fnt.lzo", 0xbfffe11c)   = -1 ENOENT (No such file or directory)
stat64("lat0-16.fnt.Z", 0xbfffe11c)     = -1 ENOENT (No such file or directory)
stat64("lat0-16.fnt.bz2", 0xbfffe11c)   = -1 ENOENT (No such file or directory)
stat64("lat0-16.psfu", 0xbfffe11c)      = -1 ENOENT (No such file or directory)
stat64("lat0-16.psfu.gz", 0xbfffe11c)   = -1 ENOENT (No such file or directory)
stat64("lat0-16.psfu.lzo", 0xbfffe11c)  = -1 ENOENT (No such file or directory)
stat64("lat0-16.psfu.Z", 0xbfffe11c)    = -1 ENOENT (No such file or directory)
stat64("lat0-16.psfu.bz2", 0xbfffe11c)  = -1 ENOENT (No such file or directory)
stat64("/usr", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
stat64("/usr/share", {st_mode=S_IFDIR|0755, st_size=8192, ...}) = 0
stat64("/usr/share/consolefonts", {st_mode=S_IFDIR|0755, st_size=16384, ...}) = 0
stat64("/usr/share/consolefonts/lat0-16", 0xbfffdeac) = -1 ENOENT (No such file or directory)
stat64("/usr/share/consolefonts/lat0-16.gz", 0xbfffdeac) = -1 ENOENT (No such file or directory)
stat64("/usr/share/consolefonts/lat0-16.lzo", 0xbfffdeac) = -1 ENOENT (No such file or directory)
stat64("/usr/share/consolefonts/lat0-16.Z", 0xbfffdeac) = -1 ENOENT (No such file or directory)
stat64("/usr/share/consolefonts/lat0-16.bz2", 0xbfffdeac) = -1 ENOENT (No such file or directory)
stat64("/usr/share/consolefonts/lat0-16.psf", 0xbfffdeac) = -1 ENOENT (No such file or directory)
stat64("/usr/share/consolefonts/lat0-16.psf.gz", {st_mode=S_IFREG|0644, st_size=1982, ...}) = 0
rt_sigprocmask(SIG_BLOCK, [CHLD], [CHLD], 8) = 0
open("/usr/share/consolefonts/lat0-16.psf.gz", O_RDONLY) = 4
fstat64(4, {st_mode=S_IFREG|0644, st_size=1982, ...}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x4037f000
read(4, "\37\213\10\10\225\320^@\2\3lat0-16.psf\0u\230k\220\34U"..., 4096) = 1982
pipe([5, 6])                            = 0
pipe([7, 8])                            = 0
clone(Process 2805 attached
child_stack=0, flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, child_tidptr=0x40040b88) = 2805
[pid  2805] close(8)                    = 0
[pid  2805] close(5)                    = 0
[pid  2805] close(0)                    = 0
[pid  2805] dup2(7, 0)                  = 0
[pid  2805] close(1)                    = 0
[pid  2805] dup2(6, 1)                  = 1
[pid  2805] execve("/sbin/gunzip", ["gunzip"], [/* 64 vars */]) = -1 ENOENT (No such file or directory)
[pid  2805] execve("/bin/gunzip", ["gunzip"], [/* 64 vars */]) = 0
[pid  2805] uname({sys="Linux", node="tri-yann", ...}) = 0
[pid  2805] brk(0)                      = 0x80a5000
[pid  2805] old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40017000
[pid  2805] access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file or directory)
[pid  2805] open("/etc/ld.so.preload", O_RDONLY) = -1 ENOENT (No such file or directory)
[pid  2805] open("/etc/ld.so.cache", O_RDONLY) = 5
[pid  2805] fstat64(5, {st_mode=S_IFREG|0644, st_size=83303, ...}) = 0
[pid  2805] old_mmap(NULL, 83303, PROT_READ, MAP_PRIVATE, 5, 0) = 0x40018000
[pid  2805] close(5)                    = 0
[pid  2805] access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file or directory)
[pid  2805] open("/lib/tls/libc.so.6", O_RDONLY) = 5
[pid  2805] read(5, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\240X\1"..., 512) = 512
[pid  2805] fstat64(5, {st_mode=S_IFREG|0644, st_size=1271388, ...}) = 0
[pid  2805] old_mmap(NULL, 1281772, PROT_READ|PROT_EXEC, MAP_PRIVATE, 5, 0) = 0x4002d000
[pid  2805] old_mmap(0x4015b000, 36864, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 5, 0x12d000) = 0x4015b000
[pid  2805] old_mmap(0x40164000, 7916, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x40164000
[pid  2805] close(5)                    = 0
[pid  2805] old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40166000
[pid  2805] set_thread_area({entry_number:-1 -> 6, base_addr:0x401662a0, limit:1048575, seg_32bit:1, contents:0, read_exec_only:0, limit_in_pages:1, seg_not_present:0, useable:1}) = 0
[pid  2805] munmap(0x40018000, 83303)   = 0
[pid  2805] rt_sigaction(SIGINT, {SIG_IGN}, {SIG_DFL}, 8) = 0
[pid  2805] rt_sigaction(SIGINT, {0x804c848, [INT], SA_RESTART}, {SIG_IGN}, 8) = 0
[pid  2805] rt_sigaction(SIGTERM, {SIG_IGN}, {SIG_DFL}, 8) = 0
[pid  2805] rt_sigaction(SIGTERM, {0x804c848, [TERM], SA_RESTART}, {SIG_IGN}, 8) = 0
[pid  2805] rt_sigaction(SIGHUP, {SIG_IGN}, {SIG_DFL}, 8) = 0
[pid  2805] rt_sigaction(SIGHUP, {0x804c848, [HUP], SA_RESTART}, {SIG_IGN}, 8) = 0
[pid  2805] ioctl(0, SNDCTL_TMR_TIMEBASE or TCGETS, 0xbffff600) = -1 EINVAL (Invalid argument)
[pid  2805] read(0,  <unfinished ...>
[pid  2804] close(7)                    = 0
[pid  2804] close(6)                    = 0
[pid  2804] clone(Process 2806 attached
child_stack=0, flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, child_tidptr=0x40040b88) = 2806
[pid  2806] write(8, "\37\213\10\10\225\320^@\2", 9) = 9
[pid  2806] write(8, "\3lat0-16.psf\0u\230k\220\34U\25\200;@M\241\266\313ZZ8"..., 512) = 512
[pid  2806] write(8, "-\266\374l7{a\30\t\207R}@\4\370\340\16R\335v\347\30ak\316"..., 512) = 512
[pid  2806] write(8, "\363ET\267\263\210MN\370\23\234\357r?;b?\273\353i\374\216"..., 512) = 512
[pid  2806] read(4, "", 4096)           = 0
[pid  2806] write(8, "K\341e\360r\250\303+`\27\\\f\273\341\2250\16\257\2\v^\r"..., 437) = 437
[pid  2806] read(4, "", 4096)           = 0
[pid  2806] exit_group(0)               = ?
Process 2806 detached
[pid  2804] close(8)                    = 0
[pid  2804] close(4)                    = 0
[pid  2804] munmap(0x4037f000, 4096)    = 0
[pid  2804] fcntl64(5, F_GETFL)         = 0 (flags O_RDONLY)
[pid  2804] fstat64(5, {st_mode=S_IFIFO|0600, st_size=0, ...}) = 0
[pid  2804] mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x4037f000
[pid  2804] _llseek(5, 0, 0xbfffe09c, SEEK_CUR) = -1 ESPIPE (Illegal seek)
[pid  2804] read(5,  <unfinished ...>
[pid  2805] <... read resumed> "\37\213\10\10\225\320^@\2\3lat0-16.psf\0u\230k\220\34U"..., 32768) = 1982
[pid  2805] read(0, "", 30786)          = 0
[pid  2805] brk(0)                      = 0x80a5000
[pid  2805] brk(0x80c6000)              = 0x80c6000
[pid  2805] brk(0)                      = 0x80c6000
[pid  2805] write(1, "6\4\2\20\0\0~\303\231\231\363\347\347\377\347\347~\0\0"..., 5156 <unfinished ...>
[pid  2804] <... read resumed> "6\4\2\20\0\0~\303\231\231\363\347\347\377\347\347~\0\0"..., 4096) = 4096
[pid  2804] pipe([4, 6])                = 0
[pid  2804] clone(Process 2807 attached
child_stack=0, flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, child_tidptr=0x40040b88) = 2807
[pid  2807] close(4)                    = 0
[pid  2807] write(6, "6\4\2\20\0\0~\303\231\231\363\347\347\377\347\347~\0\0"..., 32) = 32
[pid  2807] write(6, "\0\0\0\0\0\0n\370\330\330\334\330\330\330\370n\0\0\0\0"..., 512) = 512
[pid  2807] write(6, "\0\0\0\0\0fff$\0\0\0\0\0\0\0\0\0\0\0\0\0\0ll\376lll\376"..., 512) = 512
[pid  2807] write(6, "\0\0\0\0\0\0\374fff|ffff\374\0\0\0\0\0\0<f\302\300\300"..., 512) = 512
[pid  2807] write(6, "\0\0\0\0\0\0\340``xlffff|\0\0\0\0\0\0\0\0\0|\306\300\300"..., 512) = 512
[pid  2807] write(6, "\0\0\0\0\0208l\0008l\306\306\376\306\306\306\0\0\0\0v\334"..., 512) = 512
[pid  2807] write(6, "<\30\0\0\0\0\0\0\20|\326\320\320\320\326|\20\0\0\0\0\000"..., 512) = 512
[pid  2807] write(6, "\0\0\0\0\0\0\0\0\0\0\0\37\0\0\0\0\0\0\0\0\30\30\30\30\30"..., 512) = 512
[pid  2807] read(5,  <unfinished ...>
[pid  2804] close(6)                    = 0
[pid  2804] fcntl64(4, F_GETFL)         = 0 (flags O_RDONLY)
[pid  2804] fstat64(4, {st_mode=S_IFIFO|0600, st_size=0, ...}) = 0
[pid  2804] mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40380000
[pid  2804] _llseek(4, 0, 0xbfffe09c, SEEK_CUR) = -1 ESPIPE (Illegal seek)
[pid  2804] rt_sigprocmask(SIG_SETMASK, [CHLD], NULL, 8) = 0
[pid  2804] fstat64(4, {st_mode=S_IFIFO|0600, st_size=0, ...}) = 0
[pid  2804] read(4, "6\4\2\20\0\0~\303\231\231\363\347\347\377\347\347~\0\0"..., 4096) = 3616
[pid  2804] read(4,  <unfinished ...>
[pid  2805] <... write resumed> )       = 5156
[pid  2805] exit_group(0)               = ?
Process 2805 detached
[pid  2807] <... read resumed> "\6\f\370\0\375\377\377\377H\"\377\377R\1\377\377S\1\377"..., 4096) = 1060
[pid  2807] write(6, "\0\0\0\0\0\0208l\0x\f|\314\314\314v\0\0\0\0\0\0v\334\0"..., 512 <unfinished ...>
[pid  2804] <... read resumed> "\0\0\0\0\0\0208l\0x\f|\314\314\314v\0\0\0\0\0\0v\334\0"..., 4096) = 512
[pid  2804] read(4,  <unfinished ...>
[pid  2807] <... write resumed> )       = 512
[pid  2807] write(6, "\r$\377\377\n$\377\377\221%\377\377\222%\377\377\223%\377"..., 512 <unfinished ...>
[pid  2804] <... read resumed> "\r$\377\377\n$\377\377\221%\377\377\222%\377\377\223%\377"..., 4096) = 512
[pid  2804] read(4,  <unfinished ...>
[pid  2807] <... write resumed> )       = 512
[pid  2807] write(6, "+!\377\377\306\0\377\377\307\0\377\377\310\0\377\377\311"..., 512 <unfinished ...>
[pid  2804] <... read resumed> "+!\377\377\306\0\377\377\307\0\377\377\310\0\377\377\311"..., 4096) = 512
[pid  2804] read(4,  <unfinished ...>
[pid  2807] <... write resumed> )       = 512
[pid  2807] read(5, "", 4096)           = 0
[pid  2807] write(6, "\377\0\377\377", 4 <unfinished ...>
[pid  2804] <... read resumed> "\377\0\377\377", 4096) = 4
[pid  2804] close(4)                    = 0
[pid  2804] munmap(0x40380000, 4096)    = 0
[pid  2804] ioctl(3, 0x4b72, 0xbfffe850) = 0
[pid  2804] rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
[pid  2804] --- SIGCHLD (Child exited) @ 0 (0) ---
[pid  2804] wait4(-1, [WIFEXITED(s) && WEXITSTATUS(s) == 0], 0, NULL) = 2806
[pid  2804] sigreturn()                 = ? (mask now [])
[pid  2804] ioctl(3, 0x4b68, 0xbfffe9d0) = 0
[pid  2804] ioctl(3, 0x4b67, 0xbfffea68) = 0
[pid  2804] exit_group(0)               = ?
<... write resumed> )                   = 4
read(5, "", 4096)                       = 0
exit_group(0)                           = ?
Process 2807 detached

[-- Attachment #3: console-screen.out --]
[-- Type: application/octet-stream, Size: 4461 bytes --]

Re: 2.6.5-rc2-mm2 still does not boot but it progress : seems to be console font related

[Joseph Pingenot]

Just wanted to pipe up and say "me too!"

I've been getting this error as well.  I unfortunately don't have more
  info than that.

-Joseph
-- 
trelane@digitasaru.net--------------------------------------------------
"We continue to live in a world where all our know-how is locked into
 binary files in an unknown format. If our documents are our corporate
 memory, Microsoft still has us all condemned to Alzheimer's."
    --Simon Phipps, http://theregister.com/content/4/30410.html

Re: 2.6.5-rc2-mm2 still does not boot but it progress : seems to be console font related

[Andrew Morton]

Eric Valette <eric.valette@free.fr> wrote:
>
> I have compiled a completely clean, unpatched (I mean except of course 
>  rc2-mm2) and I can still not manage to finish booting. However, this 
>  time, I get a little bit further AND system seems to hang exactly at the 
>  same place each time (which was not the case with rc2-mm1). In fact I 
>  managed to have the same behavior after removing the initramfs patches 
>  from rc2-mm1 and fixing some other things using bk snapshots diffs (SCSI 
>  st driver).
> 
>  It hangs when executing :
>  /etc/init.d/console-screen.sh after displaying:
> 
>  Setting up general console font..
> 
>  Attached is a shell trace of a working session on
>  2.6.5-rc1-mm2. It basically does (twice wonder why!) a loop of :
> 
> 
>  /usr/bin/consolechars --tty=/dev/vc/[X] -f lat0-16
> 
>  I also traced the set of system calls /usr/bin/consolechars does via ptrace.
> 
>  Of course, the same shell script, same binaries and same settings works 
>  for 2.6.5-rc1-mm2

Are you using devfs?  If so, please try the below patch (I don't see why,
but..)

Can a sysrq-T or sysrq-P trace be generated when it has died?  You may need
to alter your initscripts so they do not stick a zero in
/proc/sys/kernel/sysrq.

---

 25-akpm/drivers/char/vt.c |    7 ++++++-
 1 files changed, 6 insertions(+), 1 deletion(-)

diff -puN drivers/char/vt.c~a drivers/char/vt.c
--- 25/drivers/char/vt.c~a	2004-03-24 09:49:10.285591688 -0800
+++ 25-akpm/drivers/char/vt.c	2004-03-24 09:50:54.355770616 -0800
@@ -2471,10 +2471,13 @@ static int con_open(struct tty_struct *t
 				tty->winsize.ws_row = video_num_lines;
 				tty->winsize.ws_col = video_num_columns;
 			}
+			release_console_sem();
 			vcs_make_devfs(tty);
+			goto out;
 		}
 	}
 	release_console_sem();
+out:
 	return ret;
 }
 
@@ -2484,11 +2487,13 @@ static void con_close(struct tty_struct 
 	if (tty && tty->count == 1) {
 		struct vt_struct *vt;
 
-		vcs_remove_devfs(tty);
 		vt = tty->driver_data;
 		if (vt)
 			vc_cons[vt->vc_num].d->vc_tty = NULL;
 		tty->driver_data = 0;
+		release_console_sem();
+		vcs_remove_devfs(tty);
+		return;
 	}
 	release_console_sem();
 }

_

Re: 2.6.5-rc2-mm2 still does not boot but it progress : seems to be console font related

[Eric Valette]

Andrew Morton wrote:

> Are you using devfs?  If so, please try the below patch (I don't see why,
> but..)

Yes. I use devfs.

Sorry for the delay, I'm back, with a new , more silent, and working 
fan, after a brutal powerdown due to fan failure and CPU overheat... As 
my PC is thre years old, it takes time to find compatible fans...

The machine that boots again but unfortunately I had to hardwire a BIOS 
CMOS reset and therefore BIOS settings have been restored to factory 
default values. I hopefully restored back then to what they where but 
cannot guaranty it 100%.

I applied the patch and get the deadlock is gone *BUT* due to BIOS 
default changed (including one that may impact performance (PIC read 
caching, PCU byte grouping, ...), and the fact that it was probably a 
deadlock due to a window that may have disappearred if I did not restore 
correctly the settings, I cannot gurante that it is the fix that fixed 
the problem.

I hope that Joseph Pingenot that had the same problem can confirm it.
I erased the mail were he was confirming the symptom and I have its 
address no more. Could you double check with him that the patch indeed 
fix the problem?

Thanks for chasing the bug anyway,

-- 
    __
   /  `                   	Eric Valette
  /--   __  o _.          	6 rue Paul Le Flem
(___, / (_(_(__         	35740 Pace

Tel: +33 (0)2 99 85 26 76	Fax: +33 (0)2 99 85 26 76
E-mail: eric.valette@free.fr

Re: 2.6.5-rc2-mm2 still does not boot but it progress : seems to be console font related

[Thomas Svedberg]

I have these hangs as well, just tried 2.6.5-rc2-mm3 and they are still 
there.
However setting video=radeonfb:off as boot parameter solves the problem, 
if this can be of any help.
More info on request.

/Thomas

Eric Valette wrote:
> Andrew,
> 
> I have compiled a completely clean, unpatched (I mean except of course 
> rc2-mm2) and I can still not manage to finish booting. However, this 
> time, I get a little bit further AND system seems to hang exactly at the 
> same place each time (which was not the case with rc2-mm1). In fact I 
> managed to have the same behavior after removing the initramfs patches 
> from rc2-mm1 and fixing some other things using bk snapshots diffs (SCSI 
> st driver).

[snip]
-- 
/ Thomas
.......................................................................
  Thomas Svedberg
  Department of Applied Mechanics
  Chalmers University of Technology

  Address: SE-412 96 Göteborg, SWEDEN
  E-mail : thsv@am.chalmers.se, thsv@bigfoot.com
  Phone  : +46 31 772 1522
  Fax    : +46 31 772 3827
.......................................................................

Re: 2.6.5-rc2-mm2 still does not boot but it progress : seems to be console font related

[Brandon Low]

I was having a similar problem, and with nothing else on the system
changed, applying the patch fixed it.  Was freezing at "freeing unused
kernel memory" without the patch.

Thanks,

Brandon

On Wed, 03/24/04 at 09:52:36 -0800, Andrew Morton wrote:
> Eric Valette <eric.valette@free.fr> wrote:
> >
> > I have compiled a completely clean, unpatched (I mean except of course 
> >  rc2-mm2) and I can still not manage to finish booting. However, this 
> >  time, I get a little bit further AND system seems to hang exactly at the 
> >  same place each time (which was not the case with rc2-mm1). In fact I 
> >  managed to have the same behavior after removing the initramfs patches 
> >  from rc2-mm1 and fixing some other things using bk snapshots diffs (SCSI 
> >  st driver).
> > 
> >  It hangs when executing :
> >  /etc/init.d/console-screen.sh after displaying:
> > 
> >  Setting up general console font..
> > 
> >  Attached is a shell trace of a working session on
> >  2.6.5-rc1-mm2. It basically does (twice wonder why!) a loop of :
> > 
> > 
> >  /usr/bin/consolechars --tty=/dev/vc/[X] -f lat0-16
> > 
> >  I also traced the set of system calls /usr/bin/consolechars does via ptrace.
> > 
> >  Of course, the same shell script, same binaries and same settings works 
> >  for 2.6.5-rc1-mm2
> 
> Are you using devfs?  If so, please try the below patch (I don't see why,
> but..)
> 
> Can a sysrq-T or sysrq-P trace be generated when it has died?  You may need
> to alter your initscripts so they do not stick a zero in
> /proc/sys/kernel/sysrq.
> 
> ---
> 
>  25-akpm/drivers/char/vt.c |    7 ++++++-
>  1 files changed, 6 insertions(+), 1 deletion(-)
> 
> diff -puN drivers/char/vt.c~a drivers/char/vt.c
> --- 25/drivers/char/vt.c~a	2004-03-24 09:49:10.285591688 -0800
> +++ 25-akpm/drivers/char/vt.c	2004-03-24 09:50:54.355770616 -0800
> @@ -2471,10 +2471,13 @@ static int con_open(struct tty_struct *t
>  				tty->winsize.ws_row = video_num_lines;
>  				tty->winsize.ws_col = video_num_columns;
>  			}
> +			release_console_sem();
>  			vcs_make_devfs(tty);
> +			goto out;
>  		}
>  	}
>  	release_console_sem();
> +out:
>  	return ret;
>  }
>  
> @@ -2484,11 +2487,13 @@ static void con_close(struct tty_struct 
>  	if (tty && tty->count == 1) {
>  		struct vt_struct *vt;
>  
> -		vcs_remove_devfs(tty);
>  		vt = tty->driver_data;
>  		if (vt)
>  			vc_cons[vt->vc_num].d->vc_tty = NULL;
>  		tty->driver_data = 0;
> +		release_console_sem();
> +		vcs_remove_devfs(tty);
> +		return;
>  	}
>  	release_console_sem();
>  }
> 
> _
> 
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

Re: 2.6.5-rc2-mm2 still does not boot but it progress : seems to be console font related

[Eric Valette]

[-- Attachment #1: Type: text/plain, Size: 443 bytes --]

Thomas Svedberg wrote:
> I have these hangs as well, just tried 2.6.5-rc2-mm3 and they are still 
> there.
> However setting video=radeonfb:off as boot parameter solves the problem, 
> if this can be of any help.
> More info on request.

Yes because the console-screen.sh shell script checks for /dev/fb. Could 
you try the patceh suggested by Andrew in this thread (I'm not sure it 
is in mm3). I attached it for your convenience.

-- eric



[-- Attachment #2: console-devfs.patch --]
[-- Type: text/x-patch, Size: 817 bytes --]

diff -puN drivers/char/vt.c~a drivers/char/vt.c
--- 25/drivers/char/vt.c~a	2004-03-24 09:49:10.285591688 -0800
+++ 25-akpm/drivers/char/vt.c	2004-03-24 09:50:54.355770616 -0800
@@ -2471,10 +2471,13 @@ static int con_open(struct tty_struct *t
 				tty->winsize.ws_row = video_num_lines;
 				tty->winsize.ws_col = video_num_columns;
 			}
+			release_console_sem();
 			vcs_make_devfs(tty);
+			goto out;
 		}
 	}
 	release_console_sem();
+out:
 	return ret;
 }
 
@@ -2484,11 +2487,13 @@ static void con_close(struct tty_struct 
 	if (tty && tty->count == 1) {
 		struct vt_struct *vt;
 
-		vcs_remove_devfs(tty);
 		vt = tty->driver_data;
 		if (vt)
 			vc_cons[vt->vc_num].d->vc_tty = NULL;
 		tty->driver_data = 0;
+		release_console_sem();
+		vcs_remove_devfs(tty);
+		return;
 	}
 	release_console_sem();
 }

_

Re: 2.6.5-rc2-mm2 still does not boot but it progress : seems to be console font related

[Andrew Morton]

Brandon Low <lostlogic@gentoo.org> wrote:
>
> I was having a similar problem, and with nothing else on the system
>  changed, applying the patch fixed it.  Was freezing at "freeing unused
>  kernel memory" without the patch.

Drat, thanks.

Re: 2.6.5-rc2-mm2 still does not boot but it progress : seems to be console font related

[Thomas Svedberg]

Eric Valette wrote:
> Thomas Svedberg wrote:
> 
>> I have these hangs as well, just tried 2.6.5-rc2-mm3 and they are 
>> still there.
>> However setting video=radeonfb:off as boot parameter solves the 
>> problem, if this can be of any help.
>> More info on request.
> 
> 
> Yes because the console-screen.sh shell script checks for /dev/fb. Could 
> you try the patceh suggested by Andrew in this thread (I'm not sure it 
> is in mm3). I attached it for your convenience.

I must have missed it, an yes it fixes the problem.
(I tried it on top of -mm3)

/Thomas

> 
> ------------------------------------------------------------------------
> 
> diff -puN drivers/char/vt.c~a drivers/char/vt.c
> --- 25/drivers/char/vt.c~a	2004-03-24 09:49:10.285591688 -0800
> +++ 25-akpm/drivers/char/vt.c	2004-03-24 09:50:54.355770616 -0800
> @@ -2471,10 +2471,13 @@ static int con_open(struct tty_struct *t
>  				tty->winsize.ws_row = video_num_lines;
>  				tty->winsize.ws_col = video_num_columns;
>  			}
> +			release_console_sem();
>  			vcs_make_devfs(tty);
> +			goto out;
>  		}
>  	}
>  	release_console_sem();
> +out:
>  	return ret;
>  }
>  
> @@ -2484,11 +2487,13 @@ static void con_close(struct tty_struct 
>  	if (tty && tty->count == 1) {
>  		struct vt_struct *vt;
>  
> -		vcs_remove_devfs(tty);
>  		vt = tty->driver_data;
>  		if (vt)
>  			vc_cons[vt->vc_num].d->vc_tty = NULL;
>  		tty->driver_data = 0;
> +		release_console_sem();
> +		vcs_remove_devfs(tty);
> +		return;
>  	}
>  	release_console_sem();
>  }
> 
> _
> 
> 


.......................................................................
  Thomas Svedberg
  Department of Applied Mechanics
  Chalmers University of Technology

  Address: S-412 96 Göteborg, SWEDEN
  E-mail : thsv@bigfoot.com, thsv@am.chalmers.se
  Phone  : +46 31 772 1522
  Fax    : +46 31 772 3827
.......................................................................