Having emerged gentoo-sources 2.6.12-r3, the system was rebooted to use the new kernel and although it worked with no unusual errors in syslog, the startup time has increased substantially. Using wvdial to connect to my ISP (ppp0, 56kbps external modem) results in slow program operation, taking a couple of minutes to even begin dialling. Starting X after boot works, but it just sits there, taking its sweet time to even show the initial cursor. Went back to using 2.6.12-r2 on the same system with no problems in any area. Am including the diff of the output to /var/log/messages between -r3 and -r2. Reproducible: Always Steps to Reproduce: 1. emerge =gentoo-sources-2.6.12-r3, make kernel as normal (copying old .config over to new kernel) 2. reboot 3. watch as 1.83ghz athlon laptop pretends to be clocked at 183mhz or so Actual Results: System operated, but slower than should be expected. Expected Results: Should be about as quick to boot up and run programs as -r2 is. (emerge --info taken while running -r2, but then nothing in here should have changed between kernel revisions, save for uname) Portage 2.0.51.22-r1 (default-linux/x86/2005.0, gcc-3.4.4, glibc-2.3.5-r0, 2.6.12-gentoo-r2 i686) ================================================================= System uname: 2.6.12-gentoo-r2 i686 mobile AMD Athlon(tm) XP2500+ Gentoo Base System version 1.6.12 dev-lang/python: 2.3.5, 2.4.1-r1 sys-apps/sandbox: 1.2.9 sys-devel/autoconf: 2.13, 2.59-r7 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.5 sys-devel/binutils: 2.16.1 sys-devel/libtool: 1.5.18-r1 virtual/os-headers: 2.6.11-r2 ACCEPT_KEYWORDS="x86 ~x86" AUTOCLEAN="yes" CBUILD="i686-pc-linux-gnu" CFLAGS="-march=athlon-xp -O2 -fomit-frame-pointer -pipe" CHOST="i686-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/kde/2/share/config /usr/kde/3/share/config /usr/lib/X11/xkb /usr/share/config /var/qmail/control" CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/texmf/web2c /etc/env.d" CXXFLAGS="-march=athlon-xp -O2 -fomit-frame-pointer -pipe" DISTDIR="/usr/portage/distfiles" FEATURES="autoconfig distlocks fixpackages sandbox sfperms strict userpriv usersandbox" GENTOO_MIRRORS="ftp://ftp.mirrorservice.org/sites/www.ibiblio.org/gentoo/ ftp://mirrors1.netvisao.pt/gentoo/ ftp://ftp.gentoo-pt.org/pub/gentoo/ " LANG="en_GB.utf8" LC_ALL="en_GB.utf8" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage" SYNC="rsync://rsync.europe.gentoo.org/gentoo-portage" USE="x86 3dnow 3dnowext X a52 aac aalib acpi aim alsa apm audiofile avi bash-completion bcmath berkdb bidi bitmap-fonts bmp bzlib canna cdparanoia cdr cjk crypt curl dga directfb divx4linux dv dvd dvdr dvdread encode fbcon ffmpeg flac freewnn ftp gd gd-external ggi gif glut gnutls gpm gtk gtk2 iconv icq ieee1394 imagemagick imap ipv6 java javascript jikes joystick jpeg libcaca libwww lzo mad matroska mikmod mime ming mmx mmxext mng motif mozilla mp3 mpeg msn mythtv ncurses nls nptl ogg oggvorbis openal opengl oscar oss pam pcmcia pcre pdflib perl png python qt quicktime readline recode samba scanner sdl slang sndfile spell sse ssl tcltk tcpd tetex tga theora tiff truetype truetype-fonts type1-fonts unicode usb vcd vorbis wifi win32codecs xml xml2 xpm xv xvid yahoo yim zlib video_cards_ati userland_GNU kernel_linux elibc_glibc" Unset: ASFLAGS, CTARGET, LDFLAGS, LINGUAS diff gentoo-r2.log gentoo-r3.log (if the same configuration works on -r2, perhaps the diff is a good place to find what's troubling the system) 3c3 < chibisuke Linux version 2.6.12-gentoo-r2 (root@chibisuke) (gcc version 3.4.4 (Gentoo 3.4.4, ssp-3.4.4-1.0, pie-8.7.8)) --- > chibisuke Linux version 2.6.12-gentoo-r3 (root@chibisuke) (gcc version 3.4.4 (Gentoo 3.4.4, ssp-3.4.4-1.0, pie-8.7.8)) 32c32 < chibisuke CPU 0 irqstacks, hard=c03da000 soft=c03d9000 --- > chibisuke CPU 0 irqstacks, hard=c03d7000 soft=c03d6000 39c39 < chibisuke Memory: 450000k/457664k available (1927k kernel code, 7104k reserved, 796k data, 164k --- > chibisuke Memory: 450012k/457664k available (1920k kernel code, 7092k reserved, 791k data, 164k 90d89 < chibisuke inotify device minor=63 124a124 > chibisuke PCI: setting IRQ 0 as level-triggered 129a130 > chibisuke spurious 8259A interrupt: IRQ7. 155a157 > chibisuke ReiserFS: hda3: replayed 6 transactions in 0 seconds 164c166 < chibisuke ACPI: Thermal Zone [THRM] (97 C) --- > chibisuke ACPI: Thermal Zone [THRM] (94 C) 168a171 > chibisuke Linux video capture interface: v1.00 170a174,175 > chibisuke usbcore: registered new driver spca5xx > chibisuke /var/tmp/portage/spca5xx-20050501/work/spca5xx-20050501/drivers/usb/spca5xx.c: spca5xx driver 00.56.02 registered 214c219 < chibisuke Yenta: ISA IRQ mask 0x0018, PCI irq 11 --- > chibisuke Yenta: ISA IRQ mask 0x0098, PCI irq 11 221a227 > chibisuke ieee1394.agent: ... no drivers for IEEE1394 product 0x/0x/0x 230d235 < chibisuke ieee1394.agent: ... no drivers for IEEE1394 product 0x/0x/0x 238c243 < chibisuke dictd: :I: 11422 starting dictd 1.10.0/rf on Linux 2.6.12-gentoo-r2 Sun Jul --- > chibisuke dictd: :I: 13491 starting dictd 1.10.0/rf on Linux 2.6.12-gentoo-r3 Sun Jul 285a291,296 > chibisuke psmouse.c: TouchPad at isa0060/serio1/input0 lost synchronization, throwing 2 bytes away. > chibisuke psmouse.c: TouchPad at isa0060/serio1/input0 lost sync at byte 1 > chibisuke psmouse.c: TouchPad at isa0060/serio1/input0 lost sync at byte 1 > chibisuke psmouse.c: TouchPad at isa0060/serio1/input0 lost sync at byte 1 > chibisuke psmouse.c: TouchPad at isa0060/serio1/input0 lost sync at byte 1 > chibisuke psmouse.c: TouchPad at isa0060/serio1/input0 - driver resynched.
Are you using the same .config for both kernels? Please post a unified diff of the dmesg output (capture to files, then use diff -u) Do you have megaraid or ite8212 storage hardware?
Created attachment 62560 [details] diff -u of messages between -r2 and -r3 The .config files are the same except for what 'make oldconfig' changed (summarised): +# CONFIG_BLK_DEV_IT821X is not set -CONFIG_INOTIFY=y Attached unified diff of file output. I haven't got any such hardware, no.
You don't appear to be loading the spca5xx driver in the -r2 instance. Does it help if you don't load then when trying -r3?
This particular issue existed before that particular driver was installed: I was running -r2 when I emerged spca5xx, and it installed under -r3's modules, which I wasn't using at the time since it caused this mentioned issue. Of course, memory be darned, I removed the module anyway, and it actually sped the init up to be of similar speed to -r2. Seemed like a good start. However, wvdial still takes forever to dial out as compared to instaneously under -r2, and X still steals the display and sits there doing nothing when it is run, so it's only helped the init. I could test more programs under -r3, if it'd help.
Ok, thats quite odd as I don't see any changes inbetween -r2 and -r3 which might have such an effect. The next step is to revert the changes, one by one, recompiling/rebooting after every change, to see if there has been any difference. I understand if you don't have enough time as this is a big thing to ask.. (maybe you could just try some?) Here are the patches you need to revert one-by-one: http://dev.gentoo.org/~dsd/gentoo-sources/trunk/2.6.12/2315_ide-no-lba.patch http://dev.gentoo.org/~dsd/gentoo-sources/trunk/2.6.12/1002_linux-2.6.12.2.patch http://dev.gentoo.org/~dsd/gentoo-sources/trunk/2.6.12/4345_it8212.patch http://dev.gentoo.org/~dsd/gentoo-sources/trunk/2.6.12/4350_megaraid-update.patch http://dev.gentoo.org/~dsd/gentoo-sources/trunk/2.6.12/4351_megaraid-compatibility.patch I've listed the most likely ones at the top, so hopefully the problematic one will appear earlier. (you might like to revert the 3 at the end in the same go, as they are very unlikely to have effect.) To revert a patch: Download it, then: # cd /usr/src/linux # patch -p1 -R -i /path/to/patch Then recompile, install, reboot as normal. Then try the next patch, etc. Thanks!
By design this had to produce results, and it did; whatever problem I'm having is contained within 1002_linux-2.6.12.2.patch, as removing this patch fixes the issue, thanks. If that's the case, this might not be a gentoo issue specifically, so I pre-emptively checked bugzilla.kernel.org and found this: http://bugzilla.kernel.org/show_bug.cgi?id=4824 My symptoms aren't so severe, but the machine architecture and setup mentioned are rather similar to mine. (XP-M 2500+, Compaq Presario 2143EA with ALi IDE controller and chipset) Presuming the unapplied patch isn't too dissimilar to the vanilla distribution, the issues might be related, but I'll wait for comment before I consider going off and editing bits of kernel on a whim. :)
Ok, great! Thanks a lot for investigating that. Please re-apply the 2.6.12.2 patch: # cd /usr/src/linux # patch -p1 -i /path/to/1002_linux-2.6.12.2.patch Then, download the patch from here to a file: http://www.kernel.org/git/gitweb.cgi?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff_plain;h=44f8e1a20cf3afe10a3744bd9317808a39a242bb;hp=4a89a04f1ee21a7c1f4413f1ad7dcfac50ff9b63 And apply that one in a similar manner. (Note that this time we aren't using -R argument in the patch command) Then recompile, install, reboot, and hopefully the system will be fast again.
Yep, running -r3 and all is back up to speed now, thanks a lot. Will this be in the next gentoo-sources?
Yes
Fixed in gentoo-sources-2.6.12-r4 Fixed in genpatches-2.6.12-7
Forgot to say, thanks a lot for investigating the issue and testing those patches so quickly. Nice to have it fixed as we're hitting release deadlines real soon...
Heh, no problem, it didn't take so long. Linux is the friend of those with lots of free time. :) Cleaning up, 2.6.12-r3 verified, no recurrence in later revisions. Closing.
