I upgraded pbbuttonsd from 0.6.7 to 0.6.8 after discussion with JoseJX about an (unrelated) problem with kernel 2.6.11 Soon after the upgrade, the machine froze (no mouse movements, no hard disk activity). I rebooted using CTRL-Apple-Power, kept with my work and the machine froze again half an hour later. I backed off to 0.6.7 since Reproducible: Always Steps to Reproduce: 1.upgrade pbbuttonsd 0.6.8 2.restart pbbuttonsd 3.wait some time working Actual Results: the machine got completely frozen Expected Results: nada $ emerge info Portage 2.0.51.19 (default-linux/ppc/2005.0, gcc-3.4.3-20050110, glibc-2.3.4.20041102-r1, 2.6.11-gentoo-r4 ppc) ================================================================= System uname: 2.6.11-gentoo-r4 ppc 7455, altivec supported Gentoo Base System version 1.6.10 Python: dev-lang/python-2.3.5 [2.3.5 (#1, Feb 18 2005, 18:55:22)] dev-lang/python: 2.3.5 sys-devel/autoconf: 2.13, 2.59-r6 sys-devel/automake: 1.7.9-r1, 1.4_p6, 1.5, 1.9.5, 1.8.5-r3, 1.6.3 sys-devel/binutils: 2.15.90.0.3-r4 sys-devel/libtool: 1.5.14 virtual/os-headers: 2.6.8.1-r2 ACCEPT_KEYWORDS="ppc ~ppc" AUTOCLEAN="yes" CFLAGS="-O2 -mcpu=7450 -maltivec -mabi=altivec -pipe" CHOST="powerpc-unknown-linux-gnu" CONFIG_PROTECT="/etc /usr/kde/2/share/config /usr/kde/3.3/env /usr/kde/3.3/share/config /usr/kde/3.3/shutdown /usr/kde/3/share/config /usr/lib/X11/xkb /usr/lib/mozilla/defaults/pref /usr/share/config /var/bind /var/qmail/control" CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d" CXXFLAGS="-O2 -mcpu=7450 -maltivec -mabi=altivec -pipe" DISTDIR="/usr/portage/distfiles" FEATURES="autoaddcvs autoconfig buildpkg ccache distlocks sandbox sfperms" GENTOO_MIRRORS="http://distfiles.gentoo.org http://www.ibiblio.org/pub/Linux/distributions/gentoo http://ftp.easynet.nl/mirror/gentoo/ ftp://ftp.easynet.nl/mirror/gentoo/ ftp://mirror.switch.ch/mirror/gentoo/" LANG="es_ES.UTF-8" LC_ALL="es_ES.UTF-8" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage" SYNC="rsync://rsync.gentoo.org/gentoo-portage" USE="X aac aalib aavm accessibility adns alsa altivec apache2 bash-completion berkdb bitmap-fonts bluetooth bonobo cdr cjk crypt cups curl directfb dv dvd dvdr dvdread emacs emboss esd fam fbcon flac fortran gd gdbm gif glitz gnome gphoto2 gpm gstreamer gtk gtk2 gtkhtml guile hal howl icq imagemagick imlib ipv6 jabber java jpeg junit lcms ldap mad mono motif mozilla mozsvg mp3 mpeg mysql nas ncurses nls nptl ogg oggvorbis opengl pam pcmcia pcre pdflib perl pic plotutils png pnp postgres ppc python qt readline ruby samba scanner sdl slang snmp spell sqlite ssl svg tcltk tcpd threads tiff truetype truetype-fonts type1-fonts unicode usb v4l v4l2 xinerama xml xml2 xmms xosd xprint xv xvid yahoo zlib video_cards_radeon linguas_es_ES linguas_es linguas_en_US linguas_en" Unset: ASFLAGS, CBUILD, CTARGET, LDFLAGS
I had the "use fb to blank" enabled. This option behaves differently depending on kernel release: in 2.6.8 is seemed to do nothing, in 2.6.9 and 2.6.10 it switched backligh off but left the card on. In 2.6.11 it left the screen white Ditto WRT Alsa. Now I switched to OSS and it is working perfectly. One of these two things seem to be the cause of the computer freezing.
i don't have any problems on using pbbuttonsd-0.6.8 with kernel 2.6.11.6.(uptime 6:32h). hope it'll work for you too... here my emerge info: Portage 2.0.51.19 (default-linux/ppc/2005.0, gcc-3.4.3-20050110, glibc-2.3.4.20041102-r1, 2.6.11.6 ppc) ================================================================= System uname: 2.6.11.6 ppc 7455, altivec supported Gentoo Base System version 1.6.10 Python: dev-lang/python-2.3.5 [2.3.5 (#1, Feb 19 2005, 19:27:36)] dev-lang/python: 2.3.5 sys-devel/autoconf: 2.13, 2.59-r6 sys-devel/automake: 1.7.9-r1, 1.8.5-r3, 1.5, 1.6.3, 1.9.5, 1.4_p6 sys-devel/binutils: 2.15.90.0.3-r4 sys-devel/libtool: 1.5.14 virtual/os-headers: 2.6.8.1-r4 ACCEPT_KEYWORDS="ppc ~ppc" AUTOCLEAN="yes" CFLAGS="-O3 -pipe -mcpu=7450 -mabi=altivec -maltivec -mpowerpc-gfxopt -fno-strict-aliasing" CHOST="powerpc-unknown-linux-gnu"
I'm almost completely sure that the problem comes from the new "NoTapTyping" feature. I set it via powerprefs, as I often touch the panel while typing. Yesterday, when I had several freezes, I noticed that the last one was just after I switched from keyboard to touchpad. I unset it and I have yet to see a freeze. Also, while looking at it, I noticed that /etc/power/pmcs-config contains "-p" in all the disk related scripts. It does not look like a great default setting for moderns drivers, though I think it is not really changing much, except the dmesg is now hda: Enabling Ultra DMA 4 hdc: Enabling MultiWord DMA 2 and it used to be: hda: Set PIO timing for mode 0, reg: 0x0c50032b and nothing for the DVD Just FYI, leaving those options as "" looks safer to me for most modern hardware.
I've finally managed to replicate this on my powerbook, it does seem to be related to the NoTapTyping feature. Turning it off prevents the lockups.
There has been a new release of pbbuttonsd. I've added it to CVS already, but could you test and see if the issue with NoTapType is fixed? The bug occurs so rarely on my machine that I can't confirm. Thanks!
Tried. With NoTapTyping it stayed up for several hours. It froze when I was under batteries, typing and selecting. I sent the output of /proc/cpuinfo, /proc/dev/pmu and lsadb to the upstream developer, per instructions on #gentoo-ppc
pbbuttonsd-0.6.9 has several updates regarding the trackpad. Did you tested that one already?
Yeah, the message #6 was in regards to the new version. So it's still a problem. Santiago, can you try with the vanilla 2.6.12_rc4 kernel to see if it's still locking? There are a number of ppc fixes in this kernel.
I've added pbbuttonsd-0.6.9-r1 to portage which should contain a fix for the race condition that was causing this problem. Please give it a try and let me know how it goes. Thanks!
Sorry, another version bump, 0.6.10 final is in portage now.
0.6.10 froze with vanilla-kernel 2.6.11.7 after about 20 hours with NoTapTyping activated. It was on (low) batteries -> slow clock, powersave, which greatly increases the likelyhood of race conditions. I guess there is still something broken. It involved ALT-TAB to switch windows plus fast mouse move, which is something I do often, and could be related. Tough. I'm switching NoTapTyping off again.
You're right, there was another race condition. Although the author claims that there still is the possibility for more race conditions, it should extremely rare now. This was achieved by reducing the number of ADB messages required for NoTapTyping. Please give 0.7.1 a try and let me know how it works for you. Also, NoTapTyping is off by default now.
I'd like to mark 0.7.1 stable in the near future, if anyone is still having problems with the default configuration please let me know. I realize that NoTapTyping isn't 100% fixed, but I'll add a warning into the ebuild to let people know that this is an experimental option and that it has the potential to cause hard locks.
Warning added and ebuild marked ppc stable. Feel free to reopen if you want to revisit this issue.
