I updated udev from 149 to 150. This made microcode-ctl (I have it in the default runlevel) need a ridiculous amount of time (several minutes instead of under 1 second) to load the CPU's microcode, stopping the boot process for several minutes. Downgrading to udev-149 again fixes the problem.
Created attachment 217005 [details] emerge --info
Same here, "modprobe microcode" take several time. microcode-ctl executable run normally after module loading, so is module loading problem My emerge --info : Portage 2.2_rc61 (default/linux/amd64/10.0/desktop, gcc-4.4.2, glibc-2.11-r1, 2.6.32-gentoo-r1 x86_64) ================================================================= System uname: Linux-2.6.32-gentoo-r1-x86_64-Intel-R-_Core-TM-_i7_CPU_920_@_2.67GHz-with-gentoo-2.0.1 Timestamp of tree: Thu, 21 Jan 2010 01:30:01 +0000 app-shells/bash: 4.0_p35 dev-java/java-config: 2.1.10 dev-lang/python: 2.6.4, 3.1.1-r1 dev-python/pycrypto: 2.1.0 dev-util/cmake: 2.8.0 sys-apps/baselayout: 2.0.1 sys-apps/openrc: 0.6.0-r1 sys-apps/sandbox: 2.2 sys-devel/autoconf: 2.13, 2.65 sys-devel/automake: 1.9.6-r3, 1.10.3, 1.11.1 sys-devel/binutils: 2.20 sys-devel/gcc-config: 1.4.1 sys-devel/libtool: 2.2.6b virtual/os-headers: 2.6.30-r1 ABI="amd64" ACCEPT_KEYWORDS="amd64 ~amd64" ACCEPT_LICENSE="*" ACCEPT_PROPERTIES="*" ALSA_CARDS="ali5451 als4000 atiixp atiixp-modem bt87x ca0106 cmipci emu10k1x ens1370 ens1371 es1938 es1968 fm801 hda-intel intel8x0 intel8x0m maestro3 trident usb-audio via82xx via82xx-modem ymfpci" ALSA_PCM_PLUGINS="adpcm alaw asym copy dmix dshare dsnoop empty extplug file hooks iec958 ioplug ladspa lfloat linear meter mmap_emul mulaw multi null plug rate route share shm softvol" ANT_HOME="/usr/share/ant" APACHE2_MODULES="actions alias auth_basic authn_alias authn_anon authn_dbm authn_default authn_file authz_dbm authz_default authz_groupfile authz_host authz_owner authz_user autoindex cache dav dav_fs dav_lock deflate dir disk_cache env expires ext_filter file_cache filter headers include info log_config logio mem_cache mime mime_magic negotiation rewrite setenvif speling status unique_id userdir usertrack vhost_alias" ARCH="amd64" ASFLAGS_x86="--32" AUTOCLEAN="yes" CBUILD="x86_64-pc-linux-gnu" CDEFINE_amd64="__x86_64__" CDEFINE_x86="__i386__" CFLAGS="-O2 -march=native -msse4.1 -msse4.2 -pipe" CFLAGS_x86="-m32" CHOST="x86_64-pc-linux-gnu" CHOST_amd64="x86_64-pc-linux-gnu" CHOST_x86="i686-pc-linux-gnu" CLEAN_DELAY="5" COLLISION_IGNORE="/lib/modules" COLORTERM="gnome-terminal" CONFIG_PROTECT="/etc /usr/share/X11/xkb /var/lib/hsqldb" CONFIG_PROTECT_MASK="/etc/ca-certificates.conf /etc/env.d /etc/env.d/java/ /etc/fonts/fonts.conf /etc/gconf /etc/gentoo-release /etc/revdep-rebuild /etc/sandbox.d /etc/splash /etc/terminfo /etc/texmf/language.dat.d /etc/texmf/language.def.d /etc/texmf/updmap.d /etc/texmf/web2c" CVS_RSH="ssh" CXXFLAGS="-O2 -pipe" DEFAULT_ABI="amd64" DISPLAY=":0.0" DISTDIR="/usr/portage/distfiles" EDITOR="/usr/bin/vi" ELIBC="glibc" EMERGE_DEFAULT_OPTS="--verbose --with-bdeps=y" EMERGE_WARNING_DELAY="10" EPREFIX="" EROOT="/" FEATURES="assume-digests distlocks fixpackages news parallel-fetch preserve-libs protect-owned sandbox sfperms strict unmerge-logs unmerge-orphans userfetch" FETCHCOMMAND="/usr/bin/wget -t 5 -T 60 --passive-ftp -O "${DISTDIR}/${FILE}" "${URI}"" GCC_SPECS="" GDK_USE_XFT="1" GENTOO_MIRRORS="ftp://ftp.free.fr/mirrors/ftp.gentoo.org/ ftp://ftp.belnet.be/mirror/rsync.gentoo.org/gentoo/ ftp://mirror.switch.ch/mirror/gentoo/" GTK_IM_MODULE="uim" HISTFILE="/home/maxime/.history" HISTSIZE="1000" HOME="/home/maxime" INFOPATH="/usr/share/info:/usr/share/binutils-data/x86_64-pc-linux-gnu/2.20/info:/usr/share/gcc-data/x86_64-pc-linux-gnu/4.4.2/info" INPUT_DEVICES="evdev" JAVAC="/etc/java-config-2/current-system-vm/bin/javac" JAVACC_HOME="/usr/share/javacc/" JAVA_HOME="/etc/java-config-2/current-system-vm" JDK_HOME="/etc/java-config-2/current-system-vm" KERNEL="linux" LANG="fr_FR.utf8" LANGUAGE="fr" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" LC_COLLATE="ja_JP.utf8" LDFLAGS="-Wl,--hash-style=both" LDFLAGS_x86="-m elf_i386" LESS="-R -M --shift 5" LESSCHARSET="utf-8" LESSOPEN="|lesspipe.sh %s" LIBDIR_amd64="lib64" LIBDIR_amd64_fbsd="lib64" LIBDIR_ppc="lib32" LIBDIR_ppc64="lib64" LIBDIR_sparc32="lib32" LIBDIR_sparc64="lib64" LIBDIR_x86="lib32" LIBDIR_x86_fbsd="lib32" LINGUAS="fr ja" LOGNAME="root" MAKEOPTS="-j9" MANPATH="/etc/java-config-2/current-system-vm/man:/usr/local/share/man:/usr/share/man:/usr/share/binutils-data/x86_64-pc-linux-gnu/2.20/man:/usr/share/gcc-data/x86_64-pc-linux-gnu/4.4.2/man:/etc/java-config/system-vm/man/" MULTILIB_ABIS="amd64 x86" MULTILIB_STRICT_DENY="64-bit.*shared object" MULTILIB_STRICT_DIRS="/lib32 /lib /usr/lib32 /usr/lib /usr/kde/*/lib32 /usr/kde/*/lib /usr/qt/*/lib32 /usr/qt/*/lib /usr/X11R6/lib32 /usr/X11R6/lib" MULTILIB_STRICT_EXEMPT="(perl5|gcc|gcc-lib|binutils|eclipse-3|debug|portage)" NETBEANS="apisupport cnd groovy gsf harness ide identity j2ee java mobility nb php profiler soa visualweb webcommon websvccommon xml" OLDPWD="/home/maxime" OPENGL_PROFILE="nvidia" PAGER="/usr/bin/less" PATH="/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/opt/bin:/usr/x86_64-pc-linux-gnu/gcc-bin/4.4.2" PKGDIR="/usr/portage/packages" PORTAGE_ARCHLIST="ppc x86-openbsd ppc-openbsd ppc64 x86-winnt x86-fbsd ppc-aix alpha arm x86-freebsd s390 amd64 arm-linux x86-macos x64-openbsd ia64-hpux hppa x86-netbsd amd64-linux ia64-linux x86 sparc-solaris x64-freebsd sparc64-solaris x86-linux x64-macos sparc m68k-mint ia64 mips ppc-macos x86-interix hppa-hpux amd64-fbsd x64-solaris mips-irix m68k sh x86-solaris sparc-fbsd" PORTAGE_BINHOST_CHUNKSIZE="3000" PORTAGE_BIN_PATH="/usr/lib64/portage/bin" PORTAGE_COMPRESS_EXCLUDE_SUFFIXES="css gif htm[l]? jp[e]?g js pdf png" PORTAGE_CONFIGROOT="/" PORTAGE_DEBUG="0" PORTAGE_DEPCACHEDIR="/var/cache/edb/dep" PORTAGE_ELOG_CLASSES="log warn error" PORTAGE_ELOG_MAILFROM="portage@neurone-network.org" PORTAGE_ELOG_MAILSUBJECT="[portage] ebuild log for ${PACKAGE} on ${HOST}" PORTAGE_ELOG_MAILURI="root" PORTAGE_ELOG_SYSTEM="echo mail" PORTAGE_FETCH_CHECKSUM_TRY_MIRRORS="5" PORTAGE_FETCH_RESUME_MIN_SIZE="350K" PORTAGE_GID="250" PORTAGE_INST_GID="0" PORTAGE_INST_UID="0" PORTAGE_NICENESS="15" PORTAGE_PYM_PATH="/usr/lib64/portage/pym" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --stats --timeout=180 --exclude=/distfiles --exclude=/local --exclude=/packages" PORTAGE_RSYNC_RETRIES="3" PORTAGE_TMPDIR="/var/tmp" PORTAGE_VERBOSE="1" PORTAGE_WORKDIR_MODE="0700" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage/layman/xwing /usr/local/portage/layman/desktop-effects /usr/local/portage/local" PRELINK_PATH_MASK="/usr/lib64/klibc" PROFILE_ONLY_VARIABLES="ARCH ELIBC KERNEL USERLAND" PWD="/home/maxime" QT_IM_MODULE="uim" RESUMECOMMAND="/usr/bin/wget -c -t 5 -T 60 --passive-ftp -O "${DISTDIR}/${FILE}" "${URI}"" ROOT="/" ROOTPATH="/opt/bin:/usr/x86_64-pc-linux-gnu/gcc-bin/4.4.2" RPMDIR="/usr/portage/rpm" RUBYOPT="-rauto_gem" RUBY_TARGETS="ruby18" SANE_BACKENDS="plustek" SANE_CONFIG_DIR="/etc/sane.d" SAVEHIST="1000" SHELL="/bin/zsh" SHLVL="1" STAGE1_USE="multilib nptl nptlonly unicode" SUDO_COMMAND="/bin/zsh" SUDO_GID="1000" SUDO_UID="1000" SUDO_USER="maxime" SYMLINK_LIB="yes" SYNC="rsync://rsync.gentoo.org/gentoo-portage" TERM="xterm" TEXINPUTS="~/.latex/packages//:" USB_DEVFS_PATH="/dev/bus/usb" USE="amd64 multilib" ALSA_CARDS="ali5451 als4000 atiixp atiixp-modem bt87x ca0106 cmipci emu10k1x ens1370 ens1371 es1938 es1968 fm801 hda-intel intel8x0 intel8x0m maestro3 trident usb-audio via82xx via82xx-modem ymfpci" ALSA_PCM_PLUGINS="adpcm alaw asym copy dmix dshare dsnoop empty extplug file hooks iec958 ioplug ladspa lfloat linear meter mmap_emul mulaw multi null plug rate route share shm softvol" APACHE2_MODULES="actions alias auth_basic authn_alias authn_anon authn_dbm authn_default authn_file authz_dbm authz_default authz_groupfile authz_host authz_owner authz_user autoindex cache dav dav_fs dav_lock deflate dir disk_cache env expires ext_filter file_cache filter headers include info log_config logio mem_cache mime mime_magic negotiation rewrite setenvif speling status unique_id userdir usertrack vhost_alias" ELIBC="glibc" INPUT_DEVICES="evdev" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" LINGUAS="fr ja" RUBY_TARGETS="ruby18" SANE_BACKENDS="plustek" USERLAND="GNU" VIDEO_CARDS="nvidia" USER="root" USERLAND="GNU" USERNAME="root" USE_EXPAND="ALSA_CARDS ALSA_PCM_PLUGINS APACHE2_MODULES APACHE2_MPMS CAMERAS CROSSCOMPILE_OPTS DVB_CARDS ELIBC FCDSL_CARDS FOO2ZJS_DEVICES FRITZCAPI_CARDS INPUT_DEVICES KERNEL LCD_DEVICES LINGUAS LIRC_DEVICES MISDN_CARDS NETBEANS_MODULES QEMU_SOFTMMU_TARGETS QEMU_USER_TARGETS RUBY_TARGETS SANE_BACKENDS USERLAND VIDEO_CARDS" USE_EXPAND_HIDDEN="CROSSCOMPILE_OPTS ELIBC KERNEL USERLAND" USE_ORDER="env:pkg:conf:defaults:pkginternal:env.d" VIDEO_CARDS="nvidia" VISUAL="/usr/bin/vi" XAUTHORITY="/home/maxime/.Xauthority" XDG_CONFIG_DIRS="/etc/xdg" XDG_DATA_DIRS="/usr/local/share:/usr/share:/usr/share/gdm" XMODIFIERS="@im=uim" _="/usr/bin/emerge"
(In reply to comment #0) > I updated udev from 149 to 150. This made microcode-ctl (I have it in the > default runlevel) need a ridiculous amount of time (several minutes instead of > under 1 second) to load the CPU's microcode, stopping the boot process for > several minutes. > > Downgrading to udev-149 again fixes the problem. > Same thing on ~x86. The modprobe of the microcode module is what is taking too much time. Verified this manually. Downgrade of udev from 150 to 149 fixes the problem.
Same issue here. As a temporary hack-around, cd /lib/firmware ; mkdir intel-ucode ; cd intel-ucode ; ln -s /lib/microcode.dat 06-0f-02 removes the delay. The diff between releases 149 and 150 of udev shows a relatively large change of the code that handles firmware, though I haven’t found any specific spot there that’d cause such a difference in behaviour. The delay on my machine was exactly 60 seconds, by the way (as reported by udevmonitor). Relevant /proc/config.gz lines: CONFIG_MICROCODE=m CONFIG_MICROCODE_INTEL=y # CONFIG_MICROCODE_AMD is not set CONFIG_MICROCODE_OLD_INTERFACE=y System: Linux thing 2.6.31-tuxonice-r1 #3 SMP Tue Sep 29 14:06:32 CEST 2009 x86_64 Intel(R) Core(TM)2 CPU T5600 @ 1.83GHz GenuineIntel GNU/Linux I take it the microcode loading is being taken care of by udev and the kernel module now, and the microcode_ctl userspace tool isn’t needed anymore?
(In reply to comment #4) > cd /lib/firmware ; mkdir intel-ucode ; cd intel-ucode ; ln -s > /lib/microcode.dat 06-0f-02 Do you mean "ln -s /lib/firmware/microcode.dat 06-0f-02"? > I take it the microcode loading is being taken care of by udev and the kernel > module now, and the microcode_ctl userspace tool isn’t needed anymore? Interesting. Why do you believe this?
(In reply to comment #5) Sorry, I didn’t think through my previous comment enough. > Do you mean "ln -s /lib/firmware/microcode.dat 06-0f-02"? I did. Actually, though, the hex digits will vary for people, depending on the CPU model (check /dev/.udev/firmware-missing...), and a symlink is the wrong thing to do here. Just ‘touch c0-ff-ee’ in /lib/firmware/intel-ucode (for the right value of c0ffee) to satisfy udev. > Interesting. Why do you believe this? Because I was lazy and didn’t read the source code. :) Having done that... * microcode_ctl just reads hex numbers from one file and writes the corresponding bytes to another file. * microcode.ko does the magic of putting those bytes in the CPU. If one dumped the output of microcode_ctl in a regular file, one could just dd that on boot into the microcode device for the same effect as running microcode_ctl. In fact... * microcode.ko actually tries to get the data itself, at microcode_intel.c:428: sprintf(name, "intel-ucode/%02x-%02x-%02x", c->x86, c->x86_model, c->x86_mask); if (request_firmware(&firmware, name, device)) { * udev seems to know about that, but I’m too lazy to find out how. * udev proceeds to start its ‘firmware’ binary (formerly the ‘firmware.sh’ script) and asks it to dump /lib/firmware/intel-ucode/c0-ff-fe into /dev/cpu/microcode However, that first file doesn’t exist... on my system, anyway, which somehow causes udev to stop for a minute. If it does exist but does not contain the data the module expects (e. g. when it’s a link to microcode.dat), the module just complains in dmesg and does nothing (and udev doesn’t wait). As far as I understand things, it’d have to contain exactly the output of microcode_ctl for udev’s firmware loader to work right. And then microcode_ctl wouldn’t be needed at boot anymore. In fact, microcode_ctl could be adequately replaced by a two-line awk or perl script, in src_compile of sys-apps/microcode-data, after careful thinking about endianness... But there’s not really a reason to do that. Sorry if this was off topic, but I wanted at least to correct my earlier statements.
(In reply to comment #4) > Sorry if this was off topic, but I wanted at least to correct my earlier > statements. Thanks for the insight, which I think is relevant.
That sounds exactly like this report: http://thread.gmane.org/gmane.linux.hotplug.devel/15302 Could you check if the attached patch works?
Created attachment 218048 [details, diff] udev-151-missing-firmware.diff The patch works for me. Added as an attachment (to correct an artificially inserted newline in the post, which causes the patch to fail).
The patch works here too. No more 60 seconds delay per CPU :) Closing as FIXED.
(In reply to comment #10) > The patch works here too. No more 60 seconds delay per CPU :) > > Closing as FIXED. > How is this fixed if the patch is not in portage?
It's fixed upstream :P
Fixed in udev-150 and udev-151 by adding -r1 for each.
