April 2011 - openSUSE Kernel - openSUSE Mailing Lists

[opensuse-kernel] [patch #2 0/5] scripts/sequence-patch.sh: Add some useful options
by Jeff Mahoney 05 Oct '11

05 Oct '11

Hi all - There are a few things I do every time I expand the source tree with sequence-patch.sh. Or, rather, things I should do and forget to do a lot. The first I always remember to do: Copy the config I want to test with to my build directory. The second I forget more often than not: Expand the kABI references into my build directory. One of the things that bites me occasionally is when there are kABI changes that go un-noticed until the automated build checker flags them and sends me an email yelling at me for my carelessness. This patch adds three options to scripts/sequence-patch.sh to take care of these automatically. --build-dir=PATH allows you to specify the build directory you'll be using. It will create the directory if it doesn't exist. It defaults to $PATCH_DIR. --config=ARCH-FLAVOR allows you to specify the config you'll use. It will copy the config to your build directory. It is not a fatal error if the config does not exist. It will just not copy one. --kabi allows you to direct sequence-patch to expand the kABI references into your build directory. Since the references are arch/flavor specific, it depends on --config=ARCH-FLAVOR. It is not a fatal error if the kABI reference doesn't exist since this is always the case in the master branch. The --build-dir PATH is subject to shell expansion. It is expanded late enough in the script that most variables should be available. There are also three new variables available when --config is used: $CONFIG contains the --config contents. $CONFIG_ARCH contains the arch part of the config. $CONFIG_FLAVOR contains the flavor part. For reference, I have this in my .bashrc: export SEQUENCE_PATCH_ARGS="--config=x86_64-desktop --build-dir=~/src/scratch/build/\$CONFIG_ARCH/linux-\$SRCVERSION-\$TAG-\$CONFIG_FLAVOR --kabi" I've also added patches for --ctags and --cscope which will automatically generate the databases for either of those tools in the build directory after the source tree is expanded. - Issues addressed in patchset #2: - Help text is now a here document for easy addition of more documentation - Explanation of which variables are allowed by --build-dir - Missing rpm/modversions script is not an error on older branches - Missing symvers files are not errors - Uses true/false for simple feature flags -Jeff -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-kernel+help(a)opensuse.org

5 13

[opensuse-kernel] openSUSE Kernel: Push Patches Upstream
by Jeff Mahoney 26 May '11

26 May '11

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi guys - We've been carrying the following patches in the openSUSE kernel for some time and there's no need for them to be openSUSE-specific. Can you review the patches below and either remove them, submit them upstream, or at least at a signed-off-by if there isn't one already so I can do so? Even though there are patches that only affect architectures which we don't support on openSUSE and can be dropped as far as openSUSE is concerned, the Factory kernel will eventually be branched for the SLE12 kernel. Getting rid of them now only makes things easier later. Alexander Graf: - - patches.arch/kvm-split-paravirt-ops-by-functionality - - patches.arch/kvm-only-export-selected-pv-ops-feature-structs - - patches.arch/kvm-split-the-KVM-pv-ops-support-by-feature - - patches.arch/kvm-replace-kvm-io-delay-pv-ops-with-linux-magic [these 4 have been disabled since 2.6.38-rc1] - - patches.fixes/kvm-ioapic.patch - - patches.fixes/kvm-macos.patch Brandon Philips: - - patches.drivers/bnx2-entropy-source.patch - - patches.drivers/e1000-entropy-source.patch - - patches.drivers/e1000e-entropy-source.patch - - patches.drivers/igb-entropy-source.patch - - patches.drivers/ixgbe-entropy-source.patch - - patches.drivers/tg3-entropy-source.patch - - patches.drivers/tg3-5785-and-57780-asic-revs-not-working.patch - - patches.suse/uvcvideo-ignore-hue-control-for-5986-0241.patch Greg Haskins: - - patches.fixes/ia64-configure-HAVE_UNSTABLE_SCHED_CLOCK-for-SGI_SN.patch Greg KH: - - patches.suse/linux-2.6.29-dont-wait-for-mouse.patch - - patches.suse/linux-2.6.29-even-faster-kms.patch - - patches.suse/linux-2.6.29-kms-after-sata.patch - - patches.suse/linux-2.6.29-jbd-longer-commit-interval.patch - - patches.suse/linux-2.6.29-touchkit.patch Hannes Reinecke: (or James Bottomley) - - patches.suse/kbd-ignore-gfx.patch - - patches.fixes/scsi-inquiry-too-short-ratelimit - - patches.suse/scsi-netlink-ml - - patches.fixes/scsi-dh-queuedata-accessors - - patches.fixes/scsi-dh-alua-retry-UA - - patches.fixes/scsi-add-tgps-setting [disabled since 2.6.39] - - patches.fixes/scsi-dh-alua-send-stpg - - patches.fixes/scsi-dh-rdac-add-stk - - patches.fixes/scsi-retry-alua-transition-in-progress - - patches.fixes/scsi-check-host-lookup-failure - - patches.drivers/megaraid-mbox-fix-SG_IO - - patches.suse/scsi-error-test-unit-ready-timeout [Kurt] - - patches.fixes/scsi-scan-blist-update - - patches.suse/no-partition-scan [Is this even necessary since we boot with 'quiet' as default now?] - - patches.suse/dm-emulate-blkrrpart-ioctl - - patches.fixes/dm-mpath-reattach-dh - - patches.suse/dm-mpath-leastpending-path-update [Has this been replaced by dm-queue-length upstream?] - - patches.suse/dm-mpath-accept-failed-paths [disabled since 2.6.38] - - patches.suse/dm-mpath-detach-existing-hardware-handler [disabled since 2.6.38] - - patches.fixes/dm-table-switch-to-readonly - - patches.suse/dm-mpath-evaluate-request-result-and-sense [disabled since 2.6.39] - - patches.fixes/dm-release-map_lock-before-set_disk_ro [added by Nikanth] - - patches.suse/dm-mpath-no-activate-for-offlined-paths [and patches.suse/mpath-fix] - - patches.suse/dm-mpath-no-partitions-feature - - Jan Beulich: - - patches.fixes/bridge-module-get-put.patch Jan Kara: - - patches.suse/readahead-request-tunables.patch - - patches.suse/raw_device_max_minors_param.diff Jeff Mahoney: [deleted a bunch] - - patches.fixes/netfilter-implement-rfc-1123-for-ftp-conntrack - - patches.fixes/proc-scsi-scsi-fix.diff - - patches.fixes/flexcop-fix-registering-braindead-stupid-names - - patches.fixes/misdn-add-support-for-group-membership-check - - patches.suse/dm-raid45-26-Nov-2009.patch [Unsure what's happening here. It hasn't been updated in ages.] Jiri Benc: - - patches.suse/panic-on-io-nmi-SLE11-user-space-api.patch [disabled since 2.6.33] [does anyone even use the binary sysctl interface for this?] Jiri Bohac: - - patches.arch/x86_64-hpet-64bit-timer.patch [disabled since 2.6.33] - - patches.suse/netfilter-ip_conntrack_slp.patch Jiri Kosina: - - patches.drivers/elousb.patch - - patches.fixes/input-add-acer-aspire-5710-to-nomux.patch - - patches.fixes/hid-add-noget-quirk-for-symboltec.patch Jiri Slaby: - - patches.suse/wireless-no-aes-select - - patches.fixes/iwlwifi-fix-tx-power-configuration-on-3945-and-4965-devices Mel Gorman / Michal Hocko: - - patches.fixes/grab-swap-token-oops - - patches.suse/files-slab-rcu.patch - - patches.suse/mm-devzero-optimisation.patch [against SLES10, issue may not exist anymore] - - patches.suse/mm-devzero-optimisation.patch - - patches.fixes/aggressive-zone-reclaim.patch [disabled since 2.6.36] Neil Brown: - - patches.fixes/nfs-slot-table-alloc - - patches.fixes/nfsd-06-sunrpc-cache-retry-cache-lookups-that-return-ETIMEDO.patch [disabled since 2.6.37] Olaf Hering: - - patches.suse/led_classdev.sysfs-name.patch [ppc] - - patches.suse/radeon-monitor-jsxx-quirk.patch [ppc] - - patches.suse/8250-sysrq-ctrl_o.patch [ppc] - - patches.suse/led_classdev.sysfs-name.patch [ppc] - - patches.arch/ppc-pegasos-console-autodetection.patch [ppc] - - patches.suse/ppc-powerbook-usb-fn-key-default.patch [ppc] - - patches.drivers/ppc64-adb [ppc] - - patches.suse/suse-ppc64-branding [ppc] - - patches.arch/ppc64-xmon-dmesg-printing.patch [ppc] - - patches.arch/ppc-prom-nodisplay.patch [ppc] - - patches.fixes/ptrace-getsiginfo [technically schwab] [ppc] - - patches.fixes/scsi-ibmvscsi-show-config.patch - - patches.fixes/scsi-ibmvscsi-module_alias.patch - - patches.drivers/ehea-modinfo.patch Olaf Kirch: - - patches.fixes/remount-no-shrink-dcache - - patches.fixes/tulip-quad-NIC-ifdown - - patches.fixes/parport-mutex [This doesn't seem like the right fix - the driver looks like it needs help.] Oliver Neukum: - - patches.fixes/sd_liberal_28_sense_invalid.diff Rafael Wysocki: - - patches.arch/x86-mcp51-no-dac [rejected by Alan Cox, Mar 2009] - - patches.fixes/tg3-fix-default-wol.patch Suresh Jayaraman: - - patches.suse/sched-revert-latency-defaults [disabled since 2.6.33, after SLE11SP1] - - patches.suse/mm-tune-dirty-limits.patch [we may want to revisit per-flavor sysctl files] Takashi Iwai: - - patches.arch/ppc-ipic-suspend-without-83xx-fix [ppc] - - patches.arch/x86-hpet-pre-read [disabled since 2.6.37] - - patches.drivers/input-Add-LED-support-to-Synaptics-device - - patches.drivers/alsa-hda-0019-Increase-default-buffer-size Thomas Renninger: - - patches.arch/x86-apic-force-bigsmp-apic-on-IBM-EXA3-4.patch - - patches.arch/acpi_thinkpad_introduce_acpi_root_table_boot_param.patch - - patches.arch/acpi_thermal_passive_blacklist.patch - - patches.arch/acpi_fix_fadt_32_bit_zero_length.patch - - patches.arch/acpi_ec_provide_non_interrupt_mode_boot_param.patch [disabled since 2.6.32] - - patches.arch/acpi_srat-pxm-rev-store.patch - - patches.arch/acpi_srat-pxm-rev-ia64.patch - - patches.arch/acpi_srat-pxm-rev-x86-64.patch - - patches.fixes/acpi_ec_sys_access_user_space_with_get_user.patch - - patches.fixes/cpufreq_ondemand_performance_optimise_default_settings.patch - - patches.arch/perf_timechart_fix_zero_timestamps.patch - - patches.fixes/intel_idle_lapic_param.patch [less than two weeks old] - - patches.fixes/intel_idle_add_flush_tlb_param.patch [less than two weeks old] Tony Jones: - - patches.suse/kdump-dump_after_notifier.patch - - patches.fixes/oprofile_bios_ctr.patch [ Up for discussion ] - - patches.suse/file-capabilities-disable-by-default.diff [never submitted] - - patches.suse/connector-read-mostly - - patches.suse/ext3-barrier-default - - patches.suse/reiserfs-barrier-default - - patches.suse/SUSE-bootsplash Thanks. - -Jeff - -- Jeff Mahoney SUSE Labs -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.17 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org/ iEYEARECAAYFAk25pVEACgkQLPWxlyuTD7JC5gCdGl2k8d2PeHQVaF5WO5Yv25V+ PE8An07vC7R5YH5iJWHhKZUJDIzuVUHG =W1WH -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-kernel+help(a)opensuse.org

20 49

[opensuse-kernel] RFC: Use tarballs for -rc and stable releases?
by Michal Marek 04 May '11

04 May '11

Hi, currently, -rc and stable release updates are done by adding patches to patches.kernel.org/. Especially the -rc1 patches are *huge* and cannot be paired with any other object in the repository, so their gzip-compressed size adds to the total repository size: $ git cat-file -p b9044e3353ff21f12a86788a18cf36ba1071b6ab:patches.kernel.org/patch-2.6.37-rc1 | gzip -c -9 | wc -c 14627847 that is almost 14 megabytes. So I would like to propose to more or less discontinue patches.kernel.org and instead use the SRCVERSION variable in rpm/config.sh. This is already used in the vanilla and linux-next branches and it works. Code changes are more easily reviewed in the kernel.git nowadays, tar-up.sh and sequence-patch also work if you don't have the tarball around (it generates one from git, which is arguably slower, but works), so I think it's worth doing it. Opinions? Michal -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-kernel+help(a)opensuse.org

2 9

[opensuse-kernel] GCC 4.6 for 12.1
by Jeff Mahoney 02 May '11

02 May '11

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi all - I know someone else mentioned this briefly WRT to Tumbleweed, but I'd like to bring up for discussion the adoption of gcc 4.6 for Factory. A bit of background. We currently maintain a kernel-trace flavor to allow people to make use of the tracing facilities. Historically, there has been overhead associated with these, which is why they were in their own flavor. Most users don't care about them, so there's no sense in slowing everyone down to satisfy them. Over time, the maintainers of these facilities have worked to lower the overhead to the point where now dynamic ftrace only adds memory overhead. It uses gcc's profiling facility to call out to a function at the beginning of every function call to allow tracking. At boot, this is a function that returns immediately but during the boot process those calls are hot-patched to nop operations so that they carry virtually zero runtime overhead. This sounds great, right? There's one last bit, and that is actually the reason why ftrace is still disabled in our regular kernel flavors. The compiler's profiling feature calls the profiling function after the function prologue. That means that the stack pointer has already been advanced and since it can advance a different amount based on the needs of each function, we don't have an easy way to get the caller's stack frame back. It's needed to resolve where the caller was called from and the only way to do that is to dedicate a register to track the start of the current stack frame. Dedicating a register for this increases register contention and forces more accesses to CPU cache or main memory, slowing everything down. So, it's not worth it when most users will never actually take advantage of the feature that requires it. For some time, I've been trying to come up with ways to work around this so that the fast path stays fast and the tracing path takes any performance hit necessary. Until recently, the performance hit would've been too big to make it worthwhile. With gcc 4.6, we have the -mfentry option which moves the call to the profiling function *before* the function's prologue. This means that the caller's stack frame is still intact and we can resolve the caller's parent without issue. I've spent the past few days working up a proof of concept that takes advantage of this and doesn't require a dedicated register. It turns out that Steven Rostedt has already posted a patch set to do exactly this and he's targeting 2.6.40 for inclusion in the mainline kernel. For now it looks like it's x86-64 only but it should be ok for i386 as well. This would go a long way to eliminate the -trace flavor from our kernel packages. The benefits are multiple: It eliminates the maintenance of another kernel flavor, but it also allows everyone to use tracing facilities without installing a separate kernel release to do so. One last bit is needed, though, and that is for gcc to allow using -pg - -mfentry with -fomit-frame-pointers. Once that bit is taken care of then we can rid the kernel of the frame pointer requirement for ftrace and eliminate the -trace flavor entirely. - -Jeff - -- Jeff Mahoney SUSE Labs -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.17 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org/ iEYEARECAAYFAk263RwACgkQLPWxlyuTD7JhyQCfXcCcDjH0dYmnoguTCe+7J2FH YWkAn3+K1qintX3acivGSmUYCDVnDuqi =3BMz -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-kernel+help(a)opensuse.org

8 13

[opensuse-kernel] Configuration error for kernel 2.6.37 in 11.4
by Larry Finger 30 Apr '11

30 Apr '11

The driver rt2800pci does not seem to be built for kernel 2.6.37. The output of zcat /proc/config.gz | grep RT2800 yields: CONFIG_RT2800PCI_PCI=y # CONFIG_RT2800PCI is not set CONFIG_RT2800USB=m # CONFIG_RT2800USB_RT30XX is not set # CONFIG_RT2800USB_RT35XX is not set # CONFIG_RT2800USB_UNKNOWN is not set CONFIG_RT2800_LIB=m The first two lines are contradictory, but I believe the second as the Ralink RT3090, which uses rt2800pci, does not work, but does when compat-wireless is loaded. If CONFIG_RT2800PCI were defined, there should also be entries for CONFIG_RT2800PCI_RT30XX, CONFIG_RT2800PCI_RT30XX, and CONFIG_RT2800PCI_UNKNOWN. Thanks, Larry -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-kernel+help(a)opensuse.org

2 4

[opensuse-kernel] Trace flavor still needed?
by Jeff Mahoney 25 Apr '11

25 Apr '11

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi all - Bnc#656091 asks about enabling UTRACE for other kernel flavors. It's too late to do it for 11.4 since enabling it modifies struct task_struct. However, most of the trace infrastructure appears to use jump labels now, which means that they're essentially free when not in use. I'd like to eliminate the trace flavor for the 12.1 release. At this point the only differences are ftrace and utrace. Has anyone done the legwork to verify the performance impact here? I suppose since it's only in Factory for now we can just change it and see what shakes out. Opinions? - -Jeff - -- Jeff Mahoney SUSE Labs -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.17 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org/ iEYEARECAAYFAk2t5QYACgkQLPWxlyuTD7LCEgCbBKbn7GcGfG2wtPSKzdyWd1X1 DOIAnijd/HP8jDvRxsgfgJvNoj34Q52t =feDe -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-kernel+help(a)opensuse.org

3 3

[opensuse-kernel] Updated kernel for 11.4
by Larry Finger 23 Apr '11

23 Apr '11

Is there a time estimate on when the 2.6.37.1 kernel in 11.4 will be updated? The reason for asking is that there is a bug that causes a kernel panic when b44 is used on systems with more than 1 GB of RAM. Commit 960ac8917 fixes this problem in 2.6.37.2. This bug is the source of the system freeze reported in Bugzilla #686492. Manually applying the patch allows the system to run. Thanks, Larry -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-kernel+help(a)opensuse.org

4 5

[opensuse-kernel] T400 Overheating
by Sankar P 15 Apr '11

15 Apr '11

Hi, Sorry for the little long mail, I wanted to explain how I got into the issue. I am using a Thinkpad T400 which feels a lot hot easily even when there is not much CPU usage. I suspected it could be a fan problem but the hardware turned to be not problematic. The machine started shutting down automatically few times because of over-heating. I became so ambitious and thought that I could write a notification system which will warn the user when the temperature exceeds a critical temperature instead of automatically shutting down. I did some googling and reading of kernel source documentation. I found that such a mechanism already exists (IIUC) (drivers/thermal/thermal_sys.c) : === case THERMAL_TRIP_CRITICAL: if (temp >= trip_temp) { if (tz->ops->notify) ret = tz->ops->notify(tz, count, trip_type); if (!ret) { printk(KERN_EMERG "Critical temperature reached (%ld C), shutting down.\n", temp/1000); orderly_poweroff(true); === However, on reading further I found that sysfs already shows the temperature settings under /sys/class/thermal (which as I understand has moved from /proc recently). Now when I saw the value for: /sys/class/thermal/cooling_device0/cur_state is 0 /sys/class/thermal/cooling_device0/max_state is 7 /sys/class/thermal/thermal_zone0/trip_point_type is critical /sys/class/thermal/thermal_zone0/temp is 70000 now my questions are: 1) Why could the cooling_device's cur_state be never going more than 0 ? I believe this could be a reason why the machine becomes hot. (I may be wrong) I tried setting the cur_state of cooling device to 5 to see if it will increase the fan speed. But trying to overwrite this file never succeeds. The documentation does not say it is a readonly value also (as opposed to max_state which is fixed constant at device registration time) 2) Are there any known standards (or userspace tools), that make use of the notification mechanism provided by the sysfs thermal settings ? (For scenarios like email someone when a machine in datacenter overheats and it may go down anytime, etc.) 3) gnome-system-monitor currently shows usage for each CPU. Is it reliable to make use of the /sys/class/thermal/thermal_zone0/temp file to indicate the value as temperature for the CPU in gnome-system-monitor ? Please let me know if I have made any glaring mistakes and I will file a bug for the T400 overheating if needed. Thanks a lot. -- Sankar P http://psankar.blogspot.com -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-kernel+help(a)opensuse.org

2 1

[opensuse-kernel] Using ext4 driver for ext2 filesystem in Factory
by Jan Kara 15 Apr '11

15 Apr '11

Hello, we are figuring out upstream how to eventually get rid of maintaing three code bases of ext2, ext3, and ext4 in parallel. These days ext4 kernel driver is able to handle both ext2 and ext3 filesystems in a backward compatible manner - i.e., you can mount, access, modify the filesystem with ext4 driver and then still be able to use it with the old ext2 driver. So I'd think that we could try changing config of our kernel in Factory so that ext4 driver is used for ext2 filesystem and see whether something breaks or not. Do people have any opinion on this? Honza -- Jan Kara <jack(a)suse.cz> SUSE Labs, CR -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-kernel+help(a)opensuse.org

9 14

[opensuse-kernel] What is the purpose of kernel-base?
by Stefan Seyfried 14 Apr '11

14 Apr '11

Hi, see $SUBJECT It seems totally useless. I seemed to remember that it was supposed to be used for virtualized guests, but it does not even have virtio included, so that can't be it... Anyone here who can enlighten me? -- Stefan Seyfried "Dispatch war rocket Ajax to bring back his body!" -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-kernel+help(a)opensuse.org

2 2