I was wondering whether we could apply the 2.6.18.8 patch to the linux tree? Assuming that that patch makes 2.6.18 better... It should be fairly simple. Only two hunks dont apply (both in net/core/skbuff.c) and those changes seem to already have been applied anyway. Stefan Here the dry-run: patching file Documentation/dontdiff patching file Documentation/sysctl/vm.txt patching file Makefile patching file arch/alpha/Kconfig patching file arch/arm/kernel/calls.S patching file arch/i386/Kconfig.cpu patching file arch/i386/kernel/alternative.c patching file arch/i386/kernel/cpu/cpufreq/acpi-cpufreq.c patching file arch/i386/kernel/microcode.c patching file arch/i386/kernel/smpboot.c patching file arch/i386/mm/boot_ioremap.c patching file arch/i386/pci/irq.c patching file arch/ia64/kernel/acpi.c Hunk #1 succeeded at 777 (offset 6 lines). patching file arch/ia64/kernel/numa.c patching file arch/ia64/kernel/topology.c patching file arch/ia64/sn/kernel/bte.c patching file arch/m32r/kernel/entry.S patching file arch/powerpc/Kconfig Hunk #1 succeeded at 740 (offset 11 lines). patching file arch/powerpc/configs/pseries_defconfig patching file arch/powerpc/kernel/traps.c patching file arch/ppc/kernel/traps.c patching file arch/s390/Kconfig patching file arch/s390/lib/Makefile patching file arch/s390/lib/div64.c patching file arch/s390/lib/uaccess.S patching file arch/s390/lib/uaccess64.S patching file arch/sh/kernel/process.c patching file arch/sparc/kernel/entry.S patching file arch/sparc64/kernel/central.c patching file arch/sparc64/kernel/entry.S patching file arch/sparc64/kernel/of_device.c patching file arch/sparc64/kernel/pci_common.c patching file arch/sparc64/kernel/pci_iommu.c patching file arch/sparc64/kernel/pci_sabre.c patching file arch/sparc64/kernel/prom.c patching file arch/sparc64/kernel/time.c patching file arch/sparc64/mm/init.c patching file arch/um/Kconfig patching file arch/um/Makefile-x86_64 patching file arch/um/include/common-offsets.h patching file arch/um/include/kern_util.h patching file arch/um/include/sysdep-i386/kernel-offsets.h patching file arch/um/include/sysdep-x86_64/kernel-offsets.h patching file arch/um/os-Linux/process.c patching file arch/um/os-Linux/sys-i386/tls.c patching file arch/um/os-Linux/tls.c patching file arch/um/sys-x86_64/stub_segv.c patching file arch/x86_64/ia32/ptrace32.c patching file arch/x86_64/kernel/entry.S patching file arch/x86_64/kernel/pci-calgary.c patching file arch/x86_64/kernel/process.c patching file arch/x86_64/kernel/setup.c patching file arch/x86_64/kernel/setup64.c patching file arch/x86_64/kernel/time.c patching file block/elevator.c patching file block/ll_rw_blk.c patching file block/scsi_ioctl.c patching file drivers/block/DAC960.c patching file drivers/block/cciss.c patching file drivers/block/cpqarray.c patching file drivers/char/agp/generic.c Hunk #1 succeeded at 1020 (offset -22 lines). patching file drivers/char/agp/intel-agp.c patching file drivers/char/hw_random/intel-rng.c patching file drivers/char/ipmi/ipmi_si_intf.c patching file drivers/char/isicom.c patching file drivers/char/mem.c Hunk #1 succeeded at 618 (offset 2 lines). patching file drivers/char/rtc.c patching file drivers/char/watchdog/sc1200wdt.c patching file drivers/clocksource/scx200_hrt.c patching file drivers/cpufreq/cpufreq_stats.c patching file drivers/i2c/chips/ds1337.c patching file drivers/ide/pci/generic.c patching file drivers/ide/ppc/pmac.c patching file drivers/ieee1394/ohci1394.c patching file drivers/infiniband/core/mad.c patching file drivers/infiniband/hw/mthca/mthca_cq.c patching file drivers/infiniband/hw/mthca/mthca_mad.c patching file drivers/infiniband/hw/mthca/mthca_qp.c patching file drivers/infiniband/hw/mthca/mthca_srq.c patching file drivers/infiniband/ulp/ipoib/ipoib_ib.c patching file drivers/infiniband/ulp/srp/ib_srp.c patching file drivers/infiniband/ulp/srp/ib_srp.h patching file drivers/input/mouse/psmouse-base.c patching file drivers/isdn/capi/capidrv.c patching file drivers/isdn/hisax/config.c patching file drivers/isdn/i4l/isdn_common.c patching file drivers/isdn/icn/icn.c patching file drivers/isdn/isdnloop/isdnloop.c patching file drivers/isdn/pcbit/drv.c patching file drivers/macintosh/via-pmu-backlight.c patching file drivers/md/dm-crypt.c patching file drivers/md/dm-snap.c patching file drivers/md/md.c patching file drivers/md/multipath.c patching file drivers/md/raid10.c patching file drivers/media/Kconfig patching file drivers/media/dvb/b2c2/flexcop-fe-tuner.c patching file drivers/media/dvb/dvb-core/dvb_net.c patching file drivers/media/dvb/frontends/cx24123.c patching file drivers/media/dvb/frontends/dvb-pll.c patching file drivers/media/dvb/frontends/lgdt330x.c patching file drivers/media/video/cx88/cx88-cards.c patching file drivers/media/video/cx88/cx88-dvb.c patching file drivers/media/video/ks0127.c patching file drivers/media/video/msp3400-driver.c patching file drivers/media/video/msp3400-driver.h patching file drivers/media/video/msp3400-kthreads.c patching file drivers/media/video/pvrusb2/Kconfig patching file drivers/media/video/pvrusb2/pvrusb2-ctrl.c patching file drivers/media/video/pvrusb2/pvrusb2-hdw-internal.h patching file drivers/media/video/pvrusb2/pvrusb2-hdw.c patching file drivers/media/video/pvrusb2/pvrusb2-v4l2.c patching file drivers/media/video/saa7134/saa7134-dvb.c patching file drivers/media/video/tuner-simple.c patching file drivers/media/video/tuner-types.c patching file drivers/media/video/tveeprom.c patching file drivers/media/video/usbvideo/quickcam_messenger.h patching file drivers/media/video/video-buf.c patching file drivers/media/video/videodev.c patching file drivers/net/bonding/bond_main.c patching file drivers/net/e1000/e1000_main.c patching file drivers/net/forcedeth.c patching file drivers/net/lp486e.c patching file drivers/net/mv643xx_eth.c patching file drivers/net/sky2.c patching file drivers/net/sky2.h patching file drivers/net/sunhme.c patching file drivers/net/tg3.c patching file drivers/net/wireless/bcm43xx/bcm43xx.h patching file drivers/net/wireless/bcm43xx/bcm43xx_debugfs.c patching file drivers/net/wireless/bcm43xx/bcm43xx_debugfs.h patching file drivers/net/wireless/bcm43xx/bcm43xx_dma.c patching file drivers/net/wireless/bcm43xx/bcm43xx_dma.h patching file drivers/net/wireless/bcm43xx/bcm43xx_leds.c patching file drivers/net/wireless/bcm43xx/bcm43xx_main.c patching file drivers/net/wireless/bcm43xx/bcm43xx_main.h patching file drivers/net/wireless/bcm43xx/bcm43xx_phy.c patching file drivers/net/wireless/bcm43xx/bcm43xx_pio.c patching file drivers/net/wireless/bcm43xx/bcm43xx_sysfs.c patching file drivers/net/wireless/bcm43xx/bcm43xx_wx.c patching file drivers/net/wireless/bcm43xx/bcm43xx_xmit.h patching file drivers/net/wireless/zd1211rw/zd_chip.c patching file drivers/pci/pci-sysfs.c patching file drivers/pci/quirks.c patching file drivers/pcmcia/ds.c patching file drivers/rtc/rtc-max6902.c patching file drivers/rtc/rtc-pcf8563.c patching file drivers/scsi/aic7xxx/aic7xxx_osm.c patching file drivers/scsi/sata_mv.c patching file drivers/scsi/scsi_lib.c patching file drivers/serial/serial_core.c patching file drivers/serial/serial_cs.c patching file drivers/usb/class/usblp.c patching file drivers/usb/core/devio.c patching file drivers/usb/core/notify.c patching file drivers/usb/core/usb.h patching file drivers/usb/gadget/ether.c patching file drivers/usb/input/hid-core.c patching file drivers/usb/input/hid-input.c patching file drivers/usb/input/hid.h patching file drivers/usb/input/usbtouchscreen.c patching file drivers/video/fbmem.c patching file drivers/video/fbsysfs.c patching file drivers/video/nvidia/nv_hw.c patching file drivers/video/nvidia/nv_setup.c patching file drivers/video/nvidia/nv_type.h patching file drivers/video/nvidia/nvidia.c patching file fs/buffer.c patching file fs/cifs/CHANGES patching file fs/cifs/file.c patching file fs/cifs/inode.c patching file fs/compat.c patching file fs/ext2/super.c patching file fs/ext3/super.c patching file fs/fuse/dir.c patching file fs/fuse/file.c patching file fs/fuse/inode.c patching file fs/hfs/super.c patching file fs/jbd/commit.c patching file fs/jfs/jfs_imap.c patching file fs/nfs/dir.c patching file fs/nfsd/nfs2acl.c patching file fs/splice.c patching file fs/sysfs/file.c patching file include/Kbuild patching file include/asm-alpha/Kbuild patching file include/asm-arm/elf.h patching file include/asm-arm/page.h patching file include/asm-arm/unistd.h patching file include/asm-arm26/Kbuild patching file include/asm-cris/Kbuild patching file include/asm-cris/arch-v10/Kbuild patching file include/asm-cris/arch-v32/Kbuild patching file include/asm-cris/byteorder.h patching file include/asm-cris/elf.h patching file include/asm-cris/page.h patching file include/asm-cris/posix_types.h patching file include/asm-cris/unistd.h patching file include/asm-generic/Kbuild patching file include/asm-generic/Kbuild.asm patching file include/asm-generic/audit_change_attr.h patching file include/asm-generic/audit_dir_write.h patching file include/asm-h8300/page.h patching file include/asm-i386/Kbuild patching file include/asm-ia64/Kbuild patching file include/asm-ia64/numa.h patching file include/asm-m32r/page.h patching file include/asm-m32r/ptrace.h patching file include/asm-m32r/sigcontext.h patching file include/asm-m32r/signal.h patching file include/asm-m32r/unistd.h patching file include/asm-m32r/user.h patching file include/asm-m68knommu/page.h patching file include/asm-powerpc/Kbuild patching file include/asm-powerpc/current.h patching file include/asm-powerpc/ptrace.h patching file include/asm-s390/Kbuild patching file include/asm-s390/div64.h patching file include/asm-sh/page.h patching file include/asm-sh/ptrace.h patching file include/asm-sh64/page.h patching file include/asm-sh64/shmparam.h patching file include/asm-sh64/signal.h patching file include/asm-sh64/user.h patching file include/asm-sparc/Kbuild patching file include/asm-sparc/page.h patching file include/asm-sparc/unistd.h patching file include/asm-sparc64/Kbuild patching file include/asm-sparc64/futex.h patching file include/asm-sparc64/page.h patching file include/asm-sparc64/shmparam.h patching file include/asm-sparc64/unistd.h patching file include/asm-um/Kbuild patching file include/asm-v850/page.h patching file include/asm-v850/param.h patching file include/asm-x86_64/Kbuild patching file include/asm-x86_64/system.h patching file include/linux/Kbuild patching file include/linux/byteorder/Kbuild patching file include/linux/dvb/Kbuild patching file include/linux/mmzone.h patching file include/linux/netfilter/Kbuild patching file include/linux/netfilter_arp/Kbuild patching file include/linux/netfilter_bridge/Kbuild patching file include/linux/netfilter_ipv4.h patching file include/linux/netfilter_ipv4/Kbuild patching file include/linux/netfilter_ipv6/Kbuild patching file include/linux/nfsd/Kbuild patching file include/linux/raid/Kbuild patching file include/linux/scx200.h patching file include/linux/serial_core.h patching file include/linux/stddef.h patching file include/linux/sunrpc/Kbuild patching file include/linux/swap.h patching file include/linux/sysctl.h Hunk #1 succeeded at 199 (offset 8 lines). patching file include/linux/tc_act/Kbuild patching file include/linux/tc_ematch/Kbuild patching file include/linux/ufs_fs.h patching file include/media/cx2341x.h patching file include/mtd/Kbuild patching file include/rdma/Kbuild patching file include/scsi/Kbuild patching file include/sound/Kbuild patching file include/video/Kbuild patching file init/Kconfig patching file kernel/module.c patching file kernel/posix-cpu-timers.c patching file kernel/sched.c patching file kernel/softirq.c patching file kernel/sysctl.c patching file kernel/taskstats.c patching file lib/audit.c patching file mm/memory.c Hunk #1 succeeded at 1131 (offset 27 lines). Hunk #3 succeeded at 1178 (offset 27 lines). Hunk #4 succeeded at 1657 (offset 96 lines). patching file mm/migrate.c patching file mm/mincore.c patching file mm/msync.c patching file mm/page_alloc.c Hunk #1 succeeded at 1687 (offset 14 lines). Hunk #3 succeeded at 2026 (offset 14 lines). Hunk #5 succeeded at 2337 (offset 14 lines). patching file mm/shmem.c patching file mm/slab.c patching file mm/truncate.c patching file mm/vmscan.c patching file mm/vmstat.c patching file net/bluetooth/cmtp/capi.c patching file net/bluetooth/hci_sock.c patching file net/bluetooth/rfcomm/tty.c patching file net/bridge/br_ioctl.c patching file net/bridge/netfilter/ebtables.c patching file net/core/dev.c Hunk #1 succeeded at 1528 (offset 50 lines). patching file net/core/skbuff.c Hunk #1 succeeded at 643 (offset 5 lines). Hunk #2 FAILED at 1951. Hunk #3 FAILED at 1962. 2 out of 3 hunks FAILED -- saving rejects to file net/core/skbuff.c.rej patching file net/core/sock.c patching file net/dccp/ipv6.c patching file net/decnet/af_decnet.c patching file net/ieee80211/softmac/ieee80211softmac_io.c patching file net/ieee80211/softmac/ieee80211softmac_scan.c patching file net/ipv4/ipvs/ip_vs_core.c patching file net/ipv4/netfilter.c patching file net/ipv4/netfilter/arp_tables.c patching file net/ipv4/netfilter/ip_conntrack_helper_h323.c patching file net/ipv4/netfilter/ip_nat_standalone.c patching file net/ipv4/netfilter/ip_tables.c patching file net/ipv4/netfilter/iptable_mangle.c patching file net/ipv4/route.c patching file net/ipv4/tcp.c patching file net/ipv4/tcp_cubic.c patching file net/ipv4/tcp_input.c patching file net/ipv4/udp.c patching file net/ipv4/xfrm4_policy.c patching file net/ipv6/ip6_flowlabel.c patching file net/ipv6/ipv6_sockglue.c patching file net/ipv6/netfilter/ip6_tables.c patching file net/ipv6/tcp_ipv6.c patching file net/ipv6/udp.c patching file net/irda/irttp.c patching file net/netfilter/Kconfig patching file net/sched/act_gact.c patching file net/sched/act_police.c patching file net/sched/cls_api.c patching file net/sched/cls_basic.c patching file net/sched/sch_api.c patching file net/sched/sch_generic.c patching file net/sctp/input.c patching file net/sunrpc/svcsock.c patching file security/seclvl.c patching file sound/core/control.c patching file sound/core/hwdep.c patching file sound/core/info.c patching file sound/core/rtctimer.c patching file sound/pci/au88x0/au88x0.c patching file sound/pci/emu10k1/emu10k1_main.c patching file sound/ppc/keywest.c patching file sound/usb/usx2y/usbusx2yaudio.c patching file sound/usb/usx2y/usx2yhwdeppcm.c _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On 10/3/07, Stefan Berger <stefanb@us.ibm.com> wrote:> > I was wondering whether we could apply the 2.6.18.8 patch to the linux tree? > Assuming that that patch makes 2.6.18 better... > > It should be fairly simple. Only two hunks dont apply (both in > net/core/skbuff.c) and those changes seem to already have been applied > anyway.I have been running 2.6.18.8 with Xen-3.1.0/Xen-3.1.1-rc1/Xen3.1.1-rc2 and the patch series I just exclude net-gso-6-linear-segmentation.patch. The only problem I encountered is anything larger than 512MB in domainU might have such error: Sep 27 01:46:43 XXX kernel: ------------[ cut here ]------------ Sep 27 01:46:43 XXX kernel: kernel BUG at /usr/src/xen-3.1.1-rc1-src/linux-2.6.18.8-xen/arch/i386/mm/hypervisor.c:167! Sep 27 01:46:43 XXX kernel: invalid opcode: 0000 [#1] Sep 27 01:46:43 XXX kernel: SMP Sep 27 01:46:43 XXX kernel: Modules linked in: ipv6 xt_state xt_tcpudp iptable_nat ip_nat iptable_mangle ip_conntrack_ftp ip_conntrack_irc ip_conntrack nfnetlink ipt_REJECT ipt_LOG iptable_filter ip_tables x_tables dm_mirror dm_mod Sep 27 01:46:43 XXX kernel: CPU: 0 Sep 27 01:46:43 XXX kernel: EIP: 0061:[<c0114f1e>] Not tainted VLI Sep 27 01:46:43 XXX kernel: EFLAGS: 00010282 (2.6.18.8-xenU #3) Sep 27 01:46:43 XXX kernel: EIP is at xen_pgd_pin+0x6e/0x80 Sep 27 01:46:43 XXX kernel: eax: ffffffea ebx: eb243ed8 ecx: 00000001 edx: 00000000 Sep 27 01:46:43 XXX kernel: esi: 00007ff0 edi: 00000000 ebp: ebef3000 esp: eb243ed8 Sep 27 01:46:43 XXX kernel: ds: 007b es: 007b ss: 0069 Sep 27 01:46:43 XXX kernel: Process nagios (pid: 22284, ti=eb242000 task=eaa5f550 task.ti=eb242000) Sep 27 01:46:43 XXX su(pam_unix)[22112]: session closed for user XXXXX Sep 27 01:46:43 XXX kernel: Stack: 00000002 00043233 2cdb0000 0059b600 ec126b80 c010f90f ec126bd0 c010f952 Sep 27 01:46:43 XXX kernel: ec79e7ac 00000000 c011c15c ecbecd48 ebf6d800 c029de68 ed6037f4 ec126bb4 Sep 27 01:46:43 XXX kernel: ebef30c0 ebef34a0 eb243fbc bf92b960 00000011 00000000 ebef3000 ec2ddb40 Sep 27 01:46:43 XXX kernel: Call Trace: Sep 27 01:46:43 XXX kernel: [<c010f90f>] __pgd_pin+0x2f/0x40 Sep 27 01:46:43 XXX kernel: [<c010f952>] mm_pin+0x32/0x50 Sep 27 01:46:43 XXX kernel: [<c011c15c>] copy_process+0x101c/0x11c0 Sep 27 01:46:43 XXX kernel: [<c029de68>] sock_wfree+0x38/0x40 Sep 27 01:46:43 XXX kernel: [<c011c5fd>] do_fork+0x7d/0x1f0 Sep 27 01:46:43 XXX kernel: [<c022aa7e>] evtchn_do_upcall+0xbe/0x100 Sep 27 01:46:43 XXX kernel: [<c0102f0c>] sys_fork+0x2c/0x30 Sep 27 01:46:43 XXX kernel: [<c0105127>] syscall_call+0x7/0xb Sep 27 01:46:43 XXX kernel: Code: d2 be f0 7f 00 00 e8 42 c4 fe ff 85 c0 78 1c 8b 5c 24 0c 8b 74 24 10 83 c4 14 c3 66 90 39 c2 72 c5 0f 0b 22 00 d8 fb 31 c0 eb bb <0f> 0b a7 00 44 17 32 c0 eb da 90 8d b4 26 00 00 00 00 89 c2 83 Sep 27 01:46:43 XXX kernel: EIP: [<c0114f1e>] xen_pgd_pin+0x6e/0x80 SS:ESP 0069:eb243ed8 I didn''t try on 2.6.18 on domainU more than 512MB though so can''t really comment on whether this is just happening to kernel 2.6.18.8. Sorry if this is out topic. Thanks. Kindest regards, Giam Teck Choon _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On Wed, Oct 03, 2007 at 07:41:36PM +0800, Teck Choon Giam wrote:> On 10/3/07, Stefan Berger <stefanb@us.ibm.com> wrote: > > > > I was wondering whether we could apply the 2.6.18.8 patch to the linux tree? > > Assuming that that patch makes 2.6.18 better... > > > > It should be fairly simple. Only two hunks dont apply (both in > > net/core/skbuff.c) and those changes seem to already have been applied > > anyway. > > I have been running 2.6.18.8 with Xen-3.1.0/Xen-3.1.1-rc1/Xen3.1.1-rc2 > and the patch series I just exclude > net-gso-6-linear-segmentation.patch. The only problem I encountered > is anything larger than 512MB in domainU might have such error: > > Sep 27 01:46:43 XXX kernel: ------------[ cut here ]------------ > Sep 27 01:46:43 XXX kernel: kernel BUG at > /usr/src/xen-3.1.1-rc1-src/linux-2.6.18.8-xen/arch/i386/mm/hypervisor.c:167!Do you have CONFIG_HIGHPTE enabled? CONFIG_HIGHPTE have caused problems on 2.6.20 and 2.6.21 rebases of Xen 3.1 on Fedora, on pgd_pin() and other page table updating hypercalls, because the kernel somehow was trying to map page table entries as writeable. ''xm dmesg'' should give more information on the cause of this Oops.> > I didn''t try on 2.6.18 on domainU more than 512MB though so can''t > really comment on whether this is just happening to kernel 2.6.18.8.I haven''t reproduced my CONFIG_HIGHPTE problems on 2.6.18-xen-3.1.0, also. I don''t know if it was a problem of our Xen rebase or the problem is present on the 2.6.18 xenbits tree. Enabling CONFIG_HIGHPTE under Xen is not recommended, anyway. Even if it was working, performance will probably suffer. -- Eduardo _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On Tue, 2007-10-02 at 16:43 -0400, Stefan Berger wrote:> > I was wondering whether we could apply the 2.6.18.8 patch to the linux > tree? Assuming that that patch makes 2.6.18 better...Might as well. Applied and pushed. Cheers, Ian. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
xen-devel-bounces@lists.xensource.com wrote on 10/03/2007 07:41:36 AM:> On 10/3/07, Stefan Berger <stefanb@us.ibm.com> wrote: > > > > I was wondering whether we could apply the 2.6.18.8 patch to the linuxtree?> > Assuming that that patch makes 2.6.18 better... > > > > It should be fairly simple. Only two hunks dont apply (both in > > net/core/skbuff.c) and those changes seem to already have been applied > > anyway. > > I have been running 2.6.18.8 with Xen-3.1.0/Xen-3.1.1-rc1/Xen3.1.1-rc2 > and the patch series I just exclude > net-gso-6-linear-segmentation.patch. The only problem I encountered > is anything larger than 512MB in domainU might have such error: > > Sep 27 01:46:43 XXX kernel: ------------[ cut here ]------------ > Sep 27 01:46:43 XXX kernel: kernel BUG at >/usr/src/xen-3.1.1-rc1-src/linux-2.6.18.8-xen/arch/i386/mm/hypervisor.c:167!> Sep 27 01:46:43 XXX kernel: invalid opcode: 0000 [#1] > Sep 27 01:46:43 XXX kernel: SMP > Sep 27 01:46:43 XXX kernel: Modules linked in: ipv6 xt_state xt_tcpudp > iptable_nat ip_nat iptable_mangle ip_conntrack_ftp ip_conntrack_irc > ip_conntrack nfnetlink ipt_REJECT ipt_LOG iptable_filter ip_tables > x_tables dm_mirror dm_mod > Sep 27 01:46:43 XXX kernel: CPU: 0 > Sep 27 01:46:43 XXX kernel: EIP: 0061:[<c0114f1e>] Not tainted VLI > Sep 27 01:46:43 XXX kernel: EFLAGS: 00010282 (2.6.18.8-xenU #3) > Sep 27 01:46:43 XXX kernel: EIP is at xen_pgd_pin+0x6e/0x80 > Sep 27 01:46:43 XXX kernel: eax: ffffffea ebx: eb243ed8 ecx: > 00000001 edx: 00000000 > Sep 27 01:46:43 XXX kernel: esi: 00007ff0 edi: 00000000 ebp: > ebef3000 esp: eb243ed8 > Sep 27 01:46:43 XXX kernel: ds: 007b es: 007b ss: 0069 > Sep 27 01:46:43 XXX kernel: Process nagios (pid: 22284, ti=eb242000 > task=eaa5f550 task.ti=eb242000) > Sep 27 01:46:43 XXX su(pam_unix)[22112]: session closed for user XXXXX > Sep 27 01:46:43 XXX kernel: Stack: 00000002 00043233 2cdb0000 0059b600 > ec126b80 c010f90f ec126bd0 c010f952 > Sep 27 01:46:43 XXX kernel: ec79e7ac 00000000 c011c15c ecbecd48 > ebf6d800 c029de68 ed6037f4 ec126bb4 > Sep 27 01:46:43 XXX kernel: ebef30c0 ebef34a0 eb243fbc bf92b960 > 00000011 00000000 ebef3000 ec2ddb40 > Sep 27 01:46:43 XXX kernel: Call Trace: > Sep 27 01:46:43 XXX kernel: [<c010f90f>] __pgd_pin+0x2f/0x40 > Sep 27 01:46:43 XXX kernel: [<c010f952>] mm_pin+0x32/0x50 > Sep 27 01:46:43 XXX kernel: [<c011c15c>] copy_process+0x101c/0x11c0 > Sep 27 01:46:43 XXX kernel: [<c029de68>] sock_wfree+0x38/0x40 > Sep 27 01:46:43 XXX kernel: [<c011c5fd>] do_fork+0x7d/0x1f0 > Sep 27 01:46:43 XXX kernel: [<c022aa7e>] evtchn_do_upcall+0xbe/0x100 > Sep 27 01:46:43 XXX kernel: [<c0102f0c>] sys_fork+0x2c/0x30 > Sep 27 01:46:43 XXX kernel: [<c0105127>] syscall_call+0x7/0xb > Sep 27 01:46:43 XXX kernel: Code: d2 be f0 7f 00 00 e8 42 c4 fe ff 85 > c0 78 1c 8b 5c 24 0c 8b 74 24 10 83 c4 14 c3 66 90 39 c2 72 c5 0f 0b > 22 00 d8 fb 31 c0 eb bb <0f> 0b a7 00 44 17 32 c0 eb da 90 8d b4 26 00 > 00 00 00 89 c2 83 > Sep 27 01:46:43 XXX kernel: EIP: [<c0114f1e>] xen_pgd_pin+0x6e/0x80 > SS:ESP 0069:eb243ed8 > > I didn''t try on 2.6.18 on domainU more than 512MB though so can''t > really comment on whether this is just happening to kernel 2.6.18.8.Do you know how to trigger this error? I tried a rather simple 2.6.18.8 domain(CS 217 + 2.6.18.8 patch with those 2 hunks cut out) with 256 MB and then 728 MB and needless to say it does not appear there.> > Sorry if this is out topic.No, it''s not. :-) Stefan> > Thanks. > > Kindest regards, > Giam Teck Choon > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xensource.com > http://lists.xensource.com/xen-devel_______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On 10/3/07, Eduardo Habkost <ehabkost@redhat.com> wrote:> Do you have CONFIG_HIGHPTE enabled? CONFIG_HIGHPTE have caused problems > on 2.6.20 and 2.6.21 rebases of Xen 3.1 on Fedora, on pgd_pin() and > other page table updating hypercalls, because the kernel somehow was > trying to map page table entries as writeable.# grep CONFIG_HIGHPTE /boot/config-2.6.18.8-xen0 CONFIG_HIGHPTE=y So yes it is enabled.> ''xm dmesg'' should give more information on the cause of this Oops.Unfortunately I don''t look at the xm dmesg during the Oops closely and now the system has been running on xen-3.1.1-rc2. I guess I will try to reproduce and post related xm dmesg and oops message when it happen when I have the time to test on my other test system after getting my new RAM (waiting for 6 x 1GB RAM).> I haven''t reproduced my CONFIG_HIGHPTE problems on 2.6.18-xen-3.1.0, > also. I don''t know if it was a problem of our Xen rebase or the problem > is present on the 2.6.18 xenbits tree. > > Enabling CONFIG_HIGHPTE under Xen is not recommended, anyway. Even if > it was working, performance will probably suffer.Ok, noted and will try on disabling CONFIG_HIGHPTE and try whether can I reproduce the oops :) Thanks ;) Kindest regards, Giam Teck Choon _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On 10/3/07, Stefan Berger <stefanb@us.ibm.com> wrote:> Do you know how to trigger this error? I tried a rather simple 2.6.18.8 > domain(CS 217 + 2.6.18.8 patch with those 2 hunks cut out) with 256 MB and > then 728 MB and needless to say it does not appear there.This error on some domainUs will shown if it is running lots of IO or taking more memory and started to swap a lot. On some domainU where hardly using much IO or memory, won''t produce such oops accordingly to my experience. However on a live domainU hosting less than 50 virtual domains can produce such oops within 2 weeks time. So it depends on memory usage and IO/load I guess... sorry for not able to provide you an exact way to reproduce such oops :( I will try to use stress (http://weather.ou.edu/~apw/projects/stress/) and see whether can I reproduce such oops on my test system after I get my new RAM which should be within these few days. BTW, the one (domainU) which is getting the oops is allocated with 1GB memory. Thanks. Kindest regards, Giam Teck Choon _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On 10/3/07, Eduardo Habkost <ehabkost@redhat.com> wrote:> On Wed, Oct 03, 2007 at 07:41:36PM +0800, Teck Choon Giam wrote: > > On 10/3/07, Stefan Berger <stefanb@us.ibm.com> wrote: > > > > > > I was wondering whether we could apply the 2.6.18.8 patch to the linux tree? > > > Assuming that that patch makes 2.6.18 better... > > > > > > It should be fairly simple. Only two hunks dont apply (both in > > > net/core/skbuff.c) and those changes seem to already have been applied > > > anyway. > > > > I have been running 2.6.18.8 with Xen-3.1.0/Xen-3.1.1-rc1/Xen3.1.1-rc2 > > and the patch series I just exclude > > net-gso-6-linear-segmentation.patch. The only problem I encountered > > is anything larger than 512MB in domainU might have such error: > > > > Sep 27 01:46:43 XXX kernel: ------------[ cut here ]------------ > > Sep 27 01:46:43 XXX kernel: kernel BUG at > > /usr/src/xen-3.1.1-rc1-src/linux-2.6.18.8-xen/arch/i386/mm/hypervisor.c:167! > > > Do you have CONFIG_HIGHPTE enabled? CONFIG_HIGHPTE have caused problems > on 2.6.20 and 2.6.21 rebases of Xen 3.1 on Fedora, on pgd_pin() and > other page table updating hypercalls, because the kernel somehow was > trying to map page table entries as writeable. > > ''xm dmesg'' should give more information on the cause of this Oops. > > > > > > I didn''t try on 2.6.18 on domainU more than 512MB though so can''t > > really comment on whether this is just happening to kernel 2.6.18.8. > > I haven''t reproduced my CONFIG_HIGHPTE problems on 2.6.18-xen-3.1.0, > also. I don''t know if it was a problem of our Xen rebase or the problem > is present on the 2.6.18 xenbits tree. > > Enabling CONFIG_HIGHPTE under Xen is not recommended, anyway. Even if > it was working, performance will probably suffer. > > -- > EduardoWow, this appears (at least in my quick test today) to be the cause of the oops I keep hitting, reported in the thread titled "Hunting down an oops in Xen 3.1.0''s 2.6.18 kernel". I would *highly* recommend the attached trivial patch to disable HIGHPTE to avoid this. -- Michael Marineau Oregon State University mike@marineau.org _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On 10/10/07 21:32, "Michael Marineau" <mike@marineau.org> wrote:> Wow, this appears (at least in my quick test today) to be the cause of > the oops I keep hitting, reported in the thread titled "Hunting down > an oops in Xen 3.1.0''s 2.6.18 kernel". I would *highly* recommend the > attached trivial patch to disable HIGHPTE to avoid this.Yes, I can easily repro this with CONFIG_HIGHPTE and ''vmalloc=700M'' on dom0''s cmdline (to force most dom0 memory to be highmem). Then I balloon down to 128 megabytes and a crash happens very soon after. We should fix this rather than disable the option. -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On 11/10/07 15:44, "Keir Fraser" <Keir.Fraser@cl.cam.ac.uk> wrote:> On 10/10/07 21:32, "Michael Marineau" <mike@marineau.org> wrote: > >> Wow, this appears (at least in my quick test today) to be the cause of >> the oops I keep hitting, reported in the thread titled "Hunting down >> an oops in Xen 3.1.0''s 2.6.18 kernel". I would *highly* recommend the >> attached trivial patch to disable HIGHPTE to avoid this. > > Yes, I can easily repro this with CONFIG_HIGHPTE and ''vmalloc=700M'' on > dom0''s cmdline (to force most dom0 memory to be highmem). Then I balloon > down to 128 megabytes and a crash happens very soon after. We should fix > this rather than disable the option.Ah, it was an easy one. Fixed by linux-2.6.18-xen.hg c/s 252:e797297402. -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On 10/11/07, Keir Fraser <Keir.Fraser@cl.cam.ac.uk> wrote:> On 11/10/07 15:44, "Keir Fraser" <Keir.Fraser@cl.cam.ac.uk> wrote: > > > On 10/10/07 21:32, "Michael Marineau" <mike@marineau.org> wrote: > > > >> Wow, this appears (at least in my quick test today) to be the cause of > >> the oops I keep hitting, reported in the thread titled "Hunting down > >> an oops in Xen 3.1.0''s 2.6.18 kernel". I would *highly* recommend the > >> attached trivial patch to disable HIGHPTE to avoid this. > > > > Yes, I can easily repro this with CONFIG_HIGHPTE and ''vmalloc=700M'' on > > dom0''s cmdline (to force most dom0 memory to be highmem). Then I balloon > > down to 128 megabytes and a crash happens very soon after. We should fix > > this rather than disable the option. > > Ah, it was an easy one. Fixed by linux-2.6.18-xen.hg c/s 252:e797297402. > > -- KeirSweet, that was fast :-) I''ll give it a test once I see the changeset show up on xenbits, it doesn''t appear to be there yet. -- Michael Marineau Oregon State University mike@marineau.org _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On 11/10/07 17:21, "Michael Marineau" <mike@marineau.org> wrote:>> Ah, it was an easy one. Fixed by linux-2.6.18-xen.hg c/s 252:e797297402. >> >> -- Keir > > Sweet, that was fast :-) I''ll give it a test once I see the changeset > show up on xenbits, it doesn''t appear to be there yet.Pull from http://xenbits.xensource.com/staging/linux-2.6.18-xen.hg That''s a pre-testing tree. The main tree is just a time-delayed (post-testing) version of it. -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On 10/11/07, Keir Fraser <Keir.Fraser@cl.cam.ac.uk> wrote:> > > > On 11/10/07 17:21, "Michael Marineau" <mike@marineau.org> wrote: > > >> Ah, it was an easy one. Fixed by linux-2.6.18-xen.hg c/s 252:e797297402. > >> > >> -- Keir > > > > Sweet, that was fast :-) I''ll give it a test once I see the changeset > > show up on xenbits, it doesn''t appear to be there yet. > > Pull from http://xenbits.xensource.com/staging/linux-2.6.18-xen.hg > > That''s a pre-testing tree. The main tree is just a time-delayed > (post-testing) version of it. > > -- KeirYup, this seems to resolve it. Although, why is the change only for l1 updates? I would think the same change would be required for l2 and l3. I know have hit an oops on a l2 update at least once, it is just rare. -- Michael Marineau Oregon State University mike@marineau.org _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
On 11/10/07 18:41, "Michael Marineau" <mike@marineau.org> wrote:>> That''s a pre-testing tree. The main tree is just a time-delayed >> (post-testing) version of it. >> >> -- Keir > > Yup, this seems to resolve it. Although, why is the change only for l1 > updates? I would think the same change would be required for l2 and > l3. I know have hit an oops on a l2 update at least once, it is just > rare.Only ptes can be allocated in high memory. -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel