2.6.13-mm2

2.6.13-mm2

[Andrew Morton]

ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/

(kernel.org propagation is slow.  There's a temp copy at
http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)



- Added Andi's x86_64 tree, as separate patches

- Added a driver for TI acx1xx cardbus wireless NICs

- Large revamp of pcmcia suspend handling

- Largeish v4l and DVB updates

- Significant parport rework

- Many tty drivers still won't compile

- Lots of framebuffer driver updates

- There are still many patches here for 2.6.14.  We're doing pretty well
  with merging up the subsystem trees.  ia64 and CIFS are still pending. 
  x86_64 and several of Greg's trees (especially USB) aren't merged yet.




Changes since 2.6.13-mm1:


 linus.patch
 git-acpi.patch
 git-cifs.patch
 git-cpufreq.patch
 git-ia64.patch
 git-audit.patch
 git-input.patch
 git-jfs.patch
 git-libata-all.patch
 git-mtd.patch
 git-netdev-all.patch
 git-nfs.patch
 git-nfs-oops-fix.patch
 git-ocfs2.patch
 git-scsi-iscsi.patch

 Subsystem trees

-non-booting-g5-fix.patch -dvb-saa7134-dvb-must-select-tda1004x.patch
-tpm_infineon-bugfix-in-pnpacpi-handling.patch
-move-truncate_inode_pages-into-delete_inode.patch
-update-filesystems-for-new-delete_inode-behavior.patch
-fix-possible-null-pointer-access-acpi_pci_irq_disable.patch
-gregkh-driver-driver-bind-fix.patch
-gregkh-driver-driver-link-device-and-class.patch
-gregkh-driver-floppy-cmos-attribute.patch
-gregkh-driver-floppy-cmos-attribute-tidy.patch
-gregkh-driver-driver-docs-whitespace.patch
-gregkh-driver-driver-docs-permissions.patch
-gregkh-driver-driver-handle-sysdev-suspend-failure.patch
-gregkh-driver-driver-kfree-check-cleanup.patch
-gregkh-driver-klist-fix-semantics.patch
-gregkh-i2c-i2c-max6875-simplify.patch
-gregkh-i2c-i2c-max6875-documentation-update.patch
-gregkh-i2c-i2c-max6875-fix-build-error.patch
-gregkh-i2c-i2c-nforce2-cleanup.patch
-gregkh-i2c-i2c-max6875-kobj_to_i2c_client.patch
-gregkh-i2c-i2c-max6875-documentation-cleanup.patch
-gregkh-i2c-i2c-max6875-cleanup.patch
-gregkh-i2c-i2c-hwmon-soften-lm75.patch
-gregkh-i2c-i2c-hwmon-lm78-j.patch
-gregkh-i2c-i2c-hwmon-document-w83627ehg.patch
-gregkh-i2c-i2c-hwmon-class-02.patch
-gregkh-i2c-i2c-hwmon-class-03.patch
-gregkh-i2c-i2c-hwmon-split-01.patch
-gregkh-i2c-i2c-hwmon-split-02.patch
-gregkh-i2c-i2c-hwmon-split-03.patch
-gregkh-i2c-i2c-hwmon-split-04.patch
-gregkh-i2c-i2c-hwmon-split-05.patch
-gregkh-i2c-i2c-hwmon-split-06.patch
-gregkh-i2c-i2c-hwmon-split-07.patch
-gregkh-i2c-i2c-hwmon-split-08.patch
-gregkh-i2c-i2c-hwmon-split-09.patch
-gregkh-i2c-i2c-w83792d-01.patch
-gregkh-i2c-i2c-w83792d-02.patch
-gregkh-i2c-i2c-w83792d-03.patch
-gregkh-i2c-i2c-refactor-message.patch
-gregkh-i2c-i2c-hwmon-tag-superio-functions-__init.patch
-gregkh-i2c-i2c-inline-i2c_adapter_id.patch
-gregkh-i2c-i2c-i2c-algo-pca-busy-bus-fix.patch
-gregkh-i2c-i2c-documentation-typo.patch
-gregkh-i2c-i2c-core-debug-cleanup.patch
-gregkh-i2c-hwmon-move-SENSORS_LIMIT.patch
-gregkh-i2c-hwmon-lm85-cleanups.patch
-gregkh-i2c-hwmon-split-round2-01.patch
-gregkh-i2c-hwmon-split-round2-02.patch
-gregkh-i2c-hwmon-split-round2-03.patch
-gregkh-i2c-hwmon-split-round2-04.patch
-gregkh-i2c-hwmon-split-round2-05.patch
-gregkh-i2c-hwmon-split-round2-06.patch
-gregkh-i2c-hwmon-split-round2-07.patch
-gregkh-i2c-hwmon-split-round2-08.patch
-gregkh-i2c-hwmon-split-round2-09.patch
-gregkh-i2c-hwmon-split-round2-10.patch
-gregkh-i2c-hwmon-split-round2-11.patch
-gregkh-i2c-hwmon-vid-table-update.patch
-gregkh-i2c-i2c-rewrite-i2c_probe.patch
-gregkh-i2c-i2c-centralize-24rf08-corruption-prevention.patch
-gregkh-i2c-i2c-kill-i2c_algorithm-stuff-01.patch
-gregkh-i2c-i2c-kill-i2c_algorithm-stuff-02.patch
-gregkh-i2c-i2c-kill-i2c_algorithm-stuff-03.patch
-gregkh-i2c-i2c-kill-i2c_algorithm-stuff-04.patch
-gregkh-i2c-i2c-kill-i2c_algorithm-stuff-05.patch
-gregkh-i2c-i2c-kill-i2c_algorithm-stuff-06.patch
-gregkh-i2c-i2c-kill-i2c_algorithm-stuff-07.patch
-gregkh-i2c-i2c-outdated-i2c_adapter-comment.patch
-gregkh-i2c-hwmon-maintainer.patch
-gregkh-i2c-i2c-drop-i2c_clientname.patch
-gregkh-i2c-w1-netlink-callbacks.patch
-ia64-cpuset-build_sched_domains-mangles-structures.patch
-git-audit-ppc64-fix.patch
-input-usb-ignore-logitech-vendor-usages.diff.patch
-input-usb-hid-simulation-usages.diff.patch
-input-i8042-fix-irq-printk.diff.patch
-input-hid-more-consumer-usages.diff.patch
-input-atkbd-allow-0x7f-scancode.diff.patch
-input-hid-variable-max-buffer-size.diff.patch
-input-hid-logitech-ultra-x-media-remote.diff.patch
-input-i8042-mux-blacklist-fsc-t3010.diff.patch
-input-iforce-wait-event.diff.patch
-hid-core-wireless-security-lock-blacklist-entry.patch
-hiddev-output-reports-are-dropped-when-hidiocsreport-is.patch
-apple-powerbook-usb-keyboard-hid-fix.patch
-kbuild-fix-make-clean-damaging-hg-repos.patch
-git-net-gregkh-i2c-w1-netlink-callbacks-fix.patch
-fix-minor-bug-in-sungemh.patch
-netlink-log-protocol-failures.patch
-mbio_bus-pm_message_t-fix.patch
-fix-klist-semantics-for-lists-which-have-elements-removed.patch
-git-scsi-misc-sr-fix.patch
-mpt-fusion-dv-fixes.patch
-watchdog-new-sbc8360-driver.patch
-sparsemem-extreme.patch
-ppc64-fix-sparsemem-extreme.patch
-sparsemem-extreme-implementation.patch
-sparsemem-extreme-hotplug-preparation.patch
-sparsemem-extreme-hotplug-preparation-fix.patch
-mm-consolidate-get_order.patch
-swap-update-swapfile-i_sem-comment.patch
-swap-correct-swapfile-nr_good_pages.patch
-swap-move-destroy_swap_extents-calls.patch
-swap-swap-extent-list-is-ordered.patch
-swap-show-span-of-swap-extents.patch
-swap-swap-unsigned-int-consistency.patch
-swap-swap-unsigned-int-consistency-warning-fix.patch
-swap-freeing-update-swap_listnext.patch
-swap-get_swap_page-drop-swap_list_lock.patch
-swap-scan_swap_map-restyled.patch
-swap-scan_swap_map-drop-swap_device_lock.patch
-swap-scan_swap_map-latency-breaks.patch
-swap-swap_lock-replace-listdevice.patch
-swap-update-swsusp-use-of-swap_info.patch
-swap-update-swsusp-use-of-swap_info-fix.patch
-delete-from_swap_cache-bug_ons.patch
-rmap-dont-test-rss.patch
-proc-pid-numa_maps-to-show-on-which-nodes-pages-reside.patch
-mm-comment-rmap.patch
-mm-micro-optimise-rmap.patch
-mm-cleanup-rmap.patch
-mm-remap-zero_page-mappings.patch
-mm-remove-atomic.patch
-vm-add-capabilites-check-to-set_zone_reclaim.patch
-vm-zone-reclaim-atomic-ops-cleanup.patch
-mm-fix-madvise-vma-merging.patch
-comment-typo-fix.patch
-vm-slabc-spelling-correction.patch
-shmem_populate-avoid-an-useless-check-and-some-comments.patch
-remove-implied-vm_ops-check.patch
-correct-_page_file-comment.patch
-arm-allow-for-arch-specific-ioremap_max_order.patch
-hugetlb-add-pte_huge-macro.patch
-hugetlb-move-stale-pte-check-into-huge_pte_alloc.patch
-hugetlb-check-pd_present-in-huge_pte_offset.patch
-remove-hugetlb_clean_stale_pgtable-and-fix-huge_pte_alloc.patch
-slab-consolidate-kmem_bufctl_t.patch
-x86-ptep-clear-optimization.patch
-x86-ptep-clear-optimization-fix.patch
-x86_64-ptep-clear-optimization.patch
-mm-slabc-prefetchw-the-start-of-new-allocated-objects.patch
-slab-removes-local_irq_save-local_irq_restore-pair.patch
-proc-pid-smaps.patch
-smaps-say-vma-not-map.patch
-smaps-use-new-ptwalks.patch
-smaps-say-kb-not-kb.patch
-smaps-print-more-fields.patch
-smaps-reading-fix.patch
-vm-add-page_state-info-to-per-node-meminfo.patch
-drivers-net-ne3210c-cleanups.patch
-net-add-driver-for-the-nic-on-cell-blades.patch
-ipw2200-build-fix.patch
-net-update-the-spider_net-driver.patch
-mdio_bus_exit-in-discarded-section-textexit.patch
-kconfig-kxgettext-message-fix.patch
-kconfig-kxgettext-eol-fix.patch
-kconfig-linuxpot-for-all-arch.patch
-selinux-reduce-memory-use-by-avtab.patch
-selinux-endian-notations.patch
-generic-vfs-fallback-for-security-xattrs.patch
-arch-ppc-kernel-ppc_ksymsc-remove-unused-define-export_symtab_strops.patch
-ppc32-remove-board-support-for-adir.patch
-ppc32-remove-board-support-for-ash.patch
-ppc32-remove-board-support-for-beech.patch
-ppc32-remove-defconfig-for-cedar.patch
-ppc32-remove-board-support-for-k2.patch
-ppc32-remove-board-support-for-mcpn765.patch
-ppc32-remove-board-support-for-menf1.patch
-ppc32-remove-board-support-for-oak.patch
-ppc32-remove-board-support-for-rainier.patch
-ppc32-remove-board-support-for-redwood.patch
-ppc32-remove-board-support-for-sm850.patch
-ppc32-remove-board-support-for-spd823ts.patch
-ppc32-remove-board-support-for-pcore.patch
-pcp32-fix-asm-ppc-dma-mappingh-sparse-warning.patch
-ppc32-add-usb-support-to-ibm-stb04xxx-platforms.patch
-ppc32-added-support-for-the-book-e-style-watchdog-timer.patch
-ppc32-add-ppc_sys-descriptions-for-powerquicc-ii-devices.patch
-ppc32-add-phy-excluded-features-to-ocp_func_emac_data.patch
-cpm_uart-fix-2nd-serial-port-on-mpc8560-ads.patch
-cpm_uart-use-schedule_timeout-instead-of-direct-call-to.patch
-cpm_uart-fix-baseaddress-for-smc-1-and-2.patch
-ppc32-cleaned-up-global-namespace-of-book-e-watchdog.patch
-ppc32-add-440gx-revf-cputable-entry.patch
-ppc32-removed-find_namec.patch
-ppc32-add-cputable-entry-for-440sp-rev-a.patch
-ppc32-dont-sleep-in-flush_dcache_icache_page.patch
-ppc32-fix-emac-tx-channel-assignments-for-npe405h.patch
-ppc32-fix-bamboo-and-luan-build-warnings.patch
-ppc32-disable-ibm405_err77-and-ibm405_err51-workarounds-for-405ep.patch
-ppc32-ppc_sys-system-on-chip-identification-additions.patch
-ppc32-add-config_hz.patch
-ppc32-add-support-for-marvell-ev64360bp-board.patch
-ppc32-defconfig-for-marvell-ev64360bp-board.patch
-ppc32-fix-wundef-warning-for-config_8xx.patch
-ppc32-added-pci-support-mpc83xx.patch
-ppc32-re-order-cputable-for-750cxe-dd24-entry.patch
-ppc32-add-cputable-entry-for-750cxe-dd24-gekko.patch
-ppc32-move-4xx-phy_mode_xxx-defines-to-ibm_ocph.patch
-ppc32-add-dcr_base-field-to-ocp_func_mal_data.patch
-ppc32-l2-cache-prefetch-fixes-on-745x.patch
-ppc32-export-cacheable_memcpy.patch
-ppc64-update-xmon-helptext.patch
-ppc64-add-vmx-save-flag-to-vpa.patch
-ppc64-replace-schedule_timeout-with-msleep_interruptible.patch
-frv-remove-export-of-strtok.patch
-mips-remove-obsolete-giu-function-call-for-vr41xx.patch
-mips-update-irq-handling-for-vr41xx.patch
-mips-change-system-type-name-in-proc-for-vr41xx.patch
-mips-remove-vr4181-support.patch
-mips-remove-vr4181-support-fix.patch
-more-vr4181-removal.patch
-mips-remove-hp-laserjet-remains.patch
-dec-pmag-aa-framebuffer-update.patch
-dec-pmag-ba-frame-buffer-update.patch
-dec-pmagb-b-framebuffer-update.patch
-mips-add-support-for-qemu-system-architecture.patch
-mips-technologies-pci-id-bits.patch
-mips-add-tanbac-vr4131-multichip-module.patch
-mips-add-default-select-configs-for-vr41xx.patch
-mips-add-default-select-configs-for-vr41xx-fix.patch
-mips-remove-vrc4171-config.patch
-mips-changed-from-vr41xx-to-vr4100-series-in-kconfig.patch
-mips-cleanup-32-64-bit-configuration.patch
-mips-nuke-trailing-whitespace.patch
-mips-fix-coherency-configuration.patch
-mips-add-pcibios_select_root.patch
-mips-add-pcibios_bus_to_resource.patch
-mips-add-more-sys_support__kernel-and.patch
-mips-fix-build-warnings.patch
-mips-remove-timexh-for-vr41xx.patch
-mips-kludge-envdev-to-build-for-64-bit-mips-with-32-bit-compat.patch
-arch-sh64-kconfig-doesnt-need-its-own-log_buf_shift.patch
-x86-compress-the-stack-layout-of-do_page_fault.patch
-x86-compress-the-stack-layout-of-do_page_fault-fix.patch
-hpet-use-read_timer_tsc-only-when-cpu-has-tsc.patch
-x86-fix-efi-memory-map-parsing.patch
-vm86-honor-tf-bit-when-emulating-an-instruction.patch
-kdump-save-parameter-segment-in-protected-mode-x86.patch
-x86-automatically-enable-bigsmp-when-we-have-more-than-8-cpus-2.patch
-x86-automatically-enable-bigsmp-when-we-have-more-than-8-cpus-2-tidy.patch
-x86-add-the-check-for-all-the-cores-in-a-package-in-cache-information.patch
-via-vt8237-apic-bypass-deassertion-quirk.patch
-es7000-platform-update-i386.patch
-i386-clean-up-vdso-alignment-padding.patch
-i386-inline-asm-cleanup.patch
-i386-inline-asm-cleanup-kexec-fix.patch
-i386-arch-cleanup-seralize-msr.patch
-i386-arch-cleanup-seralize-msr-fix.patch
-i386-inline-assembler-cleanup-encapsulate-descriptor-and-task-register-management.patch
-i386-inline-assembler-cleanup-encapsulate-descriptor-and-task-register-management-fix.patch
-i386-generate-better-code-around-descriptor-update-and-access-functions.patch
-i386-load_tls-fix.patch
-i386-use-set_pte-macros-in-a-couple-places-where-they-were-missing.patch
-i386-fix-incorrect-tss-entry-for-ldt.patch
-x86-more-asm-cleanups.patch
-x86-privilege-cleanup.patch
-x86-make-iopl-explicit.patch
-x86-remove-redundant-tss-clearing.patch
-x86-introduce-a-write-acessor-for-updating-the-current-ldt.patch
-x86-nmi-better-support-for-debuggers.patch
-x86-nmi-better-support-for-debuggers-fix.patch
-i386-encapsulate-copying-of-pgd-entries.patch
-i386-boottime-for_each_cpu-broken.patch
-i386-boottime-for_each_cpu-broken-fix.patch
-x86-x86_64-deferred-handling-of-writes-to-proc-irq-xx-smp_affinitypatch-added-to-mm-tree.patch
-x86_64-print-processor-number-in-show_regs.patch
-x86_64-fix-off-by-one-in-e820_mapped.patch
-x86_64-prefetchw-can-fall-back-to-prefetch-if-3dnow.patch
-x86_64-create-sysfs-entries-for-cpu-only-for-present-cpus.patch
-x86_64dont-call-enforce_max_cpus-when-hotplug-is-enabled.patch
-unify-x86-x86-64-semaphore-code.patch
-remove-busywait-in-refrigerator.patch
-isa-dma-suspend-for-i386-2.patch
-isa-dma-suspend-for-x86_64-2.patch
-suspend-update-documentation.patch
-swsusp-fix-remaining-u32-vs-pm_message_t-confusion.patch
-swsusp-fix-remaining-u32-vs-pm_message_t-confusion-2.patch
-swsusp-switch-pm_message_t-to-struct.patch
-swsusp-switch-pm_message_t-to-struct-pmac_zilog-fix.patch
-swsusp-switch-pm_message_t-to-struct-ppc32-fixes.patch
-swsusp-switch-pm_message_t-to-struct-chipsfb-fixes.patch
-swsusp-switch-pm_message_t-to-struct-mesh-fixes-2.patch
-fix-pm_message_t-stuff-in-mm-tree-netdev.patch
-encrypt-suspend-data-for-easy-wiping.patch
-swsusp-prevent-disks-from-spinning-down-and-up.patch
-swsusp-simpler-calculation-of-number-of-pages-in-pbe-list.patch
-swsusup-with-dm-crypt-mini-howto.patch
-swsusp-add-locking-to-software_resume.patch
-swsusp-fix-error-handling-and-cleanups.patch
-reconfigure-msi-registers-after-resume.patch
-pm-fix-process-freezing.patch
-pm-cleanup-sys-power-disk.patch
-3c59x-pm-fixes.patch
-add-suspend-resume-for-timer.patch
-arch-cris-kconfigdebug-use-lib-kconfigdebug.patch
-uml-remove-debugging-code-from-page-fault-path.patch
-uml-rename-kconfig-files-to-be-like-the-other-arches.patch
-ptrace-i386-fix-syscall-audit-interaction-with-singlestep.patch
-uml-support-ptrace-adds-the-host-sysemu-support-for-uml-and-general-usage.patch
-uml-support-reorganize-ptrace_sysemu-support.patch
-uml-support-add-ptrace_sysemu_singlestep-option-to-i386.patch
-sysemu-fix-sysaudit--singlestep-interaction.patch
-uml-support-sysemu-slight-cleanup-and-speedup.patch
-uml-workaround-gdb-problems-on-debugging.patch
-uml-fix-sigwinch-handler-race-while-waiting-for-signals.patch
-uml-fixes-performance-regression-in-activate_mm-and-thus-exec.patch
-uml-fault-handler-micro-cleanups.patch
-uml-fix-signal-frame-copy_user.patch
-uml-fix-a-macro-typo.patch
-uml-error-path-cleanup.patch
-uml-build-cleanup.patch
-uml-remove-libc-reference-in-build.patch
-uml-mark-smp-on-uml-x86_64-as-broken.patch
-uml-remove-duplicated-exports.patch
-uml-uml-i386-is-i386-when-running-on-x86_64.patch
-uml-tlb-operation-batching.patch
-uml-merge-duplicated-page-table-code.patch
-xtensa-replace-extern-inline-with-static-inline.patch
-xtensa-delete-accidental-file.patch
-s390-machine-check-handler-bugs.patch
-s390-debug-feature-changes.patch
-s390-deadlock-in-dasd_devmap.patch
-s390-64-bit-diag250-support.patch
-s390-reipl-fix-and-extern-static-inline.patch
-s390-pfault-interrupt-race.patch
-s390-crypto-driver-update.patch
-s390-compat-system-calls.patch
-s390-spinlock-corner-case.patch
-s390-disconnected-3270-console.patch
-futex_wake_op-pthread_cond_signal-speedup.patch
-detect-soft-lockups.patch
-detect-soft-lockups-export-touch_softlockup_watchdog.patch
-mtd-stop-the-nand-functions-triggering-false-softlockup-reports.patch
-relayfs.patch
-relayfs-relayfs_remove-fix.patch
-relayfs-api-cleanup.patch
-relayfs-add-read-support.patch
-relayfs-add-read-support-fix.patch
-relayfs-upgraded-read-implementation.patch
-relayfs-update-documentation.patch
-relayfs-update-documentation-fix.patch
-yenta-make-topic95-bridges-work-with-16bit-cards.patch
-kallsyms-change-compression-algorithm.patch
-more-__read_mostly-variables.patch
-provide-better-printk-support-for-smp-machines.patch
-provide-better-printk-support-for-smp-machines-tidy.patch
-nmi-update-nmi-users-of-rcu-to-use-new-api.patch
-nmi-update-nmi-users-of-rcu-to-use-new-api-documentation.patch
-inotify-speedup.patch
-use-select-in-sound-isa-kconfig.patch
-use-select-in-sound-isa-kconfig-fix.patch
-compat-be-more-consistent-about-id_t.patch
-fs-jbd-cleanups.patch
-strip-local-symbols-from-kallsyms.patch
-kill-bio-bi_set.patch
-s390-fix-invalid-kmalloc-flags.patch
-fix-invalid-kmalloc-flags-gfp_dma-alone.patch
-remove-register_ioctl32_conversion-and-unregister_ioctl32_conversion.patch
-clean-up-the-old-digi-support-and-rescue-it.patch
-support-powering-sharp-zaurus-sl-5500-lcd-up-and-down.patch
-radix-tree-remove-unnecessary-indirections-and-clean-up.patch
-radix_tag_get-differentiate-between-no-present-node-and-tag-unset-cases.patch
-radix_tag_get-differentiate-between-no-present-node-and-tag-unset-cases-fix.patch
-auxiliary-vector-cleanups.patch
-auxiliary-vector-cleanups-fix.patch
-pnp-consolidate-kmalloc-wrappers.patch
-fix-sound-arm-makefile-for-locality-of-reference.patch
-disk-quotas-fail-when-etc-mtab-is-symlinked-to-proc-mounts.patch
-disk-quotas-fail-when-etc-mtab-is-symlinked-to-proc-mounts-tidy.patch
-add-init=-warning-to-init-mainc.patch
-remove-a-dead-extern-in-memc.patch
-remove-misleading-comment-above-sys_brk.patch
-move-m68k-rtc-drivers-over-to-initcalls.patch
-move-68360serialc-over-use-initcalls.patch
-remove-pipe-definitions.patch
-sonypi-spic-initialisation-fix.patch
-sonypi-remove-obsolete-event.patch
-optimize-writer-path-in-time_interpolator_get_counter.patch
-pnp-make-pnp_dbg-conditional-directly-on-config_pnp_debug.patch
-readahead-reset-cahe_hit-earlier.patch
-meye-use-dma-mapping-constants.patch
-sunrpc-cache_register-can-use-wrong-module-reference.patch
-ipc-add-generic-struct-ipc_ids-seq_file-iteration.patch
-ipc-convert-proc-sysvipc-to-generic-seq_file-interface.patch
-flush-icache-early-when-loading-module.patch
-speedup-fat-filesystem-directory-reads-2.patch
-pnpacpi-fix-irq-and-64-bit-address-decoding.patch
-modified-firmware_classc-to-support-no-hotplug.patch
-dell_rbu-new-dell-bios-update-driver.patch
-dell_rbu-new-dell-bios-update-driver-fix.patch
-dcdbas-add-dell-systems-management-base-driver-with-sysfs-support.patch
-fsnotify-hook-on-removexattr-too.patch
-create_workqueue_thread-signedness-fix.patch
-proc-link-count-fix.patch
-add-rdinit-parameter-to-pick-early-userspace-init.patch
-cleanup-of-deadline_dispatch_requests.patch
-parport-add-netmos-9805-support.patch
-fs-kconfig-quota-help-text-updates.patch
-jffs-jffs2-remove-wrong-function-prototypes.patch
-i386-buildc-write-out-larger-system-size-to-bootsector.patch
-check_irq_per_cpu-to-avoid-dead-code-in-__do_irq.patch
-fix-function-macro-name-collision-on-i386-oprofile.patch
-remove-asm-hdregh.patch
-3c59x-read-current-link-status-from-phy.patch
-delete-unused-do_nanosleep-declaration.patch
-clean-up-missing-overflow-check-in-get_blkdev_list.patch
-console-blanking-locking-fix.patch
-do_notify_parent_cldstop-cleanup.patch
-arm26-one-g-is-enough-for-everyone.patch
-largefile-support-for-accounting.patch
-fs-remove-redundant-timespec_equal-test-in-update_atime.patch
-remove-a-redundant-variable-in-sys_prctl.patch
-remove-filef_maxcount.patch
-remove-the-second-arg-of-do_timer_interrupt.patch
-fix-cramfs-making-duplicate-entries-in-inode-cache.patch
-fix-cramfs-making-duplicate-entries-in-inode-cache-tidy.patch
-fix-send_sigqueue-vs-thread-exit-race.patch
-adapt-scripts-ver_linux-to-new-util-linux-version-strings.patch
-futex-remove-duplicate-code.patch
-additions-to-dataread_mostly-section.patch
-ntp-ntp-helper-functions.patch
-blk-use-blk_queue_xxx-functions-to-set-parameters.patch
-convert-proc-devices-to-use-seq_file-interface-fix.patch
-tty-layer-buffering-revamp.patch
-pipe-remove-redundant-fifo_poll-abstraction.patch
-remove-verify_area-remove-verify_area-from-various-uaccessh-headers.patch
-remove-verify_area-remove-or-edit-references-to-verify_area-in-documentation.patch
-remove-verify_area-remove-fs-umsdos-notes-as-it-only-contain-a-verify_area-related-note.patch
-optimise-64bit-unaligned-access-on-32bit-kernel.patch
-vt-fix-possible-memory-corruption-in-complement_pos.patch
-hpet-fix-drift-and-url.patch
-isdn_v110-warning-fix.patch
-tpm-fix-tpm_atmelc-on-ich6.patch
-create-asm-generic-fcntlh.patch
-consildate-asm-ppc-fcntlh.patch
-clean-up-the-open-flags.patch
-clean-up-the-fcntl-operations.patch
-clean-up-struct-flock-definitions.patch
-clean-up-struct-flock64-definitions.patch
-consolidate-the-asm-ppc-fcntlh-files-into-asm-powerpc.patch
-inotify-fix-event-loss-on-hardlinked-files.patch
-sunrpc-print-unsigned-integers-in-stats.patch
-block-cfq-refcounting-fix.patch
-remove-ia_attr_flags.patch
-namei-cleanup.patch
-use-get_fs_struct-in-proc.patch
-fix-enum-pid_directory_inos-in-proc-basec.patch
-remove-duplicated-code-from-proc-and-ptrace.patch
-remove-duplicated-sys_open32-code-from-64bit-archs.patch
-cifs_create-fix.patch
-dmi-remove-uneeded-function.patch
-dmi-remove-old-debugging-code.patch
-dmi-make-dmi_string-behave-like-strdup.patch
-dmi-add-onboard-devices-discovery.patch
-dmi-add-onboard-devices-discovery-fix.patch
-ipmi-use-dmi_find_device.patch
-fix-dmi_check_system.patch
-introduce-and-use-kzalloc.patch
-introduce-and-use-kzalloc-make-kcalloc-a-static-inline.patch
-ia64-convert-kcalloc-to-kzalloc.patch
-ppc64-convert-kcalloc-to-kzalloc.patch
-input-convert-kcalloc-to-kzalloc.patch
-usb-convert-kcalloc-to-kzalloc.patch
-drivers-convert-kcalloc-to-kzalloc.patch
-fs-convert-kcalloc-to-kzalloc.patch
-alsa-convert-kcalloc-to-kzalloc.patch
-ipmi-add-per-channel-ipmb-addresses.patch
-ipmi-high-res-timer-support-fixes.patch
-ipmi-watchdog-nmi-interaction-fixes.patch
-ipmi-allow-userland-to-include-ipmih.patch
-ipmi-oem-flag-handling-and-hacks-for-some-dell-machines.patch
-ipmi-clean-up-versioning-of-the-ipmi-driver.patch
-ipmi-fix-panic-ipmb-response.patch
-hfs-remove-debug-code.patch
-hfs-show_options-support.patch
-hfs-nls-support.patch
-sd-initialize-sd-cards.patch
-sd-read-only-switch.patch
-sd-read-only-switch-coding-style-fix.patch
-sd-read-only-switch-mmc-sd-init-order-fix.patch
-sd-read-only-switch-mmc-sd-ro-style-fix.patch
-sd-scr-register.patch
-sd-scr-register-fix-a-bit-byte-counting-error-in-the-mmc-sd-code.patch
-sd-scr-register-mmc-sd-scr-style-fixpatch.patch
-sd-scr-in-sysfs.patch
-sd-4-bit-bus.patch
-sd-copyright-notice.patch
-add-write-protection-switch-handling-to-the-pxa-mmc-driver.patch
-mmc-wbsd-secure-digital-support.patch
-mmc-conditional-scr-sysfs-entry.patch
-corgi-keyboard-fix-a-couple-of-compile-errors.patch
-corgi-keyboard-add-some-power-management-code.patch
-corgi-keyboard-code-tidying.patch
-corgi-touchscreen-allow-the-driver-to-share-the-pmu.patch
-corgi-touchscreen-code-cleanup--fixes.patch
-corgi-touchscreen-fix-a-pmu-bug.patch
-w100fb-rewrite-for-platform-independence.patch
-w100fb-rewrite-for-platform-independence-fix.patch
-w100fb-update-corgi-platform-code-to-match-new.patch
-input-add-a-new-switch-event-type.patch
-corgi-add-keyboard-and-touchscreen-device-definitions.patch
-corgi-add-mmc-sd-write-protection-switch-handling.patch
-kjournald-missing-jfs_unmount-check.patch
-fix-jbd-race-in-t_forget-list-handling.patch
-make-ll_rw_block-wait-for-buffer-lock.patch
-change-ll_rw_block-calls-in-jbd.patch
-change-ll_rw_block-calls-in-reiser.patch
-change-ll_rw_block-calls-in-ufs.patch
-change-hfs-to-not-use-ll_rw_block.patch
-fix-race-in-do_get_write_access.patch
-smsc-ircc2-whitespace-fixes.patch
-smsc-ircc2-formatting-fixes.patch
-smsc-ircc2-drop-dim-macro-in-favor-of-array_size.patch
-smsc-ircc2-remove-typedefs.patch
-smsc-ircc2-dont-pass-iobase-around.patch
-smsc-ircc2-add-to-sysfs-as-platform-device-new-pm.patch
-smsc-ircc2-pm-cleanup-do-not-close-device-when-suspending.patch
-smsc-ircc2-use-netdev_priv.patch
-smsc-ircc2-dont-use-void-where-specific-type-will-do.patch
-fix-smsc_ircc_init-return-value.patch
-kprobes-prevent-possible-race-conditions-generic.patch
-kprobes-prevent-possible-race-conditions-i386-changes.patch
-kprobes-prevent-possible-race-conditions-x86_64-changes.patch
-kprobes-prevent-possible-race-conditions-ppc64-changes.patch
-kprobes-prevent-possible-race-conditions-ia64-changes.patch
-kprobes-prevent-possible-race-conditions-sparc64-changes.patch
-kprobes-prevent-possible-race-conditions-sparc64-changes-fix.patch
-kprobes-ia64-fix-race-when-break-hits-and-kprobe-not-found.patch
-kprobes-fix-handling-of-simultaneous-probe-hit-unregister.patch
-pivot_root-circular-reference-fix-3.patch
-indycam--vino-drivers.patch
-dvb-clarify-description-text-for-dvb-bt8xx-in-kconfig.patch
-dvb-clarify-description-text-for-dvb-bt8xx-in-kconfig-fix.patch
-dvb-lgdt330x-check-callback-fix.patch
-yenta-auto-tune-ene-bridges-for-cb-cards.patch
-pcmcia-yenta-dont-mess-with-bridge-control-register.patch
-suspend-update-warnings-in-documentation.patch
-mm-filemapc-make-sync_page_range_nolock-static.patch
-mm-filemapc-make-generic_file_direct_io-static.patch

 Merged

+ip_conntrack_netbios_ns-build-fix.patch

 net build fix

+clear-task_struct-fs_excl-on-fork.patch

 CFQ fix

+i386-single-node-sparsemem-fix.patch

 sparsemem fix

+i386-config_acpi_srat-typo-fix.patch

 Kconfig fix

+ppc64-fix-oops-for-config_numa.patch

 pcp64 oops fix

+ppc32-fix-kconfig-mismerge.patch

 ppc32 fix

+acpi-should-depend-on-not-select-pci.patch

 ACPI fix

+sound-support-for-vaio-ra826g-hda.patch

 Vaio sound driver fix

+gregkh-driver-sysfs-strip_leading_trailing_whitespace-fix.patch

 Fix gregkh-driver-sysfs-strip_leading_trailing_whitespace.patch

+input-keyboard_tasklet-dont-touch-leds-of-already-grabed-device.patch

 input driver fix

-git-jfs-fixup.patch

 Unneeded

+fix-split-include-dependency.patch

 build fix

-ipw2100-remove-by-hand-function-entry-exit-debugging.patch
-ieee80211_module-build-fixes.patch
-ieee80211_tx-build-fix.patch
-ieee80211_rx-build-fix.patch
-ieee80211_crypt-build-fix.patch
-ieee80211_crypt_ccmp-build-fix.patch
-ieee80211_crypt_wep-build-fix.patch
-ieee80211_crypt_tkip-build-fix.patch

 Unneeded

-ocfs2-prep.patch

 Dropped this - maintaining the
 move-truncate_inode_pages-into-delete_inode.patch patch separately was a
 pita.  Simply pull it in from git-ocfs2.patch

+txx9-serial-update.patch

 MIPS serial driver fix

+arch-pci_find_device-remove-alpha-kernel-sys_alcorc.patch
+arch-pci_find_device-remove-alpha-kernel-sys_sioc.patch
+arch-pci_find_device-remove-frv-mb93090-mb00-pci-frvc.patch
+arch-pci_find_device-remove-frv-mb93090-mb00-pci-irqc.patch
+arch-pci_find_device-remove-ppc-kernel-pcic.patch
+arch-pci_find_device-remove-ppc-platforms-85xx-mpc85xx_cds_commonc.patch
+arch-pci_find_device-remove-sparc64-kernel-ebusc.patch
+arch-pci_find_device-remove-sparc64-kernel-ebusc-fix.patch

 PCI cleanups

+pci-block-config-access-during-bist.patch
+pci-block-config-access-during-bist-update.patch
+pci-block-config-access-during-bist-update-2.patch
+pci-block-config-access-during-bist-fix-42.patch
+ipr-block-config-access-during-bist.patch
+pci-unhide-smbus-on-compaq-evo-n620c.patch

 Don't let userspace access PCI registers while we're issuing a BIST.

+usbdevice_fs-header-breakage.patch

 USB fix

+w83977f-watchdog-driver.patch

 New watchdog driver

+x86_64-defconfig-update.patch
+x86_64-smpboot-write-around.patch
+x86_64-genapic-write-around.patch
+x86_64-remove-esr-disable.patch
+x86_64-remove-apic-errata.patch
+x86_64-huge-cpu-apicids.patch
+x86_64-calibrate-enable-irqs.patch
+x86_64-smp-call-single-cleanup.patch
+x86_64-srat-apicid.patch
+x86_64-apic-version.patch
+x86_64-pda-up-align.patch
+x86_64-numa-k8-nodeid.patch
+x86_64-intel-srat.patch
+x86_64-apic-bsp-id-up.patch
+x86_64-swiotlb-bounce.patch
+x86_64-pgdat-order.patch
+x86_64-aperture-swiotlb.patch
+x86_64-swiotlb-force.patch
+x86_64-mce-lockup.patch
+x86_64-mce-cmd-line.patch
+x86_64-up-apicid.patch
+x86_64-pci-gart-node-opt.patch
+x86_64-e820-off-by-one.patch
+x86_64-reserved-check.patch
+x86_64-flatmem-end-pfn.patch
+x86_64-mce-smp-resume.patch
+x86_64-pci-pxm.patch
+x86_64-dma32.patch
+x86_64-dma32-iommu.patch
+x86_64-dma32-ia64-compat.patch
+x86_64-print-uts-version.patch
+x86_64-sendfile-fix.patch
+x86_64-scalable-tlb-flush.patch
+x86_64-simnow-console.patch
+x86_64-pda-cleanup.patch
+x86_64-timex-config.patch
+x86_64-tlb-flush-array.patch
+x86_64-cpu-data-possible.patch
+x86_64-vm-holes-reserved.patch
+x86_64-idle-poll-fix.patch
+x86_64-early-page-typo.patch
+x86_64-setup-merge.patch
+x86_64-drop-disable-tsc.patch
+x86_64-show_mem-printk.patch
+x86_64-syscall-clobber.patch
+x86_64-pfn-valid-off-by-one.patch
+x86_64-nodemap-extend.patch
+x86_64-irq-bug.patch
+x86_64-hotplug-typo.patch
+x86_64-trampoline-free.patch
+x86_64-extern-inline.patch
+x86_64-lowest-pri.patch
+x86_64-vsyscall-gcc4.patch
+x86_64-srat-overlap-error.patch
+x86_64-dma-sync-range.patch
+x86_64-physflat-intel.patch

 x86_64 tree

+add-sem_is_read-write_locked-fix.patch
+add-sem_is_read-write_locked-fix-2.patch
+add-sem_is_read-write_locked-fix-3.patch

 Fix add-sem_is_read-write_locked.patch

-use-mm_counter-macros-for-nr_pte-since-its-also-under-ptl.patch
-page-fault-patches-introduce-pte_xchg-and-pte_cmpxchg.patch
-page-fault-patches-optional-page_lock-acquisition-in.patch
-page-fault-patches-optional-page_lock-acquisition-in-vs-use-mm_counter-macros-for-nr_pte-since-its-also-under-ptl.patch
-page-fault-patches-optional-page_lock-acquisition-in-nicety.patch
-page-fault-patches-no-pagetable-lock-in-do_anon_page.patch

 Drop these - we'll test Hugh's approach to fixing page_table_lock contention.

+swaptoken-tuning-fix-2.patch

 Fix swaptoken-tuning.patch

+memory-hotplug-prep-kill-local_mapnr.patch
+memory-hotplug-prep-break-out-zone-initialization.patch
+memory-hotplug-prep-break-out-zone-initialization-fix.patch
+memory-hotplug-prep-__section_nr-helper.patch
+memory-hotplug-prep-__section_nr-helper-fix.patch
+memory-hotplug-prep-fixup-bad_range.patch
+memory-hotplug-locking-node_size_lock.patch
+memory-hotplug-locking-zone-span-seqlock.patch
+memory-hotplug-sysfs-and-add-remove-functions.patch
+memory-hotplug-move-section_mem_map-alloc-to-sparsec.patch
+memory-hotplug-call-setup_per_zone_pages_min-after-hotplug.patch
+memory-hotplug-i386-addition-functions.patch
+memory-hotplug-i386-addition-functions-warning-fix.patch
+memory-hotplug-ppc64-specific-hot-add-functions.patch

 Memory hot-unplug support

+3c59x-convert-to-use-of-pci-iomap-api.patch
+pcnet32-set_ringparam-implementation.patch
+pcnet32-set_ringparam-implementation-tidy.patch
+skge-kconfig-help-text-typo-fix.patch

 netdev fixes

+acx1xx-wireless-driver.patch
+acx1xx-wireless-driver-usb-is-bust.patch
+acx1xx-allow-modular-build.patch
+acx1xx-wireless-driver-spy_offset-went-away.patch

 Driver for TI acx1xx wireless cards (a bit rough.  I happen to have one of
 these cards.  I got it to ping something).

+ppc32-make-perfmono-config_e500-specific.patch

 ppc32 fix

+mips-add-tanbac-tb0287-support.patch

 MIPS device support

+i386-seccomp-fix-for-auditing-ptrace.patch
+x86-cache-pollution-aware-__copy_from_user_ll.patch
+x86-cache-pollution-aware-__copy_from_user_ll-tidy.patch
+x86-cache-pollution-aware-__copy_from_user_ll-build-fix.patch
+x86-cache-pollution-aware-__copy_from_user_ll-build-fix-2.patch

 x86 updates

+x86_64-nmi-watchdog-frequency-calculation-adjustments.patch

 x86_64 NMI watchdog fix

-x86_64fix-cluster-mode-send_ipi_allbutself-to-use-get_cpu-put_cpu.patch
-x86_64dont-use-lowest-priority-when-using-physical-mode.patch
-x86_64use-common-functions-in-cluster-and-physflat-mode.patch
-x86_64-choose-physflat-for-amd-systems-only-when-8-cpus.patch

 Dropped, or in Andi's tree.

+alpha-process_reloc_for_got-confuses-r_offset-and-r_addend.patch

 Alpha fix

-pselect-ppoll-system-calls-tidy.patch
-pselect-ppoll-system-calls-fix.patch
-pselect-ppoll-system-calls-sigset_t-fix-2.patch
-pselect-ppoll-system-calls-sigset_t-fix-3.patch
-pselect-ppoll-system-calls-compat-fix.patch
-pselect-ppoll-system-calls-copy_to_user-check.patch

 Folded into pselect-ppoll-system-calls.patch

+convert-proc-devices-to-use-seq_file-interface-tidy.patch

 cleanup

-open-returns-enfile-but-creates-file-anyway-tidy.patch

 Folded into open-returns-enfile-but-creates-file-anyway.patch

+vga-text-console-and-stty-cols-rows.patch
+vga-text-console-and-stty-cols-rows-tidy.patch

 VGA fix

+autofs-fix-busy-inodes-after-umount.patch

 autofs3 fix

+fix-disassociate_ctty-vs-fork-race.patch

 Fix a race

+ide-scsi-highmem-cleanup.patch

 Clean up ide-scsi a bit

+prefetch-kernel-stacks-to-speed-up-context-switch.patch

 ia64 speedup

+bfs-fix-endianness-signedness-add-trivial-bugfix.patch

 BFS fixes

+cs89x0-add-netpoll-support.patch

 cs80x0 feature

+change-io_cancel-return-code-for-no-cancel-case.patch
+kiocb-locking-to-serialise-retry-and-cancel-2.patch

 AIO fixes

+subcpusets-fix-for-cpusets-minor-problem.patch

 cpusets fix

+remove-unnecessary-handle_irq_event-prototypes.patch

 cleanup

+deadline-cleanup-question-mark-operator.patch

 deadline cleanup

+parport-buffer-overflow-fix.patch
+parport-phase-fixes.patch
+parport-daisy-chain-end-detection-fix.patch
+parport-daisy-chain-device-id-reading-fix.patch
+parport-use-complete-slab-buffer.patch
+parport-constification.patch
+parport-debug_parport-build-fix.patch
+parport-kconfig-dependency-fixes.patch
+parport-include-fixes.patch
+parport-export-parport_get_port.patch

 parport driver fixes

+synclinkc-compiler-optimisation-fix.patch
+synclinkc-add-clear-stats.patch
+synclinkc-add-loopback-to-async-mode.patch
+synclinkmpc-fix-double-mapping-of-signals.patch
+synclinkmpc-disable-burst-transfers.patch
+synclinkmpc-add-statistics-clear.patch
+synclinkmpc-fix-async-internal-loopback.patch

 synclink update

-coverity-usb-host-ehci-dbg-null-check.patch

 Folded into coverity-udf-balloc-null-deref-fix.patch

+sharpsl-abstract-c7x0-specifics-from-corgi-ssp.patch
+sharpsl-add-cxx00-support-to-the-corgi-lcd-driver.patch
+sharpsl-abstract-c7x0-specifics-from-corgi.patch
+sharpsl-abstract-model-specifics-from-corgi.patch
+sharpsl-add-new-arm-pxa-machines-spitz-and-borzoi.patch

 Zaurus updates

+rapidio-message-interface-updates.patch
+rapidio-support-net-driver-fixes.patch

 rapidio driver updates

+dvb-email-address-update.patch
+dvb-remove-versionh-dependencies.patch
+dvb-avoid-building-empty-built-ino.patch
+dvb-core-glue-code-for-dmx_get_caps-and-dmx_set_source.patch
+dvb-core-dvb_demux-fix-continuity-counter-error-handling.patch
+dvb-core-dvb_demux-remove-unused-cruft.patch
+dvb-core-dvb_demux-remove-unsused-descramble-callbacks.patch
+dvb-core-dvb_demux-remove-more-unused-cruft.patch
+dvb-core-dvb_demux-use-init_list_head.patch
+dvb-core-dvb_demux-formatting-fixes.patch
+dvb-core-ci-timeout-fix.patch
+dvb-frontend-mt352-fix-signal-strength-reading.patch
+dvb-frontend-stv0299-pass-i2c-bus-to-pll-callback.patch
+dvb-frontend-s5h1420-fixes.patch
+dvb-frontend-stv0299-support-reading-both-ber-and-ucblocks.patch
+dvb-frontend-tda1004x-fix-snr-reading.patch
+dvb-frontend-ves1820-improve-tuning.patch
+dvb-frontend-cx24110-diseqc-fix.patch
+dvb-frontend-cx24110-another-diseqc-fix.patch
+dvb-frontend-cx24110-clean-up-timeout-handling.patch
+dvb-frontend-stv0297-qam128-tuning-improvement.patch
+dvb-frontend-or51132-remove-bogus-optimization-attempt.patch
+dvb-usb-add-twinhandtv-starbox-support.patch
+dvb-usb-dibusb-kworld-xpert-dvb-t-usb20-support.patch
+dvb-usb-removed-empty-module_init-exit-calls.patch
+dvb-usb-dtt200u-copy-frontend_ops-before-modifying.patch
+dvb-usb-dtt200u-add-proper-device-names.patch
+dvb-usb-core-change-dvb_usb_device_init-api.patch
+dvb-usb-digitv-support-for-nxt6000-demod.patch
+dvb-usb-white-space-cleanup.patch
+dvb-usb-cxusb-fixes-for-new-firmware.patch
+dvb-remove-noisy-debug-print.patch
+dvb-bt8xx-endianness-fix.patch
+dvb-bt8xx-cleanup.patch
+dvb-bt8xx-nebula-digitv-mt352-support.patch
+dvb-nebula-digitv-nxt6000-fix.patch
+dvb-dst-fix-symbol-rate-setting.patch
+dvb-dst-remove-unnecessary-code.patch
+dvb-dst-dprrintk-cleanup.patch
+dvb-dst-dprrintk-cleanup-gcc-29x-fix.patch
+dvb-dst-dprrintk-cleanup-gcc-295-fix.patch
+dvb-dst-identify-boards.patch
+dvb-dst-fix-dvb-c-tuning.patch
+dvb-dst-ci-doc-update.patch
+dvb-dst-updated-documentation.patch
+dvb-cinergyt2-remote-control-fixes.patch
+dvb-av7110-siemens-dvb-c-analog-video-input-support.patch
+dvb-budget-ci-add-support-for-tt-dvb-c-ci-card.patch
+dvb-budget-av-fixes-for-ci-interface.patch
+dvb-budget-av-enable-frontend-on-knc1-plus-cards.patch
+dvb-av7110-disable-superflous-firmware-handshake.patch
+dvb-av7110-conditionally-disable-workaround-for-broken-firmware.patch
+dvb-ttpci-av7110-rc5-remote-control-support.patch
+dvb-ttpci-add-pci-ids-for-old-siemens-tt-dvb-c-card.patch
+dvb-saa7146-i2c-vs-sysfs-fix.patch
+dvb-ttusb-budget-use-time_after_eq.patch

 DVB updates

+pcmcia-avoid-macro-usage.patch
+pcmcia-tiny-yenta_socketc-cleanup.patch
+pcmcia-warn-on-ioctl-usage.patch
+pcmcia-new-suspend-core.patch
+pcmcia-new-suspend-core-dev_to_instance-fix.patch
+pcmcia-convert-drivers-to-use-new-suspend-mechanism.patch
+pcmcia-convert-drivers-to-use-new-suspend-mechanism-spectrum_cs.patch
+pcmcia-convert-serial_cs-to-use-new-suspend-mechanism.patch
+pcmcia-use-runtime-suspend-resume-support-to-unify-all-suspend-code-paths.patch
+pcmcia-unified-device-removal-code-path.patch
+pcmcia-convert-drivers-to-use-unified-removal-code-path.patch
+pcmcia-remove-old-two-step-removal-mechanism.patch
+pcmcia-remove-unused-dev_list-in-drivers.patch
+pcmcia-unified-probe-code-path.patch
+pcmcia-convert-drivers-to-use-new-probe-mechanism.patch
+pcmcia-yenta-dont-mess-with-bridge-control-register.patch
+yenta-auto-tune-ene-bridges-for-cb-cards.patch

 Largeish pcmcia update

+nfsacl-solaris-vxfs-compatibility-fix.patch

 NFS fix

+spinlock-consolidation-sparc64-fix.patch

 Fix spinlock-consolidation-ia64-fix.patch

+scheduler-cache-hot-autodetect.patch

 Bring this scheduler feature back (it does way too much printk()ing still)

+sched-use-cached-variable-in-sys_sched_yield.patch

 scheduler microoptimisation

+m68k-introduce-task_thread_info.patch
+m68k-introduce-setup_thread_stack-end_of_stack.patch
+m68k-thread_info-header-cleanup.patch
+m68k-m68k-specific-thread_info-changes.patch
+m68k-convert-thread-flags-to-use-bit-fields.patch
+add-stack-field-to-task_struct.patch
+add-stack-field-to-task_struct-ia64-fix.patch
+rename-allocfree_thread_info-to-allocfree_thread_stack.patch
+use-end_of_stack.patch
+change-thread_info-access-to-stack.patch
+use-task_thread_info.patch

 Replace task_struct.thread_info with task_struct.stack and do m68k fixes
 sort-of in the middle.

+v4l-common-part-updates-and-tuner-additions.patch
+v4l-common-part-updates-and-tuner-additions-gcc-29x-fix.patch
+v4l-bttv-updates-and-card-additions.patch
+v4l-bttv-updates-and-card-additions-fix.patch
+v4l-cx88-updates-and-card-additions.patch
+v4l-cx88-updates-and-card-additions-gcc-295-fix.patch
+v4l-saa7134-updates-and-board-additions.patch
+v4l-changes-the-prefix-of-msp34xx-and-error-while.patch
+v4l-syncs-tveeprom-tuners-list-with-the-list-from.patch
+v4l-correct-lg-ntsc-taln-mini-tuner-takeover.patch
+v4l-add-saa713x-card-65-kworld-v-stream-studio-tv-terminator.patch
+v4l-add-saa713x-card-66-yuan-tun-900-saa7135.patch
+v4l-cx88-dvb-incorrectly-reporting-fixed-and.patch
+v4l-normalize-whitespace-and-comments-in-tuner.patch
+v4l-change-lg-tdvs-h062f-from-ntsc-to-atsc.patch
+v4l-some-error-treatment-implemented-at-resume.patch
+v4l-the-microtune-4049fm5-uses-an-if-frequency-of.patch
+v4l-include-linux-configh-no-longer-needed.patch
+v4l-correct-the-amux-for-composite-and-s-video.patch
+v4l-print-warning-if-pal=-or-secam=-argument-is.patch
+v4l-added-some-missing-parameter-descriptions-at.patch
+v4l-makes-the-input-event-device-for-ir-matchable.patch
+v4l-include-saa6588-compiler-option-and-files.patch
+v4l-removed-kernel-version-dependency-from.patch
+v4l-tvaudio-cleanup-and-better-degug-messages.patch
+v4l-tvaudio-cleanup-and-better-degug-messages-gcc-29c-fix.patch
+v4l-tveeprom-improved-to-support-newer-hauppage-cards.patch
+v4l-tveeprom-improved-to-support-newer-hauppage-cards-gcc-29x-fix.patch

 v4l updates

+fix-kernel-oops-with-cf-cards.patch

 Fix compact-flash bug

+fbdev-geode-updates-fix.patch

 Fix fbdev-geode-updates.patch

+sisfb-update-resurrect-makefile.patch
+sisfb-update-makefile.patch

 Fix sisfb-update.patch

+fbdev-add-vesa-coordinated-video-timings-cvt-support.patch
+nvidiafb-use-cvt-to-get-mode-for-digital-displays.patch
+savagefb-make-mode_option-available-when-compiled-as.patch
+fbcon-stop-cursor-timer-if-console-is-inactive.patch
+nvidiafb-fixed-mirrored-characters-in-big-endian-machines.patch
+fbdev-initialize-var-structure-in-calc_mode_timings.patch
+pxafb-add-hsync-time-reporting-hook.patch
+fbcon-break-up-bit_putcs-into-its-component-functions.patch
+fbcon-break-up-bit_putcs-into-its-component-functions-fix.patch
+i810fb-add-i2c-ddc-support.patch
+i810fb-add-i2c-ddc-support-fix.patch
+i810fb-add-i2c-ddc-support-fix-fix.patch
+i810fb-add-i2c-ddc-support-Makefile-fix.patch
+i810fb-stop-lcd-displays-from-flickering.patch
+quiet-non-x86-option-rom-warnings.patch
+s3c2410fb-arm-s3c2410-framebuffer-driver.patch
+s3c2410fb-platform-support-for-arm-s3c2410-framebuffer.patch

 fbdev driver updates

+drivers-md-raid1c-make-a-function-static.patch
+md-choose-better-default-offset-for-bitmap.patch
+md-use-queue_hardsect_size-instead-of-block_size-for-md-superblock-size-calc.patch
+md-add-information-about-superblock-version-to-proc-mdstat.patch
+md-report-spare-drives-in-proc-mdstat.patch
+md-make-sure-the-new-sb_size-is-set-properly-device-added-without-pre-existing-superblock.patch
+md-really-get-sb_size-setting-right-in-all-cases.patch
+md-fix-raid10-assembly-when-too-many-devices-are-missing.patch
+md-fix-bug-when-raid10-rebuilds-without-enough-drives.patch

 RAID updates

+documentation-sparse-snapshot-url.patch

 Documentation fix

+fuse-more-flexible-caching.patch

 fuse update

+mm-filemapc-make-two-functions-static.patch
+mm-filemapc-make-sync_page_range_nolock-static.patch
+mm-filemapc-make-generic_file_direct_io-static.patch

 scoping fixes

+lib-sortc-small-cleanups.patch
+drivers-net-wan-possible-cleanups.patch
+remove-acpi-s4bios-support.patch
+fs-cramfs-uncompressc-should-include-linux-cramfs_fsh.patch
+i386-x86_64-make-get_cpu_vendor-static.patch

 cleanuplets



All 874 patches:

ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/patch-list

Re: 2.6.13-mm2

[Benoit Boissinot]

On 9/8/05, Andrew Morton <akpm@osdl.org> wrote:
> 
> ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> 

>  git-cifs.patch

it adds a new compilation warning with gcc-4:
fs/cifs/cifsglob.h:335: warning: type qualifiers ignored on function
return type

The following patch fixes it (removes the const qualifier)

Signed-off-by: Benoit Boissinot <benoit.boissinot@ens-lyon.org>


--- ./fs/cifs/cifsglob.h	2005-09-08 14:50:34.000000000 +0200
+++ ./fs/cifs/cifsglob.h.new	2005-09-08 15:02:50.000000000 +0200
@@ -331,7 +331,7 @@ CIFS_SB(struct super_block *sb)
 	return sb->s_fs_info;
 }
 
-static inline const char CIFS_DIR_SEP(const struct cifs_sb_info *cifs_sb)
+static inline char CIFS_DIR_SEP(const struct cifs_sb_info *cifs_sb)
 {
 	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS)
 		return '/';

Re: 2.6.13-mm2

[Christoph Hellwig]

On Thu, Sep 08, 2005 at 05:30:42AM -0700, Andrew Morton wrote:
> -ocfs2-prep.patch
> 
>  Dropped this - maintaining the
>  move-truncate_inode_pages-into-delete_inode.patch patch separately was a
>  pita.  Simply pull it in from git-ocfs2.patch

Could you please just send the  move truncate_inode_pages into delete_inode
patch to Linus ASAP instead?  It's useful without OCFS2, I have some changes
that need it pending, and the reiser4 folks also need it to clean some
of their mess up (if they're still interested..)

Re: 2.6.13-mm2

[Martin J. Bligh]

> ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> 
> (kernel.org propagation is slow.  There's a temp copy at
> http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)
> 
> 
> 
> - Added Andi's x86_64 tree, as separate patches
> 
> - Added a driver for TI acx1xx cardbus wireless NICs
> 
> - Large revamp of pcmcia suspend handling
> 
> - Largeish v4l and DVB updates
> 
> - Significant parport rework
> 
> - Many tty drivers still won't compile
> 
> - Lots of framebuffer driver updates
> 
> - There are still many patches here for 2.6.14.  We're doing pretty well
>   with merging up the subsystem trees.  ia64 and CIFS are still pending. 
>   x86_64 and several of Greg's trees (especially USB) aren't merged yet.

Build fails on x86_64, at least, with this config:
http://ftp.kernel.org/pub/linux/kernel/people/mbligh/config/abat/amd64

arch/x86_64/pci/built-in.o(.init.text+0xa88): In function `pci_acpi_scan_root':
: undefined reference to `pxm_to_node'
make: *** [.tmp_vmlinux1] Error 1
09/08/05-06:52:31 Build the kernel. Failed rc = 2
09/08/05-06:52:31 build: kernel build Failed rc = 1
09/08/05-06:52:31 command complete: (2) rc=126
Failed and terminated the run

Re: 2.6.13-mm2

[Michael Thonke]

Hello Andrew,

I found a problem regarding to multi device support (Linux software Raid).

The problem first appears now in 2.6.13-mm2, that the kernel didn't initialize the md devices.

2.6.13-mm1 works very well, and everything is okay.

Also one strange thing I found was that my SATA devices were initialized a-sync/disordered e.g

SATA1 with one hdd then something like USB and IPv4 and such and at least SATA3: with 2nd hdd.
That I've never seen this order init order before. Seems to be mixed all around.

I tried irqpoll,pci=routeirq with no success.

I can't provide some logs, because I can't grep the dmesg since it doesn't boot.

There are changes in libata driver for sata_nv? Or md driver changes that cause that?


Thanks

Best regards

--
Michael Thonke

Re: 2.6.13-mm2

[Andrew Morton]

Michael Thonke <iogl64nx@gmail.com> wrote:
>
> Hello Andrew,
> 
> I found a problem regarding to multi device support (Linux software Raid).
> 
> The problem first appears now in 2.6.13-mm2, that the kernel didn't initialize the md devices.
> 
> 2.6.13-mm1 works very well, and everything is okay.
> 
> Also one strange thing I found was that my SATA devices were initialized a-sync/disordered e.g
> 
> SATA1 with one hdd then something like USB and IPv4 and such and at least SATA3: with 2nd hdd.
> That I've never seen this order init order before. Seems to be mixed all around.
> 
> I tried irqpoll,pci=routeirq with no success.
> 
> I can't provide some logs, because I can't grep the dmesg since it doesn't boot.
> 
> There are changes in libata driver for sata_nv? Or md driver changes that cause that?
> 

There are changes to both sata_nv and to md in 2.6.13-mm2.  To isolate them
it would be great of you could apply 

ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/broken-out/linus.patch

to 2.6.13 and see if the problem still happens.  That will separate out the
md changes which are still in -mm.

Thanks.

Re: 2.6.13-mm2

[Ronny V. Vindenes]

Andrew Morton <akpm@osdl.org> writes:

> ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> 
> (kernel.org propagation is slow.  There's a temp copy at
> http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)
> 
> 
> 
> - Added Andi's x86_64 tree, as separate patches
> 
> - Added a driver for TI acx1xx cardbus wireless NICs
> 
> - Large revamp of pcmcia suspend handling
> 
> - Largeish v4l and DVB updates
> 
> - Significant parport rework
> 
> - Many tty drivers still won't compile
> 
> - Lots of framebuffer driver updates
> 
> - There are still many patches here for 2.6.14.  We're doing pretty well
>   with merging up the subsystem trees.  ia64 and CIFS are still pending. 
>   x86_64 and several of Greg's trees (especially USB) aren't merged yet.
> 

x86-64-ptrace-ia32-bp-fix.patch breaks all 32bit apps for me on Athlon64

-- 
Ronny V. Vindenes <s864@ii.uib.no>

Re: 2.6.13-mm2

[Andrew Morton]

s864@ii.uib.no (Ronny V. Vindenes) wrote:
>
> x86-64-ptrace-ia32-bp-fix.patch breaks all 32bit apps for me on Athlon64

Great, thanks muchly for working that out.

Parag, perhaps you could confirm that reverting that patch fixes things up?

Re: 2.6.13-mm2

[Parag Warudkar]

Andrew Morton wrote:

> Parag, perhaps you could confirm that reverting that patch fixes 
> things up?

Sure - reverting the x86-64-ptrace-ia32-bp-fix patch fixes it.

Roland - if seeing backtraces and register info for the failing programs 
is going to help you, please
see the thread "2.6.13-mm1 X86_64: All 32bit programs segfault"

Thanks
Parag

Re: 2.6.13-mm2

[Andi Kleen]

On Thu, Sep 08, 2005 at 07:30:01AM -0700, Martin J. Bligh wrote:
> 
> > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> > 
> > (kernel.org propagation is slow.  There's a temp copy at
> > http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)
> > 
> > 
> > 
> > - Added Andi's x86_64 tree, as separate patches
> > 
> > - Added a driver for TI acx1xx cardbus wireless NICs
> > 
> > - Large revamp of pcmcia suspend handling
> > 
> > - Largeish v4l and DVB updates
> > 
> > - Significant parport rework
> > 
> > - Many tty drivers still won't compile
> > 
> > - Lots of framebuffer driver updates
> > 
> > - There are still many patches here for 2.6.14.  We're doing pretty well
> >   with merging up the subsystem trees.  ia64 and CIFS are still pending. 
> >   x86_64 and several of Greg's trees (especially USB) aren't merged yet.
> 
> Build fails on x86_64, at least, with this config:
> http://ftp.kernel.org/pub/linux/kernel/people/mbligh/config/abat/amd64
> 
> arch/x86_64/pci/built-in.o(.init.text+0xa88): In function `pci_acpi_scan_root':
> : undefined reference to `pxm_to_node'
> make: *** [.tmp_vmlinux1] Error 1
> 09/08/05-06:52:31 Build the kernel. Failed rc = 2
> 09/08/05-06:52:31 build: kernel build Failed rc = 1
> 09/08/05-06:52:31 command complete: (2) rc=126
> Failed and terminated the run

I tried the config in my (non mm) tree and it compiled just fine.
Must be some bad interaction with another patch in -mm* or a bad 
merge.

The original patch that introduces it is
ftp://ftp.firstfloor.org/pub/ak/x86_64/x86_64-2.6.13-1/patches/pci-pxm

pxm_to_node for x86-64 is supposed to be declared in arch/x86_64/mm/srat.c

Andrew?

-Andi

Re: 2.6.13-mm2

[Roland McGrath]

I guess something else has changed since I tested the patch.  I haven't
tried -mm2, but the current Linus tree I'm having trouble getting to boot
on my x86_64 machine atm ("soft lockup" in the e1000 driver setup).


Thanks,
Roland

Re: 2.6.13-mm2

[Grant Coady]

On Thu, 8 Sep 2005 05:30:42 -0700, Andrew Morton <akpm@osdl.org> wrote:

>
>ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/

Hi Andrew,

After this error:

  CC      drivers/parport/parport_pc.o
drivers/parport/parport_pc.c:2511: error: via_686a_data causes a section type conflict
drivers/parport/parport_pc.c:2520: error: via_8231_data causes a section type conflict
drivers/parport/parport_pc.c:2705: error: parport_pc_superio_info causes a section type conflict
drivers/parport/parport_pc.c:2782: error: cards causes a section type conflict
make[2]: *** [drivers/parport/parport_pc.o] Error 1
make[1]: *** [drivers/parport] Error 2
make: *** [drivers] Error 2

got this:

grant@sempro:/opt/linux/linux-2.6.13-mm2a$ make menuconfig
  HOSTCC  scripts/kconfig/conf.o
  HOSTCC  scripts/kconfig/kxgettext.o
  HOSTCC  scripts/kconfig/mconf.o
  HOSTCC  scripts/kconfig/zconf.tab.o
  HOSTLD  scripts/kconfig/mconf
  HOSTCC  scripts/lxdialog/checklist.o
  HOSTCC  scripts/lxdialog/inputbox.o
  HOSTCC  scripts/lxdialog/lxdialog.o
  HOSTCC  scripts/lxdialog/menubox.o
  HOSTCC  scripts/lxdialog/msgbox.o
  HOSTCC  scripts/lxdialog/textbox.o
  HOSTCC  scripts/lxdialog/util.o
  HOSTCC  scripts/lxdialog/yesno.o
  HOSTLD  scripts/lxdialog/lxdialog
scripts/kconfig/mconf arch/i386/Kconfig
Warning! Found recursive dependency: HOSTAP IEEE80211 NET_RADIO HOSTAP IEEE80211_CRYPT_WEP CRYPTO CRYPTO_ANUBIS
Warning! Found recursive dependency: HOSTAP IEEE80211 NET_RADIO HOSTAP IEEE80211_CRYPT_WEP CRYPTO CRYPTO_MD4
Warning! Found recursive dependency: HOSTAP IEEE80211 NET_RADIO HOSTAP IEEE80211_CRYPT_WEP CRYPTO CRYPTO_MD5
Warning! Found recursive dependency: HOSTAP IEEE80211 NET_RADIO HOSTAP IEEE80211_CRYPT_WEP CRYPTO CRYPTO_AES_X86_64
Warning! Found recursive dependency: NET_RADIO HOSTAP IEEE80211 NET_RADIO HERMES TMD_HERMES
Warning! Found recursive dependency: HOSTAP IEEE80211 NET_RADIO HOSTAP HOSTAP_PCI
Warning! Found recursive dependency: NET_RADIO HOSTAP IEEE80211 NET_RADIO WAVELAN
#
# using defaults found in .config
#
from this: http://bugsplatter.mine.nu/test/boxen/sempro/config-2.6.13-mm2a.gz
and: http://bugsplatter.mine.nu/test/boxen/sempro/config-2.6.13-mm2b.gz
when I tried again, slightly different config.  I don't do wireless networking, 
clueless ;-)

Grant.

Re: 2.6.13-mm2

[Andrew Morton]

Grant Coady <grant_lkml@dodo.com.au> wrote:
>
> On Thu, 8 Sep 2005 05:30:42 -0700, Andrew Morton <akpm@osdl.org> wrote:
> 
> >
> >ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> 
> Hi Andrew,
> 
> After this error:
> 
>   CC      drivers/parport/parport_pc.o
> drivers/parport/parport_pc.c:2511: error: via_686a_data causes a section type conflict
> drivers/parport/parport_pc.c:2520: error: via_8231_data causes a section type conflict
> drivers/parport/parport_pc.c:2705: error: parport_pc_superio_info causes a section type conflict
> drivers/parport/parport_pc.c:2782: error: cards causes a section type conflict
> make[2]: *** [drivers/parport/parport_pc.o] Error 1
> make[1]: *** [drivers/parport] Error 2
> make: *** [drivers] Error 2

Yes, gcc 4.x doesn't like the consts for some reason.


diff -puN drivers/parport/parport_pc.c~a drivers/parport/parport_pc.c
--- devel/drivers/parport/parport_pc.c~a	2005-09-09 02:32:49.000000000 -0700
+++ devel-akpm/drivers/parport/parport_pc.c	2005-09-09 02:33:35.000000000 -0700
@@ -2509,7 +2509,7 @@ static int __devinit sio_ite_8872_probe 
 static int __devinitdata parport_init_mode = 0;
 
 /* Data for two known VIA chips */
-static const struct parport_pc_via_data via_686a_data __devinitdata = {
+static struct parport_pc_via_data via_686a_data __devinitdata = {
 	0x51,
 	0x50,
 	0x85,
@@ -2518,7 +2518,7 @@ static const struct parport_pc_via_data 
 	0xF0,
 	0xE6
 };
-static const struct parport_pc_via_data via_8231_data __devinitdata = {
+static struct parport_pc_via_data via_8231_data __devinitdata = {
 	0x45,
 	0x44,
 	0x50,
@@ -2699,7 +2699,7 @@ enum parport_pc_sio_types {
 };
 
 /* each element directly indexed from enum list, above */
-static const struct parport_pc_superio {
+static struct parport_pc_superio {
 	int (*probe) (struct pci_dev *pdev, int autoirq, int autodma,
 		      const struct parport_pc_via_data *via);
 	const struct parport_pc_via_data *via;
@@ -2763,7 +2763,7 @@ enum parport_pc_pci_cards {
 
 /* each element directly indexed from enum list, above 
  * (but offset by last_sio) */
-static const struct parport_pc_pci {
+static struct parport_pc_pci {
 	int numports;
 	struct { /* BAR (base address registers) numbers in the config
                     space header */
_



> Warning! Found recursive dependency: HOSTAP IEEE80211 NET_RADIO HOSTAP IEEE80211_CRYPT_WEP CRYPTO CRYPTO_ANUBIS
> Warning! Found recursive dependency: HOSTAP IEEE80211 NET_RADIO HOSTAP IEEE80211_CRYPT_WEP CRYPTO CRYPTO_MD4
> Warning! Found recursive dependency: HOSTAP IEEE80211 NET_RADIO HOSTAP IEEE80211_CRYPT_WEP CRYPTO CRYPTO_MD5
> Warning! Found recursive dependency: HOSTAP IEEE80211 NET_RADIO HOSTAP IEEE80211_CRYPT_WEP CRYPTO CRYPTO_AES_X86_64
> Warning! Found recursive dependency: NET_RADIO HOSTAP IEEE80211 NET_RADIO HERMES TMD_HERMES
> Warning! Found recursive dependency: HOSTAP IEEE80211 NET_RADIO HOSTAP HOSTAP_PCI
> Warning! Found recursive dependency: NET_RADIO HOSTAP IEEE80211 NET_RADIO WAVELAN

Yup, Jeff knows about that..

Re: 2.6.13-mm2

[Andrew Morton]

Andi Kleen <ak@suse.de> wrote:
>
>  > arch/x86_64/pci/built-in.o(.init.text+0xa88): In function `pci_acpi_scan_root':
>  > : undefined reference to `pxm_to_node'
>  > make: *** [.tmp_vmlinux1] Error 1
>  > 09/08/05-06:52:31 Build the kernel. Failed rc = 2
>  > 09/08/05-06:52:31 build: kernel build Failed rc = 1
>  > 09/08/05-06:52:31 command complete: (2) rc=126
>  > Failed and terminated the run
> 
>  I tried the config in my (non mm) tree and it compiled just fine.

You must have mucked it up.

>  Must be some bad interaction with another patch in -mm* or a bad 
>  merge.

Nope.

>  The original patch that introduces it is
>  ftp://ftp.firstfloor.org/pub/ak/x86_64/x86_64-2.6.13-1/patches/pci-pxm
> 
>  pxm_to_node for x86-64 is supposed to be declared in arch/x86_64/mm/srat.c

pxm_to_node is *defined* in arch/x86_64/mm/srat.c, which is enabled by
CONFIG_ACPI_NUMA.

pxm_to_node is declared in include/asm-x86_64/numa.h

pxm_to_node is referenced in arch/i386/pci/acpi.c, under CONFIG_NUMA.

Consequently CONFIG_ACPI_NUMA=n, CONFIG_NUMA=y will fail to link.

Also x86 compilation of arch/i386/pci/acpi.c with CONFIG_NUMA=y will
generate an `implicit declaration of function' warning and will fail to
link.



Also, x86_64-srat-overlap-error.patch adds this forward decl in
arch/x86_64/mm/srat.c:

int node_to_pxm(int n);

Please, either give it static scope or, if it really needs global scope (it
doesn't), put the declaration in the right place?

Re: 2.6.13-mm2

[Andi Kleen]

On Friday 09 September 2005 12:41, Andrew Morton wrote:
> Andi Kleen <ak@suse.de> wrote:
> >  > arch/x86_64/pci/built-in.o(.init.text+0xa88): In function 
`pci_acpi_scan_root':
> >  > : undefined reference to `pxm_to_node'
> >  >
> >  > make: *** [.tmp_vmlinux1] Error 1
> >  > 09/08/05-06:52:31 Build the kernel. Failed rc = 2
> >  > 09/08/05-06:52:31 build: kernel build Failed rc = 1
> >  > 09/08/05-06:52:31 command complete: (2) rc=126
> >  > Failed and terminated the run
> >
> >  I tried the config in my (non mm) tree and it compiled just fine.
>
> You must have mucked it up.

Martin put up the wrong .config. I fixed it now in my tree.
The problem was that he didn't enable ACPI_NUMA, just NUMA
and the acpi.c code only checked _NUMA

>
>
> Also, x86_64-srat-overlap-error.patch adds this forward decl in
> arch/x86_64/mm/srat.c:
>
> int node_to_pxm(int n);
>
> Please, either give it static scope or, if it really needs global scope (it
> doesn't), put the declaration in the right place?

I'll make it static.

-Andi

Re: 2.6.13-mm2

[Grant Coady]

Hi Andrew, Marko,
On Fri, 9 Sep 2005 02:43:36 -0700, Andrew Morton <akpm@osdl.org> wrote:

>Grant Coady <grant_lkml@dodo.com.au> wrote:
>>
>> On Thu, 8 Sep 2005 05:30:42 -0700, Andrew Morton <akpm@osdl.org> wrote:
>> 
>> >
>> >ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
>> 
>> Hi Andrew,
>> 
>> After this error:
>> 
>>   CC      drivers/parport/parport_pc.o
>> drivers/parport/parport_pc.c:2511: error: via_686a_data causes a section type conflict
>> drivers/parport/parport_pc.c:2520: error: via_8231_data causes a section type conflict
>> drivers/parport/parport_pc.c:2705: error: parport_pc_superio_info causes a section type conflict
>> drivers/parport/parport_pc.c:2782: error: cards causes a section type conflict
>> make[2]: *** [drivers/parport/parport_pc.o] Error 1
>> make[1]: *** [drivers/parport] Error 2
>> make: *** [drivers] Error 2
>
>Yes, gcc 4.x doesn't like the consts for some reason.

Not using gcc 4.x, Slackware-10.1+ with Gnu C 3.3.6
>
>diff -puN drivers/parport/parport_pc.c~a drivers/parport/parport_pc.c
[...]
Thank you, compile completed :o)  Bonus!  It booted too.

Grant.

Re: 2.6.13-mm2

[Marko Kohtala]

On 9/9/05, Grant Coady <grant_lkml@dodo.com.au> wrote:
> Hi Andrew, Marko,
> On Fri, 9 Sep 2005 02:43:36 -0700, Andrew Morton <akpm@osdl.org> wrote:
> >Grant Coady <grant_lkml@dodo.com.au> wrote:
> >> On Thu, 8 Sep 2005 05:30:42 -0700, Andrew Morton <akpm@osdl.org> wrote:
> >>
> >> >ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> >>
> >> Hi Andrew,
> >>
> >> After this error:
> >>
> >>   CC      drivers/parport/parport_pc.o
> >> drivers/parport/parport_pc.c:2511: error: via_686a_data causes a section type conflict
> >> drivers/parport/parport_pc.c:2520: error: via_8231_data causes a section type conflict
> >> drivers/parport/parport_pc.c:2705: error: parport_pc_superio_info causes a section type conflict
> >> drivers/parport/parport_pc.c:2782: error: cards causes a section type conflict
> >> make[2]: *** [drivers/parport/parport_pc.o] Error 1
> >> make[1]: *** [drivers/parport] Error 2
> >> make: *** [drivers] Error 2
> >
> >Yes, gcc 4.x doesn't like the consts for some reason.

And this was documented in linux/init.h: __devinitdata can not be
const. My mistake. Thanks for fixing it.

Re: 2.6.13-mm2

[Michael Thonke]

Hello Andrew,

>There are changes to both sata_nv and to md in 2.6.13-mm2.  To isolate them
>
>it would be great of you could apply 
>
>ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/broken-out/linus.patch
>
I applied and tested it with 2.6.13 vanilla kernel + linux.patch as suggested, but so far no problem as with 2.6.13-mm2.

I also played with the git-snapshots 2.6.13-git[1-9] no problem here. I think the problem is

somewhere else we have to pay attention, too. MD Raidlevels [0,1] failed to start with 2.6.13-mm2.

Raid0 config:

2x 20GB Partitiontype 0xFD "Linux Raid autodetect"

64k Chunksize, persistent superblock.

little output from mdadm

/dev/md2:

        Version : 00.90.02

  Creation Time : Sun Jun 26 19:14:45 2005

     Raid Level : raid0

     Array Size : 40001536 (38.15 GiB 40.96 GB)

   Raid Devices : 2

  Total Devices : 2

Preferred Minor : 2

    Persistence : Superblock is persistent

    Update Time : Sun Jun 26 19:14:45 2005

          State : clean

 Active Devices : 2

Working Devices : 2

 Failed Devices : 0

  Spare Devices : 0

     Chunk Size : 64K

           UUID : c53fa0d8:9d85875b:efb82dde:11c6617c

         Events : 0.1

    Number   Major   Minor   RaidDevice State

       0       8       23        0      active sync   /dev/sdb7

       1       8        6        1      active sync   /dev/sda6

    

Raid1 config.

2x 15GB Partitiontype 0xFD "Linux Raid autodetect"

        chunksize 128k.


I have no idea where I should look, to resolve this behavior.


>to 2.6.13 and see if the problem still happens.  That will separate out the
>md changes which are still in -mm.
>
>Thanks.
>
>  
>
As for all the time, I'm willing to test to glue the problem out.

Thanks

Best regards

--
Michael Thonke

Re: 2.6.13-mm2

[Manuel Lauss]

Hello,

Andrew Morton wrote:

> +i810fb-add-i2c-ddc-support.patch
> +i810fb-add-i2c-ddc-support-fix.patch
> +i810fb-add-i2c-ddc-support-fix-fix.patch
> +i810fb-add-i2c-ddc-support-Makefile-fix.patch

compiled with CONFIG_FB_I810_I2C = n and CONFIG_FB_I810 = y
it oopses at boot in file drivers/video/i810/i810_main.c:1884

...
Kernel command line: root=/dev/hda7 video=i810fb:xres:1024,yres:768,bpp:8,hsync1:40,hsync2:80,vsync1:60,vsync2:60,extvga,vram:4,accel,mtrr
...
Unable to handle kernel NULL pointer dereference at virtual address 00000054
  printing eip:
c02543c0
*pde = 00000000
Oops: 0000 [#1]
last sysfs file:
Modules linked in:
CPU:    0
EIP:    0060:[<c02543c0>]    Not tainted VLI
EFLAGS: 00010286   (2.6.13-mm2)
EIP is at i810fb_find_init_mode+0x53/0x93
eax: c113ddd4   ebx: c1194000   ecx: c04be2dd   edx: c1194000
esi: c1194008   edi: c113ddd4   ebp: c1194240   esp: c113ddcc
ds: 007b   es: 007b   ss: 0068
Process swapper (pid: 1, threadinfo=c113c000 task=c7cd6a30)
Stack: 00000000 00000008 00000400 00000300 00000000 00001000 00000000 00000000
        00000008 00000000 00000000 00000000 00000000 00000000 00000000 00000000
        00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
Call Trace:
  [<c01ecb01>] fb_alloc_cmap+0x8d/0xa0
  [<c025491b>] i810fb_init_pci+0x10a/0x23c
  [<c026dd9e>] __driver_attach+0x0/0x33
  [<c01e1198>] pci_call_probe+0xa/0xc
  [<c01e11c8>] __pci_device_probe+0x2e/0x3f
  [<c01e11f7>] pci_device_probe+0x1e/0x30
  [<c026dcfb>] driver_probe_device+0x31/0x82
  [<c026ddc1>] __driver_attach+0x23/0x33
  [<c026d60f>] bus_for_each_dev+0x35/0x59
  [<c026dde2>] driver_attach+0x11/0x13
  [<c026dd9e>] __driver_attach+0x0/0x33
  [<c026d993>] bus_add_driver+0x52/0x92
  [<c026e0a2>] driver_register+0x2f/0x34
  [<c01e1394>] pci_register_driver+0x64/0x74
  [<c0254b51>] i810fb_init+0x2f/0x36
  [<c049a676>] do_initcalls+0x49/0x8e
  [<c0100269>] init+0x0/0x107
  [<c010028b>] init+0x22/0x107
  [<c0101281>] kernel_thread_helper+0x5/0xb
Code: 02 00 00 f3 ab 8d 73 08 b9 a0 00 00 00 89 f2 89 e0 89 e7 e8 5f 8a f8 ff 8b 0d 78 8e 4f c0 85 c9 74 1d ff 73 20 89 da 89 f8 6a 00 <ff> 35 54 00 00 00 ff 35 20 00 00 00 e8 37
  <0>Kernel panic - not syncing: Attempted to kill init!


A few debug printks suggest the pointer "specs" is NULL.

with CONFIG_FB_I810_I2C = y it survives boot, but does not work, i.e. on
this laptop there is no EDID and bios table seems borked/missing;
the driver also does no longer honour the commandline parameters
(xres/yres v/hsync1/2) and simply defaults to 640x480.

Built with CONFIG_I810_FB = m, it does _nothing_ when insmod'ded, not even
print the i810fb banner in dmesg; lsmod shows the module is there.

Unapplying the above mentioned patches makes it work again

Thanks,

-- 
  Manuel Lauss

Re: 2.6.13-mm2

[Antonino A. Daplas]

Manuel Lauss wrote:
> Hello,
> 
> Andrew Morton wrote:
> 
>> +i810fb-add-i2c-ddc-support.patch
>> +i810fb-add-i2c-ddc-support-fix.patch
>> +i810fb-add-i2c-ddc-support-fix-fix.patch
>> +i810fb-add-i2c-ddc-support-Makefile-fix.patch
> 
> compiled with CONFIG_FB_I810_I2C = n and CONFIG_FB_I810 = y
> it oopses at boot in file drivers/video/i810/i810_main.c:1884
> 
> ...
> Kernel command line: root=/dev/hda7
> video=i810fb:xres:1024,yres:768,bpp:8,hsync1:40,hsync2:80,vsync1:60,vsync2:60,extvga,vram:4,accel,mtrr
> 
> ...
> Unable to handle kernel NULL pointer dereference at virtual address
> 00000054
>  printing eip:
> c02543c0
> *pde = 00000000
> Oops: 0000 [#1]
> last sysfs file:
> Modules linked in:
> CPU:    0
> EIP:    0060:[<c02543c0>]    Not tainted VLI
> EFLAGS: 00010286   (2.6.13-mm2)
> EIP is at i810fb_find_init_mode+0x53/0x93
> eax: c113ddd4   ebx: c1194000   ecx: c04be2dd   edx: c1194000
> esi: c1194008   edi: c113ddd4   ebp: c1194240   esp: c113ddcc
> ds: 007b   es: 007b   ss: 0068
> Process swapper (pid: 1, threadinfo=c113c000 task=c7cd6a30)
> Stack: 00000000 00000008 00000400 00000300 00000000 00001000 00000000
> 00000000
>        00000008 00000000 00000000 00000000 00000000 00000000 00000000
> 00000000
>        00000000 00000000 00000000 00000000 00000000 00000000 00000000
> 00000000
> Call Trace:
>  [<c01ecb01>] fb_alloc_cmap+0x8d/0xa0
>  [<c025491b>] i810fb_init_pci+0x10a/0x23c
>  [<c026dd9e>] __driver_attach+0x0/0x33
>  [<c01e1198>] pci_call_probe+0xa/0xc
>  [<c01e11c8>] __pci_device_probe+0x2e/0x3f
>  [<c01e11f7>] pci_device_probe+0x1e/0x30
>  [<c026dcfb>] driver_probe_device+0x31/0x82
>  [<c026ddc1>] __driver_attach+0x23/0x33
>  [<c026d60f>] bus_for_each_dev+0x35/0x59
>  [<c026dde2>] driver_attach+0x11/0x13
>  [<c026dd9e>] __driver_attach+0x0/0x33
>  [<c026d993>] bus_add_driver+0x52/0x92
>  [<c026e0a2>] driver_register+0x2f/0x34
>  [<c01e1394>] pci_register_driver+0x64/0x74
>  [<c0254b51>] i810fb_init+0x2f/0x36
>  [<c049a676>] do_initcalls+0x49/0x8e
>  [<c0100269>] init+0x0/0x107
>  [<c010028b>] init+0x22/0x107
>  [<c0101281>] kernel_thread_helper+0x5/0xb
> Code: 02 00 00 f3 ab 8d 73 08 b9 a0 00 00 00 89 f2 89 e0 89 e7 e8 5f 8a
> f8 ff 8b 0d 78 8e 4f c0 85 c9 74 1d ff 73 20 89 da 89 f8 6a 00 <ff> 35
> 54 00 00 00 ff 35 20 00 00 00 e8 37
>  <0>Kernel panic - not syncing: Attempted to kill init!
> 
> 
> A few debug printks suggest the pointer "specs" is NULL.
> 

Ah, yes, sorry about that.  Can you try this patch?

Fix kernel oops when CONFIG_FB_I810_I2C is set to 'n'.

Signed-off-by: Antonino Daplas <adaplas@pol.net>

diff --git a/drivers/video/i810/i810_main.c b/drivers/video/i810/i810_main.c
--- a/drivers/video/i810/i810_main.c
+++ b/drivers/video/i810/i810_main.c
@@ -1830,7 +1830,7 @@ static void __devinit i810fb_find_init_m
 {
 	struct fb_videomode mode;
 	struct fb_var_screeninfo var;
-	struct fb_monspecs *specs = NULL;
+	struct fb_monspecs *specs = &info->monspecs;
 	int found = 0;
 #ifdef CONFIG_FB_I810_I2C
 	int i;
@@ -1853,12 +1853,11 @@ static void __devinit i810fb_find_init_m
 	if (!err)
 		printk("i810fb_init_pci: DDC probe successful\n");
 
-	fb_edid_to_monspecs(par->edid, &info->monspecs);
+	fb_edid_to_monspecs(par->edid, specs);
 
-	if (info->monspecs.modedb == NULL)
+	if (specs->modedb == NULL)
 		printk("i810fb_init_pci: Unable to get Mode Database\n");
 
-	specs = &info->monspecs;
 	fb_videomode_to_modelist(specs->modedb, specs->modedb_len,
 				 &info->modelist);
 	if (specs->modedb != NULL) {

Re: 2.6.13-mm2

[Manuel Lauss]

> Ah, yes, sorry about that.  Can you try this patch?
> 
> Fix kernel oops when CONFIG_FB_I810_I2C is set to 'n'.
> 
> Signed-off-by: Antonino Daplas <adaplas@pol.net>
> 
> diff --git a/drivers/video/i810/i810_main.c b/drivers/video/i810/i810_main.c
> --- a/drivers/video/i810/i810_main.c
> +++ b/drivers/video/i810/i810_main.c
> @@ -1830,7 +1830,7 @@ static void __devinit i810fb_find_init_m
>  {
>  	struct fb_videomode mode;
>  	struct fb_var_screeninfo var;
> -	struct fb_monspecs *specs = NULL;
> +	struct fb_monspecs *specs = &info->monspecs;
>  	int found = 0;
>  #ifdef CONFIG_FB_I810_I2C
>  	int i;
> @@ -1853,12 +1853,11 @@ static void __devinit i810fb_find_init_m
>  	if (!err)
>  		printk("i810fb_init_pci: DDC probe successful\n");
>  
> -	fb_edid_to_monspecs(par->edid, &info->monspecs);
> +	fb_edid_to_monspecs(par->edid, specs);
>  
> -	if (info->monspecs.modedb == NULL)
> +	if (specs->modedb == NULL)
>  		printk("i810fb_init_pci: Unable to get Mode Database\n");
>  
> -	specs = &info->monspecs;
>  	fb_videomode_to_modelist(specs->modedb, specs->modedb_len,
>  				 &info->modelist);
>  	if (specs->modedb != NULL) {

Thanks, it boots now, but doesnt set video mode. Spews
a bunch of "i810fb: invalid video mode" lines and defaults
to 640x480.

With I2C enabled:
i810-i2c: Probe DDC1 Bus
i810-i2c: Unable to read EDID block.
i810-i2c: Unable to read EDID block.
i810-i2c: Unable to read EDID block.
i810-i2c: Probe DDC2 Bus
i810-i2c: Unable to read EDID block.
i810-i2c: Unable to read EDID block.
i810-i2c: Unable to read EDID block.
i810-i2c: Probe DDC3 Bus
i810-i2c: Getting EDID from BIOS
i810fb_init_pci: DDC probe successful
i810fb_init_pci: Unable to get Mode Database
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
i810fb: invalid video mode
ringbuffer lockup!!!
IIR     : 0x0000
EIR     : 0x0000
PGTBL_ER: 0x0000
IPEIR   : 0x0001
IPEHR   : 0x0000
Console: switching to colour frame buffer device 80x25
I810FB: fb0         : Intel(R) 815 (Internal Graphics with AGP) Framebuffer Device v0.9.0
I810FB: Video RAM   : 4096K
I810FB: Monitor     : H: 40-80 KHz V: 60-60 Hz
I810FB: Mode        : 640x400-8bpp@69Hz

-- 
  Manuel Lauss

Re: 2.6.13-mm2

[Dominik Karall]

[-- Attachment #1: Type: text/plain, Size: 1039 bytes --]

On Thursday 08 September 2005 14:30, Andrew Morton wrote:
> ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13
>-mm2/

I have problems using NFS with 2.6.13-mm2, it failes to start, but works with 
2.6.13-ck1 (so pure 2.6.13 should work too, as there are no nfs related 
changes in -ck, I think).
Following messages appear in /var/log/messages:

Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
rpc.statd[15041]: Version 1.0.7 Starting
NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery directory
NFSD: recovery directory /var/lib/nfs/v4recovery doesn't exist
NFSD: starting 90-second grace period
portmap[15048]: connect from 127.0.0.1 to set(nfs): request from unprivileged 
port
nfsd[15046]: nfssvc: Permission denied


with 2.6.13-ck1:

Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
rpc.statd[16126]: Version 1.0.7 Starting
NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery directory
NFSD: starting 90-second grace period

hth,
dominik

[-- Attachment #2: Type: application/pgp-signature, Size: 316 bytes --]

Re: 2.6.13-mm2

[Antonino A. Daplas]

Manuel Lauss wrote:
> 
>> Ah, yes, sorry about that.  Can you try this patch?
>>
>> Fix kernel oops when CONFIG_FB_I810_I2C is set to 'n'.
>>
>> Signed-off-by: Antonino Daplas <adaplas@pol.net>
>>
>> diff --git a/drivers/video/i810/i810_main.c
>> b/drivers/video/i810/i810_main.c
>> --- a/drivers/video/i810/i810_main.c
>> +++ b/drivers/video/i810/i810_main.c
>> @@ -1830,7 +1830,7 @@ static void __devinit i810fb_find_init_m
>>  {
>>      struct fb_videomode mode;
>>      struct fb_var_screeninfo var;
>> -    struct fb_monspecs *specs = NULL;
>> +    struct fb_monspecs *specs = &info->monspecs;
>>      int found = 0;
>>  #ifdef CONFIG_FB_I810_I2C
>>      int i;
>> @@ -1853,12 +1853,11 @@ static void __devinit i810fb_find_init_m
>>      if (!err)
>>          printk("i810fb_init_pci: DDC probe successful\n");
>>  
>> -    fb_edid_to_monspecs(par->edid, &info->monspecs);
>> +    fb_edid_to_monspecs(par->edid, specs);
>>  
>> -    if (info->monspecs.modedb == NULL)
>> +    if (specs->modedb == NULL)
>>          printk("i810fb_init_pci: Unable to get Mode Database\n");
>>  
>> -    specs = &info->monspecs;
>>      fb_videomode_to_modelist(specs->modedb, specs->modedb_len,
>>                   &info->modelist);
>>      if (specs->modedb != NULL) {
> 
> Thanks, it boots now, but doesnt set video mode. Spews
> a bunch of "i810fb: invalid video mode" lines and defaults
> to 640x480.

Set CONFIG_FB_I810_I2C to n.  Your display does not have an EDID block.
Even getting the EDID from the BIOS failed.

Tony

Re: 2.6.13-mm2

[Antonino A. Daplas]

Manuel Lauss wrote:
> Console: switching to colour frame buffer device 80x25
> I810FB: fb0         : Intel(R) 815 (Internal Graphics with AGP)
> Framebuffer Device v0.9.0
> I810FB: Video RAM   : 4096K
> I810FB: Monitor     : H: 40-80 KHz V: 60-60 Hz
> I810FB: Mode        : 640x400-8bpp@69Hz
> 

One more thing, vfmin and vfmax are set to 60 and 60.  This does
not give a lot of room for the calculation.  It's possible that
the GTF might calculate for 59.9 Hz, but since calculation are done
in integers it returns as 59.

So, give a little room for vfmin and vfmax, such as 59 and 60.

Tony

Re: 2.6.13-mm2

[Andrew Morton]

Dominik Karall <dominik.karall@gmx.net> wrote:
>
> On Thursday 08 September 2005 14:30, Andrew Morton wrote:
>  > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13
>  >-mm2/
> 
>  I have problems using NFS with 2.6.13-mm2, it failes to start, but works with 
>  2.6.13-ck1 (so pure 2.6.13 should work too, as there are no nfs related 
>  changes in -ck, I think).
>  Following messages appear in /var/log/messages:
> 
>  Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
>  rpc.statd[15041]: Version 1.0.7 Starting
>  NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery directory
>  NFSD: recovery directory /var/lib/nfs/v4recovery doesn't exist
>  NFSD: starting 90-second grace period
>  portmap[15048]: connect from 127.0.0.1 to set(nfs): request from unprivileged 
>  port
>  nfsd[15046]: nfssvc: Permission denied
> 
> 
>  with 2.6.13-ck1:
> 
>  Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
>  rpc.statd[16126]: Version 1.0.7 Starting
>  NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery directory
>  NFSD: starting 90-second grace period

OK.  We don't have many nfsd patches at all in 2.6.13-mm2.  But there are
quite a few sunrpc changes.  Plus I have a few random nfs patches which
should be merged up or dropped.

In short: dunno.  Relevant people cc'ed ;)

Re: 2.6.13-mm2

[J.A. Magallon]

On 09.08, Andrew Morton wrote:
> 
> ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> 
> (kernel.org propagation is slow.  There's a temp copy at
> http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)
> 
> 

I can not ifup an interface while iptables is using it.
Is this expected behaviour ?
There is a possible bug (IMHO) in Mandrake initscripts, that start iptables
before network interfaces, but this had always worked.

Any ideas ?

--
J.A. Magallon <jamagallon()able!es>     \               Software is like sex:
werewolf!able!es                         \         It's better when it's free
Mandriva Linux release 2006.0 (Cooker) for i586
Linux 2.6.13-jam3 (gcc 4.0.1 (4.0.1-5mdk for Mandriva Linux release 2006.0))

Re: 2.6.13-mm2

[Andrew Morton]

"J.A. Magallon" <jamagallon@able.es> wrote:
>
> 
> On 09.08, Andrew Morton wrote:
> > 
> > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> > 
> > (kernel.org propagation is slow.  There's a temp copy at
> > http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)
> > 
> > 
> 
> I can not ifup an interface while iptables is using it.
> Is this expected behaviour ?

Maybe it's expected, but breaking existing userspace is a serious issue.

> There is a possible bug (IMHO) in Mandrake initscripts, that start iptables
> before network interfaces, but this had always worked.
> 
> Any ideas ?

Please always cc netdev@vger.kernel.org on networking matters.

Re: 2.6.13-mm2

[Patrick McHardy]

Andrew Morton wrote:
> "J.A. Magallon" <jamagallon@able.es> wrote:
> 
>>I can not ifup an interface while iptables is using it.
>>Is this expected behaviour ?
> 
> Maybe it's expected, but breaking existing userspace is a serious issue.

No, its not expected.

>>There is a possible bug (IMHO) in Mandrake initscripts, that start iptables
>>before network interfaces, but this had always worked.
>>
>>Any ideas ?

What's happening when you try to set the interface up? Please
provide output of ifup and strace of the failing command. Thanks.

Re: 2.6.13-mm2

[J.A. Magallon]

On 09.11, Patrick McHardy wrote:
> Andrew Morton wrote:
> > "J.A. Magallon" <jamagallon@able.es> wrote:
> > 
> >>I can not ifup an interface while iptables is using it.
> >>Is this expected behaviour ?
> > 
> > Maybe it's expected, but breaking existing userspace is a serious issue.
> 
> No, its not expected.
> 
> >>There is a possible bug (IMHO) in Mandrake initscripts, that start iptables
> >>before network interfaces, but this had always worked.
> >>
> >>Any ideas ?
> 
> What's happening when you try to set the interface up? Please
> provide output of ifup and strace of the failing command. Thanks.

werewolf:~# ifdown eth0
werewolf:~# service iptables start
Applying iptables firewall rules: 
                                                                [  OK  ]
werewolf:~# iptables -v -t nat -L
Chain PREROUTING (policy ACCEPT 2 packets, 156 bytes)
 pkts bytes target     prot opt in     out     source               destination         

Chain POSTROUTING (policy ACCEPT 5 packets, 300 bytes)
 pkts bytes target     prot opt in     out     source               destination         
    0     0 MASQUERADE  all  --  any    eth0    anywhere             anywhere            

Chain OUTPUT (policy ACCEPT 5 packets, 300 bytes)
 pkts bytes target     prot opt in     out     source               destination         
werewolf:~# iptables -v -t filter -L
Chain INPUT (policy ACCEPT 257 packets, 51631 bytes)
 pkts bytes target     prot opt in     out     source               destination         

Chain FORWARD (policy DROP 0 packets, 0 bytes)
 pkts bytes target     prot opt in     out     source               destination         
    0     0 ACCEPT     all  --  eth0   eth1    anywhere             anywhere            state RELATED,ESTABLISHED 
    0     0 ACCEPT     all  --  eth1   eth0    anywhere             anywhere            

Chain OUTPUT (policy ACCEPT 251 packets, 51163 bytes)
 pkts bytes target     prot opt in     out     source               destination         

werewolf:~# ifup eth0

Determining IP information for eth0...Operation failed.
 failed.

I traced the problem to pump, and I did a diff between strace of pump
when it works and when it doesnt (witout and with iptables started):

 socket(PF_FILE, SOCK_STREAM, 0)         = 3
 connect(3, {sa_family=AF_FILE, path="/var/run/pump.sock"}, 20) = 0
 write(3, "\0\0\0\0eth0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\300"..., 4280) = 4280
-read(3, "\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4280) = 4280
-exit_group(0)                           = ?
-Process 7931 detached
+read(3, "\1\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4280) = 4280
+socket(PF_FILE, SOCK_STREAM, 0)         = 4
+connect(4, {sa_family=AF_FILE, path="/var/run/pump.sock"}, 20) = 0
+write(4, "\0\0\0\0eth0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\300"..., 4280) = 4280
+read(4, "\1\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4280) = 4280
+write(2, "Operation failed.\n", 18Operation failed.
+)     = 18
+exit_group(1)                           = ?
+Process 7822 detached

pump seems to write something in the socket, try to read it again and gets
different results.

Note, my iptables are modular and I did not unload the modules, just stopped
them with 'service iptables stop'. Digging further, if I just do
iptables -t nat -F, pump works again.

Hope this helps.

--
J.A. Magallon <jamagallon()able!es>     \               Software is like sex:
werewolf!able!es                         \         It's better when it's free
Mandriva Linux release 2006.0 (Cooker) for i586
Linux 2.6.13-jam3 (gcc 4.0.1 (4.0.1-5mdk for Mandriva Linux release 2006.0))

Re: 2.6.13-mm2

[J.A. Magallon]

On 09.11, J.A. Magallon wrote:
> 
> On 09.11, Patrick McHardy wrote:
> > Andrew Morton wrote:
> > > "J.A. Magallon" <jamagallon@able.es> wrote:
> > > 
> > >>I can not ifup an interface while iptables is using it.
> > >>Is this expected behaviour ?
> > > 
> > > Maybe it's expected, but breaking existing userspace is a serious issue.
> > 
> > No, its not expected.
> > 
> > >>There is a possible bug (IMHO) in Mandrake initscripts, that start iptables
> > >>before network interfaces, but this had always worked.
> > >>
> > >>Any ideas ?
> > 
> > What's happening when you try to set the interface up? Please
> > provide output of ifup and strace of the failing command. Thanks.
> 
> werewolf:~# ifdown eth0
> werewolf:~# service iptables start
> Applying iptables firewall rules: 
>                                                                 [  OK  ]
> werewolf:~# iptables -v -t nat -L
> Chain PREROUTING (policy ACCEPT 2 packets, 156 bytes)
>  pkts bytes target     prot opt in     out     source               destination         
> 
> Chain POSTROUTING (policy ACCEPT 5 packets, 300 bytes)
>  pkts bytes target     prot opt in     out     source               destination         
>     0     0 MASQUERADE  all  --  any    eth0    anywhere             anywhere            
> 
> Chain OUTPUT (policy ACCEPT 5 packets, 300 bytes)
>  pkts bytes target     prot opt in     out     source               destination         
> werewolf:~# iptables -v -t filter -L
> Chain INPUT (policy ACCEPT 257 packets, 51631 bytes)
>  pkts bytes target     prot opt in     out     source               destination         
> 
> Chain FORWARD (policy DROP 0 packets, 0 bytes)
>  pkts bytes target     prot opt in     out     source               destination         
>     0     0 ACCEPT     all  --  eth0   eth1    anywhere             anywhere            state RELATED,ESTABLISHED 
>     0     0 ACCEPT     all  --  eth1   eth0    anywhere             anywhere            
> 
> Chain OUTPUT (policy ACCEPT 251 packets, 51163 bytes)
>  pkts bytes target     prot opt in     out     source               destination         
> 
> werewolf:~# ifup eth0
> 
> Determining IP information for eth0...Operation failed.
>  failed.
> 
> I traced the problem to pump, and I did a diff between strace of pump
> when it works and when it doesnt (witout and with iptables started):
> 
>  socket(PF_FILE, SOCK_STREAM, 0)         = 3
>  connect(3, {sa_family=AF_FILE, path="/var/run/pump.sock"}, 20) = 0
>  write(3, "\0\0\0\0eth0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\300"..., 4280) = 4280
> -read(3, "\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4280) = 4280
> -exit_group(0)                           = ?
> -Process 7931 detached
> +read(3, "\1\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4280) = 4280
> +socket(PF_FILE, SOCK_STREAM, 0)         = 4
> +connect(4, {sa_family=AF_FILE, path="/var/run/pump.sock"}, 20) = 0
> +write(4, "\0\0\0\0eth0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\300"..., 4280) = 4280
> +read(4, "\1\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 4280) = 4280
> +write(2, "Operation failed.\n", 18Operation failed.
> +)     = 18
> +exit_group(1)                           = ?
> +Process 7822 detached
> 
> pump seems to write something in the socket, try to read it again and gets
> different results.
> 
> Note, my iptables are modular and I did not unload the modules, just stopped
> them with 'service iptables stop'. Digging further, if I just do
> iptables -t nat -F, pump works again.
> 
> Hope this helps.
> 

And I also get this on syslog:

Sep 11 02:56:58 werewolf kernel: MASQUERADE: eth0 ate my IP address
Sep 11 02:56:58 werewolf kernel: MASQUERADE: eth0 ate my IP address


--
J.A. Magallon <jamagallon()able!es>     \               Software is like sex:
werewolf!able!es                         \         It's better when it's free
Mandriva Linux release 2006.0 (Cooker) for i586
Linux 2.6.13-jam3 (gcc 4.0.1 (4.0.1-5mdk for Mandriva Linux release 2006.0))

Re: 2.6.13-mm2

[Patrick McHardy]

[-- Attachment #1: Type: text/plain, Size: 547 bytes --]

J.A. Magallon wrote:
> And I also get this on syslog:
> 
> Sep 11 02:56:58 werewolf kernel: MASQUERADE: eth0 ate my IP address
> Sep 11 02:56:58 werewolf kernel: MASQUERADE: eth0 ate my IP address

Thanks, I'm pretty sure its caused by this patch. The problem is that
pump uses a regular UDP socket (some other dhcp clients use AF_PACKET
sockets), and packet sent by it are also handled by iptables. The
MASQUERADE rule can't find a local IP address and drops the packet.
I'm not sure how to fix it yet, reverting the patch is not a good
option.


[-- Attachment #2: x --]
[-- Type: text/plain, Size: 1219 bytes --]

[NETFILTER]: Don't exclude local packets from MASQUERADING

Increases consistency in source-address selection.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

---
commit 9baa5c67ff4ce57b6b9f68c90714a1bb876fccd7
tree 27f2c48e12e1bb5e3e6d5f8320651c213892ed20
parent fb13ab2849074244a51ae5147483610529a29ced
author Patrick McHardy <kaber@trash.net> Sun, 14 Aug 2005 17:32:50 -0700
committer David S. Miller <davem@sunset.davemloft.net> Mon, 29 Aug 2005 15:58:36 -0700

 net/ipv4/netfilter/ipt_MASQUERADE.c |    5 -----
 1 files changed, 0 insertions(+), 5 deletions(-)

diff --git a/net/ipv4/netfilter/ipt_MASQUERADE.c b/net/ipv4/netfilter/ipt_MASQUERADE.c
--- a/net/ipv4/netfilter/ipt_MASQUERADE.c
+++ b/net/ipv4/netfilter/ipt_MASQUERADE.c
@@ -86,11 +86,6 @@ masquerade_target(struct sk_buff **pskb,
 
 	IP_NF_ASSERT(hooknum == NF_IP_POST_ROUTING);
 
-	/* FIXME: For the moment, don't do local packets, breaks
-	   testsuite for 2.3.49 --RR */
-	if ((*pskb)->sk)
-		return NF_ACCEPT;
-
 	ct = ip_conntrack_get(*pskb, &ctinfo);
 	IP_NF_ASSERT(ct && (ctinfo == IP_CT_NEW || ctinfo == IP_CT_RELATED
 	                    || ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY));

Re: 2.6.13-mm2

[J.A. Magallon]

On 09.11, Patrick McHardy wrote:
> J.A. Magallon wrote:
> > And I also get this on syslog:
> > 
> > Sep 11 02:56:58 werewolf kernel: MASQUERADE: eth0 ate my IP address
> > Sep 11 02:56:58 werewolf kernel: MASQUERADE: eth0 ate my IP address
> 
> Thanks, I'm pretty sure its caused by this patch. The problem is that
> pump uses a regular UDP socket (some other dhcp clients use AF_PACKET
> sockets), and packet sent by it are also handled by iptables. The
> MASQUERADE rule can't find a local IP address and drops the packet.
> I'm not sure how to fix it yet, reverting the patch is not a good
> option.
> 
> 

> [NETFILTER]: Don't exclude local packets from MASQUERADING
> 
> Increases consistency in source-address selection.
> 
> Signed-off-by: Patrick McHardy <kaber@trash.net>
> Signed-off-by: David S. Miller <davem@davemloft.net>
> 
> ---
> commit 9baa5c67ff4ce57b6b9f68c90714a1bb876fccd7
> tree 27f2c48e12e1bb5e3e6d5f8320651c213892ed20
> parent fb13ab2849074244a51ae5147483610529a29ced
> author Patrick McHardy <kaber@trash.net> Sun, 14 Aug 2005 17:32:50 -0700
> committer David S. Miller <davem@sunset.davemloft.net> Mon, 29 Aug 2005 15:58:36 -0700
> 
>  net/ipv4/netfilter/ipt_MASQUERADE.c |    5 -----
>  1 files changed, 0 insertions(+), 5 deletions(-)
> 
> diff --git a/net/ipv4/netfilter/ipt_MASQUERADE.c b/net/ipv4/netfilter/ipt_MASQUERADE.c
> --- a/net/ipv4/netfilter/ipt_MASQUERADE.c
> +++ b/net/ipv4/netfilter/ipt_MASQUERADE.c
> @@ -86,11 +86,6 @@ masquerade_target(struct sk_buff **pskb,
>  
>  	IP_NF_ASSERT(hooknum == NF_IP_POST_ROUTING);
>  
> -	/* FIXME: For the moment, don't do local packets, breaks
> -	   testsuite for 2.3.49 --RR */
> -	if ((*pskb)->sk)
> -		return NF_ACCEPT;
> -
>  	ct = ip_conntrack_get(*pskb, &ctinfo);
>  	IP_NF_ASSERT(ct && (ctinfo == IP_CT_NEW || ctinfo == IP_CT_RELATED
>  	                    || ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY));
> 

Thanks, reverting this made things work again.

Are you confident in fixing this shortly, or should I just drop pump ?

--
J.A. Magallon <jamagallon()able!es>     \               Software is like sex:
werewolf!able!es                         \         It's better when it's free
Mandriva Linux release 2006.0 (Cooker) for i586
Linux 2.6.13-jam3 (gcc 4.0.1 (4.0.1-5mdk for Mandriva Linux release 2006.0))

Re: 2.6.13-mm2

[Patrick McHardy]

J.A. Magallon wrote:
> On 09.11, Patrick McHardy wrote:
> 
>>[NETFILTER]: Don't exclude local packets from MASQUERADING
>>
> Thanks, reverting this made things work again.
> 
> Are you confident in fixing this shortly, or should I just drop pump ?

I should have a fix within the next couple of days.

Re: 2.6.13-mm2

[Rafael J. Wysocki]

[-- Attachment #1: Type: text/plain, Size: 607 bytes --]

On Thursday, 8 of September 2005 14:30, Andrew Morton wrote:
> 
> ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> 
> (kernel.org propagation is slow.  There's a temp copy at
> http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)

Could you please reintroduce the yenta-free_irq-on-suspend.patch (attached)
into -mm?  My box does not resume from disk without it.

Greetings,
Rafael


-- 
- Would you tell me, please, which way I ought to go from here?
- That depends a good deal on where you want to get to.
		-- Lewis Carroll "Alice's Adventures in Wonderland"

[-- Attachment #2: yenta-free_irq-on-suspend.patch --]
[-- Type: text/x-diff, Size: 1362 bytes --]


From: Daniel Ritz <daniel.ritz@gmx.ch>

Resume doesn't seem to work without.

Signed-off-by: Daniel Ritz <daniel.ritz@gmx.ch>
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
---

 drivers/pcmcia/yenta_socket.c |    9 +++++++++
 1 files changed, 9 insertions(+)

diff -puN drivers/pcmcia/yenta_socket.c~yenta-free_irq-on-suspend drivers/pcmcia/yenta_socket.c
--- devel/drivers/pcmcia/yenta_socket.c~yenta-free_irq-on-suspend	2005-07-28 01:05:52.000000000 -0700
+++ devel-akpm/drivers/pcmcia/yenta_socket.c	2005-07-28 01:05:52.000000000 -0700
@@ -1107,6 +1107,8 @@ static int yenta_dev_suspend (struct pci
 		pci_read_config_dword(dev, 17*4, &socket->saved_state[1]);
 		pci_disable_device(dev);
 
+		free_irq(dev->irq, socket);
+
 		/*
 		 * Some laptops (IBM T22) do not like us putting the Cardbus
 		 * bridge into D3.  At a guess, some other laptop will
@@ -1132,6 +1134,13 @@ static int yenta_dev_resume (struct pci_
 		pci_enable_device(dev);
 		pci_set_master(dev);
 
+		if (socket->cb_irq)
+			if (request_irq(socket->cb_irq, yenta_interrupt,
+			                SA_SHIRQ, "yenta", socket)) {
+				printk(KERN_WARNING "Yenta: request_irq() failed on resume!\n");
+				socket->cb_irq = 0;
+			}
+
 		if (socket->type && socket->type->restore_state)
 			socket->type->restore_state(socket);
 	}
_

Re: 2.6.13-mm2

[Andrew Morton]

"Rafael J. Wysocki" <rjw@sisk.pl> wrote:
>
> > 
>  > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
>  > 
>  > (kernel.org propagation is slow.  There's a temp copy at
>  > http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)
> 
>  Could you please reintroduce the yenta-free_irq-on-suspend.patch (attached)
>  into -mm?  My box does not resume from disk without it.

No probs.

Daniel, do you remember why we decided to drop it?  What should we do about
this?  Thanks.

Re: 2.6.13-mm2

[Hugh Dickins]

On Sun, 11 Sep 2005, Andrew Morton wrote:
> "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
> > 
> >  Could you please reintroduce the yenta-free_irq-on-suspend.patch (attached)
> >  into -mm?  My box does not resume from disk without it.
> 
> No probs.
> 
> Daniel, do you remember why we decided to drop it?  What should we do about
> this?  Thanks.

I remember well.  My laptop does not APM resume from RAM with it.
I've just rechecked and that's still the case.  I did try various patches
from Rafael to help him work it out, but it remained a puzzle.
And I admit, it is his turn to resume this month.

Hugh

Re: 2.6.13-mm2

[Daniel Ritz]

On Sunday 11 September 2005 21.36, Andrew Morton wrote:
> "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
> >
> > > 
> >  > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> >  > 
> >  > (kernel.org propagation is slow.  There's a temp copy at
> >  > http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)
> > 
> >  Could you please reintroduce the yenta-free_irq-on-suspend.patch (attached)
> >  into -mm?  My box does not resume from disk without it.
> 
> No probs.
> 
> Daniel, do you remember why we decided to drop it?  What should we do about
> this?  Thanks.
> 

yeah, there was a long discussion about it. see:
	http://marc.theaimsgroup.com/?t=112275164900002&r=1&w=4
the reason being that it breaks APM suspend on Hugh Dickins' (added to cc:) laptop.
Linus was quite clear about why reverting...
	http://marc.theaimsgroup.com/?l=linux-kernel&m=112278810115252&w=4

we should look at both problems in detail:
- with APM it seems to break because the bridge gives interrupt before the
  handler is installed.
- with ACPI i think some _other_ device gives the interrupts too early. but
  when all devices on the interrupt unregister the irq is disabled and the
  problem is hidden.

i don't think we can do mutch about the APM case...

so Rafael, your /proc/interrupts, lspci -vvv and dmesg, please.

rgds
-daniel

Re: 2.6.13-mm2

[Martin J. Bligh]

Finally got my damned x440 box back - won't build -mm2 (-mm1 is fine)

arch/i386/kernel/srat.c:141: #error "MAX_NR_ZONES != 3, chunk_to_zone requires review"
make[1]: *** [arch/i386/kernel/srat.o] Error 1
make: *** [arch/i386/kernel] Error 2
09/11/05-00:57:13 Build the kernel. Failed rc = 2
09/11/05-00:57:13 build: kernel build Failed rc = 1
09/11/05-00:57:13 command complete: (2) rc=126

x86_64-dma32.patch:-#define MAX_NR_ZONES                3       /* Sync this wi
h ZONES_SHIFT */
x86_64-dma32.patch:-#define ZONES_SHIFT         2       /* ceil(log2(MAX_NR_ZON
S)) */
x86_64-dma32.patch:+#define MAX_NR_ZONES                4       /* Sync this wi
h ZONES_SHIFT */
x86_64-dma32.patch:+#define ZONES_SHIFT         3       /* ceil(log2(MAX_NR_ZON
S)) */

Andi, does that need changing on ia32 as well as x86_64, or are you
just missing some ifdefs? Looks to me like the rest of the patch is
specific to x86_64.

M.

Re: 2.6.13-mm2

[Martin J. Bligh]

Looks like the non-fixed version of Ingo's patch is back.

http://test.kernel.org/12554/debug/console.log

Hangs after the cpu cache auto-detect thing again on NUMA-Q.
I thought Dave Wilder debugged and fixed this? Something to do with
restoring the flags on a different CPU to that on which they were saved.

Re: 2.6.13-mm2

[Andi Kleen]

On Sun, Sep 11, 2005 at 08:07:25PM -0700, Martin J. Bligh wrote:
> Finally got my damned x440 box back - won't build -mm2 (-mm1 is fine)
> 
> arch/i386/kernel/srat.c:141: #error "MAX_NR_ZONES != 3, chunk_to_zone requires review"
> make[1]: *** [arch/i386/kernel/srat.o] Error 1
> make: *** [arch/i386/kernel] Error 2
> 09/11/05-00:57:13 Build the kernel. Failed rc = 2
> 09/11/05-00:57:13 build: kernel build Failed rc = 1
> 09/11/05-00:57:13 command complete: (2) rc=126
> 
> x86_64-dma32.patch:-#define MAX_NR_ZONES                3       /* Sync this wi
> h ZONES_SHIFT */
> x86_64-dma32.patch:-#define ZONES_SHIFT         2       /* ceil(log2(MAX_NR_ZON
> S)) */
> x86_64-dma32.patch:+#define MAX_NR_ZONES                4       /* Sync this wi
> h ZONES_SHIFT */
> x86_64-dma32.patch:+#define ZONES_SHIFT         3       /* ceil(log2(MAX_NR_ZON
> S)) */
> 
> Andi, does that need changing on ia32 as well as x86_64, or are you
> just missing some ifdefs? Looks to me like the rest of the patch is
> specific to x86_64.

It should be a straight forward fix - the new zone is empty on i386.
Ok I reviewed chunk_to_zone and it should be ok with the new empty
zone. So just the appended patch should work. Can you test?

-AndI

Make i386 compile again with fourth DMA32 zone

The code should deal with an additiona empty zone, so fix up the
#error.

Signed-off-by: Andi Kleen <ak@suse.de>

Index: linux/arch/i386/kernel/srat.c
===================================================================
--- linux.orig/arch/i386/kernel/srat.c
+++ linux/arch/i386/kernel/srat.c
@@ -137,8 +137,8 @@ static void __init parse_memory_affinity
 		 "enabled and removable" : "enabled" ) );
 }
 
-#if MAX_NR_ZONES != 3
-#error "MAX_NR_ZONES != 3, chunk_to_zone requires review"
+#if MAX_NR_ZONES != 4
+#error "MAX_NR_ZONES != 4, chunk_to_zone requires review"
 #endif
 /* Take a chunk of pages from page frame cstart to cend and count the number
  * of pages in each zone, returned via zones[].

Re: 2.6.13-mm2

[Martin J. Bligh]

>> Andi, does that need changing on ia32 as well as x86_64, or are you
>> just missing some ifdefs? Looks to me like the rest of the patch is
>> specific to x86_64.
> 
> It should be a straight forward fix - the new zone is empty on i386.
> Ok I reviewed chunk_to_zone and it should be ok with the new empty
> zone. So just the appended patch should work. Can you test?

Will do. but did you actually mean to enable it on both arches? didn't
look like it, but maybe you did.
 
> -AndI
> 
> Make i386 compile again with fourth DMA32 zone
> 
> The code should deal with an additiona empty zone, so fix up the
># error.
> 
> Signed-off-by: Andi Kleen <ak@suse.de>
> 
> Index: linux/arch/i386/kernel/srat.c
> ===================================================================
> --- linux.orig/arch/i386/kernel/srat.c
> +++ linux/arch/i386/kernel/srat.c
> @@ -137,8 +137,8 @@ static void __init parse_memory_affinity
>  		 "enabled and removable" : "enabled" ) );
>  }
>  
> -#if MAX_NR_ZONES != 3
> -#error "MAX_NR_ZONES != 3, chunk_to_zone requires review"
> +#if MAX_NR_ZONES != 4
> +#error "MAX_NR_ZONES != 4, chunk_to_zone requires review"
>  #endif
>  /* Take a chunk of pages from page frame cstart to cend and count the number
>   * of pages in each zone, returned via zones[].
> 
> 
> 
> 
> 

Re: 2.6.13-mm2

[Andi Kleen]

On Sun, Sep 11, 2005 at 11:09:59PM -0700, Martin J. Bligh wrote:
> >> Andi, does that need changing on ia32 as well as x86_64, or are you
> >> just missing some ifdefs? Looks to me like the rest of the patch is
> >> specific to x86_64.
> > 
> > It should be a straight forward fix - the new zone is empty on i386.
> > Ok I reviewed chunk_to_zone and it should be ok with the new empty
> > zone. So just the appended patch should work. Can you test?
> 
> Will do. but did you actually mean to enable it on both arches? didn't
> look like it, but maybe you did.

The zone is just empty on i386. That could have been avoided
with some ifdefs, but I didn't see any sense because an empty 
zone shouldn't hurt anybody.

-Andi

Re: 2.6.13-mm2

[Rafael J. Wysocki]

Hi,

On Sunday, 11 of September 2005 22:08, Daniel Ritz wrote:
> On Sunday 11 September 2005 21.36, Andrew Morton wrote:
> > "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
> > >
> > > > 
> > >  > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> > >  > 
> > >  > (kernel.org propagation is slow.  There's a temp copy at
> > >  > http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)
> > > 
> > >  Could you please reintroduce the yenta-free_irq-on-suspend.patch (attached)
> > >  into -mm?  My box does not resume from disk without it.
> > 
> > No probs.
> > 
> > Daniel, do you remember why we decided to drop it?  What should we do about
> > this?  Thanks.
> > 
> 
> yeah, there was a long discussion about it. see:
> 	http://marc.theaimsgroup.com/?t=112275164900002&r=1&w=4
> the reason being that it breaks APM suspend on Hugh Dickins' (added to cc:) laptop.
> Linus was quite clear about why reverting...
> 	http://marc.theaimsgroup.com/?l=linux-kernel&m=112278810115252&w=4
> 
> we should look at both problems in detail:
> - with APM it seems to break because the bridge gives interrupt before the
>   handler is installed.
> - with ACPI i think some _other_ device gives the interrupts too early. but
>   when all devices on the interrupt unregister the irq is disabled and the
>   problem is hidden.
> 
> i don't think we can do mutch about the APM case...
> 
> so Rafael, your /proc/interrupts, lspci -vvv and dmesg, please.

rafael@albercik:~> cat /proc/interrupts


> 
> rgds
> -daniel
> 
> 

-- 
- Would you tell me, please, which way I ought to go from here?
- That depends a good deal on where you want to get to.
		-- Lewis Carroll "Alice's Adventures in Wonderland"

Re: 2.6.13-mm2

[Rafael J. Wysocki]

Hi,

(continuing the unfinished message)

On Sunday, 11 of September 2005 22:08, Daniel Ritz wrote:
> On Sunday 11 September 2005 21.36, Andrew Morton wrote:
> > "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
> > >
> > > > 
> > >  > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> > >  > 
> > >  > (kernel.org propagation is slow.  There's a temp copy at
> > >  > http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)
> > > 
> > >  Could you please reintroduce the yenta-free_irq-on-suspend.patch (attached)
> > >  into -mm?  My box does not resume from disk without it.
> > 
> > No probs.
> > 
> > Daniel, do you remember why we decided to drop it?  What should we do about
> > this?  Thanks.
> > 
> 
> yeah, there was a long discussion about it. see:
> 	http://marc.theaimsgroup.com/?t=112275164900002&r=1&w=4
> the reason being that it breaks APM suspend on Hugh Dickins' (added to cc:) laptop.
> Linus was quite clear about why reverting...
> 	http://marc.theaimsgroup.com/?l=linux-kernel&m=112278810115252&w=4
> 
> we should look at both problems in detail:
> - with APM it seems to break because the bridge gives interrupt before the
>   handler is installed.
> - with ACPI i think some _other_ device gives the interrupts too early. but
>   when all devices on the interrupt unregister the irq is disabled and the
>   problem is hidden.
> 
> i don't think we can do mutch about the APM case...
> 
> so Rafael, your /proc/interrupts, lspci -vvv and dmesg, please.

rafael@albercik:~> cat /proc/interrupts
           CPU0       
  0:    2549476          XT-PIC  timer
  1:       3841          XT-PIC  i8042
  2:          0          XT-PIC  cascade
  5:         10          XT-PIC  ehci_hcd:usb3
  8:          0          XT-PIC  rtc
  9:      13808          XT-PIC  acpi
 10:     120111          XT-PIC  NVidia nForce3, SysKonnect SK-98xx
 11:      74195          XT-PIC  ohci_hcd:usb1, ohci_hcd:usb2, yenta, yenta
 12:       4733          XT-PIC  i8042
 14:      90733          XT-PIC  ide0
 15:     144590          XT-PIC  ide1
NMI:        879 
LOC:    2549489 
ERR:          6
MIS:          0

rafael@albercik:~> /sbin/lspci -vvv 
0000:00:00.0 Host bridge: nVidia Corporation nForce3 Host Bridge (rev a4)
	Subsystem: ASUSTeK Computer Inc.: Unknown device 80c5
	Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 0
	Region 0: Memory at e8000000 (32-bit, prefetchable) [size=128M]
	Capabilities: <available only to root>

0000:00:01.0 ISA bridge: nVidia Corporation nForce3 LPC Bridge (rev f6)
	Subsystem: ASUSTeK Computer Inc.: Unknown device 80c5
	Control: I/O+ Mem+ BusMaster+ SpecCycle+ MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap- 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 0

0000:00:01.1 SMBus: nVidia Corporation nForce3 SMBus (rev a4)
	Subsystem: ASUSTeK Computer Inc.: Unknown device 80c5
	Control: I/O+ Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Interrupt: pin A routed to IRQ 0
	Region 4: I/O ports at 5000 [size=64]
	Region 5: I/O ports at 5040 [size=64]
	Capabilities: <available only to root>

0000:00:02.0 USB Controller: nVidia Corporation nForce3 USB 1.1 (rev a5) (prog-if 10 [OHCI])
	Subsystem: ASUSTeK Computer Inc.: Unknown device 1858
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 0 (750ns min, 250ns max)
	Interrupt: pin A routed to IRQ 11
	Region 0: Memory at febfb000 (32-bit, non-prefetchable) [size=4K]
	Capabilities: <available only to root>

0000:00:02.1 USB Controller: nVidia Corporation nForce3 USB 1.1 (rev a5) (prog-if 10 [OHCI])
	Subsystem: ASUSTeK Computer Inc.: Unknown device 1858
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 0 (750ns min, 250ns max)
	Interrupt: pin B routed to IRQ 11
	Region 0: Memory at febfc000 (32-bit, non-prefetchable) [size=4K]
	Capabilities: <available only to root>

0000:00:02.2 USB Controller: nVidia Corporation nForce3 USB 2.0 (rev a2) (prog-if 20 [EHCI])
	Subsystem: ASUSTeK Computer Inc.: Unknown device 1859
	Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 0 (750ns min, 250ns max)
	Interrupt: pin C routed to IRQ 5
	Region 0: Memory at febfdc00 (32-bit, non-prefetchable) [size=256]
	Capabilities: <available only to root>

0000:00:06.0 Multimedia audio controller: nVidia Corporation nForce3 Audio (rev a2)
	Subsystem: ASUSTeK Computer Inc.: Unknown device 1853
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 0 (500ns min, 1250ns max)
	Interrupt: pin A routed to IRQ 10
	Region 0: I/O ports at e800 [size=256]
	Region 1: I/O ports at ec00 [size=128]
	Region 2: Memory at febff000 (32-bit, non-prefetchable) [size=4K]
	Capabilities: <available only to root>

0000:00:08.0 IDE interface: nVidia Corporation nForce3 IDE (rev a5) (prog-if 8a [Master SecP PriP])
	Subsystem: ASUSTeK Computer Inc.: Unknown device 185a
	Control: I/O+ Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 0 (750ns min, 250ns max)
	Region 4: I/O ports at ffa0 [size=16]
	Capabilities: <available only to root>

0000:00:0a.0 PCI bridge: nVidia Corporation nForce3 PCI Bridge (rev a2) (prog-if 00 [Normal decode])
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
	Status: Cap- 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR+ <PERR-
	Latency: 0
	Bus: primary=00, secondary=02, subordinate=0a, sec-latency=128
	I/O behind bridge: 0000b000-0000dfff
	Memory behind bridge: f8a00000-feafffff
	Prefetchable memory behind bridge: 30000000-33ffffff
	BridgeCtl: Parity+ SERR+ NoISA+ VGA- MAbort- >Reset- FastB2B-

0000:00:0b.0 PCI bridge: nVidia Corporation nForce3 AGP Bridge (rev a4) (prog-if 00 [Normal decode])
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
	Status: Cap- 66Mhz+ UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 64
	Bus: primary=00, secondary=01, subordinate=01, sec-latency=64
	I/O behind bridge: 0000f000-00000fff
	Memory behind bridge: f6900000-f89fffff
	Prefetchable memory behind bridge: c6800000-e67fffff
	BridgeCtl: Parity+ SERR+ NoISA- VGA+ MAbort- >Reset- FastB2B-

0000:00:18.0 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] HyperTransport Technology Configuration
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap+ 66Mhz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Capabilities: <available only to root>

0000:00:18.1 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] Address Map
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-

0000:00:18.2 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] DRAM Controller
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-

0000:00:18.3 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] Miscellaneous Control
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-

0000:01:00.0 VGA compatible controller: nVidia Corporation NV31M [GeForce FX Go5650] (rev a1) (prog-if 00 [VGA])
	Subsystem: ASUSTeK Computer Inc.: Unknown device 1852
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 64 (1250ns min, 250ns max)
	Interrupt: pin A routed to IRQ 11
	Region 0: Memory at f7000000 (32-bit, non-prefetchable) [size=16M]
	Region 1: Memory at d0000000 (32-bit, prefetchable) [size=256M]
	Expansion ROM at f89e0000 [disabled] [size=128K]
	Capabilities: <available only to root>

0000:02:00.0 Ethernet controller: Marvell Technology Group Ltd. Gigabit Ethernet Controller (rev 13)
	Subsystem: ASUSTeK Computer Inc. Marvell 88E8001 Gigabit Ethernet Controller (Asus)
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr- Stepping- SERR+ FastB2B-
	Status: Cap+ 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 64 (5750ns min, 7750ns max), cache line size 40
	Interrupt: pin A routed to IRQ 10
	Region 0: Memory at feaf8000 (32-bit, non-prefetchable) [size=16K]
	Region 1: I/O ports at d800 [size=256]
	Expansion ROM at feac0000 [disabled] [size=128K]
	Capabilities: <available only to root>

0000:02:01.0 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev ab)
	Subsystem: ASUSTeK Computer Inc.: Unknown device 1854
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap+ 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 168
	Interrupt: pin A routed to IRQ 11
	Region 0: Memory at fd200000 (32-bit, non-prefetchable) [size=4K]
	Bus: primary=02, secondary=03, subordinate=06, sec-latency=176
	Memory window 0: 30000000-31fff000 (prefetchable)
	Memory window 1: fc600000-fd1ff000
	I/O window 0: 0000b000-0000bfff
	I/O window 1: 0000c000-0000cfff
	BridgeCtl: Parity- SERR- ISA- VGA- MAbort- >Reset- 16bInt+ PostWrite+
	16-bit legacy interface ports at 0001

0000:02:01.1 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev ab)
	Subsystem: ASUSTeK Computer Inc.: Unknown device 1854
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap+ 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 168
	Interrupt: pin B routed to IRQ 11
	Region 0: Memory at fa200000 (32-bit, non-prefetchable) [size=4K]
	Bus: primary=02, secondary=07, subordinate=0a, sec-latency=176
	Memory window 0: 32000000-33fff000 (prefetchable)
	Memory window 1: f9600000-fa1ff000
	I/O window 0: 0000d000-0000d1ff
	I/O window 1: 0000d400-0000d5ff
	BridgeCtl: Parity- SERR- ISA- VGA- MAbort- >Reset- 16bInt+ PostWrite+
	16-bit legacy interface ports at 0001

0000:02:01.2 FireWire (IEEE 1394): Ricoh Co Ltd R5C552 IEEE 1394 Controller (rev 03) (prog-if 10 [OHCI])
	Subsystem: ASUSTeK Computer Inc.: Unknown device 1857
	Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
	Status: Cap+ 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 64 (500ns min, 1000ns max)
	Interrupt: pin C routed to IRQ 10
	Region 0: Memory at feafd000 (32-bit, non-prefetchable) [size=2K]
	Capabilities: <available only to root>

0000:02:01.3 System peripheral: Ricoh Co Ltd: Unknown device 0576 (rev 01)
	Subsystem: ASUSTeK Computer Inc.: Unknown device 185b
	Control: I/O- Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
	Status: Cap+ 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Interrupt: pin D routed to IRQ 11
	Region 0: Memory at feafd800 (32-bit, non-prefetchable) [size=256]
	Capabilities: <available only to root>

0000:02:01.4 System peripheral: Ricoh Co Ltd: Unknown device 0592
	Subsystem: ASUSTeK Computer Inc.: Unknown device 185c
	Control: I/O- Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
	Status: Cap+ 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Interrupt: pin D routed to IRQ 11
	Region 0: Memory at feafdc00 (32-bit, non-prefetchable) [size=256]
	Capabilities: <available only to root>

0000:02:02.0 Network controller: Broadcom Corporation BCM4306 802.11b/g Wireless LAN Controller (rev 03)
	Subsystem: ASUSTeK Computer Inc.: Unknown device 120f
	Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B-
	Status: Cap+ 66Mhz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 64
	Interrupt: pin A routed to IRQ 11
	Region 0: Memory at feafe000 (32-bit, non-prefetchable) [size=8K]
	Capabilities: <available only to root>

Greetings,
Rafael


-- 
- Would you tell me, please, which way I ought to go from here?
- That depends a good deal on where you want to get to.
		-- Lewis Carroll "Alice's Adventures in Wonderland"

Re: 2.6.13-mm2

[Rafael J. Wysocki]

On Monday, 12 of September 2005 12:06, Rafael J. Wysocki wrote:
> Hi,
> 
> (continuing the unfinished message)
> 
> On Sunday, 11 of September 2005 22:08, Daniel Ritz wrote:
> > On Sunday 11 September 2005 21.36, Andrew Morton wrote:
> > > "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
> > > >
> > > > > 
> > > >  > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> > > >  > 
> > > >  > (kernel.org propagation is slow.  There's a temp copy at
> > > >  > http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)
> > > > 
> > > >  Could you please reintroduce the yenta-free_irq-on-suspend.patch (attached)
> > > >  into -mm?  My box does not resume from disk without it.
> > > 
> > > No probs.
> > > 
> > > Daniel, do you remember why we decided to drop it?  What should we do about
> > > this?  Thanks.
> > > 
> > 
> > yeah, there was a long discussion about it. see:
> > 	http://marc.theaimsgroup.com/?t=112275164900002&r=1&w=4
> > the reason being that it breaks APM suspend on Hugh Dickins' (added to cc:) laptop.
> > Linus was quite clear about why reverting...
> > 	http://marc.theaimsgroup.com/?l=linux-kernel&m=112278810115252&w=4
> > 
> > we should look at both problems in detail:
> > - with APM it seems to break because the bridge gives interrupt before the
> >   handler is installed.
> > - with ACPI i think some _other_ device gives the interrupts too early. but
> >   when all devices on the interrupt unregister the irq is disabled and the
> >   problem is hidden.
> > 
> > i don't think we can do mutch about the APM case...
> > 
> > so Rafael, your /proc/interrupts, lspci -vvv and dmesg, please.
> 
> rafael@albercik:~> cat /proc/interrupts
]-- snip --[

BTW, please have a look at:
http://bugzilla.kernel.org/show_bug.cgi?id=4416#c36
and
http://bugzilla.kernel.org/show_bug.cgi?id=4416#c37

Greetings,
Rafael


-- 
- Would you tell me, please, which way I ought to go from here?
- That depends a good deal on where you want to get to.
		-- Lewis Carroll "Alice's Adventures in Wonderland"

Re: 2.6.13-mm2

[Martin J. Bligh]

--On Monday, September 12, 2005 07:01:22 +0200 Andi Kleen <ak@muc.de> wrote:

> On Sun, Sep 11, 2005 at 08:07:25PM -0700, Martin J. Bligh wrote:
>> Finally got my damned x440 box back - won't build -mm2 (-mm1 is fine)
>> 
>> arch/i386/kernel/srat.c:141: #error "MAX_NR_ZONES != 3, chunk_to_zone requires review"
>> make[1]: *** [arch/i386/kernel/srat.o] Error 1
>> make: *** [arch/i386/kernel] Error 2
>> 09/11/05-00:57:13 Build the kernel. Failed rc = 2
>> 09/11/05-00:57:13 build: kernel build Failed rc = 1
>> 09/11/05-00:57:13 command complete: (2) rc=126
>> 
>> x86_64-dma32.patch:-#define MAX_NR_ZONES                3       /* Sync this wi
>> h ZONES_SHIFT */
>> x86_64-dma32.patch:-#define ZONES_SHIFT         2       /* ceil(log2(MAX_NR_ZON
>> S)) */
>> x86_64-dma32.patch:+#define MAX_NR_ZONES                4       /* Sync this wi
>> h ZONES_SHIFT */
>> x86_64-dma32.patch:+#define ZONES_SHIFT         3       /* ceil(log2(MAX_NR_ZON
>> S)) */
>> 
>> Andi, does that need changing on ia32 as well as x86_64, or are you
>> just missing some ifdefs? Looks to me like the rest of the patch is
>> specific to x86_64.
> 
> It should be a straight forward fix - the new zone is empty on i386.
> Ok I reviewed chunk_to_zone and it should be ok with the new empty
> zone. So just the appended patch should work. Can you test?
> 
> -AndI
> 
> Make i386 compile again with fourth DMA32 zone
> 
> The code should deal with an additiona empty zone, so fix up the
># error.
> 
> Signed-off-by: Andi Kleen <ak@suse.de>
> 
> Index: linux/arch/i386/kernel/srat.c
> ===================================================================
> --- linux.orig/arch/i386/kernel/srat.c
> +++ linux/arch/i386/kernel/srat.c
> @@ -137,8 +137,8 @@ static void __init parse_memory_affinity
>  		 "enabled and removable" : "enabled" ) );
>  }
>  
> -#if MAX_NR_ZONES != 3
> -#error "MAX_NR_ZONES != 3, chunk_to_zone requires review"
> +#if MAX_NR_ZONES != 4
> +#error "MAX_NR_ZONES != 4, chunk_to_zone requires review"
>  #endif
>  /* Take a chunk of pages from page frame cstart to cend and count the number
>   * of pages in each zone, returned via zones[].
> 
> 
> 
> 
> 

Crashes on boot

http://test.kernel.org/12589/debug/console.log

May or may not be anything to do with what you were doing.

M.

Re: 2.6.13-mm2

[Dave Hansen]

On Mon, 2005-09-12 at 11:06 -0700, Martin J. Bligh wrote:
> Crashes on boot
> 
> http://test.kernel.org/12589/debug/console.log
> 
> May or may not be anything to do with what you were doing.

diff -puN arch/i386/mm/init.c~highmem-debug arch/i386/mm/init.c
--- memhotplug/arch/i386/mm/init.c~highmem-debug	2005-09-09 08:50:15.000000000 -0700
+++ memhotplug-dave/arch/i386/mm/init.c	2005-09-09 08:50:15.000000000 -0700
@@ -278,6 +278,7 @@ void __init add_one_highpage_init(struct
 {
 	if (page_is_ram(pfn) && !(bad_ppro && page_kills_ppro(pfn))) {
 		ClearPageReserved(page);
+		free_new_highpage(page);
 	} else
 		SetPageReserved(page);
 }
_


-- Dave

Re: 2.6.13-mm2

[Andi Kleen]

> Crashes on boot
> 
> http://test.kernel.org/12589/debug/console.log
> 
> May or may not be anything to do with what you were doing.

Easily tested by reverting dma32*. Does it help?

-Andi

Re: 2.6.13-mm2

[Rafael J. Wysocki]

On Sunday, 11 of September 2005 22:03, Hugh Dickins wrote:
> On Sun, 11 Sep 2005, Andrew Morton wrote:
> > "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
> > > 
> > >  Could you please reintroduce the yenta-free_irq-on-suspend.patch (attached)
> > >  into -mm?  My box does not resume from disk without it.
> > 
> > No probs.
> > 
> > Daniel, do you remember why we decided to drop it?  What should we do about
> > this?  Thanks.
> 
> I remember well.  My laptop does not APM resume from RAM with it.
> I've just rechecked and that's still the case.  I did try various patches
> from Rafael to help him work it out, but it remained a puzzle.

Still the observations are sort of interesting (just for the record):
1) The problem is due to interrupt sharing.  The IRQ is shared between
yenta and 3c59x.
2) Without the yenta driver the box resumes.
3) Without the 3c59x driver the box resumes.
4) If yenta is woken up _before_ 3c59x, the box resumes.
5) If yenta is woken up _after_ 3c59x, the box hangs.  Unfortunately,
this is the default, because of the PCI device numbers.

> And I admit, it is his turn to resume this month.

Thanks. ;-)

Greetings,
Rafael


-- 
- Would you tell me, please, which way I ought to go from here?
- That depends a good deal on where you want to get to.
		-- Lewis Carroll "Alice's Adventures in Wonderland"

Re: 2.6.13-mm2

[Martin J. Bligh]

--On Monday, September 12, 2005 20:51:20 +0200 Andi Kleen <ak@muc.de> wrote:

>> Crashes on boot
>> 
>> http://test.kernel.org/12589/debug/console.log
>> 
>> May or may not be anything to do with what you were doing.
> 
> Easily tested by reverting dma32*. Does it help?

No. Yet *another* bug. Sigh.

M.

Re: 2.6.13-mm2

[Andrew Morton]

"Martin J. Bligh" <mbligh@mbligh.org> wrote:
>
> 
> 
> --On Monday, September 12, 2005 20:51:20 +0200 Andi Kleen <ak@muc.de> wrote:
> 
> >> Crashes on boot
> >> 
> >> http://test.kernel.org/12589/debug/console.log
> >> 
> >> May or may not be anything to do with what you were doing.
> > 
> > Easily tested by reverting dma32*. Does it help?
> 
> No. Yet *another* bug. Sigh.
> 

You should see all the ones I fixed.

Suggest you skip -mm2 altogether.  We already know that
scheduler-cache-hot-autodetect.patch is bad, and that was dropped from mm3.

Re: 2.6.13-mm2

[Martin J. Bligh]

--Andrew Morton <akpm@osdl.org> wrote (on Monday, September 12, 2005 17:08:52 -0700):

> "Martin J. Bligh" <mbligh@mbligh.org> wrote:
>> 
>> 
>> 
>> --On Monday, September 12, 2005 20:51:20 +0200 Andi Kleen <ak@muc.de> wrote:
>> 
>> >> Crashes on boot
>> >> 
>> >> http://test.kernel.org/12589/debug/console.log
>> >> 
>> >> May or may not be anything to do with what you were doing.
>> > 
>> > Easily tested by reverting dma32*. Does it help?
>> 
>> No. Yet *another* bug. Sigh.
>> 
> 
> You should see all the ones I fixed.

Yeah, sorry, I know. Don't know how you cope at all.
 
> Suggest you skip -mm2 altogether.  We already know that
> scheduler-cache-hot-autodetect.patch is bad, and that was dropped from mm3.

-mm3 looks much better, seems to work on ia32 and x86_64 at least.

PPC64 is still broken by the tty cleanup stuff. there were some fixes
going back and forth ... hopefully someone will gather them and push
to you soon, would nice to be testing that again.

drivers/char/hvc_console.c: In function `hvc_poll':
drivers/char/hvc_console.c:600: error: `count' undeclared (first use in this function)
drivers/char/hvc_console.c:600: error: (Each undeclared identifier is reported only once
drivers/char/hvc_console.c:600: error: for each function it appears in.)
drivers/char/hvc_console.c:636: error: structure has no member named `flip'

Thanks for everything ... testing just makes one bitter ;-)

M.

Re: 2.6.13-mm2

[Chuck Ebbert]

On Thu, 08 Sep 2005 at 20:26:51 -0400, Parag Warudkar wrote:

> Andrew Morton wrote:
> 
> > Parag, perhaps you could confirm that reverting that patch fixes 
> > things up?
> 
> Sure - reverting the x86-64-ptrace-ia32-bp-fix patch fixes it.

 It looks to me like that patch corrupts ebp.

 This one works for me, though ebp still appears wrong to a 32-bit
debugger on syscall exit trace.  Maybe that doesn't matter so much.

 Test program follows (it will create a file named "__ptfile__"), then patch.

================================================================================

/* ptrace test program 
 *
 * Tests if ptrace can modify syscall arg6 (ebp) in ia32 mode.
 */
#define _GNU_SOURCE
#include <stdio.h>
#include <unistd.h>
#include <fcntl.h>
#include <sys/ptrace.h>
#include <sys/wait.h>
#include <sys/mman.h>
#include <asm/user.h>

int parent, child, status, traces;
struct user_regs_struct regs;
char zero = '0', one = '1';

void dump_regs(struct user_regs_struct *regs)
{
	printf("eax = %08X (%d), orig_eax = %08X (%d)\n",
		(unsigned int)regs->eax,
		(unsigned int)regs->eax,
		(unsigned int)regs->orig_eax,
		(unsigned int)regs->orig_eax);
	printf("ebx = %08X, ecx = %08X, edx = %08X\n",
		(unsigned int)regs->ebx,
		(unsigned int)regs->ecx,
		(unsigned int)regs->edx);
	printf("esi = %08X, edi = %08X, ebp = %08X\n",
		(unsigned int)regs->esi,
		(unsigned int)regs->edi,
		(unsigned int)regs->ebp);
}

void do_parent()
{
again:
	waitpid(child, &status, 0);
	if (WIFSTOPPED(status)) {
		if (traces && traces <= 2 ) { /* skip first, then do two */
			puts("child stopped");
			ptrace(PTRACE_GETREGS, child, 0, &regs);
			dump_regs(&regs);
			if (regs.orig_eax == 192 && traces == 1) { /* mmap2 */
				if (regs.ebp != 0)
					puts("tracer: arg6 was not 0: "
					     "ptrace is broken!");
				else {
					regs.ebp = 1; /* change arg 6 */
					ptrace(PTRACE_SETREGS, child, 0, &regs);
				}
			} else if (traces == 1) /* first syscall wasn't mmap2 */
				puts("unexpected problem, ignore the result!");
		}
		ptrace(PTRACE_SYSCALL, child, 0, 0);
		traces++;
	}
	if (!WIFEXITED(status))
		goto again;
}

void do_child()
{
	int f, i;
	void *a;
	int works = 1;

	f = open("__ptfile__", O_RDWR | O_CREAT | O_TRUNC, S_IREAD | S_IWRITE);
	for (i = 0; i < 4096; i++) /* a page of '0' */
		write(f, &zero, 1);
	for (i = 0; i < 4096; i++) /* a page of '1' */
		write(f, &one, 1);
	close(f);
	
	f = open("__ptfile__", O_RDONLY);
	ptrace(PTRACE_TRACEME, 0, 0, 0);
	kill(getpid(), SIGUSR1);
	a = mmap(0, 4096, PROT_READ, MAP_SHARED, f, 0); /* map first page */
	if (*(char *)a != one)
		works = 0; /* got first page: arg6 was unchanged */
	munmap(a, 4096);
	close(f);
	remove("__ptfile__");
	
	if (works)
		puts("ptrace works");
	else
		puts("mmap args didn't change: ptrace is broken!");
}

int main(int argc, char * const argv[])
{
	parent = getpid();
	child = fork();

	if (child)
		do_parent();
	else
		do_child();

	return 0;
}
================================================================================

Signed-off-by: Chuck Ebbert <76306.1226@compuserve.com>
Original-Patch-By: Roland McGrath <roland@redhat.com>

When the 32-bit vDSO is used to make a system call, the %ebp register for
the 6th syscall arg has to be loaded from the user stack (where it's pushed
by the vDSO user code).  The native i386 kernel always does this before
stopping for syscall tracing, so %ebp can be seen and modified via ptrace
to access the 6th syscall argument.  The x86-64 kernel fails to do this,
presenting the stack address to ptrace instead.  This makes the %rbp value
seen by 64-bit ptrace of a 32-bit process, and the %ebp value seen by a
32-bit caller of ptrace, both differ from the native i386 behavior.

This patch fixes the problem by putting the word loaded from the user stack
into %rbp before calling syscall_trace_enter, and reloading the 6th syscall
argument from there afterwards (so ptrace can change it).  This makes the
behavior match that of i386 kernels.

 arch/x86_64/ia32/ia32entry.S |   19 ++++++-------------
 1 files changed, 6 insertions(+), 13 deletions(-)

--- 2.6.13-64.orig/arch/x86_64/ia32/ia32entry.S
+++ 2.6.13-64/arch/x86_64/ia32/ia32entry.S
@@ -102,20 +102,16 @@ sysenter_do_call:	
 	.byte	0xf, 0x35
 
 sysenter_tracesys:
+	xchgl	%r9d,%ebp
 	SAVE_REST
 	CLEAR_RREGS
+	movq	%r9,R9(%rsp)
 	movq	$-ENOSYS,RAX(%rsp)	/* really needed? */
 	movq	%rsp,%rdi        /* &pt_regs -> arg1 */
 	call	syscall_trace_enter
 	LOAD_ARGS ARGOFFSET  /* reload args from stack in case ptrace changed it */
 	RESTORE_REST
-	movl	%ebp, %ebp
-	/* no need to do an access_ok check here because rbp has been
-	   32bit zero extended */ 
-1:	movl	(%rbp),%r9d
-	.section __ex_table,"a"
-	.quad 1b,ia32_badarg
-	.previous
+	xchgl	%ebp,%r9d
 	jmp	sysenter_do_call
 	CFI_ENDPROC
 
@@ -183,20 +179,17 @@ cstar_do_call:	
 	sysretl
 	
 cstar_tracesys:	
+	xchgl %r9d,%ebp
 	SAVE_REST
 	CLEAR_RREGS
+	movq %r9,R9(%rsp)
 	movq $-ENOSYS,RAX(%rsp)	/* really needed? */
 	movq %rsp,%rdi        /* &pt_regs -> arg1 */
 	call syscall_trace_enter
 	LOAD_ARGS ARGOFFSET  /* reload args from stack in case ptrace changed it */
 	RESTORE_REST
+	xchgl %ebp,%r9d
 	movl RSP-ARGOFFSET(%rsp), %r8d
-	/* no need to do an access_ok check here because r8 has been
-	   32bit zero extended */ 
-1:	movl	(%r8),%r9d
-	.section __ex_table,"a"
-	.quad 1b,ia32_badarg
-	.previous
 	jmp cstar_do_call
 				
 ia32_badarg:
__
Chuck

Re: 2.6.13-mm2

[Parag Warudkar]

On Fri, 2005-09-16 at 20:36 -0400, Chuck Ebbert wrote:
> It looks to me like that patch corrupts ebp.
> 
>  This one works for me, though ebp still appears wrong to a 32-bit
> debugger on syscall exit trace.  Maybe that doesn't matter so much. 

Did you compile the program under a x86_64 kernel, distro? I ask because
it doesn't compile for me. On my machine asm-i386/user.h and
asm-x86_64/user.h both do not contain a user_regs_struct definition with
x86 registers.

Another thing I noticed was that with the patch %eax was zero for the
failing program although I didn't see how the patch affects %eax.

Parag

Re: 2.6.13-mm2

[Parag Warudkar]

On Sat, 2005-09-17 at 00:17 -0400, Parag Warudkar wrote:
> On my machine asm-i386/user.h and
> asm-x86_64/user.h both do not contain a user_regs_struct definition
> with
> x86 registers. 

Never mind - Ubuntu seems to have broken compat headers. (Kernel version
of asm-i386/user.h is appropriate.)

P

Re: 2.6.13-mm2

[Daniel Ritz]

On Monday 12 September 2005 12.09, Rafael J. Wysocki wrote:
> On Monday, 12 of September 2005 12:06, Rafael J. Wysocki wrote:
> > Hi,
> > 
> > (continuing the unfinished message)
> > 
> > On Sunday, 11 of September 2005 22:08, Daniel Ritz wrote:
> > > On Sunday 11 September 2005 21.36, Andrew Morton wrote:
> > > > "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
> > > > >
> > > > > > 
> > > > >  > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm2/
> > > > >  > 
> > > > >  > (kernel.org propagation is slow.  There's a temp copy at
> > > > >  > http://www.zip.com.au/~akpm/linux/patches/stuff/2.6.13-mm2.bz2)
> > > > > 
> > > > >  Could you please reintroduce the yenta-free_irq-on-suspend.patch (attached)
> > > > >  into -mm?  My box does not resume from disk without it.
> > > > 
> > > > No probs.
> > > > 
> > > > Daniel, do you remember why we decided to drop it?  What should we do about
> > > > this?  Thanks.
> > > > 
> > > 
> > > yeah, there was a long discussion about it. see:
> > > 	http://marc.theaimsgroup.com/?t=112275164900002&r=1&w=4
> > > the reason being that it breaks APM suspend on Hugh Dickins' (added to cc:) laptop.
> > > Linus was quite clear about why reverting...
> > > 	http://marc.theaimsgroup.com/?l=linux-kernel&m=112278810115252&w=4
> > > 
> > > we should look at both problems in detail:
> > > - with APM it seems to break because the bridge gives interrupt before the
> > >   handler is installed.
> > > - with ACPI i think some _other_ device gives the interrupts too early. but
> > >   when all devices on the interrupt unregister the irq is disabled and the
> > >   problem is hidden.
> > > 
> > > i don't think we can do mutch about the APM case...
> > > 
> > > so Rafael, your /proc/interrupts, lspci -vvv and dmesg, please.
> > 
> > rafael@albercik:~> cat /proc/interrupts
> ]-- snip --[
> 
> BTW, please have a look at:
> http://bugzilla.kernel.org/show_bug.cgi?id=4416#c36
> and
> http://bugzilla.kernel.org/show_bug.cgi?id=4416#c37
> 

interesting. i'd say we get interrupt storms from usb which then hurt when
yenta has it's handler installed but usb has not. usb/hcd-pci.c frees the
irq on suspend...so it may be enough not to do that (survives suspend-to-ram
and suspend-to-disk here. yes, restore too :)

could you give that a tree w/o any free_irq-patches for yenta and co?

rgds
-daniel

diff --git a/drivers/usb/core/hcd-pci.c b/drivers/usb/core/hcd-pci.c
--- a/drivers/usb/core/hcd-pci.c
+++ b/drivers/usb/core/hcd-pci.c
@@ -242,7 +242,9 @@ int usb_hcd_pci_suspend (struct pci_dev 
 	case HC_STATE_SUSPENDED:
 		/* no DMA or IRQs except when HC is active */
 		if (dev->current_state == PCI_D0) {
+#if 0
 			free_irq (hcd->irq, hcd);
+#endif
 			pci_save_state (dev);
 			pci_disable_device (dev);
 		}
@@ -374,6 +376,7 @@ int usb_hcd_pci_resume (struct pci_dev *
 
 	hcd->state = HC_STATE_RESUMING;
 	hcd->saw_irq = 0;
+#if 0
 	retval = request_irq (dev->irq, usb_hcd_irq, SA_SHIRQ,
 				hcd->irq_descr, hcd);
 	if (retval < 0) {
@@ -382,6 +385,7 @@ int usb_hcd_pci_resume (struct pci_dev *
 		usb_hc_died (hcd);
 		return retval;
 	}
+#endif
 
 	retval = hcd->driver->resume (hcd);
 	if (!HC_IS_RUNNING (hcd->state)) {

Re: 2.6.13-mm2

[Hugh Dickins]

On Sun, 18 Sep 2005, Daniel Ritz wrote:
> 
> interesting. i'd say we get interrupt storms from usb which then hurt when
> yenta has it's handler installed but usb has not. usb/hcd-pci.c frees the
> irq on suspend...so it may be enough not to do that (survives suspend-to-ram
> and suspend-to-disk here. yes, restore too :)

Will make no difference to my case: I have no USB,
so don't even build drivers/usb/core/hcd-pci.o.

Hugh

Re: 2.6.13-mm2

[Daniel Ritz]

On Monday 19 September 2005 05.07, Hugh Dickins wrote:
> On Sun, 18 Sep 2005, Daniel Ritz wrote:
> > 
> > interesting. i'd say we get interrupt storms from usb which then hurt when
> > yenta has it's handler installed but usb has not. usb/hcd-pci.c frees the
> > irq on suspend...so it may be enough not to do that (survives suspend-to-ram
> > and suspend-to-disk here. yes, restore too :)
> 
> Will make no difference to my case: I have no USB,
> so don't even build drivers/usb/core/hcd-pci.o.

well, it's not supposed to do anything in your case. but maybe it
fixes rafaels case w/o this crappy yenta free_irq thing :)

-daniel

Re: 2.6.13-mm2

[Rafael J. Wysocki]

Hi,

[sorry for the delay]

On Sunday, 18 of September 2005 23:49, Daniel Ritz wrote:
]--snip--[
> > 
> > BTW, please have a look at:
> > http://bugzilla.kernel.org/show_bug.cgi?id=4416#c36
> > and
> > http://bugzilla.kernel.org/show_bug.cgi?id=4416#c37
> > 
> 
> interesting. i'd say we get interrupt storms from usb which then hurt when
> yenta has it's handler installed but usb has not. usb/hcd-pci.c frees the
> irq on suspend...so it may be enough not to do that (survives suspend-to-ram
> and suspend-to-disk here. yes, restore too :)
> 
> could you give that a tree w/o any free_irq-patches for yenta and co?

I've tried and it apparently works provided that _none_ of the IRQ-sharing
devices drops the IRQ on suspend.

I think that's the whole point: Either all of the devices should drop/request
IRQs on suspend/resume, or none of them should do this.  IMHO we need to
chose one of these options and call it "the right way" or there always
will be problems with this.

Greetings,
Rafael


> diff --git a/drivers/usb/core/hcd-pci.c b/drivers/usb/core/hcd-pci.c
> --- a/drivers/usb/core/hcd-pci.c
> +++ b/drivers/usb/core/hcd-pci.c
> @@ -242,7 +242,9 @@ int usb_hcd_pci_suspend (struct pci_dev 
>  	case HC_STATE_SUSPENDED:
>  		/* no DMA or IRQs except when HC is active */
>  		if (dev->current_state == PCI_D0) {
> +#if 0
>  			free_irq (hcd->irq, hcd);
> +#endif
>  			pci_save_state (dev);
>  			pci_disable_device (dev);
>  		}
> @@ -374,6 +376,7 @@ int usb_hcd_pci_resume (struct pci_dev *
>  
>  	hcd->state = HC_STATE_RESUMING;
>  	hcd->saw_irq = 0;
> +#if 0
>  	retval = request_irq (dev->irq, usb_hcd_irq, SA_SHIRQ,
>  				hcd->irq_descr, hcd);
>  	if (retval < 0) {
> @@ -382,6 +385,7 @@ int usb_hcd_pci_resume (struct pci_dev *
>  		usb_hc_died (hcd);
>  		return retval;
>  	}
> +#endif
>  
>  	retval = hcd->driver->resume (hcd);
>  	if (!HC_IS_RUNNING (hcd->state)) {
> 
> 

-- 
- Would you tell me, please, which way I ought to go from here?
- That depends a good deal on where you want to get to.
		-- Lewis Carroll "Alice's Adventures in Wonderland"

Re: 2.6.13-mm2

[Daniel Ritz]

On Friday 23 September 2005 18.52, Rafael J. Wysocki wrote:
> Hi,
> 
> [sorry for the delay]

[same :]

> 
> On Sunday, 18 of September 2005 23:49, Daniel Ritz wrote:
> ]--snip--[
> > > 
> > > BTW, please have a look at:
> > > http://bugzilla.kernel.org/show_bug.cgi?id=4416#c36
> > > and
> > > http://bugzilla.kernel.org/show_bug.cgi?id=4416#c37
> > > 
> > 
> > interesting. i'd say we get interrupt storms from usb which then hurt when
> > yenta has it's handler installed but usb has not. usb/hcd-pci.c frees the
> > irq on suspend...so it may be enough not to do that (survives suspend-to-ram
> > and suspend-to-disk here. yes, restore too :)
> > 
> > could you give that a tree w/o any free_irq-patches for yenta and co?
> 
> I've tried and it apparently works provided that _none_ of the IRQ-sharing
> devices drops the IRQ on suspend.

ok. i didn't look too close, but i think ohci-hcd does not fully disable interrupts
in it's suspend callback...needs a closer look. cc:ing linux-usb-devel...

> 
> I think that's the whole point: Either all of the devices should drop/request
> IRQs on suspend/resume, or none of them should do this.  IMHO we need to
> chose one of these options and call it "the right way" or there always
> will be problems with this.
> 

cc:ing linus since he seems to have a strong opinion about that free_irq-in-suspend
thing...not doing it for USB fixes the problem for both cases: APM suspend and ACPI
suspend...

> Greetings,
> Rafael

rgds
-daniel

> 
> 
> > diff --git a/drivers/usb/core/hcd-pci.c b/drivers/usb/core/hcd-pci.c
> > --- a/drivers/usb/core/hcd-pci.c
> > +++ b/drivers/usb/core/hcd-pci.c
> > @@ -242,7 +242,9 @@ int usb_hcd_pci_suspend (struct pci_dev 
> >  	case HC_STATE_SUSPENDED:
> >  		/* no DMA or IRQs except when HC is active */
> >  		if (dev->current_state == PCI_D0) {
> > +#if 0
> >  			free_irq (hcd->irq, hcd);
> > +#endif
> >  			pci_save_state (dev);
> >  			pci_disable_device (dev);
> >  		}
> > @@ -374,6 +376,7 @@ int usb_hcd_pci_resume (struct pci_dev *
> >  
> >  	hcd->state = HC_STATE_RESUMING;
> >  	hcd->saw_irq = 0;
> > +#if 0
> >  	retval = request_irq (dev->irq, usb_hcd_irq, SA_SHIRQ,
> >  				hcd->irq_descr, hcd);
> >  	if (retval < 0) {
> > @@ -382,6 +385,7 @@ int usb_hcd_pci_resume (struct pci_dev *
> >  		usb_hc_died (hcd);
> >  		return retval;
> >  	}
> > +#endif
> >  
> >  	retval = hcd->driver->resume (hcd);
> >  	if (!HC_IS_RUNNING (hcd->state)) {
> > 
> > 
> 

Re: 2.6.13-mm2

[Linus Torvalds]

On Wed, 28 Sep 2005, Daniel Ritz wrote:
> 
> cc:ing linus since he seems to have a strong opinion about that free_irq-in-suspend
> thing...not doing it for USB fixes the problem for both cases: APM suspend and ACPI
> suspend...

Trivial decision: if not freeing the irq fixes the problem, then please 
send a tested patch that does just that. It's what we used to do anyway,

		Linus