2.6.13-mm1
Subsystems (5)
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm1/
- Included Alan's big tty layer buffering rewrite. This breaks the build on lots of more obscure character device drivers. Patches welcome (please cc Alan).
Changes since 2.6.13-rc6-mm2:
linus.patch
git-acpi.patch
git-arm.patch
git-cpufreq.patch
git-cryptodev.patch
git-ia64.patch
git-audit.patch
git-audit-ppc64-fix.patch
git-input.patch
git-jfs-fixup.patch
git-kbuild.patch
git-libata-all.patch
git-mtd.patch
git-netdev-all.patch
git-nfs.patch
git-ocfs2.patch
git-serial.patch
git-scsi-block.patch
git-scsi-iscsi.patch
git-scsi-misc.patch
git-watchdog.patch
Subsystem trees
-preempt-race-in-getppid.patch -tpm_infineon-bugfix-in-pnpacpi-handling.patch -md-make-sure-resync-gets-started-when-array-starts.patch -gregkh-usb-usb-zd1201-kmalloc-size-fix.patch -export-machine_power_off-on-ppc64.patch
-usbnet-oops-fix.patch -x86_64-dont-oops-at-boot-when-empty-opteron-node-has-io.patch -git-acpi-ia64-fixes.patch
-git-acpi-ia64-fix-2.patch -git-drm-drm_agpsupport-warning-fix.patch -gregkh-i2c-i2c-hwmon-class-01.patch -gregkh-i2c-w1-changed_default_netlink_group.patch -apple-usb-touchpad-driver.patch -negative-timer-loop-with-lots-of-ipv4-peers.patch -ipw2100-cleanup-debug-prints.patch -gregkh-pci-pci-must_check-attributes.patch -git-scsi-misc-ibmvscsi-fix.patch -git-scsi-misc-ibmvscsi-fix-fix.patch -ppc32-fix-gcc4-warning-in-asm-ppc-timeh.patch -ppc64-remove-nested-feature-sections.patch -ppc64-allow-xmon=off.patch
-ppc64-four-level-pagetables.patch -ppc64-four-level-pagetables-fix.patch -ppc64-large-initrd-causes-kernel-not-to-boot.patch -i386-fix-incorrect-fp-signal-delivery.patch -x86_64-fix-numa-node-sizing-in-nr_free_zone_pages.patch -i810_audio-fix-release_region-misordering-in-error-exit-from-i810_probe.patch -race-condition-with-drivers-char-vtc-bug-in-vt_ioctlc.patch -nfs-nfs3-page-null-fill-in-a-short-read-situation.patch # wait -nfs-fix-xprt_bindresvport.patch -8250-serial-console-locking-bug-spelling-fix.patch
Merged
+non-booting-g5-fix.patch
ppc64 G5 fix
+dvb-saa7134-dvb-must-select-tda1004x.patch
DVB Kconfig fix
+tpm_infineon-bugfix-in-pnpacpi-handling.patch
TPM fix
+gregkh-driver-driver-bind-fix.patch
+gregkh-driver-driver-link-device-and-class.patch
Driver tree updates
-driver-core-fix-bus_rescan_devices-race.patch
+driver-core-fix-bus_rescan_devices-race-2.patch
Updated
+git-audit-ppc64-fix.patch
Fix git-audit.patch
+hiddev-output-reports-are-dropped-when-hidiocsreport-is.patch
Some USB fix
+git-jfs-fixup.patch
Fix rejects in git-jfs.patch
+ignore-all-debugging-info-sections-in-scripts-reference_discardedpl.patch
reference_discarded.pl tweak
+fix-minor-bug-in-sungemh.patch
Sungem fix
+netlink-log-protocol-failures.patch
Netlink debugging
+git-nfs-oops-fix.patch
Revert dud patch from git-nfs.patch
+gregkh-pci-pci-must_check-attributes.patch
PCI tree update
+fix-klist-semantics-for-lists-which-have-elements-removed.patch
klist fix for scsi usage
+git-scsi-misc-sr-fix.patch
Fix git-scsi-misc.patch
+gregkh-usb-usb-add-apple-touchpad-driver.patch
USB tree update
+watchdog-new-sbc8360-driver.patch
Watchdog driver
+ppc64-fix-sparsemem-extreme.patch
Fix sparsemem-extreme.patch
+swap-swap-unsigned-int-consistency-warning-fix.patch
cleanup
-add-swap-cache-mapping-comment.patch
-remove-stale-comment-from-swapfilec.patch
Were wrong
+arm-allow-for-arch-specific-ioremap_max_order.patch
ARM fix (controversial)
+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
hugetlb updates
+slab-consolidate-kmem_bufctl_t.patch
slab cleanup
+page-fault-patches-optional-page_lock-acquisition-in-nicety.patch
Clean up pagefault scalability patches
+ppp_mppe-add-ppp-mppe-encryption-module-author-address-change.patch
New email address
+mdio_bus_exit-in-discarded-section-textexit.patch
mdio section fix
+generic-vfs-fallback-for-security-xattrs.patch
SELinux stuff
+cpm_uart-use-schedule_timeout-instead-of-direct-call-to.patch +cpm_uart-fix-baseaddress-for-smc-1-and-2.patch
cpm_uart updates
+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
ppc32 updates
+frv-remove-export-of-strtok.patch
cleanup
+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
MIPS updates
+es7000-platform-update-i386.patch
es7k updates
+x86-add-mce-resume.patch
+pm-fix-process-freezing.patch
+pm-cleanup-sys-power-disk.patch
PM updates
+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
UML
+xtensa-replace-extern-inline-with-static-inline.patch +xtensa-delete-accidental-file.patch
xtensa arch
+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
s390 updates
+futex_wake_op-pthread_cond_signal-speedup.patch
Futex feature work
+relayfs-relayfs_remove-fix.patch
+relayfs-upgraded-read-implementation.patch
+relayfs-update-documentation-fix.patch
relayfs updates
-aio-add-enosys-into-sys_io_cancel.patch -aio-kiocb-locking-to-serialise-retry-and-cancel.patch
Dropped (I have it in a new AIO patch series but I took yet another look at all the AIO stuff and felt queasy)
+radix-tree-remove-unnecessary-indirections-and-clean-up.patch
radix-tree simplifications
+auxiliary-vector-cleanups-fix.patch
Fix auxiliary-vector-cleanups.patch
+dcdbas-add-dell-systems-management-base-driver-with-sysfs-support.patch
Dell Systems Management Base Driver
+futex-remove-duplicate-code.patch
Futex celanup
+additions-to-dataread_mostly-section.patch
More read-mostly variables
+ntp-ntp-helper-functions.patch
ntp cleanup
+blk-use-blk_queue_xxx-functions-to-set-parameters.patch
block layer cleanup
+convert-proc-devices-to-use-seq_file-interface.patch +convert-proc-devices-to-use-seq_file-interface-fix.patch
seq_file conversion
+tty-layer-buffering-revamp.patch
TTY buffering rewrite
+pipe-remove-redundant-fifo_poll-abstraction.patch
pipe cleanup
+ibm-hdaps-accelerometer-driver-with-probing.patch
HDAPS driver
+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
cleanups
+serial-console-touch-nmi-watchdog.patch
Poke the NMI watchdog when spewing to the serial console.
+optimise-64bit-unaligned-access-on-32bit-kernel.patch
Speedup
+vt-fix-possible-memory-corruption-in-complement_pos.patch
vt driver fix
+hpet-fix-drift-and-url.patch
hpet fix
+isdn_v110-warning-fix.patch
Fix a warning
+tpm-fix-tpm_atmelc-on-ich6.patch
TPM fix
+create-asm-generic-fcntlh.patch
+consildate-asm-ppc-fcntlh.patch
fcntl.h consolidation
+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
More code was dirty
+inotify-fix-event-loss-on-hardlinked-files.patch
inotify fix
+sunrpc-print-unsigned-integers-in-stats.patch
sunrpc fixlet
+open-returns-enfile-but-creates-file-anyway.patch +open-returns-enfile-but-creates-file-anyway-tidy.patch
Fix open() behaviour
+block-cfq-refcounting-fix.patch
CFQ fix
+remove-ia_attr_flags.patch
+namei-cleanup.patch
+use-get_fs_struct-in-proc.patch
Cleanups
+fix-enum-pid_directory_inos-in-proc-basec.patch
procfs fix
+remove-duplicated-code-from-proc-and-ptrace.patch +remove-duplicated-sys_open32-code-from-64bit-archs.patch
cleanups
+cifs_create-fix.patch
CIFS fix
+deprecate-openfoo-3.patch
Deprecate open("foo", 3) (old lilo's trigger this)
+fix-reboot-via-keyboard-controller-reset.patch
Make reboots work better with some keyboard controllers
+fix-dmi_check_system.patch
DMI fix
+mmc-conditional-scr-sysfs-entry.patch
MMC fix
-smsc-ircc2-pm-cleanup-do-not-close-device-when-suspending-fixes.patch
Folded into smsc-ircc2-pm-cleanup-do-not-close-device-when-suspending.patch
+kprobes-fix-handling-of-simultaneous-probe-hit-unregister.patch
kprobes fix
+pcmcia-yenta-dont-mess-with-bridge-control-register.patch +pcmcia-remove-unused-client_t.patch +pcmcia-remove-unused-vpp1-vpp2-and-vcc.patch +pcmcia-omap-cf-controller.patch +pcmcia-more-ids-for-ide_cs.patch +pcmcia-add-pcmcia-to-irq-information.patch
pcmcia/cardbus updates
+nfs-nfs3-page-null-fill-in-a-short-read-situation.patch
NFS fix
+sched-less-newidle-locking.patch
+sched-less-locking.patch
+sched-ht-optimisation.patch
CPU scheduler updates
-sched-dont-kick-alb-in-the-presence-of-pinned-task-fix.patch
Folded into sched-dont-kick-alb-in-the-presence-of-pinned-task.patch
+reiser4-fix-wundef-warnings.patch
reiser4 wranings
+v9fs-documentation-makefiles-configuration-fix-plan9port-example-in-v9fs.patch +v9fs-vfs-inode-operations-adjust-follow_link-and-put_link-to.patch +v9fs-9p-protocol-implementation-use-standard-kernel-byteswapping.patch +v9fs-9p-protocol-implementation-remove-sparse-bitwise-warnings.patch +v9fs-transport-modules-fix-a-problem-with-named-pipe-transport.patch +v9fs-transport-modules-cleanup-fd-transport.patch +v9fs-support-to-force-umount.patch +v9fs-readlink-extended-mode-check.patch +v9fs-fix-handling-of-malformed-9p-messages.patch
v9fs updates
+ide-clean-up-the-garbage-in-eighty_ninty_three.patch
IDE cleanup
+matroxfb-read-mga-pins-data-on-powerpc.patch +sisfb-update.patch +better-error-handing-in-savagefb_probe.patch +framebuffer-new-driver-for-cyberblade-i1-graphics.patch +framebuffer-bit_putcs-optimization-for-8x.patch +radeonfb-only-request-resources-we-need.patch
fbdev updates
+md-remove-old-cruft-from-md_kh-header-file.patch +md-limit-size-of-sb-read-written-to-appropriate-amount.patch +md-add-write-intent-bitmap-support-to-raid5.patch +md-write-intent-bitmap-support-for-raid6.patch +md-use-kthread-infrastructure-in-md.patch +md-ensure-bitmap_writeback_daemon-handles-shutdown-properly.patch +md-tidy-up-daemon-stop-start-code-in-md-bitmapc.patch
RAID updates
+docbook-fix-kernel-api-documentation-generation.patch +kdump-documentation-update.patch +vfs-update-documentation.patch
Documentation updates
+fuse-read-only-operations-follow_link-fix.patch
FUSE fix
All 1126 patches:
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm1/patch-list
Breaks build on PPC64
Lots of this:
In file included from fs/xfs/linux-2.6/xfs_linux.h:57,
from fs/xfs/xfs.h:35,
from fs/xfs/xfs_rtalloc.c:37:fs/xfs/xfs_arch.h:55:21: warning: "__LITTLE_ENDIAN" is not defined In file included from fs/xfs/xfs_rtalloc.c:50:fs/xfs/xfs_bmap_btree.h:65:21: warning: "__LITTLE_ENDIAN" is not defined
CC fs/xfs/xfs_acl.o
In file included from fs/xfs/linux-2.6/xfs_linux.h:57,
from fs/xfs/xfs.h:35,
from fs/xfs/xfs_acl.c:33:fs/xfs/xfs_arch.h:55:21: warning: "__LITTLE_ENDIAN" is not defined
Can't see anything obvious to cause that.
Then this:
CC drivers/char/hvc_console.odrivers/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' make[2]: *** [drivers/char/hvc_console.o] Error 1
make[1]: *** [drivers/char] Error 2
make: *** [drivers] Error 2
Presumably this:
diff -puN drivers/char/hvc_console.c~tty-layer-buffering-revamp drivers/char/hvc _console.c
--- 25/drivers/char/hvc_console.c~tty-layer-buffering-revamp Wed Aug 31 12:50
:55 2005
+++ 25-akpm/drivers/char/hvc_console.c Wed Aug 31 12:50:56 2005
@@ -597,10 +597,8 @@ static int hvc_poll(struct hvc_struct *h
/* Read data if any */
for (;;) {- int count = N_INBUF;
- if (count > (TTY_FLIPBUF_SIZE - tty->flip.count))
- count = TTY_FLIPBUF_SIZE - tty->flip.count;
-
+ count = tty_buffer_request_room(tty, N_INBUF);
+
/* If flip is full, just reschedule a later read */
if (count == 0) {
poll_mask |= HVC_POLL_READ;shouldn't be deleting the declaration of count.
and possibly the "flip removal" was incomplete (line 636) ???
Review
Quoted 5 lines · Martin J. Bligh
(quoted diff) +2
-3- if (count > (TTY_FLIPBUF_SIZE - tty->flip.count)) - count = TTY_FLIPBUF_SIZE - tty->flip.count; - + count = tty_buffer_request_room(tty, N_INBUF); +
Should be "int count = " yes
Quoted 6 lines
/* If flip is full, just reschedule a later read */ if (count == 0) { poll_mask |= HVC_POLL_READ;shouldn't be deleting the declaration of count.
and possibly the "flip removal" was incomplete (line 636) ???
Yep. You can remove the tty->flip.count test or use count, but at that point count is guaranteed to be > 0 I believe. Fixed both in my tree will push a new diff to Andre soon.
Also if you are tidying up all the 'read 64 chars and take a break' stuff should just go away. The kernel will buffer large chunks of data for you now. In the ideal case if you know the total pending space you can do
int len = tty_buffer_request_room(tty, len)
and it'll look to kmalloc a big enough buffer for you if the buffer pool isn't suitable. Even if that fails (its a hint) the tty layer will split the data across multiple smaller buffers for you when you use tty_insert_flip_*
So you should be able to just ram data at it as it comes off the hvc.
Alan
Quoted 10 lines
Quoted 5 lines
/* If flip is full, just reschedule a later read */ if (count == 0) { poll_mask |= HVC_POLL_READ;shouldn't be deleting the declaration of count. and possibly the "flip removal" was incomplete (line 636) ???
Yep. You can remove the
tty->flip.counttest or use count, but at that point count is guaranteed to be > 0 I believe. Fixed both in my tree will push a new diff to Andre soon.
There are at least a couple other spots where flip got missed, after fixing the count and flip problem mentioned these come up:
drivers/char/hvcs.c:459: error: structure has no member named flip' drivers/char/hvcs.c:472: error: structure has no member named flip'
Quoted 5 lines · Joel Schopp
There are at least a couple other spots where flip got missed, after fixing the count and flip problem mentioned these come up:
drivers/char/hvcs.c:459: error: structure has no member namedflip' drivers/char/hvcs.c:472: error: structure has no member namedflip'
Try the diff below although I suspect much of the extra logic can go away and something like
len = tty_buffer_request_root(tty, HVCS_BUFF_LEN);
if(len) {
len = hvc_get_chars(...., len);
tty_insert_flip_string(tty, buf, len);
}is better.
char/hvcs.c renamed from char/hvcs.c~ +4 -6
@@ -456,12 +456,11 @@/* remove the read masks */ hvcsd->todo_mask &= ~(HVCS_READ_MASK ); -if ((tty->flip.count +HVCS_BUFF_LEN ) <TTY_FLIPBUF_SIZE ) { +if (tty_buffer_request_room (tty,HVCS_BUFF_LEN ) >=HVCS_BUFF_LEN ) { got =hvc_get_chars (unit_address, &buf[0 ],HVCS_BUFF_LEN ); -for (i=0 ;got && i<got;i++) -tty_insert_flip_char (tty, buf[i],TTY_NORMAL ); +tty_insert_flip_string (tty, buf, got); }/* Give the TTY time to process the data we just sent. */
@@ -469,10 +468,9 @@ hvcsd->todo_mask |=HVCS_QUICK_READ ;spin_unlock_irqrestore (&hvcsd->lock, flags); -if (tty->flip.count) { -/* This is synch because tty->low_latency == 1 */ +/* This is synch because tty->low_latency == 1 */ +if (got)tty_flip_buffer_push (tty); - }if (!got) {/* Do this _after_ the flip_buffer_push */
Quoted 10 lines
Try the diff below although I suspect much of the extra logic can go away and something like
len = tty_buffer_request_root(tty, HVCS_BUFF_LEN); if(len) { len = hvc_get_chars(...., len); tty_insert_flip_string(tty, buf, len); }is better.
It's like whack a mole. 30 more now in drivers/serial/jsm/jsm_tty.c and
drivers/serial/icom.c
Quoted 2 lines · Joel Schopp
It's like whack a mole. 30 more now in
drivers/serial/jsm/jsm_tty.cand
drivers/serial/icom.c
I've been whacking moles for some time doing all those I can. the jsm_tty code needs major surgery and its bad it ever got into the kernel as most of the code is duplicating chunks of the tty layer and working around it. The jsm stuff is unintelligible and the docs dont appear to be public.
I'll take a look at icom.c now. I notice at least one bug already that should be dealt with - the existing code assumes that tty->flip.char_buf_ptr[0] is the first it inserted this time which may not be true as far as I can see. And it looks there if count was 0 so its undefined..
Assuming it means the first char this block then the following should do the trick, but really someone who knows wtf that code is trying to do needs to fix it - please review/test/let me know.
serial/icom.c renamed from serial/icom.c~ +5 -2
@@ -737,6 +737,7 @@ status =cpu_to_le16 (icom_port->statStg->rcv[rcv_buff].flags);while (status &SA_FL_RCV_DONE ) { +int first =-1 ;trace (icom_port,"FID_STATUS" , status); count =cpu_to_le16 (icom_port->statStg->rcv[rcv_buff].leLength);
@@ -751,15 +752,17 @@ icom_port->recv_buf_pci;/* Block copy all but the last byte as this may have status */ -if (count >0 ) +if (count >0 ) { + first = icon->recv_buf[offset];tty_insert_flip_string (tty, icon_port->recv_buf + offset, count -1 ); + } icount = &icom_port->uart_port.icount; icount->rx += count;/* Break detect logic */ if ((status &SA_FLAGS_FRAME_ERROR ) - && (tty->flip.char_buf_ptr[0 ] ==0x00 )) { + && first ==0 ) { status &= ~SA_FLAGS_FRAME_ERROR ; status |=SA_FLAGS_BREAK_DET ;trace (icom_port,"BREAK_DET" ,0 );
Keep whacking - obviously I don't have a PPC64 (*and please don't send me one*)
Alan
tty patches in 2.6.13-mm3 (was Re: 2.6.13-mm1)
Quoted 3 lines · Alan Cox
Quoted 2 lines · Joel Schopp
It's like whack a mole. 30 more now in
drivers/serial/jsm/jsm_tty.cand
drivers/serial/icom.c
Hi,
I'm not sure whether these are going in through some other channel, but I notice neither the Alan's hvcs.c or icom.c patches are in 2.6.13-mm3. In addition, hvc_console.c needs yet another...
Quoted 2 lines
Keep whacking - obviously I don't have a PPC64 (*and please don't send me one*)
I do have one, but have been gone for awhile. But am willing and eager to test (and get it working again :)
thanks,
-serge
drivers/char/hvc_console.c +2 -2
@@ -597,7 +597,7 @@ static int hvc_poll(struct hvc_struct *h/* Read data if any */ for (;;) { - count =tty_buffer_request_room (tty,N_INBUF ); +int count =tty_buffer_request_room (tty,N_INBUF );/* If flip is full, just reschedule a later read */ if (count ==0 ) {
@@ -633,7 +633,7 @@ static int hvc_poll(struct hvc_struct *htty_insert_flip_char (tty, buf[i],0 ); } -if (tty->flip.count) +if (tty_buffer_request_room (tty,1 ))tty_schedule_flip (tty);/*
Joel Schopp
Acked-byRe: tty patches in 2.6.13-mm3 (was Re: 2.6.13-mm1)
Quoted 3 lines
I'm not sure whether these are going in through some other channel, but I notice neither the Alan's
hvcs.coricom.cpatches are in 2.6.13-mm3. In addition,hvc_console.cneeds yet another...
Yeah, there is still a whole lot broken in -mm. Your patch below is a good start though.
Quoted 30 lines
Signed-off-by Serge Hallyn <[email protected]>Index:
linux-2.6.12/drivers/char/hvc_console.c
===================================================================drivers/char/hvc_console.c +2
-3@@ -597,7 +597,7 @@ static int hvc_poll(struct hvc_struct *h/* Read data if any */ for (;;) { - count =tty_buffer_request_room (tty,N_INBUF ); +int count =tty_buffer_request_room (tty,N_INBUF );/* If flip is full, just reschedule a later read */ if (count ==0 ) {@@ -633,7 +633,7 @@ static int hvc_poll(struct hvc_struct *htty_insert_flip_char (tty, buf[i],0 ); } -if (tty->flip.count) +if (tty_buffer_request_room (tty,1 ))tty_schedule_flip (tty);/* -To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Hmm, this patch itself seems to have a few typos, but so does the 2.6.13-mm{1,2,3}.
Quoted 11 lines · Alan Cox
serial/icom.c renamed from serial/icom.c~ +1
-0@@ -737,6 +737,7 @@ status =cpu_to_le16 (icom_port->statStg->rcv[rcv_buff].flags);while (status &SA_FL_RCV_DONE ) { +int first =-1 ;trace (icom_port,"FID_STATUS" , status); count =cpu_to_le16 (icom_port->statStg->rcv[rcv_buff].leLength);
serial/icom.c line 756
Quoted 6 lines
serial/icom.c renamed from serial/icom.c~ +2
-1@@ -751,15 +752,17 @@ icom_port->recv_buf_pci;/* Block copy all but the last byte as this may have status */ -if (count >0 ) +if (count >0 ) { + first = icon->recv_buf[offset];
s/icon/icom_port/ ?
serial/icom.c line 757
Quoted 1 line
serial/icom.c renamed from serial/icom.c~ +0
-0tty_insert_flip_string (tty, icon_port->recv_buf + offset, count -1 );
s/icon_port/icom_port/ ?
serial/icom.c line 768
Quoted 12 lines
serial/icom.c renamed from serial/icom.c~ +2
-1+ } icount = &icom_port->uart_port.icount; icount->rx += count;/* Break detect logic */ if ((status &SA_FLAGS_FRAME_ERROR ) - && (tty->flip.char_buf_ptr[0 ] ==0x00 )) { + && first ==0 ) { status &= ~SA_FLAGS_FRAME_ERROR ; status |=SA_FLAGS_BREAK_DET ;trace (icom_port,"BREAK_DET" ,0 );
And in 2.6.13-mm3, we have:
@@ -798,33 +792,26 @@ static void recv_interrupt(u16 port_int_
status &= icom_port->read_status_mask;
if (status & SA_FLAGS_BREAK_DET) {
- *tty->flip.flag_buf_ptr = TTY_BREAK;
+ flag = TTY_BREAK;
} else if (status & SA_FLAGS_PARITY_ERROR) {
trace(icom_port, "PARITY_ERROR", 0);
- *tty->flip.flag_buf_ptr = TTY_PARITY;
+ flag = TTY_PARITY;
} else if (status & SA_FLAGS_FRAME_ERROR)
- *tty->flip.flag_buf_ptr = TTY_FRAME;
+ flag = TTY_FRAME;
- if (status & SA_FLAGS_OVERRUN) {
- /*
- * Overrun is special, since it's
- * reported immediately, and doesn't
- * affect the current character
- */
- if (tty->flip.count < TTY_FLIPBUF_SIZE) {
- tty->flip.count++;
- tty->flip.flag_buf_ptr++;
- tty->flip.char_buf_ptr++;
- *tty->flip.flag_buf_ptr = TTY_OVERRUN;
- }
- }
}
- tty->flip.flag_buf_ptr++;
- tty->flip.char_buf_ptr++;
- tty->flip.count++;
- ignore_char:
- icom_port->statStg->rcv[rcv_buff].flags = 0;
+ tty_insert_flip_char(tty, icon_port->recv_buf + offset + count - 1, flag); ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Is this meant to be
tty_insert_flip_char(tty, *(icon_port->recv_buf + offset + count - 1), flag);
?
+
+ if (status & SA_FLAGS_OVERRUN)
+ /*
+ * Overrun is special, since it's
+ * reported immediately, and doesn't
+ * affect the current character
+ */
+ tty_insert_flip_char(tty, 0, TTY_OVERRUN);
+ignore_char:
+ icom_port->statStg->rcv[rcv_buff].flags = 0;
icom_port->statStg->rcv[rcv_buff].leLength = 0; icom_port->statStg->rcv[rcv_buff].WorkingLength =
(unsigned short int) cpu_to_le16(RCV_BUFF_SZ);
Again, sorry I didn't catch this back in -mm1 or -mm2...
thanks,
-serge
Quoted 11 lines · Martin J. Bligh
...
Lots of this:In file included from
fs/xfs/linux-2.6/xfs_linux.h:57,from fs/xfs/xfs.h:35, from fs/xfs/xfs_rtalloc.c:37:
fs/xfs/xfs_arch.h:55:21: warning: "__LITTLE_ENDIAN" is not defined In file included fromfs/xfs/xfs_rtalloc.c:50:fs/xfs/xfs_bmap_btree.h:65:21: warning: "__LITTLE_ENDIAN" is not defined
CCfs/xfs/xfs_acl.o
In file included fromfs/xfs/linux-2.6/xfs_linux.h:57,
Quoted 6 lines
from fs/xfs/xfs.h:35, from fs/xfs/xfs_acl.c:33:
fs/xfs/xfs_arch.h:55:21: warning: "__LITTLE_ENDIAN" is not definedCan't see anything obvious to cause that.
...
They are there since we added -Wundef to the CFLAGS several -mm kernels ago.
cu
Adrian
Signature · 4 lines
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon SeedPATCH 1/1 mips: remove typedef from struct flock
In reply to Andrew MortonHi,
This patch has fixed the following warning about MIPS. Please apply.
Yoichi
CC arch/mips/kernel/offset.s
In file included from include/linux/fcntl.h:4,
from include/linux/fs.h:689,
from include/linux/mm.h:15,
from arch/mips/kernel/offset.c:15:include/asm/fcntl.h:53: warning: useless keyword or type name in empty declaration
include/asm-mips/fcntl.h +1 -1
@@ -42,7 +42,7 @@ #ifndef __mips64 -typedef struct flock { +struct flock { short l_type; short l_whence; __kernel_off_t l_start;
Quoted 2 lines · Andrew Morton
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13 -mm1/
When I switch on my external harddisk, which is connected through usb, the kernel hangs. First time I did that at bootup there were a lot of backtraces printed on the screen but they did not find the way in the logfile :/ Now I switched the drive on while running and everything freezes after those messages:
usb 1-2.2: new high speed USB device using ehci_hcd and address 3 scsi2 : SCSI emulation for USB Mass Storage devices
usb-storage: device found at 3
usb-storage: waiting for device to settle before scanning
Vendor: ST325082 Model: 3A Rev: 3.02
Type: Direct-Access ANSI SCSI revision: 00
SCSI device sda: 488397168 512-byte hdwr sectors (250059 MB)
sda: assuming drive cache: write through
SCSI device sda: 488397168 512-byte hdwr sectors (250059 MB)
sda: assuming drive cache: write through
dominik
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
"Dominik" == Dominik Karall <[email protected]> writes:
Dominik> When I switch on my external harddisk, which is connected Dominik> through usb, the kernel hangs. First time I did that at Dominik> bootup there were a lot of backtraces printed on the screen Dominik> but they did not find the way in the logfile :/ Now I Dominik> switched the drive on while running and everything freezes Dominik> after those messages:
Dominik> usb 1-2.2: new high speed USB device using ehci_hcd and address 3 Dominik> scsi2 : SCSI emulation for USB Mass Storage devices Dominik> usb-storage: device found at 3
Dominik> usb-storage: waiting for device to settle before scanning
Dominik> Vendor: ST325082 Model: 3A Rev: 3.02 Dominik> Type: Direct-Access ANSI SCSI revision: 00
Dominik> SCSI device sda: 488397168 512-byte hdwr sectors (250059 MB) Dominik> sda: assuming drive cache: write through
Dominik> SCSI device sda: 488397168 512-byte hdwr sectors (250059 MB) Dominik> sda: assuming drive cache: write through
Have you updated the firmware on the USB enclosure? I have one using the Prolific chipset for both USB/Firewire and it was crappy until I upgraded the firmware on there. It made all the difference.
Also, can you use this USB enclosure on Windows or another computer? And which kernel version are you running? It's not clear if your on 2.6.13-mm1 or some other version.
More details would be good too, such as:
lsusb
cat /proc/version
What happens if you unplug the drive when the system hangs? Does it recover? And try powering up the enclosure without it being hooked to anything, then once 30 seconds have passed, hook it upto the Linux box and see what happens then. Maybe the power on stuff is doing strange things.
John
Quoted 20 lines · John Stoffel
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
"Dominik" == Dominik Karall <[email protected]> writes:
Dominik> When I switch on my external harddisk, which is connected Dominik> through usb, the kernel hangs. First time I did that at Dominik> bootup there were a lot of backtraces printed on the screen Dominik> but they did not find the way in the logfile :/ Now I Dominik> switched the drive on while running and everything freezes Dominik> after those messages:
Dominik> usb 1-2.2: new high speed USB device using
ehci_hcdand address 3 Dominik> scsi2 : SCSI emulation for USB Mass Storage devices Dominik> usb-storage: device found at 3
Dominik> usb-storage: waiting for device to settle before scanningDominik> Vendor: ST325082 Model: 3A Rev: 3.02 Dominik> Type: Direct-Access ANSI SCSI revision:00 Dominik> SCSI device sda: 488397168 512-byte hdwr sectors (250059 MB) Dominik> sda: assuming drive cache: write through
Dominik> SCSI device sda: 488397168 512-byte hdwr sectors (250059 MB) Dominik> sda: assuming drive cache: write through
Quoted 7 lines
Have you updated the firmware on the USB enclosure? I have one using the Prolific chipset for both USB/Firewire and it was crappy until I upgraded the firmware on there. It made all the difference.
Also, can you use this USB enclosure on Windows or another computer? And which kernel version are you running? It's not clear if your on 2.6.13-mm1 or some other version.
2.6.13-mm1, as mentioned in subject.
The external hdd worked with 2.6.13-rc6-mm1 and 2.6.13-ck1, which were the last versions I ran. Didn't test 2.6.13-rc6-mm2.
Quoted 3 lines
More details would be good too, such as:
lsusb
Bus 001 Device 004: ID 0840:009c Argosy Research, Inc. Device Descriptor:
bLength 18
bDescriptorType 1
bcdUSB 2.00
bDeviceClass 0 (Defined at Interface level) bDeviceSubClass 0 bDeviceProtocol 0 bMaxPacketSize0 64 idVendor 0x0840 Argosy Research, Inc. idProduct 0x009c bcdDevice 0.01 iManufacturer 1 Generic iProduct 2 USB 2.0 Mass Storage Device iSerial 3 009C0000000049BD bNumConfigurations 1
Configuration Descriptor:
bLength 9
bDescriptorType 2
wTotalLength 32
bNumInterfaces 1
bConfigurationValue 1
iConfiguration 0
bmAttributes 0xc0
Self Powered
MaxPower 100mA
Interface Descriptor:
bLength 9
bDescriptorType 4
bInterfaceNumber 0
bAlternateSetting 0
bNumEndpoints 2
bInterfaceClass 8 Mass Storage
bInterfaceSubClass 6 SCSI
bInterfaceProtocol 80 Bulk (Zip)
iInterface 0
Endpoint Descriptor:
bLength 7
bDescriptorType 5
bEndpointAddress 0x01 EP 1 OUT
bmAttributes 2
Transfer Type Bulk
Synch Type None
Usage Type Data
wMaxPacketSize 0x0200 1x 512 bytes
bInterval 0
Endpoint Descriptor:
bLength 7
bDescriptorType 5
bEndpointAddress 0x82 EP 2 IN
bmAttributes 2
Transfer Type Bulk
Synch Type None
Usage Type Data
wMaxPacketSize 0x0200 1x 512 bytes
bInterval 0
Device Qualifier (for other device speed):
bLength 10
bDescriptorType 6
bcdUSB 2.00
bDeviceClass 0 (Defined at Interface level) bDeviceSubClass 0 bDeviceProtocol 0
bMaxPacketSize0 64
bNumConfigurations 1
dominik
Dominik,
So what is the chipset inside the enclosure? Looking at your output, the 'Argosy' stuff doesn't tell me anything. You might have to open up the case to look in there to find more details.
Again, check with your vendor and see if there is newer firmware. And have you powered up the device without having it plugged into the system, then plug it in? What happens then?
John
Quoted 9 lines · John Stoffel
Dominik,
So what is the chipset inside the enclosure? Looking at your output, the 'Argosy' stuff doesn't tell me anything. You might have to open up the case to look in there to find more details.
Again, check with your vendor and see if there is newer firmware. And have you powered up the device without having it plugged into the system, then plug it in? What happens then?
Why should I check for newer firmware!? I don't understand that point of view. The device works without any problems with 2.6.13-ck1 as 2.6.13-rc6-mm1 and before kernels. So there is no need to check the firmware imho.
I don't think that it makes any difference if I power up first or plug in first. The device is recognized when I power it on, so it would be the same when I power it on and connect after that imho.
I will try to get the backtraces from the kernel, this should make debugging easier.
greets,
dominik
Dominik> Why should I check for newer firmware!? I don't understand Dominik> that point of view. The device works without any problems Dominik> with 2.6.13-ck1 as 2.6.13-rc6-mm1 and before kernels. So Dominik> there is no need to check the firmware imho.
That's on point of view. In my experience, it simplifies debugging to make sure the unit is at the latest firmware, since bugs in the enclosure's IDE driver could be causing the problem. As I said before, when my enclosure was upgraded, all my problems went away.
So that's why I was asking you to make sure your firmware was upto date. Is that so hard to understand?
Dominik> I don't think that it makes any difference if I power up Dominik> first or plug in first. The device is recognized when I power Dominik> it on, so it would be the same when I power it on and connect Dominik> after that imho.
Sure it can make a difference. If the enclosure puts out crap signals on the USB bus when it's powered up, and the older versions of the Linux kernel dealt with them because of an oversight, but now we're closer to the USB specs... it could the issue.
In any case, it's a *simple* test to do, unless you're not physically at the system with this device, or if you can't be bothered to:
1. unplug enclosure from USB.
2. power it off.
3. power it on.
4. wait 30 seconds.
5. plug in the USB cable.
6. what happens?
That tells us useful stuff.
Dominik> I will try to get the backtraces from the kernel, this should Dominik> make debugging easier.
That will help people debug this for sure.
In any case, I'm not going to be much help from here on.
PATCH 1/1 : struct dentry : place d_hash close to d_parent and d_name to speedup lookups
In reply to Andrew Mortondentry cache uses sophisticated RCU technology (and prefetching if available) but touches 2 cache lines per dentry during hlist lookup.
This patch moves d_hash in the same cache line than d_parent and d_name fields so that :
1) One cache line is needed instead of two.
2) the hlist_for_each_rcu() prefetching has a chance to bring all the needed data in advance, not only the part that includes d_hash.next.
I also changed one old comment that was wrong for 64bits.
A further optimisation would be to separate dentry in two parts, one that is mostly read, and one writen (d_count/d_lock) to avoid false sharing on SMP/NUMA but this would need different field placement depending on 32bits or 64bits platform.
PATCH 1/1 : struct dentry : place d_hash close to d_parent and d_name to speedup lookups
dentry cache uses sophisticated RCU technology (and prefetching if available) but touches 2 cache lines per dentry during hlist lookup.
This patch moves d_hash in the same cache line than d_parent and d_name fields so that :
1) One cache line is needed instead of two.
2) the hlist_for_each_rcu() prefetching has a chance to bring all the needed data in advance, not only the part that includes d_hash.next.
I also changed one old comment that was wrong for 64bits.
A further optimisation would be to separate dentry in two parts, one that is mostly read, and one writen (d_count/d_lock) to avoid false sharing on SMP/NUMA but this would need different field placement depending on 32bits or 64bits platform.
with the patch this time :)
include/linux/dcache.h +2 -2
@@ -88,8 +88,9 @@* negative */ /* * The next three fields are touched by __d_lookup. Place them here -* so they all fit in a 16-byte range, with 16-byte alignment. +* so they all fit in a cache line. */ +struct hlist_node d_hash;/* lookup hash list */ struct dentry *d_parent;/* parent directory */ struct qstr d_name;
@@ -103,7 +104,6 @@void *d_fsdata;/* fs-specific data */ struct rcu_head d_rcu;struct dcookie_struct *d_cookie;/* cookie, if any */ -struct hlist_node d_hash;/* lookup hash list */ int d_mounted;unsigned char d_iname[DNAME_INLINE_LEN_MIN ];/* small names */ };
Re: [PATCH] : struct dentry : place d_hash close to d_parent and d_name to speedup lookups
Quoted 10 lines · Eric Dumazet
dentry cache uses sophisticated RCU technology (and prefetching if available) but touches 2 cache lines per dentry during hlist lookup.
This patch moves
d_hashin the same cache line thand_parentandd_namefields so that :1) One cache line is needed instead of two.
2) thehlist_for_each_rcu()prefetching has a chance to bring all the needed data in advance, not only the part that includes d_hash.next.
Quoted 6 lines
I also changed one old comment that was wrong for 64bits.
A further optimisation would be to separate dentry in two parts, one that is mostly read, and one writen (d_count/d_lock) to avoid false sharing on SMP/NUMA but this would need different field placement depending on 32bits or 64bits platform.
Do you have performance numbers that show the benefits ? In the past, I did try some optimizations like this but found no demonstrable benefits. If it ain't broken .....
Thanks
Dipankar
Re: [PATCH] : struct dentry : place d_hash close to d_parent and d_name to speedup lookups
Quoted 9 lines · Dipankar Sarma
Quoted 7 lines · Eric Dumazet
dentry cache uses sophisticated RCU technology (and prefetching if available) but touches 2 cache lines per dentry during hlist lookup.
This patch moves
d_hashin the same cache line thand_parentandd_namefields so that :1) One cache line is needed instead of two.
2) thehlist_for_each_rcu()prefetching has a chance to bring all the needed data in advance, not only the part that includes d_hash.next.
Quoted 9 lines
Quoted 3 lines
I also changed one old comment that was wrong for 64bits.
A further optimisation would be to separate dentry in two parts, one that is mostly read, and one writen (d_count/d_lock) to avoid false sharing on SMP/NUMA but this would need different field placement depending on 32bits or 64bits platform.
Do you have performance numbers that show the benefits ? In the past, I did try some optimizations like this but found no demonstrable benefits. If it ain't broken .....
You mean... a microbenchmark ? :)
Well, the dentry cache hash table is so large that it's difficult to see the benefit, because chain length average is < 1 on most setups. I suspect the sizing strategy of this hash table was made a long time ago...
Reducing the number of cache lines touched per dentry in lookup could let us shrink the hash table and save some ram.
Facts now :
------------
On a 512MB ia32 machine, the default sizing of the hash table gives 131072 slots (0.5 MB of ram). So it's quite hard to populate the cache without eating all ram.
To see a benefit, I had to force a smaller hash size (appending in kernel params "dhash_entries=8191")
On this 512MB PIII 866MHz machine (L1_CACHE_SIZE=32), with 563887 files or directories the results are :
Remount the disk with noatime to make sure no disk access will be done
# mount -o remount /dev/hda2 / -o noatime
Populate the caches (dentry and disk buffers for directories)
# find / -name foofoo >/dev/null
# grep dentry_cache /proc/slabinfo dentry_cache 131104 131370 136 29 1 : tunables 120 60 0 : slabdata 4530 4530 0
So average chain length is now 16
then time new searches using data in cache
time / -name foofoo >/dev/null
# time find / -name foofoo
real 0m3.077s user 0m1.724s sys 0m1.344s # time find / -name foofoo
real 0m3.077s user 0m1.728s sys 0m1.348s # time find / -name foofoo
real 0m3.084s user 0m1.784s sys 0m1.276s # time find / -name foofoo
real 0m3.108s user 0m1.716s sys 0m1.348s # time find / -name foofoo
real 0m3.061s user 0m1.672s sys 0m1.384s # time find / -name foofoo
real 0m3.067s user 0m1.644s sys 0m1.424s # time find / -name foofoo
real 0m3.061s user 0m1.804s sys 0m1.252s
kernel 2.6.13 build time
------------------------
mount -o remount /dev/hda2 /
make clean
time make bzImage
real 12m21.755s user 11m5.070s sys 1m0.756s
------------------------------------------------------------------
Results with a patched kernel
# time find / -name foofoo
real 0m3.017s user 0m1.700s sys 0m1.316s # time find / -name foofoo
real 0m3.018s user 0m1.592s sys 0m1.424s # time find / -name foofoo
real 0m3.017s user 0m1.624s sys 0m1.396s # time find / -name foofoo
real 0m3.019s user 0m1.680s sys 0m1.340s # time find / -name foofoo
real 0m3.021s user 0m1.676s sys 0m1.344s # time find / -name foofoo
real 0m3.022s user 0m1.644s sys 0m1.376s # time find / -name foofoo
real 0m3.021s user 0m1.620s sys 0m1.404s
kernel 2.6.13 build time
------------------------
mount -o remount /dev/hda2 /
make clean
time make bzImage
real 12m8.814s user 11m2.129s sys 0m52.187s
Conclusions:
-----------
1) There is a clear benefit, even without a microbenchmark :)
2) As sizeof(struct dentry)=136 on ia32 UP (140 on SMP), and dentry slab is not using SLAB_HWCACHE_ALIGN, I suspect lot of dentries are not aligned on a cache line boundary, so maybe reducing DNAME_INLINE_LEN_MIN (or increasing it) so that sizeof(struct dentry)%L1_CACHE_BYTES = {0 or (L1_CACHE_BYTES/2)} would even give better results.
Thank you
Eric
2.6.13-mm1 - drivers/isdn/i4l/isdn_tty broken
In reply to Andrew Mortonhi Andrew,
hi Alan,
updating the kernel26mm package for archlinux to 2.6.13-mm1 i found the isdn-tty to be broken:
CC [M] drivers/isdn/i4l/isdn_tty.odrivers/isdn/i4l/isdn_tty.c: In function 'isdn_tty_try_read':drivers/isdn/i4l/isdn_tty.c:71: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:86: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:86: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:88: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:89: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:90: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:90: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:90: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:90: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:91: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:96: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:97: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c: In function 'isdn_tty_readmodem':drivers/isdn/i4l/isdn_tty.c:134: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:137: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:138: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:141: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:141: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:141: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:141: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:142: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:143: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:144: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:146: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c: In function 'isdn_tty_at_cout':drivers/isdn/i4l/isdn_tty.c:2372: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:2403: error: 'struct tty_struct' has no member named 'flip' drivers/isdn/i4l/isdn_tty.c:2418: error: 'struct tty_struct' has no member named 'flip' make[3]: *** [drivers/isdn/i4l/isdn_tty.o] Error 1
make[2]: *** [drivers/isdn/i4l] Error 2
make[1]: *** [drivers/isdn] Error 2
make: *** [drivers] Error 2
greetings,
Damir
Le Thursday 01 September 2005 12:55, Andrew Morton a écrit :
| ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2. |6.13-mm1/ | | - Included Alan's big tty layer buffering rewrite. This breaks the | build on lots of more obscure character device drivers. Patches | welcome (please cc Alan).
Signature · 2 lines
Gentlemen, I want you to know that I am not always right, but I am never wrong. -Samuel Goldwyn
Re: 2.6.13-mm1 - drivers/isdn/i4l/isdn_tty broken
Quoted 4 lines · Damir Perisa
hi Andrew,
hi Alan,updating the kernel26mm package for archlinux to 2.6.13-mm1 i found the isdn-tty to be broken:
isdn-tty needs updating and is complex to update. I've been talking with Karsten Keil about it by email and he will take a look soon (once I send him the diff ...)
2.6.13-mm1 - drivers/serial/jsm/jsm_tty broken too
In reply to Damir Perisai disabled the isdn subsystem temporarely and tried to recompile finding out that jsm-tty is affected too:
CC [M] drivers/serial/jsm/jsm_tty.odrivers/serial/jsm/jsm_tty.c: In function 'jsm_input':drivers/serial/jsm/jsm_tty.c:592: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:619: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:620: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:623: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:624: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:667: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:668: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:669: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:670: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:671: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:672: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:674: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:677: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:677: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:677: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:677: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:680: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:681: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:682: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:691: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:692: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:693: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:694: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:695: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:696: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:698: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:701: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:701: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:701: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:701: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:742: error: 'struct tty_struct' has no member named 'flip' drivers/serial/jsm/jsm_tty.c:742: error: 'struct tty_struct' has no member named 'flip' make[3]: *** [drivers/serial/jsm/jsm_tty.o] Error 1
make[2]: *** [drivers/serial/jsm] Error 2
make[1]: *** [drivers/serial] Error 2
make: *** [drivers] Error 2
hope that this tty breaks will be fixed in mm2
greetings,
Damir
Signature · 1 line
It would save me a lot of time if you just gave up and went mad now.
Re: 2.6.13-mm1: PCMCIA problem
In reply to Andrew MortonQuoted 2 lines · Andrew Morton
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm1/
I cannot start PCMCIA on x86-64 SuSE 9.3 on Asus L5D. Apparently, the following command:
sh -c modprobe --ignore-install firmware_class; echo 30 > /sys/class/firmware/timeout
loops forever with almost 100% of the time spent in the kernel.
AFAICS, 2.6.13-rc6-mm2 is also affected, but the mainline kernels are not.
Greetings,
Rafael
Signature · 3 lines
- 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-mm1: PCMCIA problem
Quoted 5 lines · Rafael J. Wysocki
Quoted 2 lines · Andrew Morton
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm1/
Quoted 8 lines
I cannot start PCMCIA on x86-64 SuSE 9.3 on Asus L5D. Apparently, the following command:
sh -c modprobe --ignore-install
firmware_class; echo 30 >/sys/class/firmware/timeoutloops forever with almost 100% of the time spent in the kernel.
AFAICS, 2.6.13-rc6-mm2 is also affected, but the mainline kernels are not.
OK. There are no notable firmware changes in there. While it's stuck
could you generate a kernel profile? I do:
readprofile -r
sleep 5
readprofile -n -v -m /boot/System.map | sort -n +2 | tail -40
Re: 2.6.13-mm1: PCMCIA problem
Quoted 13 lines · Andrew Morton
Quoted 12 lines · Rafael J. Wysocki
Quoted 2 lines · Andrew Morton
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm1/
I cannot start PCMCIA on x86-64 SuSE 9.3 on Asus L5D. Apparently, the following command:
sh -c modprobe --ignore-install
firmware_class; echo 30 >/sys/class/firmware/timeoutloops forever with almost 100% of the time spent in the kernel.
Quoted 8 lines
Quoted 1 line
AFAICS, 2.6.13-rc6-mm2 is also affected, but the mainline kernels are not.
OK. There are no notable firmware changes in there. While it's stuck
could you generate a kernel profile? I do:readprofile -r
sleep 5
readprofile -n -v -m/boot/System.map| sort -n +2 | tail -40
OK
I ran "rcpcmcia start" in one vt and the above in another. The result was:
20 *unknown* ffffffff802417f0 __memset 1 0.0052 ffffffff80242850 _raw_spin_lock 1 0.0031 ffffffff8026a44c acpi_ns_get_next_node 1 0.0118 ffffffff802ed040 sock_kfree_s 1 0.0156 ffffffff8010eb38 sysret_check 2 0.0241 ffffffff8010eb33 ret_from_sys_call 4 0.8000 ffffffff80240cc0 copy_user_generic 5 0.0168 ffffffff80221870 dummy_file_permission 7 0.4375 ffffffff8023f5e0 simple_strtol 14 0.2917 ffffffff80240c90 copy_from_user 15 0.3125 ffffffff802ac800 class_attr_store 17 0.3542 ffffffff8017df70 rw_verify_area 22 0.1719 ffffffff8023f500 simple_strtoul 24 0.1071 ffffffff8017ead0 sys_write 41 0.2847 ffffffff80240dea copy_user_generic_c 82 2.1579 ffffffff8017f1d0 fget_light 108 0.4500 ffffffff8010eab0 system_call 189 1.4427 ffffffff8017e890 vfs_write 220 0.5288 ffffffff801bfc40 sysfs_write_file 308 0.8370 0000000000000000 total 1062 0.0004
After I had stopped the (looping) "rcpcmcia start", I got:
1243 *unknown*
ffffffff80240dea copy_user_generic_c 1 0.0263 ffffffff802417f0 __memset 1 0.0052 0000000000000000 total 2 0.0000
Greetings,
Rafael
Signature · 3 lines
- 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-mm1: PCMCIA problem
Quoted 14 lines · Rafael J. Wysocki
Quoted 13 lines · Andrew Morton
Quoted 12 lines · Rafael J. Wysocki
Quoted 2 lines · Andrew Morton
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm1/
I cannot start PCMCIA on x86-64 SuSE 9.3 on Asus L5D. Apparently, the following command:
sh -c modprobe --ignore-install
firmware_class; echo 30 >/sys/class/firmware/timeoutloops forever with almost 100% of the time spent in the kernel.
Quoted 8 lines
Quoted 8 lines
Quoted 1 line
AFAICS, 2.6.13-rc6-mm2 is also affected, but the mainline kernels are not.
OK. There are no notable firmware changes in there. While it's stuck
could you generate a kernel profile? I do:readprofile -r
sleep 5
readprofile -n -v -m/boot/System.map| sort -n +2 | tail -40
]--snip--[
One more piece of information. This is the one that loops:
echo 30 > /sys/class/firmware/timeout
and the related profile is:
16 *unknown* ffffffff8010eb38 sysret_check 1 0.0120 ffffffff80240b40 clear_page 1 0.0175 ffffffff80356397 bad_gs 1 0.0001 ffffffff80240cc0 copy_user_generic 2 0.0067 ffffffff8010eb33 ret_from_sys_call 6 1.2000 ffffffff80221870 dummy_file_permission 7 0.4375 ffffffff80240c90 copy_from_user 9 0.1875 ffffffff8023f5e0 simple_strtol 11 0.2292 ffffffff8023f500 simple_strtoul 17 0.0759 ffffffff802ac800 class_attr_store 18 0.3750 ffffffff801bfc40 sysfs_write_file 61 0.1658 ffffffff8017ead0 sys_write 78 0.5417 ffffffff8017df70 rw_verify_area 143 1.1172 ffffffff80240dea copy_user_generic_c 144 3.7895 ffffffff8010eab0 system_call 163 1.2443 ffffffff8017f1d0 fget_light 184 0.7667 ffffffff8017e890 vfs_write 209 0.5024 0000000000000000 total 1055 0.0004
Greetings,
Rafael
Signature · 3 lines
- 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-mm1: PCMCIA problem
Hi!
Quoted 20 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines · Andrew Morton
Quoted 2 lines
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm1/
I cannot start PCMCIA on x86-64 SuSE 9.3 on Asus L5D. Apparently, the following command:
sh -c modprobe --ignore-install
firmware_class; echo 30 >/sys/class/firmware/timeoutloops forever with almost 100% of the time spent in the kernel.
Quoted 8 lines
Quoted 8 lines
Quoted 1 line
AFAICS, 2.6.13-rc6-mm2 is also affected, but the mainline kernels are not.
OK. There are no notable firmware changes in there. While it's stuck
could you generate a kernel profile? I do:readprofile -r
sleep 5
readprofile -n -v -m/boot/System.map| sort -n +2 | tail -40
Quoted 5 lines
]--snip--[
One more piece of information. This is the one that loops:
echo 30 >
/sys/class/firmware/timeout
Try echo -n ...
Signature · 1 line
if you have sharp zaurus hardware you don't need... you know my address
Re: 2.6.13-mm1: PCMCIA problem
Quoted 22 lines · Pavel Machek
Hi!
Quoted 20 lines
Quoted 19 lines
Quoted 19 lines
Quoted 12 lines · Andrew Morton
Quoted 2 lines
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm1/
I cannot start PCMCIA on x86-64 SuSE 9.3 on Asus L5D. Apparently, the following command:
sh -c modprobe --ignore-install
firmware_class; echo 30 >/sys/class/firmware/timeoutloops forever with almost 100% of the time spent in the kernel.
AFAICS, 2.6.13-rc6-mm2 is also affected, but the mainline kernels are not.
OK. There are no notable firmware changes in there. While it's stuck
could you generate a kernel profile? I do:readprofile -r
sleep 5
readprofile -n -v -m/boot/System.map| sort -n +2 | tail -40
Quoted 7 lines
Quoted 5 lines
]--snip--[
One more piece of information. This is the one that loops:
echo 30 >
/sys/class/firmware/timeoutTry echo -n ...
Well that helps but it means the SuSE's scripts have to be changed to work with 2.6.13-mm1. Specifically "/etc/modprobe.d/firmware". Is that intentional or not?
Rafael
Signature · 3 lines
- 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-mm1: PCMCIA problem
In reply to Pavel MachekQuoted 6 lines · Pavel Machek
Quoted 4 lines
One more piece of information. This is the one that loops:
echo 30 >
/sys/class/firmware/timeoutTry echo -n ...
Or revert gregkh-driver-sysfs-strip_leading_trailing_whitespace.patch. Obviously if you write 30\n and the write returns 2 then the shell will then try to write the \n. That returns zero and the shell tries again, ad infinitum.
Rant. It took me two full days to weed out and fix all the crap people sent me to get -mm1 into a state where it vaguely compiled and booted. And it's untested nonsense like this which wrecks the whole effort for many testers.
I suppose this is as good as anything....
From: Andrew Morton <[email protected]>
---
fs/sysfs/file.c | 8 +++++--- 1 files changed, 5 insertions(+), 3 deletions(-)
fs/sysfs/file.c renamed from fs/sysfs/file.c~gregkh-driver-sysfs-strip_leading_trailing_whitespace-fix +5 -3
@@ -202,13 +202,14 @@ fill_write_buffer(struct sysfs_buffer ** passing the buffer that we acquired in fill_write_buffer(). */ -static int -flush_write_buffer (struct dentry * dentry,struct sysfs_buffer * buffer,size_t count) +static int flush_write_buffer (struct dentry *dentry, +struct sysfs_buffer *buffer,size_t count_in) {struct attribute * attr =to_attr (dentry);struct kobject * kobj =to_kobj (dentry->d_parent);struct sysfs_ops * ops = buffer->ops;char *x; +size_t count = count_in;/* locate trailing white space */ while ((count >0 ) &&isspace (buffer->page[count -1 ]))
@@ -224,7 +225,8 @@ flush_write_buffer(struct dentry * dentr/* terminate the string */ x[count] ='\0' ; -return ops->store (kobj, attr, x, count); + ops->store (kobj, attr, x, count); +return count_in; }
_
Re: 2.6.13-mm1: swsusp problem (was: PCMCIA problem)
Quoted 13 lines · Andrew Morton
Quoted 6 lines · Pavel Machek
Quoted 4 lines
One more piece of information. This is the one that loops:
echo 30 >
/sys/class/firmware/timeoutTry echo -n ...
Or revert gregkh-driver-sysfs-strip_leading_trailing_whitespace.patch. Obviously if you write 30\n and the write returns 2 then the shell will then try to write the \n. That returns zero and the shell tries again, ad infinitum.
Quoted 6 lines
Rant. It took me two full days to weed out and fix all the crap people sent me to get -mm1 into a state where it vaguely compiled and booted. And it's untested nonsense like this which wrecks the whole effort for many testers.
I suppose this is as good as anything....
Thanks for the fix. :-)
Now, (using the fact that Pavel already is in the CC list ;-)) there's another issue I have with swsusp, which is broken in a funny way. Namely, after resuming from disk the box immediately goes into standby from which it can be woken up by pressing the power button, but then it oopses, goes into standby (or something similar) again, and hangs solid (unfortunately I can't get the traces right now).
Greetings,
Rafael
Signature · 3 lines
- 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-mm1: swsusp problem (was: PCMCIA problem)
Quoted 21 lines · Rafael J. Wysocki
Quoted 13 lines · Andrew Morton
Quoted 6 lines · Pavel Machek
Quoted 4 lines
One more piece of information. This is the one that loops:
echo 30 >
/sys/class/firmware/timeoutTry echo -n ...
Or revert gregkh-driver-sysfs-strip_leading_trailing_whitespace.patch. Obviously if you write 30\n and the write returns 2 then the shell will then try to write the \n. That returns zero and the shell tries again, ad infinitum.
Quoted 6 lines
Rant. It took me two full days to weed out and fix all the crap people sent me to get -mm1 into a state where it vaguely compiled and booted. And it's untested nonsense like this which wrecks the whole effort for many testers.
I suppose this is as good as anything....
Quoted 8 lines
Thanks for the fix. :-)
Now, (using the fact that Pavel already is in the CC list ;-)) there's another issue I have with swsusp, which is broken in a funny way. Namely, after resuming from disk the box immediately goes into standby from which it can be woken up by pressing the power button, but then it oopses, goes into standby (or something similar) again, and hangs solid (unfortunately I can't get the traces right now).
Sorry for the noise. This problem has been fixed by the same patch (ie suspend is initiated by writing a string to a file on sysfs etc.).
Greetings,
Rafael
Signature · 3 lines
- 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-mm1: PCMCIA problem
In reply to Andrew MortonHi!
Quoted 10 lines
Quoted 5 lines
Quoted 3 lines
One more piece of information. This is the one that loops:
echo 30 >
/sys/class/firmware/timeoutTry echo -n ...
Or revert gregkh-driver-sysfs-strip_leading_trailing_whitespace.patch. Obviously if you write 30\n and the write returns 2 then the shell will then try to write the \n. That returns zero and the shell tries again, ad infinitum.
Can you revert gregkh-driver-sysfs-strip_leading_trailing_whitespace.patch, instead?
Kernel should not provide "nice" interface. Striping trailing whitespace is wrong, just teach users to use sysfs right.
Pavel
Signature · 1 line
if you have sharp zaurus hardware you don't need... you know my address
2.6.13-mm1: broken drivers/video/sis/Makefile
In reply to Andrew MortonQuoted 7 lines · Andrew Morton
...
Changes since 2.6.13-rc6-mm2:
...
+sisfb-update.patch
...
fbdev updates
...
This patch accidentally replaces drivers/video/sis/Makefile with a toplevel Makefile.
cu
Adrian
Signature · 4 lines
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon SeedRe: 2.6.13-mm1: broken drivers/video/sis/Makefile
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Quoted 11 lines · Adrian Bunk
Quoted 7 lines · Andrew Morton
...
Changes since 2.6.13-rc6-mm2:
...
+sisfb-update.patch
...
fbdev updates
...
Quoted 2 lines
This patch accidentally replaces
drivers/video/sis/Makefilewith a toplevelMakefile.
ARGH..... that happens if you work with four trees at the same time...
My appologies. Correct Makefile-patch attached.
Thomas
- --
Thomas Winischhofer
Vienna/Austria
thomas AT winischhofer DOT net *** http://www.winischhofer.net
twini AT xfree86 DOT org
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
iD8DBQFDF43FzydIRAktyUcRAptjAKDPQeYc3v5Ulu+HKnbRINsCNcfwwgCgkWnD sJnT86TfSyX45JIW2KKRLog=
=TDOr
-----END PGP SIGNATURE-----
drivers/video/sis/Makefile +1 -1
@@ -4,4 +4,4 @@ obj-$(CONFIG_FB_SIS) += sisfb.o -sisfb-objs := sis_main.o sis_accel.o init.o init301.o +sisfb-objs := sis_main.o sis_accel.o init.o init301.o initextlfb.o
2.6.13-mm1: misc mwave issues
In reply to Andrew MortonQuoted 7 lines · Andrew Morton
...
Changes since 2.6.13-rc6-mm2:
...
git-serial.patch
...
Subsystem trees
...
This patch contains Alan's patch for fixing the compilation of drivers/char/mwave/mwavedd.c, but the driver is still marked as BROKEN which should now be undone.
The MWAVE also got a comment
# PLEASE DO NOT DO THIS - move this driver to drivers/serial
Since it seems this code is mostly unmaintained, can the
mv drivers/char/mwave drivers/serial/
be done in the git-serial tree?
Additionally, drivers/char/mwave/mwavedd.c now requires an
#include "8250.h"
for the serial8250_{,un}register_port prototypes.
TIA
Adrian
Signature · 4 lines
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon SeedRe: 2.6.13-mm1: misc mwave issues
Quoted 2 lines · Adrian Bunk
The MWAVE also got a comment
# PLEASE DO NOT DO THIS - move this driver todrivers/serial
Mwave is an interested toy - its mostly an enabled for the hardware and the services provided are not just serial but also audio etc
Quoted 3 lines · On Thu, Sep 01, 2005 at 03:55:42AM -0700 Andrew Morton wrote:
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm1/
i386-boottime-for_each_cpu-broken.patch
i386-boottime-for_each_cpu-broken-fix.patch
The SMP version of __alloc_percpu checks the cpu_possible_map before allocating memory for a certain cpu. With the above patches the BSP cpuid is never set in cpu_possible_map which breaks CONFIG_SMP on uniprocessor machines (as soon as someone tries to dereference something allocated via __alloc_percpu, which in fact is never allocated since the cpu is not set in cpu_possible_map).
The below fixes this, I'm not entirely sure about the voyager part, should the cpu_possible_map really be CPU_MASK_ALL to begin with there, Zwane?
arch/i386/kernel/smpboot.c +1 -0
@@ -1265,6 +1265,7 @@cpu_set (smp_processor_id (), cpu_online_map);cpu_set (smp_processor_id (), cpu_callout_map);cpu_set (smp_processor_id (), cpu_present_map); +cpu_set (smp_processor_id (), cpu_possible_map);per_cpu (cpu_state,smp_processor_id ()) =CPU_ONLINE ; }
arch/i386/mach-voyager/voyager_smp.c +1 -0
@@ -1910,6 +1910,7 @@ {cpu_set (smp_processor_id (), cpu_online_map);cpu_set (smp_processor_id (), cpu_callout_map); +cpu_set (smp_processor_id (), cpu_possible_map); }int __devinit
Zwane Mwaikambo
Acked-byQuoted 9 lines · Alexander Nyberg
Quoted 3 lines · On Thu, Sep 01, 2005 at 03:55:42AM -0700 Andrew Morton wrote:
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm1/
i386-boottime-for_each_cpu-broken.patch
i386-boottime-for_each_cpu-broken-fix.patch
Quoted 6 lines
The SMP version of
__alloc_percpuchecks thecpu_possible_mapbefore allocating memory for a certain cpu. With the above patches the BSP cpuid is never set incpu_possible_mapwhich breaksCONFIG_SMPon uniprocessor machines (as soon as someone tries to dereference something allocated via__alloc_percpu, which in fact is never allocated since the cpu is not set incpu_possible_map).
Yes indeed, if there is no mptable or madt we would have missed setting it.
Quoted 3 lines
The below fixes this, I'm not entirely sure about the voyager part, should the
cpu_possible_mapreally beCPU_MASK_ALLto begin with there, Zwane?
I wanted to avoid breaking it wholesale and since i don't entirely understand the voyager SMP boot sequence, i opted for the safe method. cpu_possible_map is fine because it's supposed to cover all possible processors, upto NR_CPUS.
Quoted 1 line
Signed-off-by Alexander Nyberg <[email protected]>
Thanks Alex,
Quoted 30 lines
Index:
mm/arch/i386/kernel/smpboot.c
===================================================================arch/i386/kernel/smpboot.c +1
-0@@ -1265,6 +1265,7 @@cpu_set (smp_processor_id (), cpu_online_map);cpu_set (smp_processor_id (), cpu_callout_map);cpu_set (smp_processor_id (), cpu_present_map); +cpu_set (smp_processor_id (), cpu_possible_map);per_cpu (cpu_state,smp_processor_id ()) =CPU_ONLINE ; }Index:
mm/arch/i386/mach-voyager/voyager_smp.c
===================================================================arch/i386/mach-voyager/voyager_smp.c +1
-1@@ -1910,6 +1910,7 @@ {cpu_set (smp_processor_id (), cpu_online_map);cpu_set (smp_processor_id (), cpu_callout_map); +cpu_set (smp_processor_id (), cpu_possible_map); }int __devinit -To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Quoted 2 lines · Andrew Morton
Dropped (I have it in a new AIO patch series but I took yet another look at all the AIO stuff and felt queasy)
What's the nature of the queasiness? Is it something that can be addressed by rewriting the patches, or just general worries about adding another feature? The status quo is not acceptable.
-ben
Quoted 9 lines · Benjamin LaHaise
Quoted 2 lines · Andrew Morton
Dropped (I have it in a new AIO patch series but I took yet another look at all the AIO stuff and felt queasy)
What's the nature of the queasiness? Is it something that can be addressed by rewriting the patches, or just general worries about adding another feature? The status quo is not acceptable.
Cons:
- Additional arguments to various fastpath functions
- Additional code size
- Additional code complexity
- Significantly degrades collective understanding of how the VFS works.
Pros:
- Unclear.
Quoted 3 lines · Andrew Morton
Cons:
- Additional arguments to various fastpath functions
One possibility is to split the async and sync versions by way of inline functions. That will result in more icache pressure, though, which makes it a questionable optimization.
Quoted 3 lines
- Additional code size
- Additional code complexity
These are general concerns of all new features. It is possible to split the code out into separate codepaths (the 2.4 async read/write paths were completely new), but that cuts down on code sharing between the traditional sync read/writes and async code. If that is a requirement for merging, then they certainly could be split out and grown in a separate aio module, but that leads to other maintenence headches.
The flip side of this question is that we already have O_DIRECT and all of its associated overhead, yet the vast majority of systems don't use it in the course of day to day operation. Maybe we should have config options for these things, but where does the line get drawn?
Quoted 1 line
- Significantly degrades collective understanding of how the VFS works.
That can be improved through the use of debugging options to teach people that aio_* functions should not block, which is the only real difference between async and non-async functions -- if you'd block, return -EIOCBRETRY instead.
Quoted 3 lines
Pros:
- Unclear.
We already have a number of users demanding buffered filesystem aio -- the UML patches to use aio just went in. Samba is able to make use of aio. There are lots of high performance server applications which require aio and filesystem caching (think iSCSI servers, web servers/caches). Right now the only way to do that is via threads, which brings into play a lot of other overhead. AIO is one of the last areas in which we don't provide a suitable implementation for use in normal applications, only highly specialised database users.
-ben
Hi Andrew,
it seems you dropped schedule-obsolete-oss-drivers-for-removal-version-2.patch, but there's zero mentioning of this dropping in the changelog of 2.6.13-mm1.
Can you explain why you did silently drop it?
TIA
Adrian
Signature · 4 lines
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon SeedQuoted 9 lines · Adrian Bunk
Hi Andrew,
it seems you dropped schedule-obsolete-oss-drivers-for-removal-version-2.patch, but there's zero mentioning of this dropping in the changelog of 2.6.13-mm1.
Can you explain why you did silently drop it?
It spat rejects and when I looked at the putative removal date I just didn't believe it anyway. Send a rediffed one if you like, but October 2005 is unrealistic.
Quoted 8 lines · Andrew Morton
Quoted 7 lines · Adrian Bunk
Hi Andrew,
it seems you dropped schedule-obsolete-oss-drivers-for-removal-version-2.patch, but there's zero mentioning of this dropping in the changelog of 2.6.13-mm1.
Quoted 5 lines
Quoted 1 line
Can you explain why you did silently drop it?
It spat rejects and when I looked at the putative removal date I just didn't believe it anyway. Send a rediffed one if you like, but October 2005 is unrealistic.
That the date is no longer realistic is clear. What disappoints me is that you didn't mention in the changelog of 2.6.13-mm1 where I'd have noticed it.
It semms I need my own bookkeeping of patches I sent that are in -mm to notice when they get lost. The only positive side effect of this is that I can use this to push you harder to forward some patches of me to Linus that stay unforwarded in -mm for several months...
cu
Adrian
Signature · 4 lines
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon SeedQuoted 12 lines · Adrian Bunk
Quoted 10 lines · Andrew Morton
Quoted 8 lines · Adrian Bunk
Hi Andrew,
it seems you dropped schedule-obsolete-oss-drivers-for-removal-version-2.patch, but there's zero mentioning of this dropping in the changelog of 2.6.13-mm1.
Can you explain why you did silently drop it?
Quoted 7 lines
Quoted 3 lines
It spat rejects and when I looked at the putative removal date I just didn't believe it anyway. Send a rediffed one if you like, but October 2005 is unrealistic.
That the date is no longer realistic is clear. What disappoints me is that you didn't mention in the changelog of 2.6.13-mm1 where I'd have noticed it.
Sometimes I can't be bothered getting into email threads over relatively unimportant stuff. Usually it's related to the number of bugs we have.
Quoted 2 lines
It semms I need my own bookkeeping of patches I sent that are in -mm to notice when they get lost.
This is called "quilt".
Quoted 3 lines
The only positive side effect of this is that
I can use this to push you harder to forward some patches of me to Linus that stay unforwarded in -mm for several months...
A single release cycle is 2-3 months.
I'll probably be dropping some of the patches which unexport symbols, btw. ANy ones which aren't really, really obvious. We have a process for this.
Quoted 17 lines · Andrew Morton
Quoted 16 lines · Adrian Bunk
Quoted 13 lines · Andrew Morton
Quoted 8 lines · Adrian Bunk
Hi Andrew,
it seems you dropped schedule-obsolete-oss-drivers-for-removal-version-2.patch, but there's zero mentioning of this dropping in the changelog of 2.6.13-mm1.
Can you explain why you did silently drop it?
It spat rejects and when I looked at the putative removal date I just didn't believe it anyway. Send a rediffed one if you like, but October 2005 is unrealistic.
Quoted 6 lines
Quoted 3 lines
That the date is no longer realistic is clear. What disappoints me is that you didn't mention in the changelog of 2.6.13-mm1 where I'd have noticed it.
Sometimes I can't be bothered getting into email threads over relatively unimportant stuff. Usually it's related to the number of bugs we have.
This is not about email threads.
You send a changelog when you announce a new -mm kernel.
Why didn't you simply mention that you dropped this patch due to rejects in the changelog you are sending?
Quoted 10 lines
Quoted 2 lines
It semms I need my own bookkeeping of patches I sent that are in -mm to notice when they get lost.
This is called "quilt".
Quoted 3 lines
The only positive side effect of this is that
I can use this to push you harder to forward some patches of me to Linus that stay unforwarded in -mm for several months...A single release cycle is 2-3 months.
And I'm talking about patches waiting in -mm for more than 5 months.
Quoted 2 lines
I'll probably be dropping some of the patches which unexport symbols, btw. ANy ones which aren't really, really obvious. We have a process for this.
You accept patches into -mm, and without any new issues with these patches you tell me more than five months later "I'll probably be dropping some of the patches which unexport symbols, btw."?
If this is how my work is appreciated here I'll better stop wasting part of my spare time and unsubscribe from linux-kernel.
cu
Adrian
Signature · 4 lines
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon SeedQuoted 21 lines · Andrew Morton
Quoted 12 lines · Adrian Bunk
Quoted 10 lines · Andrew Morton
Quoted 8 lines · Adrian Bunk
Hi Andrew,
it seems you dropped schedule-obsolete-oss-drivers-for-removal-version-2.patch, but there's zero mentioning of this dropping in the changelog of 2.6.13-mm1.
Can you explain why you did silently drop it?
Quoted 7 lines
Quoted 3 lines
It spat rejects and when I looked at the putative removal date I just didn't believe it anyway. Send a rediffed one if you like, but October 2005 is unrealistic.
That the date is no longer realistic is clear. What disappoints me is that you didn't mention in the changelog of 2.6.13-mm1 where I'd have noticed it.
Quoted 8 lines
Sometimes I can't be bothered getting into email threads over relatively unimportant stuff. Usually it's related to the number of bugs we have.
Quoted 2 lines
It semms I need my own bookkeeping of patches I sent that are in -mm to notice when they get lost.
This is called "quilt".
I'm wondering if it would be too much trouble to have a mm-drops list similar to the mm-commits list.
I also like to keep track of what patches of mine get accepted and subsequently dropped.
What I'm thinking is that it seems you have the mails to mm-commit pretty much automated (I may be wrong, but it seems that way to me). If they are indeed automated, then how hard would it be to set your end up to automatically send a mail to the same people who got the original mm-commits mail + send it to a central mm-drops list that those of us who care about this could subscribe to?
As far as I'm concerned the mails wouldn't even need to contain a reason (although one would of course be nice) - just a mail stating the fact that patch xyz was dropped from the mm tree would be great.
Signature · 3 lines
Jesper Juhl <[email protected]> Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html Plain text mails only, please http://www.expita.com/nomime.html
Quoted 3 lines · Jesper Juhl
I'm wondering if it would be too much trouble to have a mm-drops list
similar to the mm-commits list.
Well I was sending drop messages to mm-commits, but lots of people went "Waah, why did you drop my patch?". A few hours after they'd been cc'ed as the patch went in to Linus! So then I was asked to include an explanation with the drop message and that all got too hard so I turned them off.
<turns them back on again>
Quoted 2 lines
I also like to keep track of what patches of mine get accepted and subsequently dropped.
As I say, the way to do this is via your quilt series file.
Quoted 10 lines · Andrew Morton
Quoted 3 lines · Jesper Juhl
I'm wondering if it would be too much trouble to have a mm-drops list
similar to the mm-commits list.Well I was sending drop messages to mm-commits, but lots of people went "Waah, why did you drop my patch?". A few hours after they'd been cc'ed as the patch went in to Linus! So then I was asked to include an explanation with the drop message and that all got too hard so I turned them off.
If patches dropped due to being merged in mainline were then commented with a simple "merged in mainline" note, surely that would keep the "Waah .." mails out of your mailbox. :-)
Quoted 7 lines
<turns them back on again>
Quoted 2 lines
I also like to keep track of what patches of mine get accepted and subsequently dropped.
As I say, the way to do this is via your quilt series file.
Hmm, I've been looking at quilt, but never really got to the point of actually starting to use it - guess I should get started on that.
Signature · 3 lines
Jesper Juhl <[email protected]> Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html Plain text mails only, please http://www.expita.com/nomime.html
Andrew wrote"
Quoted 4 lines
So then I was asked to include an explanation
with the drop message and that all got too hard so I turned them off.<turns them back on again>
Dang it, Andrew. It didn't have to be hard. Just adding a boiler plate sentence to all the drop messages saying something like:
If I just sent the patch to Linus, that is probably why I dropped it here.
That should be enough of a clue for most folks.
Signature · 3 lines
I won't rest till it's the best ...
Programmer, Linux Scalability
Paul Jackson <[email protected]> 1.925.600.0401Quoted 9 lines · Paul Jackson
Andrew wrote"
Quoted 4 lines
So then I was asked to include an explanation
with the drop message and that all got too hard so I turned them off.<turns them back on again>
Dang it, Andrew. It didn't have to be hard.
I got three unhappy emails and turned it off again ;)
Review
Quoted 8 lines
Just adding a
boiler plate sentence to all the drop messages saying something like:(quoted diff) +0
-0If I just sent the patch to Linus, that is probably why I dropped it here. That should be enough of a clue for most folks.
OK..
Quoted 7 lines · Paul Jackson
Andrew wrote"
Quoted 4 lines
So then I was asked to include an explanation
with the drop message and that all got too hard so I turned them off.<turns them back on again>
Quoted 8 lines
Dang it, Andrew. It didn't have to be hard. Just adding a boiler plate sentence to all the drop messages saying something like:
If I just sent the patch to Linus, that is probably why I dropped it here.
That should be enough of a clue for most folks.
I agree completely. Something like that would be just fine for the patches that have been sent on to Linus.
Signature · 3 lines
Jesper Juhl <[email protected]> Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html Plain text mails only, please http://www.expita.com/nomime.html
Quoted 2 lines · Jesper
Something like that would be just fine for the
patches that have been sent on to Linus.
No - not just the patches sent to Linus - that's a burden on Andrew to separate things out.
Andrew can put the boiler plate statement on _all_ drop messages
Quoted 2 lines
If I just sent the patch to Linus, that is probably why I dropped it here.
At the very least, he gets to cuss us out for not being able to read, instead of not being able to associate the 'patch to Linus' message with the 'drop' message a few hours later.
Signature · 3 lines
I won't rest till it's the best ...
Programmer, Linux Scalability
Paul Jackson <[email protected]> 1.925.600.0401Quoted 3 lines · On Thu, Sep 01, 2005 at 03:55:42AM -0700 Andrew Morton wrote:
ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.13/2.6.13-mm1/
I got:
<7>Dead loop on netdevice eth0, fix it urgently!
When using netconsole and printing out some information from kernel to console.
The box uses:
[email protected]/eth0,[email protected]/
0000:00:0f.0 Ethernet controller: Linksys NC100 Network Everywhere Fast Ethernet 10/100 (rev 11)
Relevant config:CONFIG_NET_TULIP=y
# CONFIG_DE2104X is not setCONFIG_TULIP=yCONFIG_TULIP_MWI=y
# CONFIG_TULIP_MMIO is not setCONFIG_TULIP_NAPI=y
Matt, on another box I got some irq off hangs that went away when removing netconsole from the .config on a box with 3c59x. Is this known? The problem is getting backtraces when netconsole is active, but the last thing I see before the box goes is that some carrier is up...