Summary of changes from v2.5.43 to v2.5.44
============================================
<anton@samba.org>
ppc64: fix dump_stack
<anton@samba.org>
ppc64: sigcontext_struct -> sigcontext, from Stephen Rothwell
<anton@samba.org>
ppc64: updates for Ingo's signal changes
<anton@samba.org>
ppc64: add might_sleep to semaphore code
<anton@samba.org>
ppc64: move pci_device_to_OF_node so radeonfb can get at it
<anton@samba.org>
ppc64: remove pciconfig_iobase, its broken when IO resources are > 4GB
<anton@samba.org>
ppc64: update in_atomic definition
<anton@samba.org>
ppc64: only enable eeh if something supports it
<anton@samba.org>
ppc64: dont mark openpic_setup_lock as __initdata, we need it for cpu add
<jgrimm@touki.austin.ibm.com>
sctp: in sendmsg: on err, only free asoc if init failed. (jgrimm)
In sendmsg(), the lower layers own the association cleanup after
the ASSOCIATE call has succeeded, so the error path should _only_
free the asoc if we've failed to pass this responsibility to
the lower layer (otherwise we double-free the asoc).
<anton@samba.org>
ppc64: clean up exception table code, we werent taking the modlist_lock
<anton@samba.org>
ppc64: Only do an exception check if an 0x380 ends up in do_page_fault.
<anton@samba.org>
ppc64: some xmon fixes for SLB faults, also store breakpoint address in a long
<anton@samba.org>
ppc64: local_irq_restore was missing a gcc barrier.
<anton@samba.org>
ppc64: reduce stack usage of prom_instantiate_rtas
<anton@samba.org>
ppc64: update defconfig
<anton@samba.org>
ppc64: early printk support from Todd Inglett
<anton@samba.org>
ppc64: merge status indicator update from 2.4
<anton@samba.org>
ppc64: use generic debugger hooks in smp_call_function, busy loop instead of udelay
<jgrimm@touki.austin.ibm.com>
sctp: Fix restart address add prevention logic (jgrimm)
The logic was opposite as needed, disallowing valid restart and
allowing hijack.
<anton@samba.org>
ppc64: disable ancient select syscalls
<anton@samba.org>
ppc64: merge some xmon updates from ppc32
<anton@samba.org>
ppc64: update to match recent kbuild changes
<anton@samba.org>
ppc64: ipc fixes from Peter Bergner and 64bit types updates from 2.4
<anton@samba.org>
ppc64: fix Makefile so it actually works
<anton@samba.org>
ppc64: remove 64bit old uname syscall
<daisy@teetime.dynamic.austin.ibm.com>
sctp: Fix bug 611919 - should ignore the cwnd value for fast retransmit.
<anton@samba.org>
ppc64: limit NR_CPUS to 32 temporarily, need to hunt down the 500kB of bloat it causes
<anton@samba.org>
ppc64: Handle some broken ioctls that do _IO(,,sizeof(struct foo)) instead of just struct foo.
<anton@samba.org>
ppc64: Dont allow us to recursively call printk, spotted by Milton Miller
<anton@samba.org>
ppc64: fix an IPC bug
<jgrimm@touki.austin.ibm.com>
sctp: Various invalid address check fixes. (jgrimm)
1) Need to check inside INIT chunks for addresses, not just
INIT-ACK otherwise we don't recognize restart.
2) The address check logic isn't being excercised early enough
in the restart. Per the impl-guide it needs to happen as early
as INIT processing in COOKIE-ECHOED state.
3) Eliminate more uses of deprecated sctpParam_t
<Matt_Domsch@dell.com>
EDD: x86 BIOS Enhanced Disk Drive support
The major changes implemented in this patch:
arch/i386/boot/setup.S - int13 real mode calls store results in empty_zero_page
arch/i386/kernel/setup.c - copy results from empty_zero_page to local storage
arch/i386/kernel/edd.c - module exports results via driverfs
x86 systems suffer from a disconnect between what BIOS believes is the
boot disk, and what Linux thinks BIOS thinks is the boot disk. This
manifests itself in multi-disk systems - it's quite possible to
install a distribution, only to fail on reboot - the disk installed to
is not the disk BIOS is booting from. Dell restricts our possible
standard factory installed Linux offerings to "disks on no more than
one controller" to avoid this problem, but mechanisms now exist to
solve it and allow such configurations.
BIOS Enhanced Disk Device Services (EDD) 3.0 provides the ability for
disk adapter BIOSs to tell the OS what it believes is the boot disk.
While this isn't widely implemented in BIOSs yet, it's time that Linux
received support to be ready as BIOSs with this feature do become
available. At a minimum, LSI MegaRAID cards support this today.
EDD works by providing the bus (PCI, PCI-X, ISA, InfiniBand, PCI
Express, or HyperTransport) location (e.g. PCI 02:01.0) and interface
(ATAPI, ATA, SCSI, USB, 1394, FibreChannel, I2O, RAID, SATA) location
(e.g. SCSI ID 5 LUN 0) information for each BIOS int13 device.
The patch below creates CONFIG_EDD, that when defined, makes the
BIOS int13 calls to retrieve and store this information. The data is
copied to a safe place in setup.c, and exported via driverfs.
Here's a sample driverfs tree with two BIOS int13 devices - dev 80 has
incorrect PCI bus information, thus no symlinks are made, but as much
info as possible is presented. Dev 81 has correct PCI and SCSI
information, thus symlinks are made to the actual disc device.
/driverfs
|-- bios
| |-- int13_dev80
| | |-- extensions
| | |-- host_bus
| | |-- info_flags
| | |-- interface
| | |-- raw_data
| | |-- sectors
| | `-- version
| `-- int13_dev81
| |-- extensions
| |-- host_bus
| |-- info_flags
| |-- interface
| |-- pci_dev -> ../../root/pci2/02:0c.0/03:00.0/04:00.0
| |-- raw_data
| |-- disc -> ../../root/pci2/02:0c.0/03:00.0/04:00.0/scsi4/4:0:0:0
| |-- sectors
| `-- version
|-- bus
| |-- scsi
| | |-- devices
| | | |-- 4:0:0:0 -> ../../../root/pci2/02:0c.0/03:00.0/04:00.0/scsi4/4:0:0:0
| | `-- drivers
| | `-- sd
`-- root
|-- pci2
| |-- 02:0c.0
| | |-- 03:00.0
| | | |-- 04:00.0
| | | | |-- irq
| | | | |-- name
| | | | |-- power
| | | | |-- resource
| | | | `-- scsi4
| | | | |-- 4:0:0:0
| | | | | |-- 4:0:0:0::p1
| | | | | | |-- kdev
| | | | | | |-- name
| | | | | | |-- power
| | | | | | `-- type
| | | | | |-- 4:0:0:0:disc
| | | | | | |-- kdev
| | | | | | |-- name
| | | | | | |-- power
| | | | | | `-- type
| | | | | |-- name
| | | | | |-- power
| | | | | `-- type
(Yes, the 'bios' top-level directory isn't the right place,
and Patrick has promised to make something there in the future,
at which point this can be moved.)
The 'raw_data' file contains the full set of information returned by BIOS
with extra error reporting. This exists for vendor BIOS debugging purposes.
The 'host-bus' file contains the PCI (or ISA, HyperTransport, ...)
identifying information, as BIOS knows it.
The 'interface' file contains the SCSI (or IDE, USB, ...) identifying
information, as BIOS knows it.
The 'extensions' file lists the BIOS EDD extensions per spec.
The 'info_flags' file lists the BIOS EDD device information flags per spec.
The 'sectors' file reports the number of sectors BIOS believes this
device has.
The 'version' file lists the EDD version. To have device path
information, this must be 0x30 or above. Earlier EDD versions exist
without the device path - as much information as is available is presented.
At most 6 BIOS devices are reported, as that fills the space that's
left in the empty_zero_page. In general you only care about device
80h, though for software RAID1 knowing what 81h is might be useful also.
Known issues:
- module unload leaves a directory around. Seems related to
creating symlinks in that directory. Seen on kernel 2.5.41.
- refcounting of struct device objects could be improved.
TODO:
- Add IDE and USB disk device support
- when driverfs model of discs and partitions changes,
update symlink accordingly.
- Get symlink creator helper functions exported from
drivers/base instead of duplicating them here.
- move edd.[ch] to better locations if/when one is decided
I'd also like to acknowledge the help and comments received from Greg
KH and Patrick Mochel. This isn't something driverfs was originally
conceived to handle, their assistance has been invaluable.
Please pull from:
BK:
http://mdomsch.bkbits.net/linux-2.5-edd-tolinus
Patch (against 2.5.41+BK-current):
http://domsch.com/linux/edd30/edd-driverfs-6.patch
http://domsch.com/linux/edd30/edd-driverfs-6.patch.sign
Thanks,
Matt
--
Matt Domsch
Sr. Software Engineer, Lead Engineer, Architect
Dell Linux Solutions www.dell.com/linux
Linux on Dell mailing lists @ http://lists.us.dell.com
<anton@samba.org>
ppc64: Fix for copy_tofrom_user from Paul Mackerras
<anton@samba.org>
ppc64: Fix semctl return code, also msgrcv/msgsnd fixes courtesy of Andi
<sam@mars.ravnborg.org>
kbuild: Distributed clean infrastructure
Today there is a huge list of files in the top-level Makefile that is
deleted during make clean and make mrproper.
This patch add infrastructure to get rid of this centralised list.
Within a makefile simply use:
clean-files := files-to-be-deleted
or eventually
clean-rule := command to be executed to delete files
Files specified by host-progs and EXTRA_TARGETS are deleted during cleaning,
and the same is all *.[oas] .*.cmd .*.tmp .*.d in the visited directories.
Deleting core files is moved down to mrporper time
Patches utilising this and the centralised list will dismiss.
Based on a concept originally made by Kai Germaschewski
<sam@mars.ravnborg.org>
scsi+aic7xxx: Utilise distributed clean
List files to be deleted during make clean where they are created
<sam@mars.ravnborg.org>
drivers/{atm,char,pci,video,zorro}: ditributed clean
Move list of files to be deleted during make clean out where
they are made. host-progs files taken care of automagically
<sam@mars.ravnborg.org>
drivers/net/hamradio/soundmodem: distributed clean
Move list of files out where it belongs
<sam@mars.ravnborg.org>
kbuild: Distributed clean, misc.
o Move sound/oss file list to sound/oss/Makefile
o Remove files non-existing in the tree (khttp,net/802/submenu)
o scripts/* are handled by scripts makefile
o Do not delete .config*, be more explicit
o Add MC* - files generated by Menuconfig in toplevel dir
<sam@mars.ravnborg.org>
docbook: Makefile cleanup
o Removed special rules for JBD, covered by the general mechanishm
o Use $(obj)/ instead of Documentation/DocBook
o Introduced usage of Distributed clean
o No longer delete *~ files in top-level directory during clean
<anton@samba.org>
ppc64: fix from ppc32 for sugsuspend bug
<anton@samba.org>
ppc64: Fix pgd_index overflow in free_pgtables and move stack up to 2^41
<anton@samba.org>
ppc64: fix ptrace GETREGS/SETREGS
<sridhar@dyn9-47-18-140.beaverton.ibm.com>
sctp: Fixes a couple of sctp_peeloff() issues.
* data/notifications waiting in the parent socket's receive queue
should be moved to the peeled-off socket's receive queue.
* sendmsg() to a peer already on an association that has been
peeled off should fail and not create a new association.
<kai@tp1.ruhr-uni-bochum.de>
Move kallsyms section next to other read-only sections
As noted by Keith Owens, for some setups it may make sense to have
all read-only sections grouped together.
<kai@tp1.ruhr-uni-bochum.de>
Clean up arch/i386/vmlinux.lds.S
Add some comments indicating read-only, writeable and freed-after-init
sections.
Move .data_nosave, .data.page_aligned, .data.cacheline_aligned next
to the "normal" data area, the only difference this should make,
apart from being more logical, is a more truthful printk() during boot.
<kai@tp1.ruhr-uni-bochum.de>
kbuild: Add build dep for UML
Using .S -> .s needs scripts/fixdep, which is not necessarily there
so early during the build, so add a dependency on it.
<kai@tp1.ruhr-uni-bochum.de>
kbuild: More cleaning work...
o Having .config depend on all Config.in's via a find is
annoying, since it causes a long break before "make" actually
gets going. Comment it out for now.
o Fix some "make -j" issues with generating module versions.
o Unify "make distclean" and "make mrproper", they were doing nearly the
same thing anyway.
o Use the new "make clean" infrastructure for arch/i386/boot/ and
scripts/
o Reorganize the "make clean" stuff a little, do the
find -name '*.[oas]' | xargs rm -f from the top-level Makefile again
and only use the per-subdir rule for special cases like generated files,
host-progs and the like.
<anton@samba.org>
ppc64: add missing include
<davem@nuts.ninka.net>
[NET]: Kill final traces of csum_partial_copy_fromuser.
<johannes@erdfelt.com>
[PATCH] 2.5 uhci remove correct proc directory
This patch fixes a typo in the names used to remove the proc directory
that uhci uses.
The longer term fix of splitting out (or dropping outright) the
debugging code will come soon after the important bugs and features are
merged.
<johannes@erdfelt.com>
[PATCH] 2.5 uhci remove urb from lists on error
I'm applying Dan's patches by hand to double check everything and I
wanted to get a couple of the trivial ones out of the way first.
This patch fixes a bug where errors on the submit path wouldn't remove
URB's from the HC linked list.
If we fail adding the URB to the schedule, we need to make
sure that we remove it from the urb_list. Thanks to
Dan Streetman for finding and fixing this bug.
<mdharm-usb@one-eyed-alien.net>
[PATCH] Fixes to CB/CBI and compilation problems...
This patch:
(a) updates some comments
(b) adds some debugging
(c) fixes CB and CBI data stages
(d) adds some missing {'s and }'s -- only dumb luck let the old
version compile.
<randy.dunlap@verizon.net>
[PATCH] 2.5.42 HID-BP menu
HID-BP confuses people, we know that.
I don't want to see the HID-BP drivers removed,
but I think it would be a good idea to make them more
difficult to enable, unless someone knows what they
are looking for. (Basically it would reduce support
incidents a lot.)
The second way puts HID-BP in its own sub-menu and
updates the Help file. I think that this is the better
choice, so unless there is a great hue and cry,
please apply the "hidbpmenu-2542.patch" file.
<johannes@erdfelt.com>
[PATCH] 2.5 uhci proc path
This is a patchlet that updates a comment to use the correct path to the
proc debugging file.
<rob@osinvestor.com>
[NET]: Remove final traces of csum_partial_copy.
<johannes@erdfelt.com>
[PATCH] 2.5 uhci breadth first traversal for low speed
Woops, my fault. I forgot to send you this patch which needs to be
applied before the big one.
It's from Dan as well and switches low speed control to use breadth
first traversal to make it more fair.
Don't make low speed control use depth first. That isn't
particularly fair. Thanks to Dan Streetman for bringing
this up and the original patch.
<johannes@erdfelt.com>
[PATCH] 2.5 uhci control and interrupt queuing
This is 95% Dan's patch, but I made some small changes. The changes I've
made relative to Dan's patch is:
Drop concept of skeleton TD. After Dan's patch, this was reduced to one
entry in the form of skel_term_td. That wasn't even a skeleton TD in the
first place. This simplifies the code a little bit.
Minor formatting tweaks.
Pass on USB bandwidth changes for now.
Pass on Interrupt auto-resubmit changes for now.
Use complete_list_lock, not complete_list as the lock.
Use frame_list_lock, not fame_list_lock.
Reorganize skeleton QH's to match the order in the schedule. This cleaned
up some debugging code and made the list more logical.
Update some obsolete documentation.
Pass on code to check for race conditions and fix up. It was racy as well.
Dan, do you compile with SMP? Those two locking typos would have
generated a warning and/or error.
I've done some light testing and haven't found any problems yet. I
haven't tested the control or interrupt queuing for lack of anything to
test with yet.
Anyway, please apply.
Merge in Dan's patch to add interrupt and control queuing support. Summary of changes:
Add queuing support for Interrupt and Control message in addition to Bulk.
This resulted in some merging of code.
Fix a queuing bug when moving a child into the parent position.
Update documentation.
Update debugging code.
<anton@samba.org>
[PATCH] AIO
Hi,
I started looking at AIO on 2.5 for ppc64. Ive got a small patch which:
1. fixes a dprintk.
2. from what I can see, __aio_get_req returns NULL not -EAGAIN on failure
so fix comment.
3. we were setting okay in the failure path.
So far I get Im getting 6 passes, 5 fails, any idea how that compares
with x86?
Anton
===== fs/aio.c 1.20 vs edited =====
<bcrl@redhat.com>
get rid of a double free in aio.c introduced by a merge mistake.
<jgmyers@netscape.com>
[PATCH] aio updates
Please apply.
fix uninitialized variable causing incorrect timeout
make sys_io_cancel(), not cancel method, initialize most of returned result
minor aio_cancel_all() optimization
<davem@redhat.com>
[PATCH] De-bloat linux/fs/aio.c
This saves 2K of object code on sparc64.
Please consider and push around, thanks Ben.
<kuznet@ms2.inr.ac.ru>
[TCP]: Handle passive resets correctly in SYN-RECV.
<ddstreet@ieee.org>
[PATCH] uhci: slight docs update
Slight change - the int qh order is actually opposite of the example.
<ddstreet@ieee.org>
[PATCH] uhci: remove qh from qh->list
I think the qh needs to be removed from its qh->list, or else
uhci_remove_qh will incorrectly change the previous endpoint's qh->link,
undoing what was just done in uhci_delete_queued_urb.
<davem@nuts.ninka.net>
[SPARC64]: On broken cheetah, enable p-cache around large copies.
<maxk@qualcomm.com>
[NET]: Export sockfd_lookup.
<sridhar@dyn9-47-18-140.beaverton.ibm.com>
sctp: VTAG checks for ABORT & SHUTDOWN_COMPLETE chunks (ardelle.fan)
<sfr@canb.auug.org.au>
[SPARC]: Move over to generic siginfo.
<sridhar@dyn9-47-18-140.beaverton.ibm.com>
sctp: Fixes Bug#623286 - zero vtag in SHUTDOWN_COMPLETE chunk (samudrala)
<sfr@canb.auug.org.au>
[SPARC]: arch specific copy_siginfo_to_user no longer needed.
<kuznet@mops.inr.ac.ru>
[NET]: Prepare for zerocopy NFS and IPSEC.
- Import va10-hwchecksum-2.5.36.patch
- Import va11-udpsendfile-2.5.36.patch
- Implement new encapsulation friendly ipv4 output path.
<davem@nuts.ninka.net>
[NET]: Cleanup now that sockfd_lookup/sockfd_put are exported.
- Delete redefinitions of sockfd_{lookup,put}
- Fix socket fd leaks in route ioctl32 code.
<davem@nuts.ninka.net>
arch/sparc64/solaris/socket.c: Kill more sockfd_{lookup,put} redefinitions.
<davem@nuts.ninka.net>
net/ipv4/udp.c: proto sendpage returns int not size_t.
<davem@nuts.ninka.net>
net/bluetooth/bnep/sock.c: Kill another sockfd_lookup re-implementation.
<acme@conectiva.com.br>
o ipv4: udp seq_file support: produce only one record per seq_show
<davem@nuts.ninka.net>
net/ipv4/ip_proc.c: Fix 64-bit warnings.
<davem@nuts.ninka.net>
arch/sparc64/kernel/time.c: Include linux/profile.h
<davem@nuts.ninka.net>
arch/sparc/kernel/time.c: Include linux/profile.h
<davem@nuts.ninka.net>
[SPARC]: Export memchr.
<anton@samba.org>
ppc64: defconfig update
<davem@nuts.ninka.net>
[NET]: Apply missed parts of csum_partial_copy killing patch.
<davem@nuts.ninka.net>
arch/{i386,sh}/lib/Makefile: Kill old-checksum.o
<davem@nuts.ninka.net>
[SPARC]: Add sys_lookup_dcookie.
<rmk@flint.arm.linux.org.uk>
[ARM] time.c needs to include profile.h
time.c handles the kernel profiling, and references prof_buffer/
prof_len/prof_shift.
<rmk@flint.arm.linux.org.uk>
[ARM] Convert ARM makefiles to new kbuild (Sam Ravnborg, Kai, rmk)
<rmk@flint.arm.linux.org.uk>
[ARM] Kill compiler warning in decompressor.
<rmk@flint.arm.linux.org.uk>
[SERIAL] Fix missing parens in serial_core.h
<trond.myklebust@fys.uio.no>
[PATCH] Fix NFS typos in 2.5.43...
The following patch fixes 2 obvious typos. Thanks to davem and George
Anzinger for pointing them out.
<trond.myklebust@fys.uio.no>
[PATCH] Fix 'long long' != u64 in NFSv4 debugging printks
Thanks again to davem...
<rmk@flint.arm.linux.org.uk>
[ARM] Update ARM SA1111 pci_pool implementation
2.5.43 removed the mem_flags parameter to pci_pool_create. The ARM
SA1111 implementation needs to follow for the usbo hci implementation.
<willy@debian.org>
[PATCH] Fix file locking yield()
Use cond_resched() instead, and add a comment to explain.
<akropel1@rochester.rr.com>
[PATCH] 2.5.43: cpqarray compile fix
This fixes a recent dyslexicism in cpqarray.
<akpm@digeo.com>
[PATCH] ia32 uniprocessor compile fixes
Fix a couple of things for no-io-apic builds:
- mpparse.c wants BAD_APICID
- mpparse.c (an perhaps others) want Dprintk()
<akpm@digeo.com>
[PATCH] fix the build for CONFIG_MD
Use 'p', not 'j'
<akpm@digeo.com>
[PATCH] disable 64-bit sector_t on ppc32
From Peter Chubb
OK, make it X86 only. 64-bit architectures now get it for free;
if/when PPC32 (the only other 32-bit architecture I can seriously imagine
needing it, except possibly s390) fixes do_div() and needs large block
support, then enable it back there again.
There isn't as far as I know a flag I can test in a config.in file
that distinguishes 64 from 32-bit architectures. The intention was to
enable the choice of CONFIG_LBD where it made sense; it doesn't make
sense for 64-bit architectures; and rather than enumerate all the
32-bit architectures, I just chose two that I knew were used in
servers currently.
<akpm@digeo.com>
[PATCH] fix oops in reiserfs_ioctl()
Leftovers from the kmap_atomic conversion
<akpm@digeo.com>
[PATCH] fix a VM lockup
Patch from Hugh Dickins.
There's a loop in there which goes infinite if the active list is full
of zero-ref pages.
<akpm@digeo.com>
[PATCH] simple_rename() link count fix
From Hugh Dickins
simple_rename, like shmem_rename until I fixed that a few weeks ago,
got the parent directory's link count wrong in the exceptional case
of renaming a directory in place of an existing empty directory.
<akpm@digeo.com>
[PATCH] make filemap_sync static
From Christpoh Hellwig.
Make filemap_sync() static, and not exported to modules
<akpm@digeo.com>
[PATCH] don't allocate an extra page in vmalloc
From Marcus Alanen <marcus@infa.abo.fi>
vmalloc is allocating one more page than necessary.
<akpm@digeo.com>
[PATCH] don't make writers wait on their writeback in page reclaim
I had some code in page reclaim which is designed to catch the caller
of write(2) in page reclaim and force him to wait on writeout of his
own spindle's pages.
It works too well - there are large stalls in things like `cp
large-file foo' on small-memory machines.
So remove it, and rely on the other, more predictable throttling
schemes.
<akpm@digeo.com>
[PATCH] uninline somethings in fs/*.c
- Don't inline mpage_writepage().
Even though it had but a single call site, uninlining this function
took mpage.o from 4529 bytes of text to 3985. gcc-2.95.3.
- Various buffer.c tweaks. 13233 bytes down to 12977
<akpm@digeo.com>
[PATCH] uninline the ia32 highmem functions
With gcc-2.95.3, text size goes from 2,399,190 to 2,390,004 - nearly
10 kbytes.
<neilb@cse.unsw.edu.au>
[PATCH] kNFSd: Reorganise rpc program version management.
An rpc service like NFSD knows about a number of
versions (2,3,4) and needs to know the max and min valid
version numbers, and also the maximum xdr buffer size used
by any version.
These max/mins are currently computed at compile time which means
we need to include all the xdr*.h headers into nfssvc.c just for
computing a couple of numbers.
With this patch, these max/mins are calculated at run time from
information in the svc_version structure which has been extended to
include per-version xdrsize.
Also the nfsd_version2, nfsd_version3, and nfsd_version4
structures are moved from nfssvc.c to their more correct home
of nfs*proc.c
Code gets neated, xdr*.h are no-longer included in
nfssvc.c, and the ugly ifdef can be removed fom xdr4.h
<neilb@cse.unsw.edu.au>
[PATCH] kNFSd: Export news symbols in sunrpc
The new caching and auth code needs symbols to be exported.
<neilb@cse.unsw.edu.au>
[PATCH] kNFSd: Fix problems when releasing cache item.
When releasing a cache item we always release anything
that is part of the key, but we ONLY release the content
is if it is valid. i.e. if CACHE_VALID but not CACHE_NEGATIVE.
<neilb@cse.unsw.edu.au>
[PATCH] kNFSd: Change names of some exported functions
... as generic names probably aren't such a good idea.
Also sort the related EXPORT_SYMBOLS.
A 'qword' is a 'quoted word' and can be added, added as hex,
or got (getted??).
<hch@sgi.com>
[PATCH] kNFSd: minor knfsd cleanups
o don't define __NO_VERSION__ (no use in 2.4/2.5)
o don't include <linux/version.h>
<hch@sgi.com>
[PATCH] kNFSd: switch knfsd to vfs_read/vfs_write
Switch knfsd to vfs_read/vfs_write to work on aio-only filesystems.
This also gets stuff like the LSM checks and mandatory lock checking
for free.
<andros@citi.umich.edu>
[PATCH] kNFSd: NFSv4 patch for new setclientid, setclientid_confirm
This brings the linux-2.5.42 tree up to spec WRT setclientid and
setclientid_confirm
<neilb@cse.unsw.edu.au>
[PATCH] kNFSd: Unregister export caches in proper order
As the expkey cache references entries in export cache,
it should be flushed first.
Also, the auth_unix caches should go last.
<neilb@cse.unsw.edu.au>
[PATCH] md - 1 of 1 - Register mergeable function for linear so requests don't cross device boundries
We abstract out "which_dev" which finds which device a sector is on
and us it to make sure both ends of a to-be-merged request will
be on the same device.
<willy@debian.org>
[PATCH] Allow compilation with -ffunction-sections
If you compile the kernel with -ffunction-sections, each function gets
put in a section .text.function_name. This collides with our current use
of .text.init. So here's a patch which converts x86 to use .init.text
instead.
I've tested it on x86 and it still frees 120k of ram, so it seems to work.
Other architectures will need to change their vmlinux.lds appropriately,
and may need other changes (arm, m68k seem to use .text.init verbatim).
<maxk@qualcomm.com>
Fix files that escaped CONFIG_BLUEZ_XXX -> CONFIG_BT_XXX cleanup.
Mostly arch specific default configs.
<kai@tp1.ruhr-uni-bochum.de>
ISDN: new xmit handling for ISDN net interfaces
Instead of using locking per-channel, just lock the entire ISDN net work
interface as an entity, getting rid of weird locking.
Basically, ::hard_start_xmit() is already serialized by the network core,
so we could just rely on that. However, we want to send keep alive
frames, PPP messages from ipppd and such things directly without going
through the network stack, so we additionally lock
isdn_net_local->xmit_lock in hard_start_xmit(), which is taken in
the other paths dealing with transmitting frames as well.
<kai@tp1.ruhr-uni-bochum.de>
ISDN: Move generic xmit/recv handling into isdn_net_lib.c
Last step in separating generic / interface-type specific code,
the former is now all in isdn_net_lib.c, the specific code in
isdn_net.c / isdn_ppp.c / isdn_concap.c
<kai@tp1.ruhr-uni-bochum.de>
ISDN: Unclutter isdn_net.h
Definitions which is only needed in one source file should go there and not
in a common header.
<kai@tp1.ruhr-uni-bochum.de>
ISDN/PPP: Adapt sync-PPP
Make sync PPP compile with the latest changes to the ISDN net device
layer - PPP negotiations work again after this patch.
<kai@tp1.ruhr-uni-bochum.de>
ISDN/PPP: Rename struct 'ippp_struct' to 'struct ipppd'
and move the declaration from include/linux/isdn_ppp.h
into drivers/isdn/i4l/isdn_ppp.c
<kai@tp1.ruhr-uni-bochum.de>
ISDN/PPP: Move state from ipppd to isdn_net_dev/isdn_net_local
A lot of the state in struct ipppd actually belongs to
isdn_net_dev or isdn_net_local, making it more easily accessible at a
most places, and also removing the ambiguity whether it's link-
or bundle-specific.
<kai@tp1.ruhr-uni-bochum.de>
ISDN/PPP: Move CCP related stuff into isdn_ppp_ccp.c
Create the new files isdn_ppp_ccp.[hc] which deal with PPP CCP
(compression control protocol) related stuff. Move most of the CCP
related stuff from isdn_ppp.c into isdn_ppp_ccp.c, putting CCP
state into a new "struct ippp_ccp".
<kai@tp1.ruhr-uni-bochum.de>
ISDN: Named initializers for isdn_bsdcomp.c
<kai@tp1.ruhr-uni-bochum.de>
ISDN: Bug fixes for the isdnloop driver
Fix missed STAT_BSENT acks and don't use skb_clone() for the skb which is
sent back, the network layer gets confused by that.
<kai@tp1.ruhr-uni-bochum.de>
ISDN/PPP: Remove bogus header handling
I cannot see now reason why ISDN PPP devices should have a fake
::header routine which reserves a couple of bytes just to reclaim them
later during hard_start_xmit(), and since it works just as well without
it...
<kai@tp1.ruhr-uni-bochum.de>
ISDN/PPP: PPP header cleanups
Use a shared function for generating the PPP header, use the
{put,get}_u{8,16,32} helpers for endian-safe access.
<kai@tp1.ruhr-uni-bochum.de>
ISDN/PPP: CCP flags handling
Let isdn_ppp_ccp.c take care of keeping state / flags by itself.
<kai@tp1.ruhr-uni-bochum.de>
ISDN/PPP: Move rest of CCP reset handling into isdn_ppp_ccp.c
Basically, all what was missing was a common way to xmit frames
given the isdn_net_dev / isdn_net_local. We have this now by means
of function pointers (still room for improvement there, though), so
now let's do things from isdn_ppp_ccp.c.
<kai@tp1.ruhr-uni-bochum.de>
ISDN: Start refcounting for per-ipppd data
We had introduced ipppd_get/put() earlier, so now add a real reference
count.
<kai@tp1.ruhr-uni-bochum.de>
ISDN/PPP: Reference struct ipppd directly
Instead of just recording the slot number and then retrieving
the ipppd via that, we can now just keep a pointer and get a
reference which makes sure that it does not go away until
we drop the reference.
<kai@tp1.ruhr-uni-bochum.de>
ISDN/PPP: dynamically allocate struct ipppd, further cleanups
o Now that all the infrastructure is in place, struct ipppd can
easily be allocated when /dev/ipppX is opened and freed
in the destructor.
o Separate the mix of state and flags in struct ipppd::state.
We only have three states, open, assigned and connected, and
additionally two flags to jump out of poll(), returning POLLHUP /
POLLIN.
<kai@tp1.ruhr-uni-bochum.de>
ISDN/PPP: cosmetics
No (well, hardly any ;) code changes, only moving all /dev/ipppX related
code next to each other and some indenting changes.
<kai@tp1.ruhr-uni-bochum.de>
ISDN/PPP: clean up ipppd_write() and ipppd_ioctl()
plus other small cleanups, in particular getting rid of
isdn_ppp_{receive,send}_ccp() and calling into the CCP code directly.
<mochel@osdl.org>
driver model: make device_unregister() only mark device as !present.
There are races WRT walking the lists of the buses and classes and the
device being unregistered while it's the current node we're accessing. If
the node was deleted, ->next would point to itself, and we'd sit in an
inifinite loop (until that memory was freed from underneath us and we Oops'd).
This changes device_unregister() to only mark the device as not present. put_device()
deletes the device from all the lists it belongs to, while holding device_lock,
then drops it to call device_del().
This should guarantee that everything in the device is valid until the last reference
goes away, regardless of whether it's still physically present or not.
<mochel@osdl.org>
driver model: make driverfs an implicit initcall.
<bcrl@redhat.com>
adds more aio exports for filesystems
<lord@sgi.com>
XFS: busylist fixups
xfs_alloc_put_freelist marks the blocks it gets as "busy". However, this
function is used in two different ways:
1. xfs_alloc_fix_freelist calls xfs_alloc_put_freelist to add new
blocks to the freelist (that is, blocks it allocated with
xfs_alloc_ag_vextent).
2. xfs_alloc_delrec calls xfs_alloc_put_freelist to free deleted
b+tree blocks.
In case 1, the blocks should not be on the busy list; in case 2, they
should.
Modid: 2.5.x-xfs:slinx:129987a
<lord@sgi.com>
XFS: fix previous change, the wrong xfs_alloc_mark_busy call was removed
Modid: 2.5.x-xfs:slinx:130045a
<lord@sgi.com>
XFS: fix byte ordering issues with earlier allocator fix
Modid: 2.5.x-xfs:slinx:130071a
<hch@sgi.com>
XFS: add synch I/O entry points
Modid: 2.5.x-xfs:slinx:130135a
<lord@sgi.com>
XFS: fix xfs build on big endian architectures and cleanup
Modid: 2.5.x-xfs:slinx:130142a
<hch@lab343.munich.sgi.com>
XFS: fix a makefile comment
<mochel@osdl.org>
driver model: change bus refcounting scheme to match devices'.
Based on recent changes to device refcounting/unregistration.
struct bus_type gets a ->present field, which is set when the bus is
registered. bus_unregister() is added, which only clears ->present and
calls put_bus(). get_bus() checks the ->present and returns NULL if clear.
This allows all current references to buses to keep valid structures, while
preventing any new references to the bus structure.
<mochel@osdl.org>
driver model: make driver refcounting similar to devices'.
In the spirit of devices and buses, change driver refcounting model to
match the way that devices and buses are done.
struct device_driver gets a ->present field, which is set on registration
and cleared in driver_unregister(). get_device() checks the state of this
flag and returns NULL if cleared.
Note that the horribly wrong remove_driver() is deprecated and simply BUG()s
when called. Please convert callers to use driver_unregister(). Updates to
callers will be coming soon.
Note also that this still doesn't fix the race in which a driver module can
be removed while the refcount on a driver > 1. Near future work should help
to remedy it, but no solutions are guaranteed..
<mochel@osdl.org>
driver model: change class reference counting to be like devices'.
device classes join the club of devices, buses, and drivers. They get a
->present flag, which is set on registration and cleared on unregistration.
They also get get_devclass() and put_devclass(), which, you guessed it,
bump the reference count. get_...() of course checks the present flag and
returns NULL if clear.
<davem@nuts.ninka.net>
drivers/net/3c59x.c: Do not forget to set AddUDPChksum.
<mochel@osdl.org>
driver model: replace rwlock in struct bus_type with a rwsem.
Synchronize all walks of the device and driver lists of a bus with an rwsem wrapped
around the entire iterator, instead of using device_lock and dropping it after we
grabbed each node.
Note this also prevents deadlock when walking the list of drivers and calling
get_driver(), since get_driver() tries to take device_lock while we already have it
held.
<mochel@osdl.org>
driver model: make sure device has driver when we add it to the class.
There is a chance devclass_add_device() could get passed a device with no driver, so
we need to check that.
<mochel@osdl.org>
driver model: simplify device/driver binding.
- move list walking and matching to bus.c (since it's a function of the bus driver)
- do specialized walks of the bus's lists when binding; no more callbacks being passed
to bus_for_each_*.
- take rwsem when adding and removing both devices and drivers. lists of each are now
fully protected by that rwsem. It's also taken before we walk each list.
- move calls of device_{de,at}tach() to bus_{add,remove}_device() and calls of
driver_{de,at}tach() to bus_{add,remove}_driver().
<mochel@osdl.org>
driver model: make sure we call device_unregister() and not put_device()
when removing system and platform devices.
<mochel@osdl.org>
driver model: add per-class rwsem and protect list accesses with it.
This is similar to struct bus_type's rwsem, though classes aren't doing
anything nearly as fancy with it. We just make sure to take it when ever
we add or remove any devices or drivers.
<davem@nuts.ninka.net>
arch/sparc{,64}/vmlinux.lds.S: Update for init section name changes.
<acme@conectiva.com.br>
o ipv4: make arp seq_file show method only produce one record per call
<kai@tp1.ruhr-uni-bochum.de>
kbuild: Speed up new "make clean/mrproper"
Do as little as possible apart from calling sub-makes and calling 'rm',
which makes things go quite a bit faster.
<kai@tp1.ruhr-uni-bochum.de>
kbuild: More "make clean" cosmetics
Make the DocBook "make clean" output nicer and use the fast variant
for arch/i386/boot as well.
<kai@tp1.ruhr-uni-bochum.de>
kbuild: another "make clean" micro-optimization
Patch by Peter Samuelson, merged to current code, actually has a noticable
impact on "make clean" speed.
<acme@conectiva.com.br>
o ipv4: remove the hack, using seq->private to hold state
<zaitcev@redhat.com>
arch/sparc/boot/Makefile: Use libs-y.
<zaitcev@redhat.com>
[sparc] Use Kai's way to make asm_offsets.h
<marcel@holtmann.org>
Use kernel crc32 for BlueZ BNEP
This patch modifies the BlueZ BNEP layer to use the CRC32 code from
the lib/ directory.
<maxk@qualcomm.com>
Users must have CAP_NET_ADMIN capability in order to create
or destroy devices.
<maxk@qualcomm.com>
Just like many other parts of the kernel Bluetooth code was abusing
typedefs for non opaque objects. This changeset cleans up L2CAP code
and headers. In addition it optimizes sendmsg for L2CAP sockets.
<andmike@us.ibm.com>
[PATCH] scsi host cleanup 1/3 (base) (corrected)
I had diffed against a tree missing a compile fix this is the correct base
patch.
This is a resend of my previous patch clean ups to the scsi_host lists.
* Made function naming consistent with rest of SCSI
* Corrected a problem with driverfs registration to early. Also
changed from put_device to device_unregister.
* Fixed a regression in my previous patch that the scsi_host
list was not sorted by host number. When we get some device
naming this hack can be removed.
* Switch scsi host template, name, host lists to struct
list_head's.
* Moved all scsi_host related register / unregister functions
into hosts.c
* Added list accessor interface and created a function similar
to driverfs bus_for_each_dev.
The full patch is available at:
http://www-124.ibm.com/storageio/patches/2.5/scsi-host
-andmike
--
Michael Anderson
andmike@us.ibm.com
hosts.c | 890 +++++++++++++++++++++++++++++++++++++++++++++++++---------------
hosts.h | 97 +++---
2 files changed, 734 insertions(+), 253 deletions(-)
<andmike@us.ibm.com>
[PATCH] scsi host cleanup 2/3 (mid lvl changes)
This is a resend of my previous patch clean ups to the scsi_host lists.
* Made function naming consistent with rest of SCSI
* Corrected a problem with driverfs registration to early. Also
changed from put_device to device_unregister.
* Fixed a regression in my previous patch that the scsi_host
list was not sorted by host number. When we get some device
naming this hack can be removed.
* Switch scsi host template, name, host lists to struct
list_head's.
* Moved all scsi_host related register / unregister functions
into hosts.c
* Added list accessor interface and created a function similar
to driverfs bus_for_each_dev.
The full patch is available at:
http://www-124.ibm.com/storageio/patches/2.5/scsi-host
-andmike
--
Michael Anderson
andmike@us.ibm.com
scsi.c | 456 +++---------------------------------------------------------
scsi_proc.c | 57 ++++---
scsi_syms.c | 5
sg.c | 6
4 files changed, 63 insertions(+), 461 deletions(-)
<andmike@us.ibm.com>
[PATCH] scsi host cleanup 3/3 (driver changes)
If you read my previous post on this patch I indicated that few of the
driver changes I was only able to compile test ( block/cciss_scsi.c,
scsi/53c700.c, scsi/pcmcia/*, scsi/wd33c93.c). The changes to the
drivers are to remove the old interfaces and possibly extra NULL inits
of struct members. These changes will need to be ok'd by there
respective maintainers. I did receive feedback from two.
I have tested these changes on ips, aic, qlogicisp, and scsi_debug
drivers.
This is a resend of my previous patch clean ups to the scsi_host lists.
* Made function naming consistent with rest of SCSI
* Corrected a problem with driverfs registration to early. Also
changed from put_device to device_unregister.
* Fixed a regression in my previous patch that the scsi_host
list was not sorted by host number. When we get some device
naming this hack can be removed.
* Switch scsi host template, name, host lists to struct
list_head's.
* Moved all scsi_host related register / unregister functions
into hosts.c
* Added list accessor interface and created a function similar
to driverfs bus_for_each_dev.
The full patch is available at:
http://www-124.ibm.com/storageio/patches/2.5/scsi-host
-andmike
--
Michael Anderson
andmike@us.ibm.com
acorn/scsi/acornscsi.c | 6 ++----
acorn/scsi/arxescsi.c | 8 ++------
acorn/scsi/cumana_2.c | 8 ++------
acorn/scsi/eesox.c | 8 ++------
acorn/scsi/powertec.c | 8 ++------
block/cciss_scsi.c | 21 ++-------------------
scsi/53c700.c | 6 ++----
scsi/aic7xxx/aic7xxx_linux_host.h | 1 -
scsi/cpqfcTSinit.c | 4 +---
scsi/fcal.c | 4 +---
scsi/ips.h | 3 ---
scsi/pcmcia/aha152x_stub.c | 3 ++-
scsi/pcmcia/fdomain_stub.c | 3 ++-
scsi/pcmcia/nsp_cs.c | 3 ++-
scsi/pcmcia/qlogic_stub.c | 3 ++-
scsi/wd33c93.c | 5 +----
16 files changed, 25 insertions(+), 69 deletions(-)
<kai@tp1.ruhr-uni-bochum.de>
kbuild: Fix temporary hack
We don't need to include .config for "make clean", but we sure do
otherwise...
(pointed out by Sam Ravnborg)
<maxk@qualcomm.com>
Just like many other parts of the kernel Bluetooth code was abusing
typedefs for non opaque objects. This Changeset cleanups Bluetoth HCI
code and headers.
<maxk@qualcomm.com>
Consistent naming for Bluetooth HCI events, commands and parameters.
Cleanup unused HCI stuff.
Optimize HCI receive path.
<maxk@qualcomm.com>
Correct RFCOMM modem status implementation.
RFCOMM DLC API fixes.
<hch@infradead.org>
[PATCH] make LSM register functions GPLonly exports
These exports have the power to change the implementations of all
syscalls and I've seen people exploiting this "feature".
Make the exports GPLonly (which some LSM folks agreed to
when it was merged initially to avoid that).
<hch@infradead.org>
[PATCH] remove LSM file_llseek hook
In the initial discussion LSM folks agreed on this, the
rationale is that lsseek itself makes no sense to
project as mmap() and pread/pwrite() allow access to any
area of the file anyway.
<mochel@osdl.org>
driver model: add struct device_driver::shutdown() method and device_present() helper.
->shutdown() is added to be explicitly called during reboot transitions by
device_shutdown(). Drivers implementing this method should use it solely to quiesce
the device. Drivers do not have to free memory allocated by the driver, as it is
implied that the system is rebooting; and, they must not fuss with the reference
counts. Devices are guaranteed not to disappear during the call, though their
validity may be checked using device_present().
<mochel@osdl.org>
driver model: protect drivers' device list accesses with bus's rwsem.
Drivers must belong to bus, and each bus has an rwsem. Instead of mucking with
the device_lock spinlock, and dropping it on each iteration of the loop, we take
the bus's lock (read, so multiple drivers can access their list at the same time)
around the entire walk of the list.
<mochel@osdl.org>
driver model: clean up bus_for_each_{dev,drv}
- make sure we check what we get back from get_bus()
- do down_read() instead of down_write(), so multiple people can do iterations
concurrently.
- Note that devices and drivers are removed from these while doing a down_write(),
so all removals will be stalled until the iterators are done.
<mochel@osdl.org>
driver model: introduce device_sem to protect global device list instead of device_lock.
- device_sem is added to make the global list walks easier (for power mgmt and shutdown), so
they don't have to take and drop device_lock a lot.
- Change list modifications to take device_sem instead of device_lock.
- Ditto for device refcount modifications.
- Kill get_device_locked(), as all the users are now gone.
<mochel@osdl.org>
driver model: power management/shutdown cleanup.
- take device_sem around all global list walks.
- don't modify refcount, as get_device()/put_device() also take device_sem
- But, make sure we check device_present() to make sure device is still around.
(Note device removal will block until device_sem is dropped, so list will remain
intact.)
- Separate out device_shutdown() walk from device_suspend() walk. Even though the
code is nearly identical, it's a lot clearer as to what is going on when they
are autonomous. It was my bad for originally putting that FIXME in there, encouraging
the consolidation.
- Add debugging hooks for my convenience. :)
- Call ->shutdown() when shutting down device, instead of ->remove(). (See ChangeSet
1.799 for description and semantics).
<dhowells@redhat.com>
[PATCH] do_generic_file_read / readahead adjustments
This does the following three things:
(1) Makes the functions in mm/readahead.c only use struct file* to pass to
readpage(). address_mapping* and file_ra_state* are used instead to keep
track of readahead stuff.
(2) Adds a new function do_generic_mapping_read() that is similar to
do_generic_file_read(), except that it uses a mapping pointer and a
readahead state pointer to access a file. The file* is only used to pass
to readpage().
(3) Turns do_generic_file_read() into an inline function in linux/fs.h that
simply wraps do_generic_mapping_read().
This should mean that it is no longer necessary to have a struct file to
access a file in this manner. Just an inode or address space should be
sufficient.
It also means alternate read-ahead structures can be maintained.
The reason I want this is that I'm writing a general cache manager for
filesystems such as AFS, NFSv4, and Lustre. Block devices are made available
to the "cache manager" by means of a filesystem that can be mounted. I'm
storing meta data in an inode in the cache, but to scan this at the moment I
need to gain a "struct file" to use with do_generic_file_read().
This involves either creating a dummy dentry and struct file (which will cause
Al Viro to come looking for me with a shotgun), or to use an extra auxilliary
filesystem mounted with do_kern_mount(), neither of which are particularly
appealing.
This patch is the alternative... it provides a function that I can pass an
address_space to. This also allows me to make use of readahead semantics
without having to reinvent them for myself.
<sridhar@dyn9-47-18-140.beaverton.ibm.com>
sctp: Fixes a bug in the calculation of highest new tsn in sack.
<thockin@freakshow.cobalt.com>
drivers/net/eepro100.c:
* mii cleanups
* use mii_check_link()
* check link every timer
<thockin@freakshow.cobalt.com>
drivers/net/mii.c: only call netif_carrier_{on,off} if there is a state change
<sparker@sun.com>
drivers/net/eepro100.c:
* use netif_msg_xxx() instead of speedo_debug
* better default debug mask, proper use of debug module param
* add ETHTOOL_[GS]MSGLVL
<sparker@sun.com>
drivers/net/eepro100.c: simplify wait_for_cmd_done(), better errors
<sparker@sun.com>
drivers/net/eepro100.c: only set priv->last_rx_time if we did work
<sparker@sun.com>
drivers/net/eepro100.c: mask the interrupt and do a small delay on close()
<thockin@freakshow.cobalt.com>
drivers/net/natsemi.c: init msg_enable in proper way
<mochel@osdl.org>
driver model: change struct device::present to enumerated value with multiple states.
Adapted from Greg KH:
- add multiple possible enumerated states a device can be in: UNINITIALIZED, INITIALIZED,
REGISTERED, and GONE.
- Check whether device is INITIALIZED or REGISTERED in device_present().
- Change struct device::current_state to ::power_state to better reflect what it is.
<mochel@osdl.org>
driver model (pci): change call of remove_driver to driver_unregister.
<mochel@osdl.org>
driver model (scsi): change calls of remove_driver() to driver_unregister().
<mochel@osdl.org>
driver model(usb): change calls of remove_driver() to driver_unregister().
<mochel@osdl.org>
driver model (arm): change calls of remove_driver() to driver_unregister().
<mochel@osdl.org>
driver model: use list_for_each_safe() when removing devices from a driver.
Duh: in driver_detach(), one of the first things we do is remove the device
from the driver's list of devices. So, we obviously cannot use ->next.
<thockin@freakshow.cobalt.com>
drivers/net/eepro100.c: eliminate speedo_intrmask
<jejb@mulgrave.(none)>
[SCSI] bring block TCQ helpers into line with new TCQ code
<jejb@mulgrave.(none)>
[SCSI 53c700] update with new TCQ code
<jejb@mulgrave.(none)>
[SCSI sd] make cache probe sensitive to unsupported mode page
<davej@codemonkey.org.uk>
net/ipv4/udp.c: Log short packets more verbosely.
<mochel@osdl.org>
add sysfs filesystem, maintaining the driverfs revision history.
<mochel@osdl.org>
sysfs: modify enough names to get it build and link w/o conflicts.
- Change all exported symbols.
- driverfs_get_inode() is not static; make it static in sysfs.
- Change name in struct filesystem_type.
- Add target to fs/Makefile.
It now shows up in /proc/partitions, and is mountable. There's nothing there,
but you can mount it!
<mochel@osdl.org>
sysfs: copy driverfs_fs.h to sysfs.h
<mochel@osdl.org>
sysfs: search replace to convert remaining names from driverfs* to sysfs*.
<mochel@osdl.org>
sysfs: clean up a couple of remaining driverfs references in sysfs.h.
<kuznet@ms2.inr.ac.ru>
[IPv4]: More output path work.
- Fix a bug noted by Maxim Giryaev (gem@asplinux.ru)
when corking on a non-SG interface
- Convert ICMP and ip_send_reply to ip_append_foo.
- Rename generic_getfrag to more namespace friendly
ip_generic_getfrag.
- Kill n_iov member from ip_reply_arg.
<mochel@osdl.org>
sysfs: do kern_mount() on init.
driverfs had some extra baggage of only doing kern_mount() when it was being used.
That's all well and good in principle, but it's always used, starting immediately
after it's initialized.
This exploits that knowledge, and just does kern_mount() from sysfs_init().
Other trivial cleanups and comment corrections.
<mochel@osdl.org>
sysfs: copy driverfs documentation to be sysfs documentation.
<jejb@mulgrave.(none)>
[SCSI sd] correct wrong prink statement
<lists@mdiehl.de>
[PATCH] usbtest updates
Various small fixes and adds ids for new test firmware.
<patmans@us.ibm.com>
[PATCH] Re: [PATCH] SCSI-2 LUN setting consolidation
This patch consolidates the setting of the LUN in byte 1 of the SCSI
command block for SCSI-2 and lower devices.
This is needed for multi-path IO (some devices can actually have differnt
LUN values for each path), but is also a nice clean up of the code.
sg.c was able to inhibit setting the value, this removes that capability -
if it is really needed it can be black listed in the device_list[] flags,
or sg could set a similiar flag.
This is patched against the latest bk (as of some time on Oct 16), and
patches clean against 2.5.43.
Tests were run with different adapters and various disks and a tape drive,
I have no SCSI-2 multi-lun devices, so the main verifcation was done using
a modified scsi_debug.c driver to emulate a multi-lun SCSI-2 device.
drivers/scsi/osst.c | 2 --
drivers/scsi/scsi.c | 6 ++++++
drivers/scsi/scsi_error.c | 12 +++++-------
drivers/scsi/scsi_ioctl.c | 16 ++++------------
drivers/scsi/scsi_scan.c | 31 ++++---------------------------
drivers/scsi/sd.c | 22 +++++-----------------
drivers/scsi/sg.c | 4 ----
drivers/scsi/sr.c | 17 ++---------------
drivers/scsi/sr_ioctl.c | 20 +-------------------
drivers/scsi/sr_vendor.c | 18 +++---------------
drivers/scsi/st.c | 2 --
include/scsi/sg.h | 2 +-
12 files changed, 31 insertions(+), 121 deletions(-)
<andmike@us.ibm.com>
[PATCH] scsi_debug new scan fix
If you use scsi_debug the patch below fixes a problem that has existed
since the updated scan code was merged in. scsi_debug previously assumed
that the scsi_device used to probe and the device permanently added to the
host_queue would be the same. This caused scsi_debug to allocate some
internal data and key it off the scsi_device pointer for later use. This
resulted in scsi_debug failing all IO post scanning. This patch corrects
this problem.
Note:
Douglas Gilbert is the maintainer of this driver.
dougg@gear.torque.net
http://www.torque.net/sg/
During Douglas Gilbert's time-off he connects when he can so it
maybe a bit until he can address this.
In the interim this patch makes scsi_debug functional.
The full patch is available at:
http://www-124.ibm.com/storageio/patches/2.5/scsi-debug
-andmike
--
Michael Anderson
andmike@us.ibm.com
scsi_debug.c | 59 +++++++++++++++++++++++++++++++----------------------------
1 files changed, 31 insertions(+), 28 deletions(-)
<acme@conectiva.com.br>
o ipv4: remove the hack, make udp seq_file functions use seq->private
<akpm@digeo.com>
[PATCH] ramdisk fix
The ramdisk address_space needs to have a non-zero
mapping->backing_dev_info.memory_backed, so it doesn't contribute to
global memory accounting.
At present it is confusing memory accounting and causing
balance_dirty_pages to go into an infinite loop.
<akpm@digeo.com>
[PATCH] 3com driver fix
From Alexey: teach the 3com ethernet driver to add checksums to
outgoing UDP frames.
<greg@kroah.com>
Cset exclude: david-b@pacbell.net|ChangeSet|20021015071026|11647
<david-b@pacbell.net>
[PATCH] LDM and driver binding
Here's a more complete patch for that driver binding problem.
Now (a) when scanning for a driver to handle a new device,
scanning stops after the first successful probe not the
first successful match; and (b) new drivers get every chance to
bind to additional devices, even after the first one succeeds.
Please merge. This will help resolve some of the USB flakies
in recent kernels. Note that (a) was noticed independently by
me and Andries Brouwer, and AFAIK this is the first patch to
also fix failure (b) which was uncovered by fixing (a).
<ddstreet@ieee.org>
[PATCH] change devio-disconnect no-driver error code
The talk about disconnect locking reminded me of this - the error code in
the no-driver case for the disconnect ioctl isn't a unique error code.
This changes the error code to what getdriver() uses, -ENODATA.
<acme@conectiva.com.br>
[PATCH] Add support for JTEC FA8101 USB to Ethernet device
o Add support for JTEC FA8101 USB to Ethernet device
<david-b@pacbell.net>
[PATCH] one liner, anti-oops
This fixes a potential oops (depending on how the HCD handles it)
from a recent patch of mine. Evidently I didn't test this bit a
device that'd show the problem, sigh. Luckily Martin Diehl was
testing and found this.
<jgrimm@touki.austin.ibm.com>
sctp: Fix small data can bypass pending rtx data bug. (jgrimm)
<trini@kernel.crashing.org>
PPC32: Update section names to .foo.text/.foo.data form.
<paulus@samba.org>
PPC32: make xchg take volatile *, remove duplicate xchg_u32 defn.
Previously we had xchg_u32 both inline in system.h and out-of-line
in arch/ppc/kernel/misc.S, and it was exported for good measure.
We only need the inline definition.
<paulus@samba.org>
PPC32: export open/read/lseek/close for the sake of ALSA modules.
<paulus@samba.org>
PPC32: Update boot wrapper Makefiles.
We can't now use make -C in conjunction with Rules.make because
the rules for making .o files break if we do. This updates the
boot wrapper Makefiles so that we don't use make -C, instead we
use call descend plus various other new features in Rules.make.
<paulus@samba.org>
PPC32: fix a typo in mpc10x.h.
<trini@kernel.crashing.org>
PPC32: improve the memory size detection for PReP machines.
<cort@fsmlabs.com>
PPC32: Fixes for Force PowerCore boards.
<danc@mvista.com>
PPC32: Add support for the SBS Palomar 4 embedded board.
<paulus@samba.org>
PPC32: Add some extra kernel debugging options.
<paulus@samba.org>
PPC32: make do_div handle all 64 bits of the dividend (not just 32).
We do this with an out-of-line assembler function, __div64_32,
for the case when the high word of the dividend is non-zero.
<paulus@samba.org>
PPC32: allow for host bridges that have a type field in the addr word.
Some host bridges that use the indirect (addr and data registers)
method of access have a field that needs to be set to 0 for accesses
to the primary bus and 1 for accesses to subordinate buses. This
change allows for that.
<ddstreet@ieee.org>
[PATCH] uhci interrupt resubmit fixes
After the interrupt queueing was added, I don't think the old way of
resetting interrupts will work anymore. This patch changes it to simply
do a full unlink and resubmission automatically. Note that since
usb_hcd_giveback_urb() is never called for a resubmitting interrupt URB,
that means whatever gets released in usb_hcd_giveback_urb() won't get
released for that URB. The only way to work around that is call
usb_hcd_giveback_urb after the user unlinks in their completion handler,
which will call the completion handler again with -ECONNRESET
status...which wouldn't be all that bad, but the drivers have to expect
it.
Hopefully the interrupt resubmission will go away soon...
<viro@math.psu.edu>
[PATCH] infrastructure
* added ->media_changed(disk). Callers of ->check_media_changed()
are using that one if present
* added ->revaldiate_disk(disk). Callers of ->revalidate()
are using that one if present
* added disk->queue. do_open() uses it if non-NULL
* added disk->private_data
* removed assignments to inode->i_data.backing_... - it belongs to
filesystem that contains device node and we don't care about it;
inode->i_mapping points to ->i_data of bdev inode and there we should
(and do) set that field.
<viro@math.psu.edu>
[PATCH] cdrom helpers
cdrom_open(), cdrom_release(), cdrom_ioctl() and cdrom_media_changed()
got an additional argument - struct cdrom_device_info *. Drivers that used
to have them as methods are using wrappers that pass right cdrom_device_info.
Existing callers (in ide-cd.c) updated. pcd.c, mcdx.c, sbpcd.c and sr.c
got ->private_data in their disks pointing to relevant objects. Since all
callers of cdrom_find_device() are gone, it got removed. Since all users
of cdrom_device_info->dev are gone, it's not set anymore (and got removed).
<viro@math.psu.edu>
[PATCH] compile fixes
DaveM spotted bogus cast sprintf in checks.c. Fixed. Several places
that unregistered device called put_device() instead.
<viro@math.psu.edu>
[PATCH] pd.c cleaned up
* switched to private queue
* got ->private_data and ->queue set
* ->open() and friends are using ->bd_disk->private_data instead of
messing with DEVICE_NR(inode->i_rdev)
* request handler uses ->rq_disk->private_data instead of DEVICE_NR(...)
* DEVICE_NR() removed.
* switched to new methods.
<viro@math.psu.edu>
[PATCH] pcd.c cleaned up
* switched to private queue
* got ->queue set
* ->open() and friends are using ->bd_disk->private_data instead of
messing with DEVICE_NR(inode->i_rdev)
* request handler uses ->rq_disk->private_data instead of DEVICE_NR(...)
* DEVICE_NR() removed.
* switched to new methods.
* got rid of CURRENT
<viro@math.psu.edu>
[PATCH] pf.c cleaned up
* switched to private queue
* got ->queue set
* ->open() and friends are using ->bd_disk->private_data instead of
messing with DEVICE_NR(inode->i_rdev)
* request handler uses ->rq_disk->private_data instead of DEVICE_NR(...)
* DEVICE_NR() removed.
* switched to new methods.
* got rid of CURRENT
<viro@math.psu.edu>
[PATCH] probes
introduced blk_register_region() and blk_unregister_region()
removed blk_set_probe()
switched to almost final variant of get_gendisk()
<viro@math.psu.edu>
[PATCH] ide.c cleaned up
* got ->private_data and ->queue set
* got blk_register_region() done in ide-probe.c
* module loading removed from ide_open() (and taken to probe)
* ->open() and friends are using ->bd_disk->private_data instead of
messing with get_info_ptr(inode->i_rdev)
* request handler uses ->rq_disk->private_data instead of DEVICE_NR(...)
* DEVICE_NR() removed.
* switched to new methods.
* get_info_ptr() killed.
* ide_xlate_1024() switched to struct block_device * (probably should
simply die).
NOTE: ide-tape pulls mind-boggling crap - it can be opened as a block
device, but you can't do any IO on it; only IDE ioctls. It happens to
work (almost by accident), but I'd really prefer to take said ioctls to
driverfs and be done with that.
NOTE: WTF are ide{disk,cd,tape,scsi,floppy} lists? They are defined,
exported and never touched by anything.
<viro@math.psu.edu>
[PATCH] md.c
AAAARGH. OK, I've got it to work with the current tree, but... I don't
like the lifetime rules - both old and new ones. Summary:
* we create a gendisk upon attempt to open or upon successful
attempt to run (which _can_ happen without open).
* we never remove them until the rmmod (or ever).
* mddev is held by gendisk.
The source of trouble is our API - ioctls on possibly unconfigured
devices when no IO is possible, potentially setting up this (or another,
or many other) devices. I _think_ we can get the old rules (mddev
exists when somebody holds it open or when it runs or when somebody is
looking at it funny), but the way I see will take more messing with
driverfs objects than I want right now.
Notice that current API is, was and will be ripe with interesting races
- it really ought to be replaced with something saner, but that would
take serious modifications of userland tools. I'll talk with Neil about
possible solution, this patch just gets it back to life.
<viro@math.psu.edu>
[PATCH] mfmhd.c
* set ->private_data and ->queue
* removed DEVICE_NR
* switched to private queue
* (partial) compile fixes - it still misses stuff from 2.5.1...
<viro@math.psu.edu>
[PATCH] loop.c
* set ->private_data and ->queue
* ->open() and friends are using ->bd_disk->private_data instead of
messing with DEVICE_NR(inode->i_rdev)
<viro@math.psu.edu>
[PATCH] sr.c
* set ->queue
* removed DEVICE_NR()
* switched request handling to ->rq_disk
<viro@math.psu.edu>
[PATCH] sd.c
* set ->queue and ->private_data
* removed DEVICE_NR and SD_PARTITION
* switched to new methods
* switched ->open() and friends to use of ->bd_disk
* switched request handling to use of ->rq_disk
* slightly cleaned up
<viro@math.psu.edu>
[PATCH] ftl.c
* switched to private queue
* set ->queue and ->private_data
* switched to new methods
* switched ->open() and friends to use of ->bd_disk
* switched request handling to use of ->rq_disk
* cleaned up
<viro@math.psu.edu>
[PATCH] aztcd.c
* switched to private queue
* set ->queue
* switched to new methods
<viro@math.psu.edu>
[PATCH] cdu31a.c
* switched to private queue
* set ->queue
* slightly cleaned up
<viro@math.psu.edu>
[PATCH] cm206.c
* switched to private queue
* set ->queue
<viro@math.psu.edu>
[PATCH] gscd.c
* switched to private queue
* set ->queue
* switched to new methods
<viro@math.psu.edu>
[PATCH] mcd.c
* switched to private queue
* set ->queue
<viro@math.psu.edu>
[PATCH] mcdx.c
* switched to private queue
* set ->queue
* switched to use of ->rq_disk
<viro@math.psu.edu>
[PATCH] optcd.c
* switched to private queue
* set ->queue
* switched to new methods
<viro@math.psu.edu>
[PATCH] sbpcd.c
* switched to private queue
* set ->queue
* switched to use of ->rq_disk
<viro@math.psu.edu>
[PATCH] sjcd.c
* switched to private queue
* set ->queue
* switched to new methods
<viro@math.psu.edu>
[PATCH] sonycd535.c
* switched to private queue
* set ->queue
* switched to new methods
* slightly cleaned up
<viro@math.psu.edu>
[PATCH] acsi.c
* switched to private queue
* set ->queue and ->private_data
* switched to new methods
* cleaned up
* switched to use of ->bd_disk and ->rq_disk
<viro@math.psu.edu>
[PATCH] ubd
* switched to private queue
* set ->queue and ->private_data
* switched to new methods
* switched to use of ->bd_disk and ->rq_disk
<viro@math.psu.edu>
[PATCH] cciss
* switched to private queue
* set ->queue and ->private_data
* switched to new methods
* switched to use of ->rq_disk
NOTE: we still have mess in open(), release() and ioctl().
<viro@math.psu.edu>
[PATCH] umem
* switched to private queue
* set ->queue and ->private_data
* switched to new methods
* switched to use of ->bd_disk
<viro@math.psu.edu>
[PATCH] swim_iop
* switched to private queue
* set ->queue and ->private_data
* switched to new methods
* switched to use of ->bd_disk
* fixed the idiocy I'd introduced there during end_request()
reshuffling - too many places got hit by search-and-replace ;-/
NOTE: swim_iop (and swim3) are broken for situations when several drives
are present. Failure mode is rather amusing - the thing does a sort of
load-balancing and if you have two drives even requests end up sent to
the first and odd - to the second ;-) Had been there since they went
into the tree. Swim3 mentions that support of multiple drives needs to
be fixed, swim_iop (derived from it) had lost even that warning.
<viro@math.psu.edu>
[PATCH] swim3
* switched to private queue
* set ->queue and ->private_data
* switched to new methods
* switched to use of ->bd_disk
<viro@math.psu.edu>
[PATCH] acorn floppy
* switched to private queue
* set ->queue and ->private_data
* switched to new methods
* compile fixes
NOTE: it still has b0rken locking - hadn't been updated in 2.5.1
<viro@math.psu.edu>
[PATCH] amiga floppy
* switched to private queue
* set ->queue and ->private_data
* switched to new methods
<viro@math.psu.edu>
[PATCH] atari floppy
* switched to private queue
* set ->queue and ->private_data
* switched to new methods
<viro@math.psu.edu>
[PATCH] DAC960
* switched to private queues
* set ->queue and ->private_data
* switched to new methods
<viro@math.psu.edu>
[PATCH] cpqarray
* switched to private queues
* set ->queue and ->private_data
* switched to new methods
<viro@math.psu.edu>
[PATCH] floppy
* switched to private queues
* set ->queue and ->private_data
* switched to new methods
<viro@math.psu.edu>
[PATCH] i2o
* switched to private queues
* set ->queue and ->private_data
* switched to new methods
<viro@math.psu.edu>
[PATCH] old methods removal
* old methods have no instances left; removed.
<davej@codemonkey.org.uk>
[PATCH] updated MAINTAINERS
<davej@codemonkey.org.uk>
[PATCH] gscd printk's wrong type
<davej@codemonkey.org.uk>
[PATCH] CONFIG_DEBUG_STACKOVERFLOW
Taken from unknown contributor in 2.4
<davej@codemonkey.org.uk>
[PATCH] bluesmoke incorrectly calls function twice.
The smp_call_function() does the check on the local cpu for us.
<davej@codemonkey.org.uk>
[PATCH] Remove duplicate config.help
(is already in drivers/char/agp/Config.help)
<davej@codemonkey.org.uk>
[PATCH] Elan BIOS quirk workaround.
Marc Singer noticed some BIOSen on the AMD Elan
alias the cpu frequency registers in a place that some
drivers probe during bootup. Result - changed CPU speed
after booting.
The fix is to unmap the alias.
Marc put a patch forward for 2.2, this one is cooked up
by me, but is currently untested but functionally equivalent.
<davej@codemonkey.org.uk>
[PATCH] Silence epat init.
<davej@codemonkey.org.uk>
[PATCH] Missing checks in sis drm
<davej@codemonkey.org.uk>
[PATCH] ens1370 uses wrong printk type
<davej@codemonkey.org.uk>
[PATCH] Typos
<davej@codemonkey.org.uk>
[PATCH] Make sbpcd look more like 2.5 driver
Most work done by Paul Gortmaker. More work needed.
<davej@codemonkey.org.uk>
[PATCH] Add printk levels
<davej@codemonkey.org.uk>
[PATCH] 64bit fixes for smbfs
By Urban Widmark
<davej@codemonkey.org.uk>
[PATCH] extern inline -> static inline
As per Documentation/SubmittingPatches
<davej@codemonkey.org.uk>
[PATCH] Correct indentation on x86-64 MTRR driver.
Somewhere along the line, indentation on this file went mental.
<greg@kroah.com>
driver core: fix up merge mess
<markh@osdl.org>
[PATCH] 2.5.43 aacraid driver
Fill devno[cid] from request instead of sc_request->sr_request since
sc_request is passed in as NULL now.
<bunk@fs.tum.de>
[PATCH] don't #include tqueue.h in rio_linux.c
<dhowells@redhat.com>
[PATCH] missing put in AFS client
This patch fixes a bug in my AFS client in which a put of an outstanding call
was missing in the unmount cleanup code. This occasionally resulted in the AFS
module becoming unremovable.
<axboe@suse.de>
[PATCH] sym2 get host
Recent scsi host changes broke sym2, it iterates over the scsi host list
itself. This patch changes sym2 to use scsi_host_hn_get() to find a scsi
host from its hostno.
<davem@redhat.com>
[PATCH] Fix scsi breakage
A recent change that went into BK-2.5 removed the next member of the
scsi host and replaced it with lists, which is cool, but this broke a
few common drivers.
Here is the fix for aic7xxx_old.
<axboe@suse.de>
[PATCH] make SCSI understand REQ_BLOCK_PC
Make SCSI layer understand REQ_BLOCK_PC requests
<axboe@suse.de>
[PATCH] make calling scsi_cmd_ioctl() part of generic cdrom_ioctl
Remove the calling of scsi_cmd_ioctl() from ide-cd. Put it in the
uniform cdrom layer, this also makes SCSI SG_IO work.
<axboe@suse.de>
[PATCH] block cleanups
o Change the blk_dump_rq_flags() a bit
o Use rq->bio as an indication of ->bio is valid or not, instead of
checking REQ_PC. Now we can have a bio on a REQ_BLOCK_PC as well.
o Don't print eio error in end_that_request_first(), if REQ_QUIET is set.
o debug check for valid bio in end_that_request_first()
o Add blk_pc_request() check REQ_BLOCK_PC, similar to blk_fs_request()
<mochel@osdl.org>
driver model: make sure we only try to bind drivers to devices that don't have a driver.
<greg@kroah.com>
USB: fix problem with removing a USB root hub.
<torvalds@penguin.transmeta.com>
Fix up sym53c8xx driver for new scsi_host_hn_get()
infrastructure.
<rddunlap@osdl.org>
[PATCH] 2.5.43 IO APIC bit fields
This recognizes & logs 2 more IO APIC bit fields and reduces UNEXPECTED
IO-APIC traffic.
Has also been acked by Maciej W. Rozycki (marco).
<torvalds@home.transmeta.com>
Fix IDE init order dependency problem, noted by Jens Axboe.
This is pretty ugly, Al will fix up properly.
<mochel@osdl.org>
driver model: fix matching bug.
If a device didn't match a driver, bus_match() returned 0, and device_attach()
would break out of the loop. So, the first driver was the only one being
checked every time.
<adam@kroptech.com>
[PATCH] cpqarray IDA_LOCK compile fix
<rddunlap@osdl.org>
[PATCH] cpia fix for older compilers
<rlievin@free.fr>
[PATCH] char tipar driver minor update
you will find a patch for the tipar char driver.
This patch fixes some typos and add a documentation entry.
<rmk@arm.linux.org.uk>
[PATCH] set_task_state() UP memory barriers
This fixes the UP set_task_state and set_current_state in 2.5.43 to
ensure that we don't re-order loads around the store for setting
task->state.
This patch is from Nicolas Pitre, nico@cam.org
(As a side note - 2.4.10 also contains the same problem, but 2.2
contain the memory barrier.)
<scott.feldman@intel.com>
e100 1/4:
* Remove broken home-grown Rx polling mode (make way for NAPI)
<torvalds@home.transmeta.com>
Make x86 UP "set_mb()" use a lighter barrier than doing a full
locked "xchg". It only needs a compiler barrier on UP.
<scott.feldman@intel.com>
e100 2/4:
* Added meaningful message for self-test failures
* Removed confusing messages from probe when link isn't detected
<scott.feldman@intel.com>
e100 3/4:
* Updated change log
* Adding notifier to track ifname change for /proc info
<scott.feldman@intel.com>
e100 4/4:
* updates to Documentation/networking/e100.txt
<jeb.j.cramer@intel.com>
e1000 1/10:
* add new pci ids
<jeb.j.cramer@intel.com>
e1000 2/10:
* Changed flow control defaults to minimize packet drops
<jeb.j.cramer@intel.com>
e1000 3/10:
* Set interrupt delay timer defaults to 0 across all adapters
<jeb.j.cramer@intel.com>
e1000 4/10:
* Newer controllers required manual phy PM enable for WoL to work
<jeb.j.cramer@intel.com>
e1000 5/10:
* 2/3 wire downshift phy bits changed in newer phy revs
<jeb.j.cramer@intel.com>
e1000 6/10:
* Adding notifier to track ifname change for /proc info
<jeb.j.cramer@intel.com>
e1000 7/10:
* Added software interrupt to ensure rx ring is cleaned in case
of previous allocation failures
* Filled netdev->mem_end member
<jeb.j.cramer@intel.com>
e1000 8/10:
* Forcing 1000/fd is easier to undo with ethtool
* Adapters supporting WoL capabilities are now an inclusive versus
exclusive list
<jeb.j.cramer@intel.com>
e1000 9/10:
* Added bullets to changelog
* Whitespace cleanup
<jeb.j.cramer@intel.com>
e1000 10/10:
* Documentation/networking/e1000.txt updates
<angus.sawyer@dsl.pipex.com>
[PATCH] Fix for raid1 against 2.5.43
A cloned bio still needs destructor set (and should clear bi_io_vec as
well), this was broken when Jens rewrote bio_alloc() recently.
This should now be OK with the mempool_alloc failing, and initialises
bi_io_vec = NULL for nr_iovecs = 0.
<torvalds@home.transmeta.com>
Make a polite version of BUG_ON() - WARN_ON() which doesn't
kill the machine.
Damn I hate people who kill the machine for no good reason.
<rddunlap@osdl.org>
[PATCH] fix sysrq typos
<ambx1@neo.rr.com>
[PATCH] PnP Rewrite V0.9 - 2.5.43
The included patch is essentially a Linux Plug and Play Support rewrite. It
contains many significant improvements, including the following:
1.) A Global Plug and Play Layer
- Now drivers do not have to worry about which plug and play
protocol they are using. Calls are made directly to the Linux
Plug and Play Layer and then forwarded to the appropriate
protocol.
- This will make it very easy to integrate ACPI PnP support when
it's ready
2.) A complete Plug and Play BIOS driver
- The Plug and Play BIOS now supports reading and writing of
resource configurations.
- It is now possible to enable disabled PNPBIOS devices. Therefore
the user can safely enable PnP OS support in their BIOS.
3.) Driver Model Integration
- The entire plug and play layer is integrated into the driver model
- The user interface is housed here
- PnP protocols are listed under the bus "pnp"
4.) A powerful global resource configuration interface
- The user can use this to activate PnP devices for legacy and
user-level drivers
- See the documentation for how to configure devices.
5.) Automatic resource allocation for needed devices
6.) A PnP device name database
And many more improvements.
This patch also adds me to the maintainers list, considering the current
PnP maintainer has been inactive for over 2 years now.
<fdavis@si.rr.com>
[IPV4]: Fix CONFIG_NET_FASTROUTE compile.
<kuznet@mops.inr.ac.ru>
[IPv4]: More output path work.
- Change {udp,raw}_sendmsg to ip_append_data, work done
by Maxim Giryaev <gem@asplinux.ru>
- Get rid of ip_build_xmit() and helper functions.
Nobody uses it anymore.
<steve@gw.chygwyn.com>
[NET]: Kill old sklist_foo interfaces.
<yoshfuji@linux-ipv6.org>
net/ipv6/mcast.c: Fix source address selection of MLD Report/Done messages.
<yoshfuji@linux-ipv6.org>
[IPV6]: Several MLD fixes
- Ignore queries for invalid addresses
- MLD for link-local addresses
<steve@chygwyn.com>
[DECNET]: Update to support timeouts.
<davem@redhat.com>
[PATCH] Fix scsi OOPS on bootup
It ends up trying to list_del() from an uninitialized
list_head.
<fdavis@si.rr.com>
[PATCH] drivers/block/xd.c compile
This fixes a 'used but not declared' compile error
<davem@nuts.ninka.net>
net/ipv6/mcast.c: Remove unused variable addr_type.
<dipankar@in.ibm.com>
[PATCH] RCU helper patchset 1/2
This first RCU helper patch adds a read_barrier_depends() primitive
to all archs which is NOP for archs that doesn't require an rmb()
for data dependent reads when writes are ordered using a wmb().
In reality, only alpha requires an rmb(), the rest are NOPs.
It is likely to be necessary in most situations that would use RCU.
Please apply.
Description :
Sometime ago, during a discussion on lock-free lookups, it was
agreed that an additional memory barrier interface,
read_barrier_depends() that is lighter than an rmb(),
is necessary to make sure that data-dependent reads are not
re-ordered over this barrier. For many processors, data-dependency
enforces order, so this interface is a NOP, but for those that don't
(like alpha), it can be a rmb().
For example, the following code would force ordering (the initial
value of "a" is zero, "b" is one, and "p" is "&a"):
CPU 0 CPU 1
b = 2;
memory_barrier();
p = &b; q = p;
read_barrier_depends();
d = *q;
because the read of "*q" depends on the read of "p" and these
two reads should be separated by a read_barrier_depends(). However,
the following code, with the same initial values for "a" and "b":
CPU 0 CPU 1
a = 2;
memory_barrier();
b = 3; y = b;
read_barrier_depends();
x = a;
does not enforce ordering, since there is no data dependency between
the read of "a" and the read of "b". Therefore, on some CPUs, such
as Alpha, "y" could be set to 3 and "x" to 0. rmb()
needs to be used here, not read_barrier_depends().
The original discussion can be found at -
http://marc.theaimsgroup.com/?t=100259422200002&r=1&w=2
Explanation of the need for read_barrier_depends()
can be found at http://lse.sf.net/locking/wmbdd.html
<dipankar@in.ibm.com>
[PATCH] RCU helper patchset 2/2
This adds a set of list macros that make handling of list protected
by RCU simpler. The interfaces added are -
list_add_rcu
list_add_tail_rcu
- Adds an element by taking care of memory barrier (wmb()).
list_del_rcu
- Deletes an element but doesn't re-initialize the pointers in
the element for supporting RCU based traversal.
list_for_each_rcu
__list_for_each_rcu
- Traversal of RCU protected list - takes care of memory barriers
transparently.
<davem@nuts.ninka.net>
[SPARC64]: Disable old cheetah pcache optimization.
<viro@math.psu.edu>
[PATCH] compile fixes
sbpcd.c: removes bogus duplicate definition of sbpcd_lock, use of
CURRENT (we are using private queue) and call of invalidate_buffers() in
->media_changed() (caller does it itself). That went in a changeset from
davej - looks like a merge problem...
sr.c: braino in ifdefed printk - s/disk/cd->disk/. Thanks to jejb
for spotting that one...
<viro@math.psu.edu>
[PATCH] jsfd converted to use of private queue
* switched to private queues
* set ->queue and ->private_data
* switched to use of ->rq_disk
<viro@math.psu.edu>
[PATCH] nbd converted to private queue
* switched to private queues
* set ->queue and ->private_data
* switched to use of ->bd_disk/->rq_disk
* merged private blocksize, etc. arrays into nbd_dev[]
* cleaned up
<viro@math.psu.edu>
[PATCH] stram switched to private queue
* switched to private queues
* set ->queue
<ak@muc.de>
[PATCH] x86-64 updates for 2.5.43
This fixes a few files that got lost with the last merge and merges
with 2.5.43/i386. Only changes architecture specific files.
It depends on one other patch (for linux/ioctl32.h) which I'm sending
separately.
Changes:
- Include missing files (pageattr.c) and Makefile changes
- Update IA32 subsystem. Various small fixes and a big merge
with sparc64.
- Change HZ to 1000
- Merge some of the 2.5.43/i386 profiling changes. No full oprofile yet.
- Fix many warnings
- Update defconfig
- Various other smaller cleanups and bugfixes.
<ak@muc.de>
[PATCH] add linux/ioctl32.h header for 2.5.43
Several ports (x86-64,ppc64,sparc64) which do 32bit ioctl emulation
have functions for drivers to let them define their own translation
handlers. So far it was a bit complicated to use because there
was no standard include file that declares the prototypes for this
stuff. Then drivers and other subsystems can start to declare their own
ioctl translation handlers.
This patch add include/linux/ioctl32.h to fix this.
The actual implementation is in arch specific code. It has been coordinated
with DaveM and Anton.
This patch is needed for the x86-64 merge I sent in separate mail.
There is currently no preprocessor symbol that the drivers can test
to see if they should use this. So far they have to check arch symbols.
This will be fixed later when all 64bit ports that need it add the dynamic
ioctl registration API too (that's ia64,s390x,mips64)
<mochel@osdl.org>
[PATCH] device removal
The problem was that when the refcount hit 0, it was unconditionally
assuming that it had been added, which is wrong. The patch below
corrects that, and fixes the Oops when loading the floppy driver.
<torvalds@home.transmeta.com>
Linux v2.5.44