Ok, there's a largish 2.5.8-pre1 patch out there in testing right now.
It syncs up with various sources over the last two weeks - notably USB,
networking, sparc, net drivers, and the beginning of the -dj tree merge.
Some of the merges are known incomplete (eg some drivers do not compile
etc), but testing is always a good idea.
Changelog appended.
Linus
PS. For people who still haven't noticed - the "I'm leaving Linux and
Transmeta" email was a April fools joke (and not written by me, btw, so
don't credit me for being funny)
-----
Summary of changes from v2.5.7 to v2.5.8-pre1
============================================
<jes@wildopensource.com> (02/03/15 1.473.4.15)
acenic net driver update:
* clean up vlan defines to dramatically reduce number of ifdefs
* re-optimize private structure across cache line boundaries
* fix typo(s) in printk/comments
<kai.reichert@udo.edu> (02/03/18 1.538)
USB printer driver
added HP DeskJet 959C to the quirks list
<paschal@rcsis.com> (02/03/18 1.539)
USB printer driver
added more printer quirks to the list
<david-b@pacbell.net> (02/03/18 1.540)
USB hcd.c, non-HS periodic transfers
This patch adds a missing "break", which prevented low and full
speed periodic transfers from getting through the "hcd" framework
to the driver (OHCI, for now). Sigh... :)
<david-b@pacbell.net> (02/03/18 1.541)
USB mem flags nonpoisonous
This patch completely punts on passing SLAB_POISON, I've
gotten burnt by it once too many. Seems like the slab code
changed somewhere. I've got a separate patch to make the
pci_pool code use CONFIG_DEBUG_SLAB, which I'll send
around.
Meanwhile, I needed this to get at least EHCI to intialize on
a 2.5.7-pre1 system. Please merge.
<david-b@pacbell.net> (02/03/18 1.542)
USB
Comment and documentation cleanups
<david-b@pacbell.net> (02/03/18 1.543)
USB
inline cleanup to save a chunk of memory in usb-ohci
<david-b@pacbell.net> (02/03/18 1.544)
USB echi and Intel ICH
This updates the Philips tweak so that it also applies
to late-model ICH chips from Intel. (Or so I'm told ... :)
That's at least three EHCI implementations known to
behave on Linux. (And one hopes VIA soon too...)
It also cleans up a few comments relating to 64bit DMA;
recent API spec updates make it look like no games are
needed with the PCI DMA mask -- it doesn't change the
segment used by pci_pool allocations, so it just needs
to get turned on if the overall system does 64bit DMA.
I've a query in to see if those Philips/Intel tweaks are
chip quirks, or just something the EHCI 1.0 spec isn't
at all clear about.
<davem@nuts.ninka.net> (02/03/18 1.537.1.1)
On sparc{,64}, use ptrace_check_attach instead of
verifying things by hand.
<davem@nuts.ninka.net> (02/03/18 1.537.1.2)
In Sparc{,64} signal handling, tsk->p_pptr --> tsk->parent
<davem@nuts.ninka.net> (02/03/18 1.537.2.1)
Fix build error on non-x86.
<david-b@pacbell.net> (02/03/18 1.545)
This updates linux/Documentation/usb/proc_usb_info.txt to:
- refer to "usbfs"
- describe the /proc/bus/usb/BBB/DDD files
- more info about the .../drivers and .../devices
- ... generally, gives more information.
This is ever so slightly forward looking in how it describes
bandwidth requirements for high speed periodic transfers,
it's expecting a bugfix patch that's in my queue. (That info
is currently broken/meaningless.)
<johannes@erdfelt.com> (02/03/18 1.546)
[PATCH] uhci.c 2.4.19-pre3 kmem_cache_alloc flags
My previous patch which cleaned up some of the spinlocks, moved one of
the spinlocks around a call to kmem_cache_alloc. It would sometimes
erroneously call it with GFP_KERNEL.
This patch fixes the problem by always calling it with GFP_ATOMIC.
Thanks to Greg for pointing this out to me.
JE
<johannes@erdfelt.com> (02/03/18 1.547)
[PATCH] uhci.c 2.4.19-pre3 erroneous completion callback
uhci.c would call the completion callback when the call to submit_urb
failed. This is a rare situation.
This patch only calls the completion handler if the URB successfully
completed immediately (as in the case of talking to the virtual root
hub).
JE
<johannes@erdfelt.com> (02/03/18 1.548)
[PATCH] uhci.c 2.4.19-pre3 interrupt deadlock
Unfortunately, I left out one line from my spinlock cleanup patch
recently.
As a result, using interrupt URB's could cause a deadlock on SMP
kernels.
This should fix the deadlock Greg reported.
JE
<ganesh@tuxtop.vxindia.veritas.com> (02/03/19 1.549)
USB ipaq driver
Don't submit urbs while holding spinlocks. Not strictly required in
2.5.x, but it's always better to do less while holding a spinlock.
Also a good idea to keep 2.{4,5}.x drivers in sync.
<david-b@pacbell.net> (02/03/19 1.550)
USB usbfs periodic endpoint/bandwidth reporting
This is an updated version of a patch I sent around a
while back. It's against 2.5.7-pre1 (so presumably is
fine on 2.5.7), and addresses feedback against that
earlier patch.
It's bugfixes, mostly for highspeed support, to what
/proc/bus/usb/devices shows:
- Shows isochronous periods correctly (logarithmic
encoding, possibly 1/2/4 microframes if highspeed)
- Likewise for high-speed interrupt periods (similar)
- Makes high bandwidth endpoints look like they
just do bigger packets (up to 3 KBytes/uframe)
- Shows highspeed bandwidth correctlly (80% reserved,
vs 90% reserved for full/low speed).
<greg@kroah.com> (02/03/19 1.551)
USB proc_usb_info.txt
documented the fact that the interval is not always reported in ms.
<davem@nuts.ninka.net> (02/03/19 1.537.1.3)
Update sparc64 defconfig.
<davem@nuts.ninka.net> (02/03/19 1.537.2.2)
Netfilter enhancement from Harald Welte and Netfilter team.
Add destroy callback to ip_conntrack_helper, to be used by L4
protocol trackers.
<oliver@oenone.homelinux.org> (02/03/19 1.552)
USB hpusbscsi driver
Port changes from 2.4:
We do request_sense ourselves to comply with
the scanner command set
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.3)
Remove obsolete confusing instructions on tcp_max_syn_backlog
from IPv4 sysctl documentation.
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.4)
Make pkt_sched.h:PSCHED_TDIFF_SAFE behave sane when measuring
large time intervals.
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.5)
Remove unused field from TCP struct open_request.
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.6)
Do not fail creating _new_ NOARP entry with EPERM.
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.7)
Old bug in skbuff.c, found by someone, but was lost.
In __pskb_pull_tail, pskb_expand_head return value test
was inverted.
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.8)
IPv4 FIB routing fixes:
- fix device leakage in multipath
- fix oops due to race by adding spinlock
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.9)
In IPv4 ICMP:
pskb_pull really may reallocate packet after the check
for 8 bytes was removed from ip_input, so set icmp
header pointer after pskb_pull call not before.
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.10)
Fix for ipv4 tunnel devices:
- do not make path mtu discovery, when it is disabled :-)
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.11)
IP input fixes:
- no need to check for pskb_may_pull() in ip_local_deliver_finish, header
is guaranteed to be at right place here.
- remove cleaing ip_summed with IP options, it was required due to
broken eth_copy_and_csum, but we do not use it any more
and it happens to break loopback
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.12)
Terrible bug in ipv4/route.c, mis-sized ip_rt_acct leads to
complete memory corruption.
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.13)
TCP Input fixes:
1) Two bugs noticed by Pasi:
- Wrong rtt update interval
- Forgot to clear retrans_stamp when entering established
state
2) Missing LAST_ACK case of processing segment text (step 7) in
tcp_rcv_state_process.
3) Remove "final cut" code, purism is good not all the times. :-)
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.14)
UDP fixes:
- respect multicast interface when connecting
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.15)
IPV6 addrconf exploit fix:
- stop external DoS attack feeding lots of IPv6 prefixes
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.16)
IPv6 neighbour discovery fixes:
- Answer to neighbour solicitations on SIT, otherwise
freebsd does not want to talk to us.
- Fix wrong structure nd_msg and... use it :-)
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.17)
TCP ipv6 fixes:
- Fix open_request lookup bug that was already fixed in ipv4
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.18)
Port of 2.2.x AF_PACKET bug fix.
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.19)
Fix bug in sch_prio.c where wrong handle was
being dumped.
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.20)
In sch_sfq.c, allow to descrease length of queue
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.21)
Add new sysctl, medium_id, to devinet.
It is used to differentiate the devices by the medium
they are attached to. It is used to change proxy_arp behavior:
the proxy arp feature is enabled for packets forwarded between
two devices attached to different media.
<kuznet@ms2.inr.ac.ru> (02/03/19 1.537.2.22)
Allow to bind to an already in use local port
during connect when the connection will still have a unique
identity. Fixes port space exhaustion, especially in web
caches.
Initial work done by Andi Kleen.
<davem@nuts.ninka.net> (02/03/19 1.537.2.23)
Fix mis-merge of TCP_LAST_ACK fix.
<davem@nuts.ninka.net> (02/03/19 1.537.2.24)
Update port-allocation changes to coincide with struct sock
splitup.
<davem@nuts.ninka.net> (02/03/19 1.537.2.25)
Update port-allocation changes to coincide with struct sock
splitup.
<davem@nuts.ninka.net> (02/03/19 1.537.2.26)
Kill unused local var in af_inet.c:inet_stream_connect
<kanojsarcar@yahoo.com> (02/03/19 1.537.1.4)
Move VPTE_BASE_foo definitions to common
header instead of scattered all over assembly files.
<eli.kupermann@intel.com> (02/03/20 1.537.3.2)
e100 net driver update:
1) This patch provides fix for "wake on arp" and "wake on unicast"
functionality when card is suspended by power management. When e100_suspend
was called for the device that is in netif_running state the load filter
command was executed in the asynchronic mode and the order of actions
required to put device into wake up enabled mode was broken.
The fix enables to execute WOL configure and load filter commands in the
synchronic mode despite of fact that device is in netif_running state. The
exec_non_cu_command uses the driver_isolated flag to identify this
situation.
2) add EXPORT_NO_SYMBOLS (yay Intel, you have come so far... :))
3) bump version to 2.0.25-pre1
<jgarzik@mandrakesoft.com> (02/03/20 1.537.3.3)
Merge ethtool initiate-nic-self-test ioctl, and support for it in e100 net drvr.
Contributed by Eli Kupermann @ Intel, modified by me.
<adilger@clusterfs.com> (02/03/20 1.537.4.1)
Add three scripts for BK users, to Documentation/BK-usage:
bzsend: good for users who want to send reviewable BK patches
bz64wrap, unbz64wrap: bzip2 uuencoding wrappers for encapulating BK patches
<tvignaud@mandrakesoft.com> (02/03/20 1.537.4.2)
Remove silly overdependency on Perl 5.6.1 in BK helper scripts.
<jgarzik@mandrakesoft.com> (02/03/20 1.537.3.4)
Add support file e100_test to e100 net driver. Missed in earlier merge.
<jgarzik@mandrakesoft.com> (02/03/20 1.537.3.5)
Merge dl2k gigabit ethernet driver update vendor:
* add rio_timer to watch rx condition
* move poll initiation to rx refill loop
* use del_timer_sync to avoid race (me)
* CodingStyle cleanups (me)
<jgarzik@mandrakesoft.com> (02/03/20 1.537.3.6)
Merge orinoco_plx wireless driver pci ids from 2.4.x.
<jgarzik@mandrakesoft.com> (02/03/20 1.537.4.3)
Update rocketport serial driver:
* remove linux 2.1.x backwards compat code (William Stinson)
* remove ENABLE_PCI define, use CONFIG_PCI instead
* no need to enclose MODULE_xxx in ifdef MODULE
<jgarzik@mandrakesoft.com> (02/03/20 1.537.4.4)
Add two AC97 codec ids to old OSS ac97_codec driver.
Contributed by Peter Christy.
<k.kasprzak@box43.pl> (02/03/20 1.537.3.7)
de620 net driver janitor fixes:
* free_irq on error
* check request_region error value
<silicon@falcon.sch.bme.hu> (02/03/20 1.537.4.5)
Update munish WAN driver to not kfree memory multiple times.
<wstinson@infonie.fr> (02/03/20 1.537.4.6)
Fix DocBook documentation for ALSA merge,
basically s#drivers/sound#sound/oss#
<jgarzik@mandrakesoft.com> (02/03/20 1.537.3.8)
Revert epic100 net driver power sequence "fix", it broke some boards.
<davem@nuts.ninka.net> (02/03/20 1.537.2.27)
Fix reverse logic in checking sock_writeable return
in UDP case. I note in passing that the TCP case is
wrong because TCP does not use sock_writeable()s
<davem@nuts.ninka.net> (02/03/20 1.537.1.5)
In sparc64/ebus, handle machines with both RIO and
non-RIO EBUSes correctly.
<greg@kroah.com> (02/03/20 1.551.1.1)
USB hub
changed the interval for checking if the device is connected yet or not.
Thanks to Itai Nahshon <nahshon@actcom.co.il> for the information.
<davem@nuts.ninka.net> (02/03/20 1.537.1.6)
On sparc64 Schizo PCI controllers, there is no inofixup
to apply during IRQ building.
<davem@nuts.ninka.net> (02/03/20 1.537.1.7)
On sparc64, handle assigning ROM and non-standard resources
properly.
<davem@nuts.ninka.net> (02/03/20 1.537.1.8)
In Sun GEM/HME drivers, if OpenBoot firmware is not
available (e.g. i386) fetch the ethernet MAC address
from the vital-product data contained in the PCI
ROM of the card.
<davem@nuts.ninka.net> (02/03/20 1.537.1.9)
Model Sparc64 pci_assign_resource more closely to the
implementation in drivers/pci/setup-res.c to make it
easier to track bugs.
Fix calculation of mem_space end on Sparc64 Schizo PCI controllers.
The decode register is set up for the whole 4GB even though the
top 2GB is reserved for DMA to/from main memory (ie. IOMMU translated)
<davem@nuts.ninka.net> (02/03/21 1.537.1.10)
Merge 2.4.x Sun GEM/HME net driver fixes.
<davem@nuts.ninka.net> (02/03/21 1.537.1.11)
Remove debugging printk while probing MAC address.
Unregister netdev on shutdown before iounmapping registers.
<davem@nuts.ninka.net> (02/03/21 1.537.1.12)
Sun HME/GEM driver probing cleanups.
<greg@kroah.com> (02/03/21 1.554)
USB visor driver
Added support for the Palm m130 device, thanks to Udo Eisenbarth
<udo.eisenbarth@web.de> for the information.
<greg@kroah.com> (02/03/21 1.555)
USB serial driver core
- Moved all manipulation of port->open_count into the core. Now the
individual driver's open and close functions are called only when the
first open() and last close() is called. Making the drivers a bit
smaller and simpler.
- Fixed a bug if a driver didn't have the owner field set.
<greg@kroah.com> (02/03/21 1.556)
USB serial drivers
changes due to open_count being handled by the usb-serial core code.
<oliver@oenone.homelinux.org> (02/03/21 1.557)
USB kaweth driver
added optus@home uep1045a driver to the list of supported devices.
<davem@nuts.ninka.net> (02/03/21 1.537.2.28)
Add missing KERN_foo printk specifiers to networking.
Based upon a patch from Denis Vlasenko.
<uzi@uzix.org> (02/03/21 1.537.1.13)
Merge 2.4.x VGER sparc32 changes into 2.5.x
<laforge@gnumonks.org> (02/03/21 1.537.2.29)
Add configure Configure.help message and
respective config option for CONFIG_IP_NF_NAT_LOCAL
<wstinson@infonie.fr> (02/03/21 1.537.1.14)
Remove explicit initialization of static vars to zero
in Sparc ports.
<cruault@724.com> (02/03/21 1.537.2.30)
Make sure outgoing ICMP and TCP resets
use the most uptodate value of ip_default_ttl sysctl.
<davem@nuts.ninka.net> (02/03/21 1.537.1.15)
Move bootstr_valid/bootstr_buf back into .data section.
Add comment explaining that why these must not be moved into the
.bss section.
<davem@nuts.ninka.net> (02/03/22 1.537.2.31)
Code (and commentary) in SYN-RECEIVED processing
assumes that it cannot be reached in the crossed SYN case.
This is wrong if the original SYNs came from a malicious packet
generator third party. This can result in a 4 minute ACK
fight if the sequence numbers are correct.
The fix is the verify the ACK before we do anything else, which
should cover all cases.
This bug was discovered by Casper Dik.
<petkan@mastika.> (02/03/22 1.558)
USB pegasus driver
fix problem which cause hotplug/unplug crash the kernel
<petkan@mastika.> (02/03/22 1.559)
USB
added rtl8150 usb ethernet driver
<davem@nuts.ninka.net> (02/03/22 1.537.2.32)
Bump TcpPassiveOpens when tcp_create_openreq_child succeeds.
We have not been bumping this since we create openreqs in TCP,
ie. some 6 years ago. :-)
<stewart@inverse.wetlogic.net> (02/03/22 1.560)
[PATCH] Re: [PATCH] hiddev code and docs cleanup
I took some time to clean up the code a little, and to add the new
calls to the documentation. See patch below.
--
Paul
<greg@kroah.com> (02/03/23 1.561)
USB visor driver
added support for the Palm i705 device.
thanks to Thomas Riemer for the information.
<johannes@erdfelt.com> (02/03/25 1.562)
[PATCH] 2.4.19-pre3 uhci.c zero packet
Don't know if you saw this on the list or not. Here's a patch which
fixes zero packet support for uhci.c
- Setup the TD correctly, we weren't shifting the null data size
- Only add the zero packet if the transfer was a multiple of the
endpoint of the size
JE
<davem@nuts.ninka.net> (02/03/25 1.537.1.16)
Merge 2.4.x sparc64 PCI IRQ routing fixes into 2.5
<uzi@uzix.org> (02/03/25 1.537.1.17)
Sparc32 cleanups.
<david-b@pacbell.net> (02/03/26 1.563)
USB ohci-hcd update
It updates the "ohci-hcd" driver to address two issues, one
of which could potentially have caused oopses. First, it takes
out calls to usb_dec_dev_use() on the submit error path; that's
already done elsewhere. (Noted by Georg Acher.) Second,
it removes some pointless diagnostics. One is for schedule
overrun interrupts, the other is for accessing the frame counter.
<petkan@mastika.lnxw.com> (02/03/26 1.564)
USB pegasus driver
semaphore cleanup and proper link detection
<davem@nuts.ninka.net> (02/03/26 1.537.1.18)
Do the slot mapping adjustment to PROM interrupt
property in pci_intmap_match even if pbm->num_pbm_intmap is
zero.
<davem@nuts.ninka.net> (02/03/26 1.537.2.33)
Fix device list locking.
Based upon a patch from Maxim Krasnyansky.
<laforge@gnumonks.org> (02/03/26 1.537.2.34)
Big netfilter newnat patch for 2.5.7:
- support for multiple expected connections
(necessary for protocols like H.323, SIP, PPTP)
- helper-definable limit of unconfirmed expectations
- timeouts for expectations
- full graph of connection relations, even after expectation confirmed
- various changes in the API towards conntrack and NAT helper
- automatic conntrack helper loading when at helper is loaded
- NAT mangling of TCP SACK in case of sequence number alteration
(no need to delete SACKPERM anymore, I hope Alexey is happy now)
<davem@nuts.ninka.net> (02/03/26 1.537.1.19)
SunHME driver updates:
- Fix SMP locking throughout
- Deal with hw bug involving lost writes to rxring base address
- Disable Never Give Up mode on transmitter, set TX attempts
limit to 16 (the default). This deals with a possible hang
when the TX deadlocks with the PHY when a jabber occurs.
- Delete some dead code.
<johannes@erdfelt.com> (02/03/27 1.565)
[PATCH] USB uhci bugfix
The issue was that the poisoning would posing the data *before* it gave
it to us as well, but I didn't clear out a pointer (qh->urbp) and later
dereferenced it.
Thanks for the report again and for the pci_pool patch so I could
troubleshoot it.
JE
<david-b@pacbell.net> (02/03/27 1.566)
USB ohci-hcd driver update
- bugfix: control endpoints can't stall
- bugfix: remove bogus intr unlink optimization,
by sharing intr/iso code
- bugfix: iso submit uses urb->interval
- removed iso urb->next ring logic
(belongs in hcd layer if anywhere)
- simplify/shorten/correct completion handling
- in debug, labels setup packets as such
- bring CVS ids back up to date
<david-b@pacbell.net> (02/03/27 1.567)
USB core sanity check
Periodically folk have run into problems where usb-ohci oopses
due to device refcount bugs ...
This is a minor patch to move the sanity check out of usb-ohci
into the generic bits of usbcore. There are comments that
suggest a path for a more comprehensive approach too.
Applies cleanly against 2.5.7 and I've been testing with it
for a while. I can't think of any reason it shouldn't also go
into 2.4, beyond the patch not applying cleanly there ... :)
<davem@nuts.ninka.net> (02/03/27 1.537.2.35)
Tigon3 net driver fixes:
- Make use of pci_unmap_xxx storage optimization.
- In tg3_get_invariants, fix cacheline_sz_reg shifts.
On 5703 chips, if latency timer is less than 64, set
it to 64.
- In tg3_test_dma, only run the actual test on 5700/5701
chips. Aparently 5702/5703 revs have some problems.
<davem@nuts.ninka.net> (02/03/27 1.537.1.20)
In SBUS probing, handle empty SBUS correctly.
<greg@kroah.com> (02/03/27 1.568)
USB serial console support added
<davem@nuts.ninka.net> (02/03/27 1.537.2.36)
Tigon3 net driver bug fix:
- ETHTOOL_GLINK handling forgets to return on success
<greg@kroah.com> (02/03/28 1.569)
USB serial config.in changes
- make serial console only selectable if the code is not a module.
- removed lots of CONFIG_EXPERMENTAL dependancies as the drivers have been stable for some time now.
<greg@kroah.com> (02/03/28 1.570)
USB uhci bug fix.
use proper GFP flag setting for submitting a urb.
<greg@kroah.com> (02/03/28 1.571)
USB io_edgeport driver update
fixes to let io_edgeport work properly as a usb serial console
<ganesh@veritas.com> (02/03/28 1.572)
USB serial core
Module count of a serial converter driver is currently not
decremented if a disconnect happens while the tty is held open.
The fix is to close the device in usb_serial_disconnect() so that
module refcounts are properly updated.
<david-b@pacbell.net> (02/03/28 1.573)
USB audio driver
- Makes "audio" set the urb->interval in its periodic URBs,
so it can submit through the hcd framework.
<david-b@pacbell.net> (02/03/28 1.574)
USB hcd driver updates
- Nitpickey bugfix to root hub config descriptors ... can't use
the same one for high and full speed, since the encoding
is different (255 ms FS == 0xff, 256 ms HS == 0x12).
- Related, force period to 1/4 second rather than doing
any sanity checking for the roothub timer (from Georg)
- Don't "giveback" urbs on submit path errors (from Georg)
... means they don't get completion callbacks
- Additional error checks on URB data (from Georg)
- Uses <linux/completion.h> for unlink synchronization
- The "already unlinking" error case is reported like other
unlinking errors (not as success)
- Ripped out urb->next handling ... it wasn't compatible
with the ISO loop model, and at this point I believe it
should be completely replaced with queuing urbs inside
of the HCDs. (Every HCD handles it for ISO, UHCI needs
a magic flag to enable it for bulk ...)
<stewart@wetlogic.net> (02/03/28 1.575)
USB hiddev interface
updated the version number due the previous changes
<davem@nuts.ninka.net> (02/03/28 1.537.1.21)
Sparc SBUS fix: Make for_all_sbusdev work with an empty SBUS.
<davem@nuts.ninka.net> (02/03/30 1.537.2.37)
net/core/sock.c needs linux/tcp.h to get at TCP state macros.
<spse@secret.org.uk> (02/04/01 1.576)
[PATCH] Update to konicawc driver
This patch against 2.5.7 fixes an oops and a memleak in the konicawc driver
and also adds an option to set the FPS.
<viro@math.psu.edu> (02/04/02 1.537.5.1)
[PATCH] initrd issue
Fix initrd problem that appeared back in 2.5.2-pre6 when kdev_t type
changed and comparison function was incorrectly converted..
<dalecki@evision-ventures.com> (02/04/02 1.537.5.2)
[PATCH] 2.5.7 IDE 23
- Support for additional Promise controller id's (PDC20276).
- Remove code duplication between do_rw_taskfile and do_taskfile.
This will evolve into a more reasonable ata_command() function
finally. The ata_taskfile function has far too many arguments, but
I favour this over having two different code paths for getting
actual data to the drive.
<dalecki@evision-ventures.com> (02/04/02 1.537.5.3)
[PATCH] 2.5.7 IDE 24
- Push BAD_DMA_DRIVE and GOOD_DMA_DRIVE to the ide-pmac.c file, since this is
the only place where those get used.
- Kill unused fields from the ide_task_s structure. In esp. we pass a task
attached to a request and not the other way around!
- Rename hwif field to channel in struct ide_drive_s.
- Move the request queue to the level where proper serialization has to happen
anyway - the channel structure.
<dalecki@evision-ventures.com> (02/04/02 1.537.5.4)
[PATCH] 2.5.7 IDE 25
- Replace the task_io_reg_t with the simple u8. There is no need to obfuscate
the code more then necessary.
- kill some unnecessary type definitions out from hdreg.h.
- Add proper attributes to register files in hdreg.h.
- Don't use raw arrays for tfRegister and hobRegister in ide_task_s. Use out
nice global structures describing the fields in them. This allows to kill
the following defines:
IDE_DATA_OFFSET
IDE_FEATURE_OFFSET
IDE_NSECTOR_OFFSET
IDE_SECTOR_OFFSET
IDE_LCYL_OFFSET
IDE_HCYL_OFFSET
IDE_SELECT_OFFSET
IDE_COMMAND_OFFSET
and many many others.
- Please have a look at the following in pdc4030.c. It couldn't have worked!
This has been fixed in one go with the above change:
memcpy(args.hobRegister, NULL, sizeof(struct hd_drive_hob_hdr));
- Kill the redundant *_REG_HOB definitions. They don't help readability in any
way.
<dalecki@evision-ventures.com> (02/04/02 1.537.5.5)
[PATCH] 2.5.7 IDE 26
- Mark all members of structures, which get jiffies assigned or involved in
ugly timeout calculations with the prefix PADAM_ for easy spotting. This is
Polish for "I'm falling down" or "This brings me to the knees" or slag
comment for "What a sh..". Please be assured that it doesn't sound vulgar.
Please grep for it to see immediately why this nomenclature is justified.
- Rename hwifs_s to ata_channel and eliminate ide_hwifs_t as well as the HWIF
macro. OK this step makes this patch rather big.
<dalecki@evision-ventures.com> (02/04/02 1.537.5.6)
[PATCH] 2.5.7 IDE 27
- Make for less terse error messages in ide-tape.c.
- Replaced all timecomparisions done by hand with all the proper timer_after()
commands.
- Remove the drive niec1 mechanisms alltogether. There are several reasons for
this:
1. The code implementing it is nonintelliglible and therefore propably
broken.
2. If we have to invent somethig about sceduling drive IO, it should be done
on the BIO level.
3. We may in fact interleave with the IO sceduling on the upper layers and
the results of two overlapping signal filters overlapped with each other can
be disasterous to the overall throughput. (In fact they *are* most of the
time.)
4. It was not working if you had intermixed modes on different drives
DMA versus PIO.
5. Our goal is to have a driver which is able to share the badwidth
properly and shouldn't needing this kind of "tuning".
- Remove unused nice2 from disk struct.
- Rename channel member of ata_channel to unit and device to dev to
just prevent wrong interpretations. This prevents constructs like
channel->channel...
<dalecki@evision-ventures.com> (02/04/02 1.537.5.7)
[PATCH] 2.5.7 IDE 28a
- Apply Pavel Macheks suspend resume double resume fix.
- Finally remove the busy field for ata_operations and replace it with
MOD_INC_USE_COUNT and MOD_DEC_USE_COUNT.
- Fix ali15xx chipset support by removing initialization differences,
apparently caused by mislead interpretation of the specs or a mismatch
between the specification and reality.
- Guard calls to ide_set_handler with checks to see whatever the previously
installed IRQ handler already served it's purpose.
- Convert timeout checks on poll_timeout to the time_before() interface.
- Consolidate the two different IRQ handlers for multi mode PIO writes into
one. The problems remain the same but at least now we will only have to
tangle one single problem.
<akpm@zip.com.au> (02/04/02 1.537.5.8)
[PATCH] ext2_fill_super breakage
In 2.5.7 there is a thinko in the allocation and initialisation
of the fs-private superblock for ext2. It's passing the wrong type
to the sizeof operator (which of course gives the wrong size)
when allocating and clearing the memory.
Lesson for the day: this is one of the reasons why this idiom:
some_type *p;
p = malloc(sizeof(*p));
...
memset(p, 0, sizeof(*p));
is preferable to
some_type *p;
p = malloc(sizeof(some_type));
...
memset(p, 0, sizeof(some_type));
I checked the other filesystems. They're OK (but idiomatically
impure). I've added a couple of defensive memsets where
they were missing.
<viro@math.psu.edu> (02/04/02 1.537.1.24)
[PATCH] romfs inode allocation
Obvious romfs fix.
<viro@math.psu.edu> (02/04/02 1.537.1.25)
[PATCH] conditional system call cleanup
This version of sys_nfsservctl() fix hadn't made DaveM complain.
<viro@math.psu.edu> (02/04/02 1.537.1.26)
[PATCH] minixfs cleanups (1/4)
Move minix-private stuff to fs/minix/minix.h.
<viro@math.psu.edu> (02/04/02 1.537.1.27)
[PATCH] minixfs cleanups (2/4)
Clean up the write_super-related code in minixfs
<viro@math.psu.edu> (02/04/02 1.537.1.28)
[PATCH] minixfs cleanups (3/4)
Kill BKL in minix/itree* (similar to ext2 patches)
<viro@math.psu.edu> (02/04/02 1.537.1.29)
[PATCH] minixfs cleanups (4/4)
Clean up minix/bitmap.c
<viro@math.psu.edu> (02/04/02 1.537.1.30)
[PATCH] set_blocksize() in JFS
Use sb_set_blocksize() in JFS instead of trying to do it by hand.
<viro@math.psu.edu> (02/04/02 1.537.1.31)
[PATCH] hfs compile fix
Fix typo
<viro@math.psu.edu> (02/04/02 1.537.1.32)
[PATCH] restoring block size upon umount
get_sb_bdev() stores original block size in ->s_old_blocksize and
kill_block_super() restores it.
This kills 99% of crap with "oh, I've mounted/umounted that device and
its behaviour had changed" (remaining 1% can be dealt in pretty similar
ways; ideally I'd like to see ioctls that get/set block size dead and
gone).
<viro@math.psu.edu> (02/04/02 1.537.1.33)
[PATCH] fsync_bdev() conversion
Bunch of places converted from fsync_dev/invalidate_buffers to
fsync_bdev/invalidate_bdev.
<viro@math.psu.edu> (02/04/02 1.537.1.34)
[PATCH] brw_kiovec() converted to struct block_device *
brw_kiovec() and ll_rw_kiovec() switched to struct block_device *.
<torvalds@penguin.transmeta.com> (02/04/02 1.537.1.35)
update version and defconfig
<davem@nuts.ninka.net> (02/04/02 1.537.2.38)
In tcp_v4_send_reset, use inet_sk to get at
ttl of tcp_socket.
<haveblue@us.ibm.com> (02/04/03 1.537.1.37)
[PATCH] BKL reduction in do_exit
Push BKL down to the (few) routines that actually need it,
remove it from the do_exit() path.
<davej@suse.de> (02/04/03 1.537.1.38)
[PATCH] Hyperthreading binfmt.
Another from 2.4, see comments for details
<davej@suse.de> (02/04/03 1.537.1.39)
[PATCH] fix broken asm constraint
Simple compile fix.
<greg@kroah.com> (02/04/03 1.537.1.40)
[PATCH] small fix for mpparse.c
Here's a very tiny bugfix for arch/i386/kernel/mpparse.c in the
2.4.19-pre2 kernel. It fixes the problem if there is an error in the
MP_processor_info() function where the mpc_apicid value is greater than
MAX_APICS, then we need to decrement the number of valid processors
before we return (the number was just incremented before the check.)
The patch was written by James Cleverdon.
<davej@suse.de> (02/04/03 1.537.1.41)
[PATCH] AGPGART capability handling cleanup
Clean up capability handling in AGPGart.
This came forward from 2.4
<davej@suse.de> (02/04/03 1.537.1.42)
[PATCH] EFI GUID partition table support.
EFI GUID partition table support from Matt Domsch
<davej@suse.de> (02/04/03 1.537.1.43)
[PATCH] about locations of various sound files.
introduction of ALSA moved some files around.
This updates various references.
<davej@suse.de> (02/04/03 1.537.1.44)
[PATCH] Support for ITE interrupt router
Support for an extra interrupt router.
<adam@nmt.edu> (02/04/03 1.537.1.45)
[PATCH] 3ware driver update for 2.5.8-pre1
Self explanatory driver update from vendor.
<davej@suse.de> (02/04/03 1.537.1.46)
[PATCH] watchdog API documentation.
<davej@suse.de> (02/04/03 1.537.1.47)
[PATCH] eicon driver was sleeping with lock held.
<davej@suse.de> (02/04/03 1.537.1.48)
[PATCH] extra codepage support.
Another forward port from 2.4
<davej@suse.de> (02/04/03 1.537.1.49)
[PATCH] AMD ELAN support.
Add support for AMD Elan.
(More ELAN patches to follow which rely on CONFIG_MELAN)
<davej@suse.de> (02/04/03 1.537.1.50)
[PATCH] Cyrix irq router tweak
Another small change from 2.4
<davej@suse.de> (02/04/03 1.537.1.51)
[PATCH] Cyclades driver region cleanup
<davej@suse.de> (02/04/03 1.537.1.52)
[PATCH] Various completion users.
These files use completion, but don't include header.
<davej@suse.de> (02/04/03 1.537.1.53)
[PATCH] Document an errata workaround in apic code.
<davej@suse.de> (02/04/03 1.537.1.54)
[PATCH] faster update_atime.
Another 2.4 forward port, original from Andrew Morton.
<davej@suse.de> (02/04/03 1.537.1.55)
[PATCH] DMI scanner update.
Various quirks from 2.4 and other sources.
<davej@suse.de> (02/04/03 1.537.1.56)
[PATCH] AMD Elan uses slightly different clock freq
<davej@suse.de> (02/04/03 1.537.1.57)
[PATCH] export rbtree routines for modules.
<davej@suse.de> (02/04/03 1.537.1.58)
[PATCH] Detect get_block() errors in block_read_full_page()
Originally from Anton Altaparmakov..
This causes errors from get_block() in block_read_full_page() to be
detected and handled properly (by setting page error flag). Without the
patch the page (or parts of the page) will contain random data on
get_block() failing without any form of error being signalled which can
be catastrophic for filesystems using block_read_full_page() for
accessing their metadata. And for normal data it would mean the user
would see random data instead of what they expected.
<davej@suse.de> (02/04/03 1.537.1.59)
[PATCH] add AMD Elan resources.
<davej@suse.de> (02/04/03 1.537.1.60)
[PATCH] Extra cards support for MOXA driver
<davej@suse.de> (02/04/03 1.537.1.61)
[PATCH] faster kiobuf init.
Originally from Intel, has been around various vendor trees for a while.
Aparently worth a noticable speed up in some applications.
<davej@suse.de> (02/04/03 1.537.1.62)
[PATCH] Fix up broken do while macros.
<davej@suse.de> (02/04/03 1.537.1.63)
[PATCH] document new address space operations.
<davej@suse.de> (02/04/03 1.537.1.64)
[PATCH] fix up broken comment delimiters.
<davej@suse.de> (02/04/03 1.537.1.65)
[PATCH] proc race on task_struct->sig
Originally from Chris Mason <mason@suse.com>..
collect_sigign_sigcatch can race against exit_sighand. I haven't been
able to reproduce it, but I think it causes the oops reported in the
'Kernel Hangs 2.4.16 on heavy io Oracle Tivolie TSM' thread.
<davej@suse.de> (02/04/03 1.537.1.66)
[PATCH] silence DVD_INVALIDATE_AGID output.
Ancient patch from Jens to hush certain drives which generate a lot of
printk noise when playing DVDs.
<davej@suse.de> (02/04/03 1.537.1.67)
[PATCH] Simple boot flag specification support.
Originally by Alan. It was in the 2.4-ac tree for ages, much tweaked
by myself and Thomas Hood. Its main purpose is for skipping the full
memory test and other time-wasting diagnostics on reboot.
<davej@suse.de> (02/04/03 1.537.1.68)
[PATCH] sonypi driver update from 2.4
<davej@suse.de> (02/04/03 1.537.1.69)
[PATCH] group #include's together in x86 ioremap.c
<davej@suse.de> (02/04/03 1.537.1.70)
[PATCH] Gameport patch for drivers/sound/mad16.c
From: Michael Haardt <michael@moria.de> for 2.4, munged by me.
the MAD16 driver is able to enable/disable its gameport, but it does
not register it in the input subsystem. The appended patch against
2.4.19-pre4 fixes that and also allows to enable to gameport via kernel
command line, not only when loaded as module.
<davej@suse.de> (02/04/03 1.537.1.71)
[PATCH] CREDITS updates
<davej@suse.de> (02/04/03 1.537.1.72)
[PATCH] Various typo fixes.
Mostly harmless but 1-2 break compilation.
<davej@suse.de> (02/04/03 1.537.1.73)
[PATCH] region handling cleanups for tpqic02
<davej@suse.de> (02/04/03 1.537.1.74)
[PATCH] UDF write support problem in 2.5.7
From Peter Osterlund <petero2@telia.com>
I can't get UDF write support to work in kernel 2.5.7 or 2.5.7-pre2.
The problem is that linux/config.h is not included, so CONFIG_UDF_RW
is undefined and the driver is compiled without write support. This
patch fixes my problem:
<davej@suse.de> (02/04/03 1.537.1.75)
[PATCH] unnecessary includes.
A few follow ups to the cleanup done circa 2.5.4
<davej@suse.de> (02/04/03 1.537.1.76)
[PATCH] Update file list in INDEX
<davej@suse.de> (02/04/03 1.537.1.77)
[PATCH] PCI IDS update.
Mostly from 2.4. Adds some new ones, and removes some dupes.
<davej@suse.de> (02/04/03 1.537.1.78)
[PATCH] Update bigphysarea URL
<davej@suse.de> (02/04/03 1.537.1.79)
[PATCH] watchdog nowayout for i810-tco
Originally from Matt Domsch.
Adds a nowayout module option to override CONFIG_WATCHDOG_NOWAYOUT
From 2.4
<davej@suse.de> (02/04/03 1.537.1.80)
[PATCH] videodev fixups / generic usercopy helper
Originally from Gerd...
I've just noticed that a hole left in the recent changes which should
allow the usb v4l drivers to unregister with open file handles. The
drivers itself handle it just fine, but video_generic_ioctl() will barf
when called on unregistered devices. Oops.
One way to fix this is to expect drivers call the helper function and
pass a pointer for the function doing the actual work, i.e. handle it
this way:
driver_ioctl(inode,file,cmd,userptr)
-> video_usercopy(inode,file,cmd,userptr,func)
copy_from_user(...)
-> func(inode,file,cmd,kernelptr);
copy_to_user(...)
Patch against 2.5.7-pre2 below. It updates videodev.[ch] and adapts
usbvideo.c to show how the driver changes will look like.
Note that this change makes the usercopy helper function a very generic
one, it probably could be used for other drivers to (as long as the API
has sane magic numbers based on _IO*(...) defines) as there is no
video4linux-related stuff in there any more. So we might think of
renaming it an moving it to some more central place (fs/ioctl.c maybe).
<davej@suse.de> (02/04/03 1.537.1.81)
[PATCH] watchdog nowayout for acquirewdt
Originally from Matt Domsch.
Adds a nowayout module option to override CONFIG_WATCHDOG_NOWAYOUT
From 2.4
<davej@suse.de> (02/04/03 1.537.1.82)
[PATCH] printk levels for vme_scc driver
<davej@suse.de> (02/04/03 1.537.1.83)
[PATCH] watchdog nowayout for eurotechwdt
Originally from Matt Domsch.
Adds a nowayout module option to override CONFIG_WATCHDOG_NOWAYOUT
From 2.4
<davej@suse.de> (02/04/03 1.537.1.84)
[PATCH] watchdog nowayout for ib700wdt
Originally from Matt Domsch.
Adds a nowayout module option to override CONFIG_WATCHDOG_NOWAYOUT
From 2.4
<davej@suse.de> (02/04/03 1.537.1.85)
[PATCH] more then enough typos.
s/more then/more than/
<davej@suse.de> (02/04/03 1.537.1.86)
[PATCH] remove dead comment
Original from William Lee Irwin III <wli@holomorphy.com>
<davej@suse.de> (02/04/03 1.537.1.87)
[PATCH] __init/__exit does nothing in prototypes
<davej@suse.de> (02/04/03 1.537.1.88)
[PATCH] reiserfs tools update.
Recommended version bump.
<davej@suse.de> (02/04/03 1.537.1.89)
[PATCH] Remove guess from bttv docs.
<davej@suse.de> (02/04/03 1.537.1.90)
[PATCH] More verbosity in VIA tweak
From 2.4
<davej@suse.de> (02/04/03 1.537.1.91)
[PATCH] watchdog nowayout for advantechwdt
Originally from Matt Domsch.
Adds a nowayout module option to override CONFIG_WATCHDOG_NOWAYOUT
From 2.4
<davej@suse.de> (02/04/03 1.537.1.92)
[PATCH] updated documentation for w9966 driver.
From 2.4
<davej@suse.de> (02/04/03 1.537.1.93)
[PATCH] extra sanity checks for mempool
<davej@suse.de> (02/04/03 1.537.1.94)
[PATCH] khttpd logs wrong debug message on leaving function.
<davej@suse.de> (02/04/03 1.537.1.95)
[PATCH] nbd compile fix.
<davej@suse.de> (02/04/03 1.537.1.96)
[PATCH] strtok -> strsep in adfs
<davej@suse.de> (02/04/03 1.537.1.97)
[PATCH] remove workaround for old binutils.
Modern binutils spits out a warning, fixing this broke older binutils.
I guess we can do this in 2.5 ?
<davej@suse.de> (02/04/03 1.537.1.98)
[PATCH] remove bogus return from mtrr driver.
It's amazing what you spot when you Lindent things.
<davej@suse.de> (02/04/03 1.537.1.99)
[PATCH] strtok->strsep in hpfs
<davej@suse.de> (02/04/03 1.537.1.100)
[PATCH] strtok->strsep in hfs
<davej@suse.de> (02/04/03 1.537.1.101)
[PATCH] typo in pci_set_mwi header
<davej@suse.de> (02/04/03 1.537.1.102)
[PATCH] strtok->strsep in autofs
<davej@suse.de> (02/04/03 1.537.1.103)
[PATCH] strtok->strsep in shmem
<davej@suse.de> (02/04/03 1.537.1.104)
[PATCH] i2c-proc wasn't checking kmalloc result
<davej@suse.de> (02/04/03 1.537.1.105)
[PATCH] compile fix for gemtek-pci radio card
<davej@suse.de> (02/04/03 1.537.1.106)
[PATCH] strtok->strsep for autofs4
<davej@suse.de> (02/04/03 1.537.1.107)
[PATCH] strtok->strsep isofs
<davej@suse.de> (02/04/03 1.537.1.108)
[PATCH] strtok->strsep in usb
<davej@suse.de> (02/04/03 1.537.1.109)
[PATCH] strtok->strsep in jfs
<davej@suse.de> (02/04/03 1.537.1.110)
[PATCH] Only offer ARM PCMCIA on ARM machines.
<davej@suse.de> (02/04/03 1.537.1.111)
[PATCH] strtok->strsep in ntfs
<davej@suse.de> (02/04/03 1.537.1.112)
[PATCH] strtok->strsep for reiserfs
<davej@suse.de> (02/04/03 1.537.1.113)
[PATCH] strtok->strsep in isdn avmb1 capifs
<