ChangeSet@1.1173, 2003-03-09 18:49:41-08:00, torvalds@home.transmeta.com Cache the MSR_IA32_SYSENTER_CS value in the per-CPU TSS (using the otherwise unused cpl1 entry for SS), so that we can avoid re-loading it on task switches if it doesn't change. ChangeSet@1.1172, 2003-03-09 18:48:06-08:00, torvalds@home.transmeta.com Fix nanosleep() problem noticed by Todd Mokros . If we raced on a timer expire, we'd get a negative timeout and think that is was a _huge_ positive timeout. ChangeSet@1.1171, 2003-03-09 11:25:32-08:00, torvalds@home.transmeta.com Fix up i387 task switching bug introduced by Andi Kleen's patch to speed it up - use the proper bitmask for clearing "used-fpu" state. ChangeSet@1.1170, 2003-03-09 10:32:51-08:00, ak@muc.de [PATCH] Fast path context switch - microoptimize FPU reload Following some changes on x86-64. When cpu_has_fxsr is defined to 1 like in many kernels unlazy_fpu can collapse to three instructions. For that inlining is a very good idea. Otherwise it's 10 instructions or so, which can be still inlined. We don't need the lock prefix to test our local thread flags state. Unfortunately test_thread_flag currently always uses test_bit which has a LOCK on SMP, but that's unnecessary. LOCK is costly on P4, so it's a good idea to avoid it. Work around this for now by testing directly. Better would be probably to define __set_bit for all architectures to not guarantee atomicity and then always use that for local thread_info accesses in linux/thread_info.h ChangeSet@1.1169, 2003-03-09 10:31:41-08:00, levon@movementarian.org [PATCH] fix oprofile on x86 > 1 counter Without this we have a choice between dropping lots of counter events for counters > 0, or getting dazed and confused. This brings it inline with the 2.4 module code. Tested on my 2-way. Also fix a typo from Steven Cole, and remove some unnecessary code ChangeSet@1.1167, 2003-03-09 08:48:42-06:00, jejb@raven.il.steeleye.com Merge raven.il.steeleye.com:/home/jejb/BK/scsi-misc-2.5 into raven.il.steeleye.com:/home/jejb/BK/scsi-for-linus-2.5 ChangeSet@1.1166, 2003-03-08 22:39:04-08:00, mbligh@aracnet.com [PATCH] NUMA scheduler fixup From: Ingo Molnar This cleans up the NUMA scheduler to have separate triggers for idle and busy rebalance ratio, preventing some complex interactions where idle cpus can cause busy ones to busy rebalance more often. The core functionality of this patch is rebalace_tick(). Tested in -mjb tree for over 1 month ... no problems at all found - retested on top of 64-bk3 with new scheduler updates. ChangeSet@1.1165, 2003-03-08 22:06:03-08:00, akpm@digeo.com [PATCH] fix a warning in eepro100.c Remove the unneeded forward declaration of eepro100_remove_one(). It is defeating the __attribute__(unused) tag on the definition of eepro100_remove_one() and is causing a "defined but not used" warning. ChangeSet@1.1164, 2003-03-08 22:05:54-08:00, akpm@digeo.com [PATCH] revert the "remove kernel_flag" patch With CONFIG_SMP=n, CONFIG_PREEMPT=y, CONFIG_DEBUG_SPINLOCK=y we get many undefined refs to kernel_flag. This is because spinlock debugging works on uniprocessor kernels now, and it stores state inside the spinlock structure to do this. Having working spinlock debugging on UP kernels seems more important than saving four bytes, so... ChangeSet@1.1163, 2003-03-08 22:05:45-08:00, akpm@digeo.com [PATCH] remove compile warning from serial console initcall Patch from: "Martin J. Bligh" This tiny patch removes the new compiler warning from my build - the new console_initcall mechanism seems to require int from console_init ... I made serial8250_console_init look like con_init ChangeSet@1.1162, 2003-03-08 22:05:36-08:00, akpm@digeo.com [PATCH] rpc_delete_timer race fix Patch from Ulrich Weigand Fix a race wherein timer_pending() returns false, but the timer handler is actually running on another CPU. We need to call del_timer_sync() regardless, so we wait for completion of the handler. ChangeSet@1.1161, 2003-03-08 22:05:27-08:00, akpm@digeo.com [PATCH] missing spin_unlock() in sysfs_remove_dir() missing spin_unlock() in sysfs_remove_dir() ChangeSet@1.1160, 2003-03-08 19:28:57-08:00, akpm@digeo.com [PATCH] raw.c: dev_t cleanup and oops fix Patch from Andries.Brouwer@cwi.nl The next patch in the dev_t series eliminates the last applied use of MAX_BLKDEV - only the definition in major.h remains. Sneaky as I am, I combine this patch with the fix for an Oops: On open, raw_open does filp->f_dentry->d_inode->i_mapping = bdev->bd_inode->i_mapping; storing a pointer to bdev stuff. But on release this pointer stayed, the block device is not referenced anymore and disappears, and the next open references undefined stuff. I checked, and this can actually cause an Oops - scenario: # raw /dev/raw/raw12 /dev/hdf # dd if=/dev/raw/raw12 of=/dev/null bs=512 count=1 # raw /dev/raw/raw12 0 0 # dd if=/dev/raw/raw12 of=/dev/null bs=512 count=1 Oops. More precisely the problem is that dentry_open does file_ra_state_init(&f->f_ra, inode->i_mapping); And file_ra_state_init uses mapping->backing_dev_info->ra_pages. Ugly, to use so much information about the inode even before the inode has been opened. In the patch below I reset i_mapping upon release of the raw device. akpm: I fixed a typo and exported default_backing_dev_info to GPL modules for this. ChangeSet@1.1159, 2003-03-08 19:28:48-08:00, akpm@digeo.com [PATCH] Disable the "Unknown IO_APIC" message This message is the source of 90% of the traffic on linux-smp and is now starting to fill up kernel bugzilla. Nobody has ever actually done anything about it and the message is wasting people's time. Ingo acked this change. ChangeSet@1.1158, 2003-03-08 19:28:39-08:00, akpm@digeo.com [PATCH] Larger buffer for /proc/interrupts display Patch from Zwane Mwaikambo Scale the buffer which is used for accumulating the /proc/interrupts output. On large SMP it overflows. ChangeSet@1.1157, 2003-03-08 19:28:30-08:00, akpm@digeo.com [PATCH] fix SMP lockup in eepro100 with ethtool on unused Patch from Jason Lunz When support for the GSET and SSET ethtool ioctls was added to eepro100.c in 2.4.20, the tx lock was overloaded to serialize their use. Unfortunately, this lock is only initialized in dev->open(), causing ethtool to deadlock the machine when used on an unconfigured eepro100 interface. The fix is to initialize the spinlock at probe time. ChangeSet@1.1156, 2003-03-08 19:28:20-08:00, akpm@digeo.com [PATCH] i386 show_interrupts() fix show_interrupts() is accidentally modifying the incoming seq_file pointer. It goes oops if seq_printf() returns non-zero. ChangeSet@1.1155, 2003-03-08 19:28:10-08:00, akpm@digeo.com [PATCH] protect 'action' in show_interrupts Patch from Zwane Mwaikambo On SMP or preemptible kernels, every instance of show_interrupts() is oopsily racy wrt request_irq() and free_irq(). Fix that up by taking the irq_desc_t's lock while walking the action list. ChangeSet@1.1154, 2003-03-08 19:27:42-08:00, akpm@digeo.com [PATCH] Extended attribute sharing and debug macro typo fixes Patch from Andreas Gruenbacher Tony Dziedzic has found two bugs in the extended attributes code. Patches with explanations are attached. 1: Extended attribute sharing on ext2/ext3 not working The mb_cache_entry_insert function constantly returns an -EBUSY error instead of 0, which causes the xattr cache that is needed by the xattr sharing mechanism on ext2/ext3 to not share anything. This patch fixes the problem. (It is possible that after applying this fix we will hit bugs in code that wasn't used before.) 2: Oops in one of the xattr debug statements: The old_bh variable is NULL if an inode that previously had no EA's assigned would share an EA block with another inode. (This was hidden by the xattr sharing bug). ChangeSet@1.1153, 2003-03-08 19:27:24-08:00, akpm@digeo.com [PATCH] fix memory leak in load_elf_binary() - Fix an error-path leak of memory and a dentry in load_elf_binary() (spotted by Oleg Drokin). - Tidy up the handling of bad executable images: return -ENOEXEC rather than 0. (We've committed suicide anyway, so probably nobody gets to see this). - Fix up the logic in load_aout_interp(). It gets itself all set up to return -ENOEXEC but there is in fact no path by which it can do this. Just return ~0UL on errors, like load_elf_interp(). ChangeSet@1.1152, 2003-03-08 19:27:07-08:00, akpm@digeo.com [PATCH] ext2: fix error-path double-free Fix a double-kfree on an ext2 error path. Reported by Oleg Drokin. ChangeSet@1.1151, 2003-03-08 19:26:43-08:00, akpm@digeo.com [PATCH] hugetlb unmap_vmas() SMP && PREEMPT fix Patch from Kevin Pedretti The unmap_vmas() logic is designed to chew away at all the pages without holding off preemption for too long. But with CONFIG_SMP=y and CONFIG_PREEMPT=y the number of pages which we batch up between rescheduling opportunities is not a multiple of HPAGE_SIZE. So unmap_vmas() ends up calling unmap_hugepage_range() with a poorly aligned&sized region, and it goes BUG. Fix that up by ensuring that we always work across hugepage regions in HPAGE_SIZE chunks. ChangeSet@1.1150, 2003-03-08 19:26:21-08:00, akpm@digeo.com [PATCH] usercopy checks in old_readdir() Fix up some unchecked copy_to_user()s ChangeSet@1.1149, 2003-03-08 19:25:58-08:00, akpm@digeo.com [PATCH] fix div-by-zero in bonding.c Patch from: Shmulik Hen The following patch fixes a division by zero bug in the bonding module that happens when transmitting through a bond with no slaves, in the XOR bonding mode. ChangeSet@1.1148, 2003-03-08 19:25:21-08:00, akpm@digeo.com [PATCH] move CONFIG_SWAP around Patch from Tom Rini Take CONFIG_SWAP out of the top-level menu into the general setup menu. Make it dependent on CONFIG_MMU and common to all architectures. ChangeSet@1.1147, 2003-03-08 19:24:48-08:00, akpm@digeo.com [PATCH] Fix copy_page_range()'s handling of invalid pages Patch from Xavier Bru If copy_page_range encounters a pte which maps an invalid pageframe it will proceed to try to add an rmap entry against that page. This causes oopses when an application which has mapped an IO device via /dev/mem forks. Fix that up by correctly skipping the page_add_rmap() for these pte's. ChangeSet@1.1146, 2003-03-08 19:24:29-08:00, akpm@digeo.com [PATCH] slab debug: symbolic output in caller tracking use print_symbol() to decode the offender's program counter. ChangeSet@1.1145, 2003-03-08 19:24:12-08:00, akpm@digeo.com [PATCH] slab debug: track caller program counter Patch from Manfred Spraul A patch that records the last kfree caller's program counter and prints that if a poison check fails. ChangeSet@1.1144, 2003-03-08 19:23:55-08:00, akpm@digeo.com [PATCH] slab use-after-free detector Patch from Petr Vandrovec Modifies check_poison function to not only verify that last byte is POISON_END, but also that all preceeding bytes are either POISON_BEFORE or POISON_AFTER bytes. ChangeSet@1.1143, 2003-03-08 19:23:39-08:00, akpm@digeo.com [PATCH] Fix vm_area_struct slab corruption Patch from Hugh Dickins Hugh's patch fixes vm_area_struct slab corruption due to mremap's move_vma mistaking how do_munmap splits vmas in one case. Neither of us are very happy with it - it is fragile, and obscure. Hugh will revisit this later, but for now it should fix up the potential memory corruption. ChangeSet@1.1142, 2003-03-08 19:23:21-08:00, akpm@digeo.com [PATCH] Make diskstats per-cpu using kmalloc_percpu Patch from Ravikiran G Thirumalai Makes the disk stats on struct gendisk per-cpu. ChangeSet@1.1141, 2003-03-08 19:23:08-08:00, akpm@digeo.com [PATCH] Allow VFS readahead to fall to zero Some workloads really, really want to have no readahead. Databases which are perfoming small synchronous I/Os against a file which has extremely poor layout. Any readahead at all is a lose here. But the current readahead code refuses to adapt that low. Fix it up so that we can indeed adaptively disable readahead altogether, and do not start it again until we have seen max_readahead()'s worth of consecutive reads. ChangeSet@1.1140, 2003-03-08 19:22:58-08:00, akpm@digeo.com [PATCH] Implement sendfile() for NFS Patch from Trond Myklebust Implement sendfile() for the NFS client. This is required for loop-on-NFS support. ChangeSet@1.1139, 2003-03-08 19:22:49-08:00, akpm@digeo.com [PATCH] fix possible latency in balance_dirty_pages() Tasks which throttle in balance_dirty_pages() will loop until the amount of dirty memory falls below the configured dirty_ratio. This exposes the possibility that one task could be stuck in there for arbitrary periods of time due to page dirtying activity by other tasks. The patch changes the logic so that tasks will break out of the loop if they have written enough pages, regardless of the current dirty memory limits. Here "enough" pages is 1.5x the number of pages which they just dirtied. If the amount of dirty memory in the machine happens to still exceed dirty_ratio (say, due to MAP_SHARED activity) then the task will again throttle after dirtying a single page. But there is now an upper limit on the time for which a single task will be captured in balance_dirty_pages(). ChangeSet@1.1138, 2003-03-08 19:22:37-08:00, akpm@digeo.com [PATCH] register_blkdev cleanups Patch from Andries.Brouwer@cwi.nl The following patch does the following: - static const char *blkdevs[MAX_BLKDEV]; disappears - get_blkdev_list, (un)register_blkdev, __bdevname are moved from block_dev.c to genhd.c - the third "fops" parameter of register_blkdev was unused; now removed everywhere - zillions of places had printk("cannot get major") upon error return from register_blkdev; removed all of these and inserted a single printk in register_blkdev. Of course the reason for the patch is that one fixed size array is eliminated. ChangeSet@1.1137, 2003-03-08 10:24:13-08:00, mbligh@aracnet.com [PATCH] 6/6 cacheline align files_lock I'm getting a lot of cacheline bounce from .text.lock.file_table due to false sharing of the cahceline. The following patch just aligns the lock in it's own cacheline. ChangeSet@1.1136, 2003-03-08 10:24:04-08:00, mbligh@aracnet.com [PATCH] 5/6 Provide basic documentation for profiling People keep asking for this info, and Andrew asked me to put it under the Documentation directory ... provides really simple instructions for taking a profile so that users can report performance changes in a useful way. ChangeSet@1.1135, 2003-03-08 10:23:57-08:00, mbligh@aracnet.com [PATCH] 4/6 Fix the type of get_zholes_size for NUMA-Q From Andy Whitcroft Fix the type of get_zholes_size for NUMA-Q ChangeSet@1.1134, 2003-03-08 10:23:49-08:00, mbligh@aracnet.com [PATCH] 3/6 Convert physnode_map to u8 From Andy Whitcroft Convert physnode_map from an int to a u8 to save cachelines. ChangeSet@1.1133, 2003-03-08 10:23:42-08:00, mbligh@aracnet.com [PATCH] 2/6 Make CONFIG_NUMA work on non-numa machines. From Andy Whitcroft A few very simple changes in order to make CONFIG_NUMA work everywhere, so the distros can build one common binary kernel for distributions. ChangeSet@1.1132, 2003-03-08 10:23:33-08:00, mbligh@aracnet.com [PATCH] 1/6 Share common physnode_map code between NUMA-Q and Summit From Andy Whitcroft Share a common physnode_map structure between NUMA-Q and Summit. ChangeSet@1.1129, 2003-03-08 07:53:17-08:00, davem@nuts.ninka.net [SPARC64]: Make sure update_process_times runs inside of irq_{enter,exit} region. ChangeSet@1.1125.3.1, 2003-03-08 07:42:49-08:00, rml@tech9.net [PATCH] no need for kernel_flag on UP This is a minor cleanup. We currently define and declare the BKL's kernel_flag spinlock on either SMP or PREEMPT, which means a UP+PREEMPT machine gets it. We only need the actual lock on SMP. ChangeSet@1.1099.1.4, 2003-03-08 09:03:45-06:00, jejb@raven.il.steeleye.com Fix loop problem in SCSI error handler sense collection Also remove previous work around so sense is collected now if the driver doesn't do it automatically. Also fix a problem where the retries get reset when the command is reissued (we only want the retries set when the command is first created). ChangeSet@1.1099.1.3, 2003-03-08 08:58:09-06:00, jejb@raven.il.steeleye.com Fix DMA to stack problem in scsi_error.c ChangeSet@1.1099.1.2, 2003-03-08 08:56:52-06:00, jejb@raven.il.steeleye.com Correct nested variable thinko in scsi_error.c ChangeSet@1.1125.2.1, 2003-03-08 09:46:47-01:00, davej@codemonkey.org.uk [WATCHDOG] amd7xx_tco updates from Zwane, and nuke EXPORT_NO_SYMBOLS ChangeSet@1.1125.1.10, 2003-03-08 08:46:06-01:00, davej@codemonkey.org.uk [CPUFREQ] Move pci define to pci_ids.h ChangeSet@1.1125.1.9, 2003-03-08 08:39:57-01:00, davej@codemonkey.org.uk [CPUFREQ] update documentation The sysfs directory where the cpufreq-related files are stored changed due to the new device interface code, so the documentation needs to be updated accordingly. Also, add some information about the reference counting and the exporting of sysfs files by the drivers. ChangeSet@1.1125.1.8, 2003-03-08 08:39:21-01:00, davej@codemonkey.org.uk [CPUFREQ] allow cpufreq drivers to export sysfs files This patch lets cpufreq drivers export per-CPU files in the cpufreq and cpu-specific sysfs directory. As an example, a file "scaling_available_frequencies" is added to the p4-clockmod.c driver. ChangeSet@1.1125.1.7, 2003-03-08 08:38:32-01:00, davej@codemonkey.org.uk [CPUFREQ] add support for ICH4-M chipset in speedstep driver Intel ICH4-M soutbridges use exactly the same register interface for SpeedStep as ICH2-M and ICH3-M southbridges -- which makes adding support for this bridge (almost) trivial ChangeSet@1.1125.1.6, 2003-03-08 08:37:39-01:00, davej@codemonkey.org.uk [CPUFREQ] updated cpufreq ref-counting and locking scheme This patch takes use of the now-working cpufreq_interface.kset and cpufreq_policy.kobj to use reference counting within the cpufreq core wherever this is more appropriate than the previous approach -- using one semaphore. Additionally, the callbacks to the driver modules are protected now. ChangeSet@1.1125.1.5, 2003-03-08 08:36:56-01:00, davej@codemonkey.org.uk [CPUFREQ] remove unneeded code - no cpufreq driver uses the frequency table helper "setpolicy" any more ("target" is much more appropriate for them anyways) - so remove that helper - all cpufreq drivers use the advanced registration process, so some compatibility code can safely be removed. ChangeSet@1.1125.1.4, 2003-03-08 08:36:30-01:00, davej@codemonkey.org.uk [CPUFREQ] fix userspace governor. Let's put the sysfs file exported by the cpufreq userspace governor also into the cpufreq/ - subdirectory of /sys/devices/sys/cpu0/. ChangeSet@1.1125.1.3, 2003-03-08 08:35:48-01:00, davej@codemonkey.org.uk [CPUFREQ] fix cpufreq core breakage(s) - update two more cpufreq-related sysfs files to the new interface code - always store the new, user-requested policy in another struct cpufreq_policy so that we can safely fall back to the old one in case something fails (this equals the behaviour before Pat's patch) - the kobject which was registered in cpufreq_add_dev was inside a variable private to this function -- so the whole cpufreq sysfs interface returned -EINVAL. ChangeSet@1.1128, 2003-03-07 23:01:14-08:00, marcel@holtmann.org [SPARC64]: Translate AUTOFS_IOC_EXPIRE_MULTI ioctl. ChangeSet@1.1125.1.2, 2003-03-07 19:37:53-08:00, torvalds@home.transmeta.com Ignore initramfs cpio file ChangeSet@1.1125.1.1, 2003-03-07 19:36:34-08:00, torvalds@home.transmeta.com Avoid warning due to missing return value ChangeSet@1.1127, 2003-03-07 19:26:50-08:00, davem@nuts.ninka.net [SPARC64]: Update defconfig. ChangeSet@1.1122.1.3, 2003-03-07 19:08:39-08:00, davem@nuts.ninka.net [SPARC64]: Use pci_remove_bus_device to delete, found by Ben Collins. ChangeSet@1.1122.1.2, 2003-03-07 18:43:11-08:00, davem@nuts.ninka.net [FRAMEBUFFER]: Convert SBUS LEO driver to new APIs. ChangeSet@1.1123.1.2, 2003-03-07 16:46:13-08:00, greg@kroah.com kbuild: handle any failures of the gen_init_cpio or initramfs image to stop the build. This also shows how to add files to the initramfs build, but is commented out. Patch originally done by Kai. ChangeSet@1.1123.1.1, 2003-03-07 16:39:06-08:00, greg@kroah.com gen_init_cpio: Add the ability to add files to the cpio image. ChangeSet@1.1122.1.1, 2003-03-07 15:37:48-08:00, davem@nuts.ninka.net [SPARC]: One too many chars in INIT_C_CC ChangeSet@1.1068.7.3, 2003-03-07 17:01:46-06:00, stevef@smfhome1.austin.rr.com Fix oops in getdfs when null path passed in on mount. Fix oops when changed readsize caused readpages problem. Add support for altering rsize so can reduce pages read across net below default of 4 ChangeSet@1.1123, 2003-03-07 12:19:30-08:00, mingo@elte.hu [PATCH] "interactivity changes", sched-2.5.64-B2 This fixes the SMP runqueue locking bug when updating the wakers priority. It also includes: - only update the priority and do a requeueing if the sleep average has changed. (this does not happen for pure CPU hogs or pure interactive tasks, so no need to requeue/recalc-prio in that case.) [All the necessary values are available at that point already, so gcc should have an easy job making this branch really cheap.] - do not do a full task activation in the migration-thread path - that is supposed to be near-atomic anyway. - fix up comments I solved the SMP locking bug by moving the requeueing outside of try_to_wake_up(). It does not matter that the priority update is not atomically done now, since the current process wont do anything inbetween. (well, it could get preempted in a preemptible kernel, but even that wont do any harm.) ChangeSet@1.1122, 2003-03-07 10:43:32-08:00, shemminger@osdl.org [PATCH] Turn off aio printk meant for debugging (2.5.64) The following messages are of interest only when debugging aio. Otherwise, they are just console clutter. ChangeSet@1.1121, 2003-03-07 10:43:25-08:00, willy@debian.org [PATCH] fs/locks.c: fix bugs - Remove broken lock accounting - Introduce __locks_delete_block() - Stop using kdevname() - Fix locks_remove_posix() ChangeSet@1.1120, 2003-03-07 10:41:00-08:00, mingo@elte.hu [PATCH] more "interactivity changes", sched-B2 - fix a (now-) bug in kernel/softirq.c, it did a wakeup outside any atomic regions, which falsely identified random processes as a non-atomic wakeup, and which causes random priority boost to be distributed. - reset the initial idle thread's priority back to PRIO_MAX after doing the wakeup_forked_process() - correct preemption relies on this. - update current->prio immediately after a backboost. - clean up effective_prio() & sleep_avg calculations so that there are fewer RT-task special cases. This has the advantage of the sleep_avg being maintained even for RT tasks - this could be advantegous for tasks that briefly enter/exit RT mode. ChangeSet@1.1119, 2003-03-07 08:42:40-08:00, green@namesys.com [PATCH] memleak in drivers/char/vt.c Seems there is a memleak on error exit path in drivers/char/vt.c, here's the patch. Found with help of smatch + enhanced unfree script. ChangeSet@1.1077.3.6, 2003-03-07 10:30:33-06:00, hch@lst.de [PATCH] remove some braindamage from drivers/scsi/pcmcia/Kconfig It contains one user-selectable option that does nothing but hiding the four pcmcia drivers if not seleted (although we have an extra menu for them!) one one that does exactly nothing. Remove those two. ChangeSet@1.1118, 2003-03-07 07:48:46-08:00, mingo@elte.hu [PATCH] NUMA scheduler breakage __activate_task() and wake_up_forked_process() should call nr_running_inc(rq) rather than doing a rq->nr_running++. Noted by Rick Lindsley ChangeSet@1.1117, 2003-03-07 07:44:05-08:00, mzyngier@freesurf.fr [PATCH] Fix arch/alpha/vmlinux.lds.S typos. The console initcall patch that went in contains a typo that prevents alpha from building. ChangeSet@1.1116, 2003-03-07 07:43:58-08:00, mzyngier@freesurf.fr [PATCH] Fix fs/binfmt_elf.c build The stack reducing patch that recently went in prevent alpha from building (missing some ELF_CORE_COPY_XFPREGS ifdefs). The excluded patch fixes it. ChangeSet@1.1115, 2003-03-07 07:43:51-08:00, zwane@linuxpower.ca [PATCH] noirqbalance still doesn't do anything This one simply sets TARGET_CPUS to cpu_callout_map instead of cpu_online_map so that when we finally do boot we actually use the other cpus for servicing interrupts. ChangeSet@1.1114, 2003-03-07 07:43:45-08:00, mbligh@aracnet.com [PATCH] revert pfn_to_nid change. This change was wrong. pfn_to_nid is a macro. ChangeSet@1.1113, 2003-03-07 07:43:37-08:00, mbligh@aracnet.com [PATCH] Fix NUMA scheduler problem after interactivity merge NODE_THRESHOLD got accidentally dropped in the interactive scheduler changes merge. This puts it back. ChangeSet@1.1112, 2003-03-07 07:43:30-08:00, david@gibson.dropbear.id.au [PATCH] Squash warning in ohci-pci.c on PowerBooks Apple PowerBooks want in ohci-pci.c for the prototype of pci_device_to_OF_node(). This patch adds it to the already present list of PowerBook specific #includes: ChangeSet@1.1111, 2003-03-07 07:43:23-08:00, david@gibson.dropbear.id.au [PATCH] Squash warnings in usb-serial.c This squashes (gcc-3.2) "label and end of compound statement deprecated" warnings in usb-serial.c. ChangeSet@1.1077.3.5, 2003-03-07 09:14:42-06:00, willy@debian.org [PATCH] Some parisc updates for SCSI The config stanza for SCSI_ZALON was in the wrong place, so move it. Make SCSI_NCR53C8XX conflict with SCSI_ZALON. Give ncr53c8xx and sym53c8xx names Fill in detach and release methods unconditionally. Call pci_set_master() in sym53c8xx. Both these drivers need interrupt.h, not sched.h ChangeSet@1.1025.19.6, 2003-03-07 01:23:06-06:00, kai@vaio.(none) ISDN/HiSax: Remove amd7930.c This has been around for many years but was never finished. No need to carry it on to yet another stable kernel. ChangeSet@1.1025.19.5, 2003-03-07 01:20:29-06:00, kai@vaio.(none) ISDN/HiSax: Get rid of CARD_* defines hisax.h defined a CARD_* macro for each CONFIG option, which is totally unnecessary, just use the CONFIG option directly. ChangeSet@1.1025.19.4, 2003-03-07 01:08:04-06:00, kai@vaio.(none) ISDN/HiSax: Remove unused ISDN_CHIP_* macros They were only defined, never used. ChangeSet@1.1025.19.3, 2003-03-07 01:04:07-06:00, kai@vaio.(none) ISDN/HiSax: config.c cleanup Consolidate allocation / freeing of various data structures into one function. ChangeSet@1.1025.19.2, 2003-03-07 00:56:03-06:00, kai@vaio.(none) ISDN/HiSax: Introduce probe() functions Just separate the setup_*() functions of hisax subdrivers into calls to *_probe() functions, a step towards using standard initialization and modularization of hisax. ChangeSet@1.1068.12.8, 2003-03-06 22:28:14-08:00, bjorn_helgaas@hp.com [PATCH] ia64: IA32 support without sysctl doesn't work Make CONFIG_SYSCTL control sys32_sysctl as well. Based on a patch from Peter Chubb. TAG: lia64-v2.5.64 ChangeSet@1.1068.12.7, 2003-03-06 22:21:44-08:00, davidm@tiger.hpl.hp.com ia64: Minor formatting/whitespace fixes in ia64-version of acpi.c. ChangeSet@1.1068.12.6, 2003-03-06 22:17:03-08:00, jbarnes@sgi.com [PATCH] ia64: ACPI fix for no PCI Andy Grover told me this should be posted here. It allows ACPI to compile even with PCI turned off. Patch against 2.5.60. ChangeSet@1.1068.12.5, 2003-03-06 22:04:54-08:00, sfr@canb.auug.org.au [PATCH] ia64: compat_sys_futex() support ia 64 part of the patch. This gives you a 32 bit version of sys_futex (hopefully). ChangeSet@1.1107.1.3, 2003-03-06 21:29:33-08:00, torvalds@home.transmeta.com Fix a very theoretical race between the new RCU lookup and concurrent renames in another directory. I doubt this can be triggered in practice, and the fix is a bit heavy-handed, but let's see if numbers can show that the simple fix doesn't show any real lock contention. ChangeSet@1.1107.1.2, 2003-03-06 21:27:20-08:00, torvalds@home.transmeta.com Fix "con_init()" function type and stale (and incorrect) declaration ChangeSet@1.1107.1.1, 2003-03-06 21:12:44-08:00, anton@samba.org [PATCH] missing include list.h must now include stddef since it uses NULL. ChangeSet@1.1068.12.4, 2003-03-06 21:11:54-08:00, davidm@tiger.hpl.hp.com ia64: Fix several small bugs/omissions from the 2.5.64 sync. ChangeSet@1.1068.1.25, 2003-03-06 20:21:40-06:00, mochel@osdl.org sysfs: Fix binary file handling. From Rusty Lynch: * makes sysfs cleanup the buffer->data allocated by the attribute write functions * fixes a bug that causes the kernel to oops when somebody attempts to write to the file. ChangeSet@1.1068.1.24, 2003-03-06 20:19:40-06:00, mochel@osdl.org sysfs: fix up directory removal, once and for all. This patch was previously integrated, but dropped erroneously by a bk merge. This fixes file cleanup during a sysfs directory removal by deleting files from ->d_subdirs as they're iterated over, then checking if we can get a reference on them. ChangeSet@1.1107, 2003-03-06 17:57:26-08:00, hpa@transmeta.com [PATCH] bootsect removal This is the "boot sector removal" patch for i386 updated for Linux 2.5.63-bk7. The only change against the 2.5.59 version is that the FDOPTS Makefile variable, which lets one create a floppy image with arbitrary kernel options, has been merged from the x86-64 version, and the comments have been updated. The patch removes the in-kernel boot sector, which these days rarely work correctly (it only supports up to 1 MB kernels, and only work on legacy floppies -- not on IDE or USB devices, nor on any kind of emulated devices like El Torito), replaces it with a placeholder stub, and sets up the Makefile targets to create floppies or floppy images using SYSLINUX. The FDOPTS Makefile variable can be used to set kernel command line options (thanks to a suggestion by Andi Kleen.) Note that the same change has already been applied to the x86-64 architecture. The last hunk of the patch corrects a comment in that architecture. ChangeSet@1.1106, 2003-03-06 17:49:37-08:00, rmk@arm.linux.org.uk [PATCH] Fix up pci_scan_bridge and friends Now we tackle pci_add_new_bus and pci_scan_bridge. The hotplug code currently uses this, but I'd like it to die off; pci_scan_bridge() should be used to scan behind bridges. This may mean hotplug needs some changes to pci_scan_bridge - if so, we need to find out what changes are required and fix it. pci_alloc_child_bus() does what pci_add_new_bus() did, except it doesn't attach the new bus to the parents list of child buses. The only way this bus can be reached from the parent bus is by scanning the parents devices list, and locating a device with a non-NULL subordinate bus. The only code which should be doing this is the PCI code. Since the new bus will have an empty list head for bus->node, we can detect unattached buses prety easily. (see pci-3.diff.) pci_scan_bridge() changes slightly - we use our new pci_scan_child_bus() function from pci-3.diff, which doesn't attach devices to the global tree. This means callers of pci_scan_child_bus() and pci_scan_bridge() (ie, hotplug) will need to call pci_bus_add_devices(). ChangeSet@1.1105, 2003-03-06 17:49:30-08:00, rmk@arm.linux.org.uk [PATCH] Convert setup-bus resource allocation to use bus->devices - Convert setup-bus.c resource allocation to scan bus->devices rather than bus->children. As noted previously, newly discovered child buses will not be on the parents list of children buses, so when we're trying to assign resources, we need to scan the bus for devices with subordinate buses rather than using the list of children buses. ChangeSet@1.1104, 2003-03-06 17:49:22-08:00, rmk@arm.linux.org.uk [PATCH] Only add devices to bus->devices while scanning... The pci_find* functions search using the following lists: bus->children (for subordinate buses) pci_root_buses (for all root buses) pci_devices (for devices) This leaves one list which we can add devices to without any drivers finding the new devices before we've finished with them. - initialise bus->node list head. - pci_scan_slot will scan the specified slot, and add the discovered devices to the bus->devices list only. These devices will not appear on the global device list, and do not show in sysfs, procfs. pci_scan_slot returns the number of functions found. If you want to find the devices, you have to scan bus->devices and look for devices where list_empty(&dev->global_list) is true. - new function "pci_bus_add_devices" adds newly discovered devices to the global device lists, and handles the sysfs and procfs stuff, making the devices available to drivers. All our buses which have an empty list head are treated as "new" (since they are not attached to the parent buses list of children) and are also added. Currently, no buses will be in this state when this function is called. - new function "pci_scan_child_bus" scans a complete bus, building a list of devices on bus->devices only, performing bus fixups via pcibios_fixup_bus() and scanning behind bridges. It does make devices externally visible. - pci_do_scan_bus retains its original behaviour - ie, it scans and makes devices available immediately. ChangeSet@1.1103, 2003-03-06 17:49:14-08:00, rmk@arm.linux.org.uk [PATCH] Eliminate stack allocation of struct pci_dev... - Eliminate the stack allocation of a struct pci_dev, and make pci_scan_slot() take a bus and a devfn argument. - Add "dev->multifunction" to indicate whether this is a multifunction device. - Run header fixups before inserting the new pci device into any device lists or announcing it to the drivers. - Convert some more stuff to use the list_for_each* macro(s). ChangeSet@1.1102, 2003-03-06 17:49:06-08:00, rmk@arm.linux.org.uk [PATCH] Separate out pci bus resource allocator - Separate out bus resource allocator (pci_bus_alloc_resource) - Provide pci_enable_bridges to setup command register for all pci bridges. ChangeSet@1.1101, 2003-03-06 17:44:51-08:00, torvalds@home.transmeta.com Merge home.transmeta.com:/home/torvalds/v2.5/alan into home.transmeta.com:/home/torvalds/v2.5/linux ChangeSet@1.1068.13.43, 2003-03-06 17:44:28-08:00, alan@lxorguk.ukuu.org.uk [PATCH] update via driver from 3.35-ac to 3.36 ChangeSet@1.1068.13.42, 2003-03-06 17:44:20-08:00, alan@lxorguk.ukuu.org.uk [PATCH] fix wrong type and statics in amd ide ChangeSet@1.1068.13.41, 2003-03-06 17:44:13-08:00, alan@lxorguk.ukuu.org.uk [PATCH] switch ide taskfile ioctls to ide_execute_command ChangeSet@1.1068.13.40, 2003-03-06 17:44:05-08:00, alan@lxorguk.ukuu.org.uk [PATCH] remove spare cast ChangeSet@1.1068.13.39, 2003-03-06 17:43:58-08:00, alan@lxorguk.ukuu.org.uk [PATCH] add sensible names to the ide iops The outbsync one has further changes to come in the next batch to sort out some platforms like PPC ChangeSet@1.1068.13.38, 2003-03-06 17:43:49-08:00, alan@lxorguk.ukuu.org.uk [PATCH] switch ide-io (core ioctls etc) to ide_execute_command Also fix the irq masking bug ChangeSet@1.1068.13.37, 2003-03-06 17:43:42-08:00, alan@lxorguk.ukuu.org.uk [PATCH] switch ide-floppy to ide_execute_command ChangeSet@1.1068.13.36, 2003-03-06 17:43:35-08:00, alan@lxorguk.ukuu.org.uk [PATCH] ide-dma This switches ide-dma to ide_execute_command and makes IDE DMA also happy when running shared IRQ ChangeSet@1.1068.13.35, 2003-03-06 17:43:28-08:00, alan@lxorguk.ukuu.org.uk [PATCH] bring ide-disk driver into line with 2.4.21pre This doesn't have the new -ac testing stuff from 2.4.21pre5-ac but it has the stuff I now trust in it. This fixes the cache flush stuff and also means ide-disk now should be safe shared IRQ ChangeSet@1.1068.13.34, 2003-03-06 17:43:18-08:00, alan@lxorguk.ukuu.org.uk [PATCH] clean up all the console inits using an initcall variant Dave Woodhouse did this a while ago and its been kicking around in my tree just fine. This gets rid of a lot of the init lists of functions still left in the kernel by having an __init type array built up for the consoles as well ChangeSet@1.1068.13.33, 2003-03-06 17:43:10-08:00, alan@lxorguk.ukuu.org.uk [PATCH] correct building of the old ide/hd.c driver ChangeSet@1.1068.13.32, 2003-03-06 17:43:02-08:00, alan@lxorguk.ukuu.org.uk [PATCH] correct irq logic for x86 We must not look for a PCI IRQ for a non native mode IDE device We must write PCI_INTERRUPT_LINE for VIA ChangeSet@1.1068.13.31, 2003-03-06 17:42:55-08:00, alan@lxorguk.ukuu.org.uk [PATCH] update PCI quirks ALi Magik requires workarounds for TV chips IDE controllers require proper handling in legacy mode PXB must be disabled on C0 450NX or the IDE will corrupt memory VIA northbridge devices require the interrupt line is written NEC_CBUS_2/3 require ISA DMA workarounds ChangeSet@1.1068.13.30, 2003-03-06 17:42:46-08:00, alan@lxorguk.ukuu.org.uk [PATCH] ALi it turns out has a 31bit audio device ChangeSet@1.1068.13.29, 2003-03-06 17:42:37-08:00, alan@lxorguk.ukuu.org.uk [PATCH] clean up 2.2 stuff in wanrouter code ChangeSet@1.1068.13.28, 2003-03-06 17:42:28-08:00, alan@lxorguk.ukuu.org.uk [PATCH] correct file names in comments in mm ChangeSet@1.1068.13.27, 2003-03-06 17:42:20-08:00, alan@lxorguk.ukuu.org.uk [PATCH] correct emu10k url ChangeSet@1.1068.13.26, 2003-03-06 17:42:12-08:00, alan@lxorguk.ukuu.org.uk [PATCH] remove 2.0/2.2 stuff from wanrouter ChangeSet@1.1068.13.25, 2003-03-06 17:42:06-08:00, alan@lxorguk.ukuu.org.uk [PATCH] remove 2.0/2.2 stuff ChangeSet@1.1068.13.24, 2003-03-06 17:41:57-08:00, alan@lxorguk.ukuu.org.uk [PATCH] add a new PCI quirk type for the ALi Magik series This chip needs latency and other setups to avoid a PCI/AGP deadlock with tv cards ChangeSet@1.1068.13.23, 2003-03-06 17:41:50-08:00, alan@lxorguk.ukuu.org.uk [PATCH] remove 2.0 and 2.2 stuff from netfilter ChangeSet@1.1068.13.22, 2003-03-06 17:41:42-08:00, alan@lxorguk.ukuu.org.uk [PATCH] remove 2.2 bits fromw anpipe ChangeSet@1.1068.13.21, 2003-03-06 17:41:33-08:00, alan@lxorguk.ukuu.org.uk [PATCH] remove 2.2 junk from efs ChangeSet@1.1068.13.20, 2003-03-06 17:41:25-08:00, alan@lxorguk.ukuu.org.uk [PATCH] Add ELF types for Hitach H8 series ChangeSet@1.1068.13.19, 2003-03-06 17:41:18-08:00, alan@lxorguk.ukuu.org.uk [PATCH] remove unused beep macro on sh ChangeSet@1.1068.13.18, 2003-03-06 17:41:09-08:00, alan@lxorguk.ukuu.org.uk [PATCH] correct BUG doc in parisc ChangeSet@1.1068.13.17, 2003-03-06 17:41:02-08:00, alan@lxorguk.ukuu.org.uk [PATCH] add serial port table for PC9800 ChangeSet@1.1068.13.16, 2003-03-06 17:40:53-08:00, alan@lxorguk.ukuu.org.uk [PATCH] kill long unused macro ChangeSet@1.1068.13.15, 2003-03-06 17:40:46-08:00, alan@lxorguk.ukuu.org.uk [PATCH] kill the now dead ide_ioreg_t ChangeSet@1.1068.13.14, 2003-03-06 17:40:38-08:00, alan@lxorguk.ukuu.org.uk [PATCH] correct BUG doc ChangeSet@1.1068.13.13, 2003-03-06 17:40:31-08:00, alan@lxorguk.ukuu.org.uk [PATCH] missed patch - static not extern inline in cia ChangeSet@1.1068.13.12, 2003-03-06 17:40:23-08:00, alan@lxorguk.ukuu.org.uk [PATCH] kill stdarg in intermezzo ChangeSet@1.1068.13.11, 2003-03-06 17:40:16-08:00, alan@lxorguk.ukuu.org.uk [PATCH] filter more 2.2 junk ChangeSet@1.1068.13.10, 2003-03-06 17:40:08-08:00, alan@lxorguk.ukuu.org.uk [PATCH] remove a pile of 2.0 and 2.2 support ChangeSet@1.1068.13.9, 2003-03-06 17:39:51-08:00, alan@lxorguk.ukuu.org.uk [PATCH] fix aic7xxx aicasm build The -ldb has to be at the end sometimes ChangeSet@1.1068.13.8, 2003-03-06 17:39:45-08:00, alan@lxorguk.ukuu.org.uk [PATCH] another typo that escaped ChangeSet@1.1068.13.7, 2003-03-06 17:39:38-08:00, alan@lxorguk.ukuu.org.uk [PATCH] more 2.0/2.2 bits ChangeSet@1.1068.13.6, 2003-03-06 17:39:29-08:00, alan@lxorguk.ukuu.org.uk [PATCH] more 2.0 crap ChangeSet@1.1068.13.5, 2003-03-06 17:39:22-08:00, alan@lxorguk.ukuu.org.uk [PATCH] look a typo 8) ChangeSet@1.1068.13.4, 2003-03-06 17:39:14-08:00, alan@lxorguk.ukuu.org.uk [PATCH] exterminate 2.2 gunk from video stuff ChangeSet@1.1068.13.3, 2003-03-06 17:39:06-08:00, alan@lxorguk.ukuu.org.uk [PATCH] remove 2.2 compatmac stuff from rio driver ChangeSet@1.1068.13.2, 2003-03-06 17:38:57-08:00, alan@lxorguk.ukuu.org.uk [PATCH] minor typo fix that got missed ChangeSet@1.1068.13.1, 2003-03-06 17:38:50-08:00, alan@lxorguk.ukuu.org.uk [PATCH] minor breakage fix from typo fixes ChangeSet@1.1100, 2003-03-06 17:25:00-08:00, linux@brodo.de [PATCH] pcmcia: it works again! platform_device_register may only be called after all class-specific device data is initialized, or else the class-type add_device call (which enables the pcmcia sockets) will fail. ChangeSet@1.1099, 2003-03-06 17:24:19-08:00, Andries.Brouwer@cwi.nl [PATCH] scsi_error fix imm.c: spelling scsi.h: remove old and now incorrect comment scsi_scan.c: remove superfluous final return scsi_error.c: apart from similar trivialities the only change: If a command fails (e.g. because it belongs to a newer SCSI version than the device), it is fed to scsi_decide_disposition(). That routine must return SUCCESS, unless the error handler should be invoked. In the situation where host_byte is DID_OK, and message_byte is COMMAND_COMPLETE, and status is CHECK_CONDITION, there is no reason at all to invoke aborts and resets. The situation is normal. I see here UNIT ATTENTION, Power on occurred and ILLEGAL REQUEST, Invalid field in cdb. The 2.5.64 code does not return SUCCESS, but it returns the return code of scsi_check_sense(), and that may be FAILED in case we do not have valid sense. ChangeSet@1.1098, 2003-03-06 17:15:02-08:00, akpm@digeo.com [PATCH] fix ioremap off by one bug Patch from Andreas Klein Fixes an off-by-one in ioremap() - local variable `last_addr' is *inclusive* - it is the last byte of the mapping, not the usual (last byte + 1). ChangeSet@1.1097, 2003-03-06 17:14:55-08:00, akpm@digeo.com [PATCH] ext3: fix error-path memory leak Fix an error-path memory leak in ext3_fill_super(). ChangeSet@1.1096, 2003-03-06 17:14:48-08:00, akpm@digeo.com [PATCH] uninlining in fs/binfmt_elf.c uninline some functions which have multiple callsites. A 5% reduction in text size. ChangeSet@1.1095, 2003-03-06 17:14:41-08:00, akpm@digeo.com [PATCH] reduce stack size: elf_core_dump() Patch from "Randy.Dunlap" This patch reduces stack size in elf_core_dump() from over 0x400 (0x4a4 e.g.) to less than 0x100 (0xb0 on a P4 with gcc 2.96). ChangeSet@1.1094, 2003-03-06 17:14:34-08:00, akpm@digeo.com [PATCH] missed unlock_kernel() in ext3+htree Patch from Alex Tomas There's a missing unlock_kernel(). ChangeSet@1.1093, 2003-03-06 17:14:27-08:00, akpm@digeo.com [PATCH] remove __pte_offset Patch from Dave Hansen ptes this time ChangeSet@1.1092, 2003-03-06 17:14:20-08:00, akpm@digeo.com [PATCH] remove __pmd_offset Patch from Dave Hansen Same thing as the __pgd_offset one, just for pmds this time to keep the naming consistent. ChangeSet@1.1091, 2003-03-06 17:14:13-08:00, akpm@digeo.com [PATCH] remove __pgd_offset Patch from Dave Hansen __pgd_offset() and pgd_offset() are completely different functions. __pgd_offset() is really just a helper to figure out which entry in a pgd an address would fall into. pgd_offset() does all the leg work and actually fetches the real pgd entry. pgd_index() is a much saner name for what __pgd_offset() does. In fact, we do this: #define __pgd_offset(address) pgd_index(address) The attached patch removes all instances of __pgd_offset and just replaces them with pgd_index. Compiles with and without PAE on x86. ChangeSet@1.1090, 2003-03-06 17:14:05-08:00, akpm@digeo.com [PATCH] fix inode reclaim imbalance Well I thought we were shooting down unused inodes a bit aggressively. With 30,000 inodes total and 1000 unused, the 1000 are getting reclaimed instantly. shrink_icache_memory() needs to return the number of reclaimable inodes to page reclaim, not the total number. ChangeSet@1.1089, 2003-03-06 17:13:58-08:00, akpm@digeo.com [PATCH] Reduced stack usage in random.c Patch from: Matthew Wilcox Reduces stack use in random.c by 350 (or maybe 700) bytes by sharing a local array between two functions. ChangeSet@1.1088, 2003-03-06 17:13:51-08:00, akpm@digeo.com [PATCH] export pfn_to_nid to modules Patch from Patricia Gaughen This function was recently uninlined for NUMAQ. But it used to be a macro, and modules need it. ChangeSet@1.1087, 2003-03-06 17:13:43-08:00, akpm@digeo.com [PATCH] cciss: fix initialization for PCI hotplug Patch from Stephen Cameron * Fix driver to wait for firmware to indicate that it is ready. (Needed for PCI hotplug case, but for normal warm/cold reboot, by the time driver inits, firmware will already be ready.) ChangeSet@1.1086, 2003-03-06 17:13:36-08:00, akpm@digeo.com [PATCH] sk98lin 64-bit divide fix This driver is doing a division on a 64 bit valus which is identical to "jiffies" and is asking for __udiv64. Cast the 64-bit value back to long. ChangeSet@1.1085, 2003-03-06 17:13:30-08:00, akpm@digeo.com [PATCH] rtc lock ranking fix Patch from: Zwane Mwaikambo Fix up a possible AB/BA deadlock identified by Dawson Engler's latest toy. ChangeSet@1.1084, 2003-03-06 17:13:22-08:00, akpm@digeo.com [PATCH] Fix nfs oops during mount Patch from Trond Myklebust This has been floating about for a while, and fixes the oops which people keep on reporting. ChangeSet@1.1083, 2003-03-06 17:13:15-08:00, akpm@digeo.com [PATCH] balance_irq lockup fix irqdesc_t.lock is taken from interrupts - it needs irq-safe locking from process context. Also timer_irq_works() needs to use an unsigned long to hold a jiffies value (Kai Bankett) ChangeSet@1.1082, 2003-03-06 17:13:08-08:00, akpm@digeo.com [PATCH] i386 IRQ balancing cleanup General coding cleanups in io_apic.c: - make pending_irq_balance_apicid[] static - reorder several functions so that forward decls are not needed and so that `inline' actually works. - 80 col fixes. ChangeSet@1.1077.3.4, 2003-03-06 18:52:50-06:00, willy@debian.org [PATCH] Remove naked GFP_DMA from drivers/scsi/sr.c In get_sectorsize(), we can already sleep in scsi_wait_req(). In get_capabilities(), we can use GFP_KERNEL because it is only called from sr_attach() which already does a GFP_KERNEL allocation. ChangeSet@1.1077.3.3, 2003-03-06 18:52:19-06:00, willy@debian.org [PATCH] Remove naked GFP_DMA from drivers/scsi/sd.c This GFP_DMA can be GFP_KERNEL because sd_init_onedisk calls sd_spinup_disk calls scsi_wait_req which can sleep. ChangeSet@1.1077.3.2, 2003-03-06 18:51:48-06:00, hch@lst.de [PATCH] update qlogic pcmcia support dito ChangeSet@1.1077.3.1, 2003-03-06 18:51:16-06:00, hch@lst.de [PATCH] update fdomain pcmcia support get it to actually compile cleanly again, switch to scsi_add_host, remove host list walking. ChangeSet@1.1068.11.7, 2003-03-06 18:27:31-06:00, kai@tp1.ruhr-uni-bochum.de do_mounts: Fix boot from ramdisk Russell King investigated a failure case I introduced: When booting with "load_ramdisk=1", we use the kernel root= parameter to determine from what device to get the contents to copy into a ramdisk and then mount that ramdisk as root. For the copy to work, /dev/root needs to point to the device to load the ramdisk from. ChangeSet@1.1075.2.1, 2003-03-06 16:17:07-08:00, davem@nuts.ninka.net [IPSEC]: Fix build when ipsec is disabled. ChangeSet@1.1068.10.6, 2003-03-07 00:16:38+00:00, rmk@flint.arm.linux.org.uk [ARM] Add generic SSP "PIO" mode driver Add SSP PIO driver - provides an interface for talking to microcontrollers and such like connected to the SA1110's SSP port. ChangeSet@1.1068.8.7, 2003-03-06 15:47:22-08:00, greg@kroah.com [PATCH] USB: unfortunatly, we can't call usb_unlink_urb() right now all of the time. The host controllers have to be fixed up before we can safely take out the check for dev->state. ChangeSet@1.1068.8.6, 2003-03-06 15:46:40-08:00, david-b@pacbell.net [PATCH] USB: track usb ch9 device state This patch merges the USB state definitions from the ARM Linux code (inside the sa1100 driver) and uses them to track what can be done with the device. That replaces the recently added "udev->present" flag with a more complete/standard state model. There are a few changes that might affect behavior if things start to go really haywire: - usb_set_address() and usb_set_configuration(), used while enumerating, handle some unlikely cases more correctly: don't allow setting address to zero (undefined behavior), and do allow un-configuring (config 0). (Adds a FIXME for an existing set-configuration bug too.) - usb_disconnect() flags the state change earlier (as soon as it's known). - usb_submit_urb() works in the states where messaging is allowed, and also enforces the "unless configured, only control traffic is legal" rule. - usb_unlink_urb() doesn't care any more about that state. (There seemed to be agreement that it must not matter.) This will help with some further cleanups in the complex of issues relating to driver removal, device removal, config changing (with driver unbind and rebind), reset, and so on. ChangeSet@1.1068.12.3, 2003-03-06 14:56:39-08:00, davidm@tiger.hpl.hp.com ia64: Hook up POSIX-timer syscalls. Take advantage of ptrace_notify(). ChangeSet@1.1068.8.5, 2003-03-06 13:55:05-08:00, stern@rowland.harvard.edu [PATCH] USB: Patch for auto-sense cmd_len This patch fixes an oversight in usb-storage whereby the command length and command buffer for an automatically-generated REQUEST-SENSE command would not be initialized properly. ChangeSet@1.1068.12.2, 2003-03-06 13:42:13-08:00, davidm@tiger.hpl.hp.com ia64: Correct the value of siginfo.si_addr for SIGSEGV signals triggered by NaT-page-consumption faults. ChangeSet@1.1068.12.1, 2003-03-06 13:34:32-08:00, davidm@tiger.hpl.hp.com ia64: Sync with 2.5.64. ChangeSet@1.1077, 2003-03-06 13:14:03-08:00, torvalds@penguin.transmeta.com Merge ChangeSet@1.1075.1.1, 2003-03-06 13:08:47-08:00, mingo@elte.hu [PATCH] "interactivity changes", sched-2.5.64-A6 This includes the following items: - a new wakeup feature: SMART_WAKE_CHILD (now unconditional) [this is a reworked version of the wakeup changes from Andrea's tree.] - scheduler parameter tunings (CHILD_PENALTY, MAX_TIMESLICE, STARVATION_LIMIT, MAX_SLEEP_AVG) - activate_task() splitup to separate the interactivity-active variant from the task-queue variant => this fixes our statistics in some cases. - finer-grained preemption: if priority is equal then the task with the longer pending timeslice wins. - make it possible to disable sync-wakeups via SYNC_WAKEUPS (default:1) - a naming cleanup (sleep_timestamp => last_run) - fix up STARVATION_LIMIT usage so that a value of 0 is recognized as 'no starvation limit'. It also merges Linus' "priority boost backmerging" ChangeSet@1.1076, 2003-03-06 12:59:29-08:00, torvalds@home.transmeta.com Make the "interactive bonus" work both ways - both the process that is actually showing interactive behaviour (sleeping) and the process that wakes up the interative process get a bonus. This means that processes that help interactive processes (usually system daemons, and X11 in particular) will themselves get some bonus. The waker only gets whatever is left over from the "real" interactive user, to avoid any bad behaviour. ChangeSet@1.1068.10.5, 2003-03-06 20:34:37+00:00, rmk@flint.arm.linux.org.uk Merge flint.arm.linux.org.uk:/usr/src/linux-bk-2.5/linux-2.5 into flint.arm.linux.org.uk:/usr/src/linux-bk-2.5/linux-2.5-rmk ChangeSet@1.1060.3.18, 2003-03-06 20:26:33+00:00, rmk@flint.arm.linux.org.uk [CPUFREQ] Make sa11x0_ppcr_to_freq return in units of kHz, not 100kHz. ChangeSet@1.1060.3.17, 2003-03-06 19:27:52+00:00, rmk@flint.arm.linux.org.uk [ARM] CPUFREQ - allow ARM to work with userspace governor. Patch from Dominik Brodowski. This patch is necessary to update the ARM drivers so that they can live with the new cpufreq governor "userspace" (which I sent to Linus a moment ago). ChangeSet@1.914.127.34, 2003-03-06 11:22:26-08:00, davidm@tiger.hpl.hp.com ia64: Make ia64_fetch_and_add() simpler to optimize so lib/rwsem.c can be optimized properly. TAG: lia64-v2.5.60 ChangeSet@1.1025.20.3, 2003-03-06 09:36:37-08:00, ink@jurassic.park.msu.ru [PATCH] alpha pcibios_claim_console_setup fix pcibios_claim_console_setup() doesn't look behind PCI-PCI bridges. Noticed by Anton Blanchard. Ivan. ChangeSet@1.1073, 2003-03-06 09:11:27-08:00, laforge@netfilter.org [NETFILTER]: Fix ipv6 build. ChangeSet@1.1072, 2003-03-06 09:08:59-08:00, jmorris@intercode.com.au [IPV4]: Fix multicast route lookups. ChangeSet@1.1060.3.16, 2003-03-06 16:49:18+00:00, rmk@flint.arm.linux.org.uk [CPUFREQ] Update ARM CPUFREQ drivers A 4 patch set from Dominik Brodowski, tested and fixed up by rmk for ARM. (1) - the system3.c cpufreq notifier couldn't possibly compile before because cpufreq_updateminmax has been undefined for ages. - clean up sa1100fb.c notifier to specify exactly what's left TBD. - double #include of cpufreq.c in drivers/pcmcia/sa1100_generic.c (2) Split up config symbols, add help text for CPU_FREQ_PROC_INTF (3) - update the Integrator CPUfreq driver so that it can get along with cpufreq policies. - modularize Integrator CPUfreq driver (all it did with !CONFIG_CPU_FREQ was to print out some debug information) (4) - update the SA11x0 CPUfreq drivers so that they can get along with cpufreq policies and governors. - update the cpufreq core so that cpufreq_get() returns something sensible during the first ->setpolicy or ->target call. ChangeSet@1.1060.3.15, 2003-03-06 16:27:13+00:00, rmk@flint.arm.linux.org.uk [ARM] Clean up ARM PCI support (bios32.c) Eliminate have_isa_bridge - use isa_bridge instead. Clean up PCI bus walking. Clean up detection of fast back to back capability. Ensure we set the bridge control register appropriately. Pin swizzle functions are never called with pin == 0. ChangeSet@1.1060.3.14, 2003-03-06 16:09:12+00:00, rmk@flint.arm.linux.org.uk [ARM] Add better PM support to SA1111 and SA11x0. This follows our existing PM interfaces. ChangeSet@1.1060.3.13, 2003-03-06 16:01:00+00:00, rmk@flint.arm.linux.org.uk [ARM] Power management updates This provides an element of core PM support to ARM, providing an APM like user space interface, and separating the core PM support from the SA1100 PM implementation. ChangeSet@1.1060.3.12, 2003-03-06 15:53:08+00:00, rmk@flint.arm.linux.org.uk [ARM] Ensure we preserve other CPSR bits when switching to SVC mode. ChangeSet@1.1060.3.11, 2003-03-06 14:58:36+00:00, rmk@flint.arm.linux.org.uk [ARM] Update mach-types to latest vesion. ChangeSet@1.1060.3.10, 2003-03-06 14:53:11+00:00, rmk@flint.arm.linux.org.uk [ARM] Make TLB instruction selection more fine-grained This eliminates some unnecessary branches and instructions in the TLB flush macros where configurations allow for this to occur. ChangeSet@1.1060.3.9, 2003-03-06 14:44:32+00:00, rmk@flint.arm.linux.org.uk [ARM] Fix ARM ide.h Kill ide_* resource handling. Add __ide_mm* macros to use our string based mmio. ChangeSet@1.1060.3.8, 2003-03-06 14:07:07+00:00, rmk@flint.arm.linux.org.uk [ARM] Kill old mkdep dependency hack. We include a header file in an indirect method, which mkdep didn't understand. Now that mkdep is gone, we don't need the hack to make mkdep do the right thing. ChangeSet@1.1060.3.7, 2003-03-06 14:03:09+00:00, rmk@flint.arm.linux.org.uk [ARM] Make ARM's pci_controller_num dev argument appear to be used. This cuts down on the number of warnings during the build. ChangeSet@1.1060.3.6, 2003-03-06 13:58:40+00:00, rmk@flint.arm.linux.org.uk [ARM] Add dummy set_pgd() implementation. ChangeSet@1.1060.3.5, 2003-03-06 13:54:01+00:00, rmk@flint.arm.linux.org.uk [ARM] Add new kernel typedefs for __kernel_{timer,clockid}_t. ChangeSet@1.1060.3.4, 2003-03-06 13:41:45+00:00, rmk@flint.arm.linux.org.uk [ARM] Always claim the timer IRQ using SA_INTERRUPT. ChangeSet@1.1060.3.3, 2003-03-06 13:18:58+00:00, rmk@flint.arm.linux.org.uk [ARM] Always call schedule_tail() after a fork() ChangeSet@1.1060.3.2, 2003-03-06 13:14:56+00:00, rmk@flint.arm.linux.org.uk [ARM] Convert ecard.c shutdown/reboot to use driver model ecard.c used to use the reboot notifier to ensure that the expansion cards are in the correct state for rebooting. Use the driver model shutdown method instead. ChangeSet@1.1060.3.1, 2003-03-06 13:01:35+00:00, rmk@flint.arm.linux.org.uk [ARM] Remove set_mac_address from acorn drivers. Acorn Ethernet drivers provided a duplicate set_mac_address implementation which reflects the generic ethernet version in drivers/net/net_init.c. Use the generic version instead. ChangeSet@1.1068.10.4, 2003-03-05 21:39:41-08:00, torvalds@home.transmeta.com Fix up some timeouts to use the proper types. (While "signed long" is the same in practice, it is technically incorrect, and the new anal type-checker complains). ChangeSet@1.1068.10.3, 2003-03-05 21:22:59-08:00, davem@nuts.ninka.net [JIFFIES]: Use typecheck in time_foo jiffies macros. ChangeSet@1.1068.10.2, 2003-03-05 21:22:16-08:00, davem@nuts.ninka.net [KERNEL]: Add typecheck macro for verifying types at compile time. ChangeSet@1.1068.11.6, 2003-03-05 20:09:49-06:00, sam@mars.ravnborg.org [PATCH] kbuild: touch-module after successfull creation only On Mon, Mar 03, 2003 at 05:11:10PM -0500, Bill Davidsen wrote: > scripts/Makefile.modinst:16: *** Uh-oh, you have stale module entries. You messed with SUBDIRS, do not complain if something goes wrong. This happens if you have encountered a compile error in a module. In this case you did not succeed the compilation of fs/binfmt_aout, and therefore no .o file can be located. kbuild assumes this is because you have messed with SUBDIRS, which is wrong. ChangeSet@1.1068.11.5, 2003-03-05 20:07:18-06:00, arnd@arndb.de [trivial] avoid a warning for each module on s390x s390x has a reference to _GLOBAL_OFFSET_TABLE_ in each module that is resolved by the module loader. This patch prevents modpost from emitting a warning about that symbol. ChangeSet@1.1068.11.4, 2003-03-05 20:05:54-06:00, sam@mars.ravnborg.org [PATCH] kbuild: Smart notation for non-verbose output Create a nice shorthand to enable the non-verbose output mode. make V=1 => Gives verbose output (default) make V=0 => Gives non-verbose output One of the reasons why people does not use KBUILD_VERBOSE=0 that much is simply the typing needed. This notation should make it acceptable to type it. The usage of "make V=0" is restricted to the command line. Anyone that wants to enable the non-verbose mode pr. default shall set KBUILD_VERBOSE in the shell. ChangeSet@1.1068.11.3, 2003-03-05 19:53:26-06:00, kai@tp1.ruhr-uni-bochum.de kbuild: Make per-cpu-check ignore __crc_ symbols. We warn when symbols end in __per_cpu, but aren't in the per-cpu section. However, checksum symbols gave false positives. ChangeSet@1.1068.11.2, 2003-03-05 19:51:34-06:00, kai@tp1.ruhr-uni-bochum.de kbuild: Add some missing FORCE Those were missing, not particularly harmful, but still needed for correctness. ChangeSet@1.1068.11.1, 2003-03-05 19:50:32-06:00, kai@tp1.ruhr-uni-bochum.de kbuild: Make build stop on vmlinux link error set -e is needed for each (continued) line. ChangeSet@1.1071, 2003-03-05 15:50:01-08:00, miyazawa@linux-ipv6.org [IPSEC]: Add full ipv6 support. Credits also to Mitsuru Kanda , YOSHIFUJI Hideaki , and Kunihiro Ishiguro. ChangeSet@1.1068.9.4, 2003-03-05 17:44:21-06:00, mochel@osdl.org cpufreq: fix compile error. Use ->dev instead of ->intf.dev in struct cpufreq_policy. Introduced whilst cleaning up part of the cpufreq code. ChangeSet@1.1068.9.3, 2003-03-05 17:38:05-06:00, mochel@osdl.org sysfs: don't complain when sysfs can't register. During startup, sysfs won't succeed in registering with itself, because it's not yet inititialized. This was causing an unnecessary error on boot. ChangeSet@1.1068.9.2, 2003-03-05 17:33:10-06:00, mochel@osdl.org driver model: fix platform_match() - Make it compare the name in struct platform_dev to the driver's name. From Dominik Brodowski ChangeSet@1.1068.9.1, 2003-03-05 17:27:00-06:00, mochel@osdl.org sysfs: Make sure root inode is executable and readable by everyone. ChangeSet@1.1068.1.22, 2003-03-05 14:25:25-08:00, levon@movementarian.org [PATCH] Fix oprofile on UP, small additional fix The below has been in -mm for a while, and has been tested on my UP and 2-way machines. OProfile was completely unsafe on UP - a spinlock is no protection against the NMI arriving and putting data into the buffer. Pretty stupid bug. This fixes it by implementing reader/writer windows in the buffer and removing the lock altogether. This patch was originally done by Will Cohen. It also fixes the oops Dave Hansen saw on 2.5.62 SMP ChangeSet@1.1068.8.4, 2003-03-05 12:06:48-08:00, greg@kroah.com [PATCH] USB: added support for radio shack device to pl2303 driver. Thanks to gene_heskett@iolinc.net for the info for this. ChangeSet@1.1068.1.21, 2003-03-05 11:57:07-08:00, andmike@us.ibm.com [PATCH] Fix SCSI error handler abort case I had my list empty checks reversed if aborting and bus device reset failed. The condition that causes the error handler to run is still unknown. ChangeSet@1.1068.8.3, 2003-03-05 11:44:45-08:00, andre.breiler@null-mx.org [PATCH] io_edgeport.c diff to fix endianess bugs attached a fix for the io_edgeport usb serial driver This diff fixes endianess issues which prevented the driver to work on bigendian machines (e.g. sparc). ChangeSet@1.1068.8.2, 2003-03-05 11:43:00-08:00, clemens@ladisch.de [PATCH] usb-midi.h: fixes for SC-8820/50 sync with Nagano's version: - protect vendors ids against multiple definitions - sort Roland device ids - fix SC-8850 cable bitmask - add quirk for the SC-8820 - add quirk for the MOTU Fastlane ChangeSet@1.1068.8.1, 2003-03-05 11:41:00-08:00, greg@kroah.com [PATCH] USB: add support for Treo devices to the visor driver. Finally... ChangeSet@1.1025.20.2, 2003-03-05 10:16:58-08:00, Jeff.Wiedemeier@hp.com [PATCH] enable setting of marvel/titan agp->type Now that the agpgart support is in, this can be enabled. /jeff ChangeSet@1.1068.1.20, 2003-03-05 08:23:08-08:00, davem@nuts.ninka.net [SPARC64]: Fix cpufreq config deps. ChangeSet@1.1068.1.18, 2003-03-05 07:36:20-08:00, davem@redhat.com [PATCH] Fix time comparison typing bugs. Many places use inappropriate types for comparing jiffies, int for example. Fix those places up. ChangeSet@1.1068.1.17, 2003-03-04 19:11:19-08:00, torvalds@home.transmeta.com Linux 2.5.64 TAG: v2.5.64