summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Documentation: Ask driver writers to provide PM supportRafael J. Wysocki2007-05-083-0/+163
| | | | | | | | | | | | | | | | | | Add a paragraph in Documentation/SubmittingDrivers requesting that the basic PM support be provided by new device drivers. Add two new documents in Documentation/power/ giving general instructions on debugging the suspend/resume functionality and testing the suspend and resume support in device drivers. Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Cc: Pavel Machek <pavel@ucw.cz> Cc: David Brownell <david-b@pacbell.net> Cc: Nigel Cunningham <ncunningham@linuxmail.org> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Greg KH <greg@kroah.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* Fix constant folding and poor optimization in byte swapping codeTrent Piepho2007-05-081-39/+69
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Constant folding does not work for the swabXX() byte swapping functions, and the C versions optimize poorly. Attempting to initialize a global variable to swab16(0x1234) or put something like "case swab32(42):" in a switch statement will not compile. It can work, swab.h just isn't doing it correctly. This patch fixes that. Contrary to the comment in asm-i386/byteorder.h, gcc does not recognize the "C" version of swab16 and turn it into efficient code. gcc can do this, just not with the current code. The simple function: u16 foo(u16 x) { return swab16(x); } Would compile to: movzwl %ax, %eax movl %eax, %edx shrl $8, %eax sall $8, %edx orl %eax, %edx With this patch, it will compile to: rolw $8, %ax I also attempted to document the maze different macros/inline functions that are used to create the final product. Signed-off-by: Trent Piepho <xyzzy@speakeasy.org> Cc: Francois-Rene Rideau <fare@tunes.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* softlockup: s/99/MAX_RT_PRIO/Oleg Nesterov2007-05-081-1/+1
| | | | | | | | | Don't use hardcoded 99 value, use MAX_RT_PRIO. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: Ingo Molnar <mingo@elte.hu> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* freezer: task->exit_state should be treated as boleanOleg Nesterov2007-05-081-3/+2
| | | | | | | | | | | | | Except for BUG_ON() checks, we should not use EXIT_XXXX defines outside of exit/wait paths. Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: Pavel Machek <pavel@ucw.cz> Cc: Ingo Molnar <mingo@elte.hu> Cc: Nick Piggin <nickpiggin@yahoo.com.au> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* ipmi: add pci remove handlingCorey Minyard2007-05-081-0/+3
| | | | | | | | | Add pci_remove handling to the driver, so it will clean up if the device is hot-removed. Signed-off-by: Corey Minyard <minyard@acm.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* ipmi: add new IPMI nmi watchdog handlingCorey Minyard2007-05-085-42/+103
| | | | | | | | | | | | | Convert over to the new NMI handling for getting IPMI watchdog timeouts via an NMI. This add config options to know if there is the ability to receive NMIs and if it has an NMI post processing call. Then it modifies the IPMI watchdog to take advantage of this so that it can know if an NMI comes in. It also adds testing that the IPMI NMI watchdog works. Signed-off-by: Corey Minyard <minyard@acm.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* ipmi: allow shared interruptsCorey Minyard2007-05-081-35/+89
| | | | | | | | | | | | | | | | The IPMI driver used enable_irq and disable_irq when it got into situations where it couldn't allocate memory; it did this to avoid having the interrupt just lock the machine when it couldn't get memory to perform the transaction to disable the interrupt. This patch modifies the driver to not use disable_irq and enable_irq. It instead sends the messages to the BMC to perform this operation. It also makes sure interrupts are cleanly disabled when the interface is shut down and cleans up some shutdown things that are no longer necessary. Signed-off-by: Corey Minyard <cminyard@mvista.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* ipmi: add powerpc openfirmware sensingCorey Minyard2007-05-081-0/+108
| | | | | | | | | | | | | | | Add support for of_platform_driver to the ipmi_si module. When loading the module, the driver will be registered to of_platform. The driver will be probed for all devices with the type ipmi. It's supporting devices with compatible settings ipmi-kcs, ipmi-smic and ipmi-bt. Only ipmi-kcs could be tested. Signed-off-by: Christian Krafft <krafft@de.ibm.com> Acked-by: Heiko J Schick <schihei@de.ibm.com> Signed-off-by: Corey Minyard <minyard@acm.org> Acked-by: Arnd Bergmann <arnd.bergmann@de.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* mm: shrink parent dentries when shrinking slabAndrew Morton2007-05-081-1/+1
| | | | | | | | | | | | | | | | Teach the dentry slab shrinker to aggressively shrink parent dentries when shrinking the dentry cache. This is done to attempt to improve the situation where the dentry slab cache gets a lot of internal fragmentation due to pages containing directory dentries. It is expected that this change will cause some of those dentries to be reaped earlier, and with less scanning. Needs careful testing. Cc: Miklos Szeredi <mszeredi@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* fix quadratic behavior of shrink_dcache_parent()Miklos Szeredi2007-05-081-37/+67
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The time shrink_dcache_parent() takes, grows quadratically with the depth of the tree under 'parent'. This starts to get noticable at about 10,000. These kinds of depths don't occur normally, and filesystems which invoke shrink_dcache_parent() via d_invalidate() seem to have other depth dependent timings, so it's not even easy to expose this problem. However with FUSE it's easy to create a deep tree and d_invalidate() will also get called. This can make a syscall hang for a very long time. This is the original discovery of the problem by Russ Cox: http://article.gmane.org/gmane.comp.file-systems.fuse.devel/3826 The following patch fixes the quadratic behavior, by optionally allowing prune_dcache() to prune ancestors of a dentry in one go, instead of doing it one at a time. Common code in dput() and prune_one_dentry() is extracted into a new helper function d_kill(). shrink_dcache_parent() as well as shrink_dcache_sb() are converted to use the ancestry-pruner option. Only for shrink_dcache_memory() is this behavior not desirable, so it keeps using the old algorithm. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Maneesh Soni <maneesh@in.ibm.com> Acked-by: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: Neil Brown <neilb@suse.de> Cc: Trond Myklebust <trond.myklebust@fys.uio.no> Cc: Christoph Hellwig <hch@lst.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* reduce size of task_struct on 64-bit machinesWilliam Cohen2007-05-082-8/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | This past week I was playing around with that pahole tool (http://oops.ghostprotocols.net:81/acme/dwarves/) and looking at the size of various struct in the kernel. I was surprised by the size of the task_struct on x86_64, approaching 4K. I looked through the fields in task_struct and found that a number of them were declared as "unsigned long" rather than "unsigned int" despite them appearing okay as 32-bit sized fields. On x86_64 "unsigned long" ends up being 8 bytes in size and forces 8 byte alignment. Is there a reason there a reason they are "unsigned long"? The patch below drops the size of the struct from 3808 bytes (60 64-byte cachelines) to 3760 bytes (59 64-byte cachelines). A couple other fields in the task struct take a signficant amount of space: struct thread_struct thread; 688 struct held_lock held_locks[30]; 1680 CONFIG_LOCKDEP is turned on in the .config [akpm@linux-foundation.org: fix printk warnings] Cc: <linux-arch@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* ext2/3/4: fix file date underflow on ext2 3 filesystems on 64 bit systemsMarkus Rechberger2007-05-083-9/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Taken from http://bugzilla.kernel.org/show_bug.cgi?id=5079 signed long ranges from -2.147.483.648 to 2.147.483.647 on x86 32bit 10000011110110100100111110111101 .. -2,082,844,739 10000011110110100100111110111101 .. 2,212,122,557 <- this currently gets stored on the disk but when converting it to a 64bit signed long value it loses its sign and becomes positive. Cc: Andreas Dilger <adilger@dilger.ca> Cc: <linux-ext4@vger.kernel.org> Andreas says: This patch is now treating timestamps with the high bit set as negative times (before Jan 1, 1970). This means we lose 1/2 of the possible range of timestamps (lopping off 68 years before unix timestamp overflow - now only 30 years away :-) to handle the extremely rare case of setting timestamps into the distant past. If we are only interested in fixing the underflow case, we could just limit the values to 0 instead of storing negative values. At worst this will skew the timestamp by a few hours for timezones in the far east (files would still show Jan 1, 1970 in "ls -l" output). That said, it seems 32-bit systems (mine at least) allow files to be set into the past (01/01/1907 works fine) so it seems this patch is bringing the x86_64 behaviour into sync with other kernels. On the plus side, we have a patch that is ready to add nanosecond timestamps to ext3 and as an added bonus adds 2 high bits to the on-disk timestamp so this extends the maximum date to 2242. Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* Allow access to /proc/$PID/fd after setuid()Alexey Dobriyan2007-05-081-0/+18
| | | | | | | | | | | | | | | | | | | | | | /proc/$PID/fd has r-x------ permissions, so if process does setuid(), it will not be able to access /proc/*/fd/. This breaks fstatat() emulation in glibc. open("foo", O_RDONLY|O_DIRECTORY) = 4 setuid32(65534) = 0 stat64("/proc/self/fd/4/bar", 0xbfafb298) = -1 EACCES (Permission denied) Signed-off-by: Alexey Dobriyan <adobriyan@openvz.org> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: James Morris <jmorris@namei.org> Cc: Chris Wright <chrisw@sous-sol.org> Cc: Ulrich Drepper <drepper@redhat.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Acked-By: Kirill Korotaev <dev@openvz.org> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Christoph Hellwig <hch@lst.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* simplify the stacktrace codeChristoph Hellwig2007-05-088-61/+28
| | | | | | | | | | | | | | | | | | | Simplify the stacktrace code: - remove the unused task argument to save_stack_trace, it's always current - remove the all_contexts flag, it's alwasy 0 Signed-off-by: Christoph Hellwig <hch@lst.de> Cc: Paul Mundt <lethal@linux-sh.org> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Andi Kleen <ak@suse.de> Cc: Akinobu Mita <akinobu.mita@gmail.com> Acked-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* block_write_full_page(): report ENOSPCAndrew Morton2007-05-081-0/+1
| | | | | | | | block_write_full_page() forgot to propagate ENPSOC into the address_space. Cc: Guillaume Chazarain <guichaz@yahoo.fr> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* Factor outstanding I/O error handlingGuillaume Chazarain2007-05-085-30/+17
| | | | | | | | | | Cleanup: setting an outstanding error on a mapping was open coded too many times. Factor it out in mapping_set_error(). Signed-off-by: Guillaume Chazarain <guichaz@yahoo.fr> Cc: Steven Whitehouse <swhiteho@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* uml: an idle system should have zero load averageJeff Dike2007-05-081-2/+2
| | | | | | | | | | | | | | The ever-vigilant users of linode.com noticed that an idle 2.6 UML has a persistent load average of ~.4. It turns out that because the UML timer handler processed softirqs before actually delivering the tick, the tick was counted in the context of the idle thread about half the time. Signed-off-by: Jeff Dike <jdike@linux.intel.com> Cc: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* uml: hostfs style fixesJeff Dike2007-05-083-171/+167
| | | | | | | | | hostfs needed some style goodness. Signed-off-by: Jeff Dike <jdike@linux.intel.com> Cc: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* uml: make hostfs_setattr() support operations on unlinked open filesAlberto Bertogli2007-05-083-44/+74
| | | | | | | | | | | | | | | | | | This patch allows hostfs_setattr() to work on unlinked open files by calling set_attr() (the userspace part) with the inode's fd. Without this, applications that depend on doing attribute changes to unlinked open files will fail. It works by using the fd versions instead of the path ones (for example fchmod() instead of chmod(), fchown() instead of chown()) when an fd is available. Signed-off-by: Alberto Bertogli <albertito@gmail.com> Signed-off-by: Jeff Dike <jdike@linux.intel.com> Cc: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* use SLAB_PANIC flag cleanupAkinobu Mita2007-05-086-27/+11
| | | | | | | | | | | | | | Use SLAB_PANIC and delete duplicated panic(). Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com> Cc: Ian Molton <spyro@f2s.com> Cc: David Howells <dhowells@redhat.com> Cc: Andi Kleen <ak@suse.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mundt <lethal@linux-sh.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* Add white list into modpost.c for memory hotplug code and ia64's machvec sectionYasunori Goto2007-05-083-2/+29
| | | | | | | | | | | | | | | | | | | | | | | | This patch is add white list into modpost.c for some functions and ia64's section to fix section mismatchs. sparse_index_alloc() and zone_wait_table_init() calls bootmem allocator at boot time, and kmalloc/vmalloc at hotplug time. If config memory hotplug is on, there are references of bootmem allocater(init text) from them (normal text). This is cause of section mismatch. Bootmem is called by many functions and it must be used only at boot time. I think __init of them should keep for section mismatch check. So, I would like to register sparse_index_alloc() and zone_wait_table_init() into white list. In addition, ia64's .machvec section is function table of some platform dependent code. It is mixture of .init.text and normal text. These reference of __init functions are valid too. Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com> Cc: Sam Ravnborg <sam@ravnborg.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* Fix section mismatch of memory hotplug related code.Yasunori Goto2007-05-085-22/+26
| | | | | | | | | This is to fix many section mismatches of code related to memory hotplug. I checked compile with memory hotplug on/off on ia64 and x86-64 box. Signed-off-by: Yasunori Goto <y-goto@jp.fujitsu.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* mm: move common segment checks to separate helper functionDmitriy Monakhov2007-05-084-74/+53
| | | | | | | | | | [akpm@linux-foundation.org: cleanup] Signed-off-by: Monakhov Dmitriy <dmonakhov@openvz.org> Cc: Christoph Hellwig <hch@lst.de> Acked-by: Anton Altaparmakov <aia21@cam.ac.uk> Acked-by: David Chinner <dgc@sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* Increase slab redzone to 64bitsDavid Woodhouse2007-05-082-21/+25
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There are two problems with the existing redzone implementation. Firstly, it's causing misalignment of structures which contain a 64-bit integer, such as netfilter's 'struct ipt_entry' -- causing netfilter modules to fail to load because of the misalignment. (In particular, the first check in net/ipv4/netfilter/ip_tables.c::check_entry_size_and_hooks()) On ppc32 and sparc32, amongst others, __alignof__(uint64_t) == 8. With slab debugging, we use 32-bit redzones. And allocated slab objects aren't sufficiently aligned to hold a structure containing a uint64_t. By _just_ setting ARCH_KMALLOC_MINALIGN to __alignof__(u64) we'd disable redzone checks on those architectures. By using 64-bit redzones we avoid that loss of debugging, and also fix the other problem while we're at it. When investigating this, I noticed that on 64-bit platforms we're using a 32-bit value of RED_ACTIVE/RED_INACTIVE in the 64-bit memory location set aside for the redzone. Which means that the four bytes immediately before or after the allocated object at 0x00,0x00,0x00,0x00 for LE and BE machines, respectively. Which is probably not the most useful choice of poison value. One way to fix both of those at once is just to switch to 64-bit redzones in all cases. Signed-off-by: David Woodhouse <dwmw2@infradead.org> Acked-by: Pekka Enberg <penberg@cs.helsinki.fi> Cc: Christoph Lameter <clameter@engr.sgi.com> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* Merge master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6Linus Torvalds2007-05-082-9/+9
|\ | | | | | | | | | | * master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6: [SCSI] esp_scsi: Fix section mismatch warnings. [VIDEO] sunxvr2500: Fix PCI device ID table.
| * [SCSI] esp_scsi: Fix section mismatch warnings.Martin Habets2007-05-071-1/+1
| | | | | | | | | | Signed-off-by: Martin Habets <errandir_news@mph.eclipse.co.uk> Signed-off-by: David S. Miller <davem@davemloft.net>
| * [VIDEO] sunxvr2500: Fix PCI device ID table.David S. Miller2007-05-071-8/+8
| | | | | | | | | | | | Noticed by Meelis Roos. Signed-off-by: David S. Miller <davem@davemloft.net>
* | Fix bluetooth HCI sysfs compileLinus Torvalds2007-05-081-2/+2
|/ | | | | | | More fallout from the removal of "struct subsystem" from the core device model. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* Merge git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6Linus Torvalds2007-05-0724-106/+1861
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6: [IA64] update memory attribute aliasing documentation & test cases [IA64] fail mmaps that span areas with incompatible attributes [IA64] allow WB /sys/.../legacy_mem mmaps [IA64] make ioremap avoid unsupported attributes [IA64] rename ioremap variables to match i386 [IA64] relax per-cpu TLB requirement to DTC [IA64] remove per-cpu ia64_phys_stacked_size_p8 [IA64] Fix example error injection program [IA64] Itanium MC Error Injection Tool: pal_mc_error_inject() interface [IA64] Itanium MC Error Injection Tool: Makefile changes [IA64] Itanium MC Error Injection Tool: Driver sysfs interface [IA64] Itanium MC Error Injection Tool: Doc and sample application [IA64] Itanium MC Error Injection Tool: Kernel configuration
| * Pull mem-attribute into release branchTony Luck2007-04-306-58/+392
| |\
| | * [IA64] update memory attribute aliasing documentation & test casesBjorn Helgaas2007-03-302-34/+284
| | | | | | | | | | | | | | | | | | | | | Updates documentation and adds some test cases. Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
| | * [IA64] fail mmaps that span areas with incompatible attributesBjorn Helgaas2007-03-301-3/+43
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Example memory map (from HP sx1000 with VGA enabled): 0x00000 - 0x9FFFF supports only WB (cacheable) access 0xA0000 - 0xBFFFF supports only UC (uncacheable) access 0xC0000 - 0xFFFFF supports only WB (cacheable) access Some versions of X map the entire 0x00000-0xFFFFF area at once. With the example above, this mmap must fail because there's no memory attribute that's safe for the entire area. Prior to this patch, we performed the mmap with a UC mapping. When X accessed the WB memory at 0xC0000, it caused an MCA. The crash can happen when mapping 0xC0000 from either /dev/mem or a /sys/.../legacy_mem file. Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
| | * [IA64] allow WB /sys/.../legacy_mem mmapsBjorn Helgaas2007-03-301-2/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Allow cacheable mmaps of legacy_mem if WB access is supported for the region. The "legacy_mem" file often contains a shadow option ROM, and some versions of X depend on this. Tim Yamin <plasm@roo.me.uk> reported that this change fixes X on a Dell PowerEdge 3250. Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
| | * [IA64] make ioremap avoid unsupported attributesBjorn Helgaas2007-03-302-10/+56
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Example memory map (from HP sx1000 with VGA enabled): 0x00000 - 0x9FFFF supports only WB (cacheable) access 0xA0000 - 0xBFFFF supports only UC (uncacheable) access 0xC0000 - 0xFFFFF supports only WB (cacheable) access pci_read_rom() indirectly uses ioremap(0xC0000) to read the shadow VGA option ROM. ioremap() used to default to a 16MB or 64MB UC kernel identity mapping, which would cause an MCA when reading 0xC0000 since only WB is supported there. X uses reads the option ROM to initialize devices. A smaller test case is: # echo 1 > /sys/bus/pci/devices/0000:aa:03.0/rom # cp /sys/bus/pci/devices/0000:aa:03.0/rom x To avoid this, we can use the same ioremap_page_range() strategy that most architectures use for all ioremaps. These page table mappings come out of the vmalloc area. On ia64, these are in region 5 (0xA... addresses) and typically use 16KB or 64KB mappings instead of 16MB or 64MB mappings. The smaller mappings give more flexibility to use the correct attributes. Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
| | * [IA64] rename ioremap variables to match i386Bjorn Helgaas2007-03-301-13/+13
| | | | | | | | | | | | | | | | | | | | | No functional change, just use the same names as i386. Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
| * | Pull percpu-dtc into release branchTony Luck2007-04-3012-48/+63
| |\ \
| | * | [IA64] relax per-cpu TLB requirement to DTCChen, Kenneth W2007-02-074-41/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Instead of pinning per-cpu TLB into a DTR, use DTC. This will free up one TLB entry for application, or even kernel if access pattern to per-cpu data area has high temporal locality. Since per-cpu is mapped at the top of region 7 address, we just need to add special case in alt_dtlb_miss. The physical address of per-cpu data is already conveniently stored in IA64_KR(PER_CPU_DATA). Latency for alt_dtlb_miss is not affected as we can hide all the latency. It was measured that alt_dtlb_miss handler has 23 cycles latency before and after the patch. The performance effect is massive for applications that put lots of tlb pressure on CPU. Workload environment like database online transaction processing or application uses tera-byte of memory would benefit the most. Measurement with industry standard database benchmark shown an upward of 1.6% gain. While smaller workloads like cpu, java also showing small improvement. Signed-off-by: Ken Chen <kenneth.w.chen@intel.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
| | * | [IA64] remove per-cpu ia64_phys_stacked_size_p8Chen, Kenneth W2007-02-078-7/+47
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | It's not efficient to use a per-cpu variable just to store how many physical stack register a cpu has. Ever since the incarnation of ia64 up till upcoming Montecito processor, that variable has "glued" to 96. Having a variable in memory means that the kernel is burning an extra cacheline access on every syscall and kernel exit path. Such "static" value is better served with the instruction patching utility exists today. Convert ia64_phys_stacked_size_p8 into dynamic insn patching. This also has a pleasant side effect of eliminating access to per-cpu area while psr.ic=0 in the kernel exit path. (fixable for per-cpu DTC work, but why bother?) There are some concerns with the default value that the instruc- tion encoded in the kernel image. It shouldn't be concerned. The reasons are: (1) cpu_init() is called at CPU initialization. In there, we find out physical stack register size from PAL and patch two instructions in kernel exit code. The code in question can not be executed before the patching is done. (2) current implementation stores zero in ia64_phys_stacked_size_p8, and that's what the current kernel exit path loads the value with. With the new code, it is equivalent that we store reg size 96 in ia64_phys_stacked_size_p8, thus creating a better safety net. Given (1) above can never fail, having (2) is just a bonus. All in all, this patch allow one less memory reference in the kernel exit path, thus reducing syscall and interrupt return latency; and avoid polluting potential useful data in the CPU cache. Signed-off-by: Ken Chen <kenneth.w.chen@intel.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
| * | | Pull error-inject into release branchTony Luck2007-04-306-0/+1406
| |\ \ \
| | * | | [IA64] Fix example error injection programTony Luck2007-02-061-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Progam accessed using /sys/devices/system/node/node0/cpu%d/err_inject/ This path only exists for CONFIG_NUMA=y systems. Better to use /sys/devices/system/cpu/cpu%d/err_inject/ which is available on all systems. Signed-off-by: Tony Luck <tony.luck@intel.com>
| | * | | [IA64] Itanium MC Error Injection Tool: pal_mc_error_inject() interfaceFenghua Yu2007-01-301-0/+33
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch implements pal_mc_error_inject() interface in kernel. Both physical mode and virtual mode are supported. Signed-off-by: Fenghua Yu <fenghua.yu@intel.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
| | * | | [IA64] Itanium MC Error Injection Tool: Makefile changesFenghua Yu2007-01-301-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch has Makefile changes. Signed-off-by: Fenghua Yu <fenghua.yu@intel.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
| | * | | [IA64] Itanium MC Error Injection Tool: Driver sysfs interfaceFenghua Yu2007-01-301-0/+293
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This kernel driver patch provides sysfs interface for user application to call pal_mc_error_inject() procedure. Signed-off-by: Fenghua Yu <fenghua.yu@intel.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
| | * | | [IA64] Itanium MC Error Injection Tool: Doc and sample applicationFenghua Yu2007-01-301-0/+1068
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch contains a documention and sample application. Since the sample application has ~1000 lines of code, it might not be suitable in a kernel documention in kenrel tree. If you think this is not good place to hold the sample application, please let me know and I'm open to other choices e.g. sourceforge etc. Signed-off-by: Fenghua Yu <fenghua.yu@intel.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
| | * | | [IA64] Itanium MC Error Injection Tool: Kernel configurationFenghua Yu2007-01-302-0/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch has kenrel configuration changes for the MC Error Injection Tool. Signed-off-by: Fenghua Yu <fenghua.yu@intel.com> Signed-off-by: Tony Luck <tony.luck@intel.com>
* | | | | Merge branch 'server-cluster-locking-api' of git://linux-nfs.org/~bfields/linuxLinus Torvalds2007-05-0715-205/+546
|\ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * 'server-cluster-locking-api' of git://linux-nfs.org/~bfields/linux: gfs2: nfs lock support for gfs2 lockd: add code to handle deferred lock requests lockd: always preallocate block in nlmsvc_lock() lockd: handle test_lock deferrals lockd: pass cookie in nlmsvc_testlock lockd: handle fl_grant callbacks lockd: save lock state on deferral locks: add fl_grant callback for asynchronous lock return nfsd4: Convert NFSv4 to new lock interface locks: add lock cancel command locks: allow {vfs,posix}_lock_file to return conflicting lock locks: factor out generic/filesystem switch from setlock code locks: factor out generic/filesystem switch from test_lock locks: give posix_test_lock same interface as ->lock locks: make ->lock release private data before returning in GETLK case locks: create posix-to-flock helper functions locks: trivial removal of unnecessary parentheses
| * | | | | gfs2: nfs lock support for gfs2Marc Eshel2007-05-072-10/+104
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add NFS lock support to GFS2. Signed-off-by: Marc Eshel <eshel@almaden.ibm.com> Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu> Acked-by: Steven Whitehouse <swhiteho@redhat.com>
| * | | | | lockd: add code to handle deferred lock requestsMarc Eshel2007-05-074-7/+40
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Rewrite nlmsvc_lock() to use the asynchronous interface. As with testlock, we answer nlm requests in nlmsvc_lock by first looking up the block and then using the results we find in the block if B_QUEUED is set, and calling vfs_lock_file() otherwise. If this a new lock request and we get -EINPROGRESS return on a non-blocking request then we defer the request. Also modify nlmsvc_unlock() to call the filesystem method if appropriate. Signed-off-by: Marc Eshel <eshel@almaden.ibm.com> Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
| * | | | | lockd: always preallocate block in nlmsvc_lock()Marc Eshel2007-05-071-23/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Normally we could skip ever having to allocate a block in the case where the client asks for a non-blocking lock, or asks for a blocking lock that succeeds immediately. However we're going to want to always look up a block first in order to check whether we're revisiting a deferred lock call, and to be prepared to handle the case where the filesystem returns -EINPROGRESS--in that case we want to make sure the lock we've given the filesystem is the one embedded in the block that we'll use to track the deferred request. Signed-off-by: Marc Eshel <eshel@almaden.ibm.com> Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
| * | | | | lockd: handle test_lock deferralsMarc Eshel2007-05-073-16/+73
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Rewrite nlmsvc_testlock() to use the new asynchronous interface: instead of immediately doing a posix_test_lock(), we first look for a matching block. If the subsequent test_lock returns anything other than -EINPROGRESS, we then remove the block we've found and return the results. If it returns -EINPROGRESS, then we defer the lock request. In the case where the block we find in the first step has B_QUEUED set, we bypass the vfs_test_lock entirely, instead using the block to decide how to respond: with nlm_lck_denied if B_TIMED_OUT is set. with nlm_granted if B_GOT_CALLBACK is set. by dropping if neither B_TIMED_OUT nor B_GOT_CALLBACK is set Signed-off-by: Marc Eshel <eshel@almaden.ibm.com> Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-29 02:10:50 +0100