- 24 Aug, 2004 40 commits
-
-
Andrew Morton authored
Fix up the logic which decides when the caller can dip into page reserves. - If the caller has realtime scheduling policy, or if the caller cannot run direct reclaim, then allow the caller to use up to a quarter of the page reserves. - If the caller has __GFP_HIGH then allow the caller to use up to half of the page reserves. - If the caller has PF_MEMALLOC then the caller can use 100% of the page reserves. Signed-off-by:
Andrew Morton <akpm@osdl.org> Signed-off-by:
Linus Torvalds <torvalds@osdl.org>
-
Nick Piggin authored
Previously the ->protection[] logic was broken. It was difficult to follow and basically didn't use the asynch reclaim watermarks (pages_min, pages_low, pages_high) properly. Now use ->protection *only* for lower-zone protection. So the allocator now explicitly uses the ->pages_low, ->pages_min watermarks and adds ->protection on top of that, instead of trying to use ->protection for everything. Pages are allocated down to (->pages_low + ->protection), once this is reached, kswapd the background reclaim is started; after this, we can allocate down to (->pages_min + ->protection) without blocking; the memory below pages_min is reserved for __GFP_HIGH and PF_MEMALLOC allocations. kswapd attempts to reclaim memory until ->pages_high is reached. Signed-off-by:
Nick Piggin <nickpiggin@yahoo.com.au> Signed-off-by:
-
Nick Piggin authored
Slightly change the writeout watermark calculations so we keep background and synchronous writeout watermarks in the same ratios after adjusting them for the amout of mapped memory. This ensures we should always attempt to start background writeout before synchronous writeout and preserves the admin's desired background-versus-forground ratios after we've auto-adjusted one of them. Signed-off-by:
Nick Piggin <nickpiggin@cyberone.com.au> Signed-off-by:
-
Hugh Dickins authored
A tmpfs user reported increasingly slow directory reads when repeatedly creating and unlinking in a mkstemp-like way. The negative dentries accumulate alarmingly (until memory pressure finally frees them), and are just a hindrance to any in-memory filesystem. simple_lookup set d_op to arrange for negative dentries to be deleted immediately. (But I failed to discover how it is that on-disk filesystems seem to keep their negative dentries within manageable bounds: this effect was gross with tmpfs or ramfs, but no problem at all with extN or reiser.) Signed-off-by:
Hugh Dickins <hugh@veritas.com> Signed-off-by:
-
Hugh Dickins authored
Clarify mmgrab by collapsing it into get_task_mm (in fork.c not inline), and commenting on the special case it is guarding against: when use_mm in an AIO daemon temporarily adopts the mm while it's on its way out. Signed-off-by:
-
Marcelo Tosatti authored
Back when we were discussing the need for a memory barrier in sync_page(), it came to me (thanks Andrea!) that the bit operations can be perfectly reordered on architectures other than x86. I think the commentary on i386 bitops.h is misleading, its worth to note that that these operations are not guaranteed not to be reordered on different architectures. clear_bit() already does that: * clear_bit() is atomic and may not be reordered. However, it does * not contain a memory barrier, so if it is used for locking purposes, * you should call smp_mb__before_clear_bit() and/or smp_mb__after_clear_bit() * in order to ensure changes are visible on other processors. Signed-off-by:
-
Hugh Dickins authored
Swapoff can make good use of a page's anon_vma and index, while it's still left in swapcache, or once it's brought back in and the first pte mapped back: unuse_vma go directly to just one page of only those vmas with the same anon_vma. And unuse_process can skip any vmas without an anon_vma (extending the hugetlb check: hugetlb vmas have no anon_vma). This just hacks in on top of the existing procedure, still going through all the vmas of all the mms in mmlist. A more elegant procedure might replace mmlist by a list of anon_vmas: but that would be more work to implement, with apparently more overhead in the common paths. Signed-off-by:
-
Hugh Dickins authored
With page_map_lock out of the way, there's no need for page_referenced and try_to_unmap to use trylocks - provided we switch anon_vma->lock and mm->page_table_lock around in anon_vma_prepare. Though I suppose it's possible that we'll find that vmscan makes better progress with trylocks than spinning - we're free to choose trylocks again if so. Try to update the mm lock ordering documentation in filemap.c. But I still find it confusing, and I've no idea of where to stop. So add an mm lock ordering list I can understand to rmap.c. Signed-off-by:
-
Hugh Dickins authored
With page_map_lock gone, how to stabilize page->mapping's anon_vma while acquiring anon_vma->lock in page_referenced_anon and try_to_unmap_anon? The page cannot actually be freed (vmscan holds reference), but however much we check page_mapped (which guarantees that anon_vma is in use - or would guarantee that if we added suitable barriers), there's no locking against page becoming unmapped the instant after, then anon_vma freed. It's okay to take anon_vma->lock after it's freed, so long as it remains a struct anon_vma (its list would become empty, or perhaps reused for an unrelated anon_vma: but no problem since we always check that the page located is the right one); but corruption if that memory gets reused for some other purpose. This is not unique: it's liable to be problem whenever the kernel tries to approach a structure obliquely. It's generally solved with an atomic reference count; but one advantage of anon_vma over anonmm is that it does not have such a count, and it would be a backward step to add one. Therefore... implement SLAB_DESTROY_BY_RCU flag, to guarantee that such a kmem_cache_alloc'ed structure cannot get freed to other use while the rcu_read_lock is held i.e. preempt disabled; and use that for anon_vma. Fix concerns raised by Manfred: this flag is incompatible with poisoning and destructor, and kmem_cache_destroy needs to synchronize_kernel. I hope SLAB_DESTROY_BY_RCU may be useful elsewhere; but though it's safe for little anon_vma, I'd be reluctant to use it on any caches whose immediate shrinkage under pressure is important to the system. Signed-off-by:
-
Hugh Dickins authored
The pte_chains rmap used pte_chain_lock (bit_spin_lock on PG_chainlock) to lock its pte_chains. We kept this (as page_map_lock: bit_spin_lock on PG_maplock) when we moved to objrmap. But the file objrmap locks its vma tree with mapping->i_mmap_lock, and the anon objrmap locks its vma list with anon_vma->lock: so isn't the page_map_lock superfluous? Pretty much, yes. The mapcount was protected by it, and needs to become an atomic: starting at -1 like page _count, so nr_mapped can be tracked precisely up and down. The last page_remove_rmap can't clear anon page mapping any more, because of races with page_add_rmap; from which some BUG_ONs must go for the same reason, but they've served their purpose. vmscan decisions are naturally racy, little change there beyond removing page_map_lock/unlock. But to stabilize the file-backed page->mapping against truncation while acquiring i_mmap_lock, page_referenced_file now needs page lock to be held even for refill_inactive_zone. There's a similar issue in acquiring anon_vma->lock, where page lock doesn't help: which this patch pretends to handle, but actually it needs the next. Roughly 10% cut off lmbench fork numbers on my 2*HT*P4. Must confess my testing failed to show the races even while they were knowingly exposed: would benefit from testing on racier equipment. Signed-off-by:
-
Hugh Dickins authored
First of a batch of five patches to eliminate rmap's page_map_lock, replace its trylocking by spinlocking, and use anon_vma to speed up swapoff. Patches updated from the originals against 2.6.7-mm7: nothing new so I won't spam the list, but including Manfred's SLAB_DESTROY_BY_RCU fixes, and omitting the unuse_process mmap_sem fix already in 2.6.8-rc3. This patch: Replace the PG_anon page->flags bit by setting the lower bit of the pointer in page->mapping when it's anon_vma: PAGE_MAPPING_ANON bit. We're about to eliminate the locking which kept the flags and mapping in synch: it's much easier to work on a local copy of page->mapping, than worry about whether flags and mapping are in synch (though I imagine it could be done, at greater cost, with some barriers). Signed-off-by:
-
Roger Luethi authored
I really wanted /proc/pid/statm to die and I still believe the reasoning is valid. As it doesn't look like that is going to happen, though, I offer this fix for the respective documentation. Note: lrs/drs fields are switched. Signed-off-by:
Roger Luethi <rl@hellgate.ch> Signed-off-by:
-
Arjan van de Ven authored
This enables apic=bigsmp automatically on some big HP machines that need it. This makes them boot without kernel parameters on a generic arch kernel. Signed-off-by:
-
William Lee Irwin III authored
We need to be able to dereference struct device in include/asm-ia64/dma-mapping.h. Signed-off-by:
-
Andi Kleen authored
Prevents some noise during boot up when no MD volumes are found. I think I picked it up from someone else, but I cannot remember from whom (sorry) Cc: Neil Brown <neilb@cse.unsw.edu.au> Signed-off-by:
-
Pete Zaitcev authored
We at Red Hat shipped a larger number of arguments for quite some time, it was required for installations on IBM mainframe (s390), which doesn't have a good way to pass arguments. There are a number of reasonable situations that go past the current limits of 8. One that comes to mind is when you want to perform a manual vnc install on a headless machine using anaconda. This requires passing in a number of parameters to get anaconda past the initial (no-gui) loader screens. Signed-off-by:
-
Suparna Bhattacharya authored
From: Chris Mason I compared the 2.6 pipetest results with the 2.4 suse kernel, and 2.6 was roughly 40% slower. During the pipetest run, 2.6 generates ~600,000 context switches per second while 2.4 generates 30 or so. aio-context-switch (attached) has a few changes that reduces our context switch rate, and bring performance back up to 2.4 levels. These have only really been tested against pipetest, they might make other workloads worse. The basic theory behind the patch is that it is better for the userland process to call run_iocbs than it is to schedule away and let the worker thread do it. 1) on io_submit, use run_iocbs instead of run_iocb 2) on io_getevents, call run_iocbs if no events were available. 3) don't let two procs call run_iocbs for the same context at the same time. They just end up bouncing on spinlocks. The first three optimizations got me down to 360,000 context switches per second, and they help build a little structure to allow optimization #4, which uses queue_delayed_work(HZ/10) instead of queue_work. That brings down the number of context switches to 2.4 levels. Adds aio_run_all_iocbs so that normal processes can run all the pending retries on the run list. This allows worker threads to keep using list splicing, but regular procs get to run the list until it stays empty. The end result should be less work for the worker threads. I was able to trigger short stalls (1sec) with aio-stress, and with the current patch they are gone. Could be wishful thinking on my part though, please let me know how this works for you. Signed-off-by: -
Suparna Bhattacharya authored
This patch tries be a little fairer across multiple io contexts in handling retries, helping make sure progress happens uniformly across different io contexts (especially if they are acting on independent queues). It splices the ioctx runlist before processing it in __aio_run_iocbs. If new iocbs get added to the ctx in meantime, it queues a fresh workqueue entry instead of handling them righaway, so that other ioctxs' retries get a chance to be processed before the newer entries in the queue. This might make a difference in a situation where retries are getting queued very fast on one ioctx, while the workqueue entry for another ioctx is stuck behind it. I've only seen this occasionally earlier and can't recreate it consistently, but may be worth including anyway. Signed-off-by:
-
Suparna Bhattacharya authored
From: Daniel McNeil <daniel@osdl.org> From: Chris Mason <mason@suse.com> AIO: retry infrastructure fixes and enhancements Reorganises, comments and fixes the AIO retry logic. Fixes and enhancements include: - Split iocb setup and execution in io_submit (also fixes io_submit error reporting) - Use aio workqueue instead of keventd for retries - Default high level retry methods - Subtle use_mm/unuse_mm fix - Code commenting - Fix aio process hang on EINVAL (Daniel McNeil) - Hold the context lock across unuse_mm - Acquire task_lock in use_mm() - Allow fops to override the retry method with their own - Elevated ref count for AIO retries (Daniel McNeil) - set_fs needed when calling use_mm - Flush workqueue on __put_ioctx (Chris Mason) - Fix io_cancel to work with retries (Chris Mason) - Read-immediate option for socket/pipe retry support Note on default high-level retry methods support ================================================ High-level retry methods allows an AIO request to be executed as a series of non-blocking iterations, where each iteration retries the remaining part of the request from where the last iteration left off, by reissuing the corresponding AIO fop routine with modified arguments representing the remaining I/O. The retries are "kicked" via the AIO waitqueue callback aio_wake_function() which replaces the default wait queue entry used for blocking waits. The high level retry infrastructure is responsible for running the iterations in the mm context (address space) of the caller, and ensures that only one retry instance is active at a given time, thus relieving the fops themselves from having to deal with potential races of that sort. Signed-off-by: -
Bjorn Helgaas authored
Add pci_enable_device()/pci_disable_device(). In the past, drivers often worked without this, but it is now required in order to route PCI interrupts correctly. Signed-off-by:
Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by:
-
Bjorn Helgaas authored
Add pci_enable_device()/pci_disable_device(). In the past, drivers often worked without this, but it is now required in order to route PCI interrupts correctly. Signed-off-by:
-
Bjorn Helgaas authored
Add pci_enable_device()/pci_disable_device(). In the past, drivers often worked without this, but it is now required in order to route PCI interrupts correctly. Signed-off-by:
-
Bjorn Helgaas authored
Add pci_enable_device()/pci_disable_device(). In the past, drivers often worked without this, but it is now required in order to route PCI interrupts correctly. Signed-off-by:
-
Bjorn Helgaas authored
Add pci_enable_device()/pci_disable_device(). In the past, drivers often worked without this, but it is now required in order to route PCI interrupts correctly. Signed-off-by:
-
Bjorn Helgaas authored
Add pci_enable_device()/pci_disable_device(). In the past, drivers often worked without this, but it is now required in order to route PCI interrupts correctly. Signed-off-by:
-
Bjorn Helgaas authored
I don't have this hardware, so this has been compiled but not tested. Add pci_enable_device()/pci_disable_device In the past, drivers often worked without this, but it is now required in order to route PCI interrupts correctly. In addition, this driver incorrectly used the IRQ value from PCI config space rather than the one in the struct pci_dev. Signed-off-by:
-
Bjorn Helgaas authored
Add pci_enable_device()/pci_disable_device(). In the past, drivers often worked without this, but it is now required in order to route PCI interrupts correctly. Signed-off-by:
-
Dave Hansen authored
When using CONFIG_NONLINEAR, a zone's mem_map isn't contiguous, and isn't allocated in the same place. This means that nonlinear doesn't really have a mem_map[] to pass into free_area_init_node() or memmap_init_zone() which makes any sense. So, this patch removes the 'struct page *mem_map' argument to both of those functions. All non-NUMA architectures just pass a NULL in there, which is ignored. The solution on the NUMA arches is to pass the mem_map in via the pgdat, which works just fine. To replace the removed arguments, a call to pfn_to_page(node_start_pfn) is made. This is valid because all of the pfn_to_page() implementations rely only on the pgdats, which are already set up at this time. Plus, the pfn_to_page() method should work for any future nonlinear-type code. Finally, the patch creates a function: node_alloc_mem_map(), which I plan to effectively #ifdef out for nonlinear at some future date. Compile tested and booted on SMP x86, NUMAQ, and ppc64. From: Jesse Barnes <jbarnes@engr.sgi.com> Fix up ia64 specific memory map init function in light of Dave's memmap_init cleanups. Signed-off-by:
Jesse Barnes <jbarnes@sgi.com> From: Dave Hansen <haveblue@us.ibm.com> Looks like I missed a couple of architectures. This patch, on top of my previous one and Jesse's should clean up the rest. From: William Lee Irwin III <wli@holomorphy.com> x86-64 wouldn't compile with NUMA support on, as node_alloc_mem_map() references mem_map outside #ifdefs on CONFIG_NUMA/CONFIG_DISCONTIGMEM. This patch wraps that reference in such an #ifdef. From: William Lee Irwin III <wli@holomorphy.com> Initializing NODE_DATA(nid)->node_mem_map prior to calling it should do. From: Dave Hansen <haveblue@us.ibm.com> Rick, I bet you didn't think your nerf weapons would be so effective in getting that compile error fixed, did you? Applying the attached patch and commenting out this line: arch/i386/kernel/nmi.c: In function `proc_unknown_nmi_panic': arch/i386/kernel/nmi.c:558: too few arguments to function `proc_dointvec' will let it compile. Signed-off-by:
-
Guillaume Thouvenin authored
Function wdtpci_init_one() in file wdt_pci.c generates a warning when compiling the watchdog driver. Signed-off-by:
-
William Lee Irwin III authored
Zwane Mwaikambo <zwane@fsmlabs.com> wrote: The following caused some fireworks whilst merging i386 cpu hotplug. any_online_cpu(0x2) returns 32 on i386 if we're forced to continue past the only set bit due to the additional find_first_bit in the find_next_bit i386 implementation. Not wanting to change current behaviour in the bitops primitives and since the NR_CPUS thing is a cpumask issue, i've opted to fix next_cpu() and first_cpu() instead. This might save a couple of lines of code. From: <akpm@osdl.org> Fix cross-arch ulong/int disaster with find_next_bit(). Signed-off-by:
-
Andi Kleen authored
This fixes various issues in the previous update, in particular a kernel without CONFIG_GART_IOMMU should boot now again, The kernel discoverys PCI BUS<->CPU affinity on AMD systems now. It is so far used by dma_alloc_coherent to allocate memory Experimental patches to add this to sysfs exist, but they're not included yet. On systems with no memory on a CPU this information may be wrong. It has a new experimental CONFIG_UNORDERED_IO option. When enabled it uses write combining for stores to device iomemory mapping. This may give better performance with some device drivers, but has a slight risk of breaking drivers (in general if a driver works on ia64,ppc64,sparc64 it should also work). Based on some discussions with Grant Grundler. It requires the driver to use memory barriers properly. I would be interested in feedback on any performance changes you're seeing. For a production system I would recommend to keep it turned off(although I run it on all my systems and haven't run into any problems yet) ACPI and Centrino speedstep is enabled now for Nocona systems. The IOMMU code does lazy merging by default now, which should be safe and may increase performance on block IO. It also avoids SAC force by default now. The machine check code has been improved again, hopefully it is good now. It will log now machine check events from before the last reset. And various other fixes. The x86-64 parts are now gcc 3.5 clean. And various other fixes - Update defconfig - Reset lost ticks on lost time warning, print RIP. - Make TASK_SIZE test for 32bit (Arjan van de Ven) - Work around bug in generic code that broke pcibus_to_cpumask - Actually fix dummy iommu code - Compile i386 acpi and speedstep-centrino cpufreq modules - Export cpu_khz - Fix compilation without GART_IOMMU - Optimize find_*_bit functions for small fields - Discover nodes near PCI busses on K8 (Travis Betak, changed by me) - Optimize gart tlb flush slightly - Add experimental CONFIG_UNORDERED_IO for unordered IO stores - Add 32bit emulation for PTRACE_GETEVENTMSG - Fix kernel_fpu_{begin,end} for preemptive kernels (Alexander Nyberg) - Readd proper check for biomerge (got lost) - Set up 32bit vsyscall page for ptrace early - Add 32bit emulation for lookup_dcookie() for oprofile - Export copy_page / clear_page - Use rex prefix in save_init_fpu fxsave (Jan Beulich) - Make it compile again - Fix handling of hwdev == NULL (= ISA/LPC devices) in swiotlb - Convert PCI DMA code to dma devices - Change IOMMU code to use dummy fallback device instead of hardcoded NULL tests everywhere. - Test iommu_sac_force instead of nommu for DAC supported macro (will cause more drivers to use DAC) - Harden non IOMMU dma_alloc_consistent code to fail less likely. - Remove use of strsep in option parsers - Remove duplicated exports (Arjan van der Ven) - Fix EFAULT checking in ptrace (John Blackwood) - Update defconfig - Remove dead URL from boot/setup.S (R.J. Wysocki) - Use compat_sigval_t instead of sigval_t32 (Al Viro) - Nanooptimization in 32bit ptregs calls - Fix gcc 3.5 compilation in mtrr.h - Pass pt_regs as pointer to avoid illegal pass by reference (for gcc 3.5) - Make set_bit take int not long (Harald Dunkel) - Avoid panic on pci_map_sg and pci_alloc_consistent overflow in GART IOMMU - Handle large lost time delays in HPET code (Suresh B. Siddha) - Work around theoretical bugs in prefetch handling (suggested by Jamie Lokier) - Remove mtrr_strings declaration for gcc 3.5 - Set KBUILD_IMAGE for make rpm (William Lee Irwin III) - Add iommu=noaperture to not touch the aperture - Clean up argument parsing for iommu= option - Export symbols for xchgadd based rwsems (still disabled) - Define iommu_bio_merge for !CONFIG_GART_IOMMU - Don't use backwards rep ; movsb for memmove - Out line bitmap search functions (saves 8k .text, from i386) - Convert bitmap search functions to 64bit accesses and optimize them a bit. - Handle corrupted page tables in page fault handler - Set iommu_merge (without force) to on by default again. - Don't do bio merging by default for iommu=merge. This should make it safe to use again - Add iommu=biomerge option to enable BIO merging (like old iommu=merge) - Fix iommu=memaper=... parsing - More MCE fixes (based on a patch by Eric Morton, heavily changed by me) - Fix check for banks causing exceptions - Allow to reinit MCEs later even after mce=off, fix wrong use of __initdata to disable at boot, but reenable later. - Log left over machine checks after boot and resume - Fix missing prototype warning with CPU_FREQ on - Fix parsing of noexec=on (Ian Hastie) - Fix warning in ia32_binfmt.c - Resync time variable cpu frequency handling with i386 - Resync msr.c with i386 - Add 0x60 level 1 intel cache descriptor (from i386) - Remove duplicated 32bit ioctls (Arnd Bergmann) - Enable -msoft-float (from i386) - Use faster version of FPU hang fix - handle the exception * a bit experimental, if you see "kernel ... math error" events in the log please report. Signed-off-by: -
Adam Kropelin authored
Adds a kernel boot parameter "lpj=NNN" which allows the operator to specify the loops-per-jiffy value. This shaves up to a quarter of a second off boot times, which are critical for embedded appliances. It's a bit thin, but the code is in __init. Signed-off-by:
Adam Kropelin <akropel1@rochester.rr.com> Signed-off-by:
-
Mika Kukkonen authored
CC [M] drivers/isdn/hisax/avm_pci.o drivers/isdn/hisax/avm_pci.c: In function `setup_avm_pcipnp': drivers/isdn/hisax/avm_pci.c:817: warning: label `ready' defined but not used Patch is big because I replaced the '} else { ... }' with 'goto ready; }' and so had to remove one level of indentation from code. Signed-off-by: -
Jeff Dike authored
This patch includes the following - updated defconfig move uml.lds.S and main.c from arch/um to arch/um/kernel per Sam's suggestions steal bitops.c from arch/i386 convert all calls to open_private_file to dentry_open Signed-off-by:
-
Jeff Dike authored
The patch below fixes a few UML-specific bugs not related to the rest of the kernel a bogus error return and some formatting in the fork code correct calculation of task.thread.kernel_stack remove a bogus panic a couple of fixes to allow UML to boot in the presence of exec-shield Signed-off-by:
-
Jeff Dike authored
The patch below brings UML up to date with interface changes and the like irq.c includes profile.h to bring in a missing definition use the cpu_{set,clear} interface use the new get_signal_to_deliver interface define instruction_pointer Signed-off-by: -
Coywolf Qi Hunt authored
This patch removes a group of unused bh functions in um. This 2.2 legacy code should be cleaned up. Signed-off-by:
Coywolf Qi Hunt <coywolf@greatcn.org> Cc: Jeff Dike <jdike@addtoit.com> Signed-off-by:
-
Paolo \'Blaisorblade\' Giarrusso authored
Update os_process_pc and os_process_parent: now a PID can be > 32768 (so increase number of digits) and make it work even with spaces in the command name. Signed-off-by:
Paolo 'Blaisorblade' Giarrusso <blaisorblade_spam@yahoo.it> Cc: Jeff Dike <jdike@addtoit.com> Signed-off-by:
-
Paolo \'Blaisorblade\' Giarrusso authored
Signed-off-by:
-
Paolo \'Blaisorblade\' Giarrusso authored
From: Oleg Drokin <green@linuxhacker.ru>, Jeff Dike <jdike@addtoit.com>, and me If size > 128K, with this patch malloc will call vmalloc; free will detect whether to call vfree or kfree or __real_free(). The 2.4 version could forget free()ing something; this has been fixed. Signed-off-by:
-
