1. 10 Sep, 2002 40 commits
    • Sam Ravnborg's avatar
      [PATCH] ftape EXPORT_SYMBOL damage clean-up · b8c2ab08
      Sam Ravnborg authored
      The reason for the ftape messup of export-objs is the usage of the
      strange FT_KSYM macro in ftape_syms.c.
      
      That exist solely for backwards compatibility for kernel 2.1.18 and older.
      
      Better clean it up.
      b8c2ab08
    • Matthew Wilcox's avatar
      [PATCH] sleeping file locks · 48b3fb54
      Matthew Wilcox authored
       - Add FL_SLEEP flag to indicate we intend to sleep and therefore desire
         to be placed on the block list.  Use it for POSIX & flock locks.
       - Remove locks_block_on.
       - Change posix_unblock_lock to eliminate a race that will appear once we
         don't use the BKL any more.
       - Update the comment for locks_same_owner() and rename it to
         posix_same_owner().
       - Change locks_mandatory_area() to allocate its lock on the stack and
         call posix_lock_file() instead of repeating that logic.
       - Rename the "caller" parameter to posix_lock_file() to "request"
         to better show that this is not to be inserted directly.
       - Redo some of the proc code a little.  Stop exposing kernel addresses
         to userspace (whoever thought _that_ was a good idea?!) and show how
         we should be printing the device name.  The last part is ifdeffed
         out to avoid breaking lslk.
       - Remove FL_BROKEN.  And there was much rejoicing.
      48b3fb54
    • Art Haas's avatar
      [PATCH] designated initializer patches for fs_nfs · af692bd6
      Art Haas authored
        Here are some patches for C99 initializers in fs/nfs. Patches
        are against 2.5.32.
      af692bd6
    • Celso González's avatar
      [PATCH] drivers_net_at1700.c save_flags unsigned check · a541bd3a
      Celso González authored
        The function save_flags must use unsigned long instead long (signed)
        This trivial patch solves the problem
      a541bd3a
    • Celso González's avatar
      [PATCH] drivers_net_pcmcia_smc91c92_cs.c · 0c87aa01
      Celso González authored
        The function save_flags must use an unsigned long parameter instead a
        long (signed) one
        This trivial patch solves the problem
      0c87aa01
    • Celso González's avatar
      [PATCH] drivers_net_pcmcia_aironet4500_cs.c save_flags unsigned check · 8cd672dc
      Celso González authored
        The function save_flags must use an unsigned long parameter instead a
        long (signed) one
        This trivial patch solves the problem
      8cd672dc
    • Brad Hards's avatar
      [PATCH] Re: header cleanup - drivers_ieee1394_sbp2.c · f5b07cc7
      Brad Hards authored
        <asm/io.h> has the normal idempotent construction on every architecture.
      
        The attached file removes the second #include.
      f5b07cc7
    • Brad Hards's avatar
      [PATCH] header cleanup - drivers_char_serial_tx3912.c · c7195a11
      Brad Hards authored
        <linux/init.h> has the normal idempotent construction.
      
        The attached file removes the second #include.
      c7195a11
    • Celso González's avatar
      [PATCH] drivers_net_de600.c save_flags unsigned check · c374f6aa
      Celso González authored
        The function save_flags must use unsigned long instead long (signed)
        This trivial patch solves the problem
      c374f6aa
    • Lucas Correia Villa Real's avatar
      [PATCH] 2.5.31_drivers_char_lp.c · bfa61ad0
      Lucas Correia Villa Real authored
        This is a trivial patch already applied in the -ac tree for the 2.4.19 kernel.
        Patch for lp.c avoid +/- operations with 0 and explicit some debug information
        as KERN_INFO or KERN_ERR.
      bfa61ad0
    • James Mayer's avatar
      4b3ceffb
    • Randy Dunlap's avatar
      [PATCH] 2.5.31 spell_typo fix · 511cc63c
      Randy Dunlap authored
      511cc63c
    • Peter Samuelson's avatar
      [PATCH] remove duplicated AGP Config.in · cc263bd2
      Peter Samuelson authored
        drivers/char/Config.in still has a complete copy of agp/Config.in.
        It's an exact cut-n-paste - the md5sums even match. (:
      cc263bd2
    • Marcus Alanen's avatar
      [PATCH] [patch 2.5] at1700 trivial · 52ae1390
      Marcus Alanen authored
        Bad error path..
      
        ret is already set to -ENODEV, no need to set them again before
        jumping out.
      52ae1390
    • Rusty Russell's avatar
      [PATCH] Designated initializers for cs46xx drivers · 3a1ed3f3
      Rusty Russell authored
      The old form of designated initializers are obsolete: we need to
      replace them with the ISO C forms before 2.6.  Gcc has always supported
      both forms anyway.
      3a1ed3f3
    • Bernhard Fischer's avatar
      1560d74b
    • Rusty Russell's avatar
      [PATCH] Designated initializers for shm · 78642ee8
      Rusty Russell authored
      The old form of designated initializers are obsolete: we need to
      replace them with the ISO C forms before 2.6.  Gcc has always supported
      both forms anyway.
      78642ee8
    • Celso González's avatar
      [PATCH] drivers_net_ni65.c save_flags unsigned check · 9bbe1a9c
      Celso González authored
        The function save_flags must use unsigned long instead long (signed)
        This trivial patch solves the problem
      9bbe1a9c
    • Celso González's avatar
      [PATCH] drivers_net_pcmcia_3c574_cs.c save_flags unsigned check · aa3ff3b2
      Celso González authored
        The function save_flags must use an unsigned long parameter instead a
        long (signed) one
      
        This trivial patch solves the problem
      aa3ff3b2
    • Brad Hards's avatar
      [PATCH] header cleanup - drivers_char_dz.c · 8a0dfac8
      Brad Hards authored
        <linux/serial.h> has the normal idempotent construction.
      
        The attached file removes the second #include.
      8a0dfac8
    • Marcus Alanen's avatar
      eefc72e9
    • Matt Domsch's avatar
      [PATCH] Domsch zip code change · df33c3f2
      Matt Domsch authored
        Trivial patch changes my zip code.  Applies to 2.4.x and 2.5.x trees.
      df33c3f2
    • Celso González's avatar
      [PATCH] drivers_net_3c505.c save_flags unsigned check · 2a44c0a3
      Celso González authored
        The function save_flags must use unsigned long instead long (signed)
        This trivial patch solves the problem
      2a44c0a3
    • Skip Ford's avatar
      [PATCH] Comment fix asm-i386_hardirq.h · 1953e4b5
      Skip Ford authored
      1953e4b5
    • Celso González's avatar
      [PATCH] drivers_net_hamradio_scc.c save_flags unsigned check · da203172
      Celso González authored
        The function save_flags must use unsigned long instead long (signed)
        This trivial patch solves the problem
      da203172
    • Celso González's avatar
      [PATCH] drivers_net_arcnet_arcnet.c save_flags unsigned check · 8d37d851
      Celso González authored
        The function save_flags must use unsigned long instead long (signed)
        This trivial patch solves the problem
      8d37d851
    • James Mayer's avatar
      [PATCH] Typos in drivers_s390_net_iucv.h · b9aa5e4d
      James Mayer authored
      b9aa5e4d
    • Celso González's avatar
      [PATCH] drivers_net_pcmcia_fmvj18x_cs.c save_flags unsigned check · 530df4b1
      Celso González authored
        The function save_flags must use an unsigned long parameter instead a
        long (signed) one
      
        This trivial patch solves the problem
      530df4b1
    • Linus Torvalds's avatar
      The scheduler should complain not just about interrupts, · 231991f4
      Linus Torvalds authored
      but also about being called whenever we're holding any
      other preemption locks.
      231991f4
    • Linus Torvalds's avatar
      atari_rootsec.h moved to fs/partitions/atari.h, but somehow the · 34fc0f54
      Linus Torvalds authored
      version in include/linux didn't get deleted.
      34fc0f54
    • Andrew Morton's avatar
      [PATCH] resurrect CONFIG_HIGHPTE · 81e0a1a6
      Andrew Morton authored
      Bill Irwin's patch to fix up pte's in highmem.
      
      With CONFIG_HIGHPTE, the direct pte pointer in struct page becomes the
      64-bit physical address of the single pte which is mapping this page.
      
      If the page is not PageDirect then page->pte.chain points at a list of
      pte_chains, which each now contain an array of 64-bit physical
      addresses of the pte's which are mapping the page.
      
      The functions rmap_ptep_map() and rmap_ptep_unmap() are used for
      mapping and unmapping the page which backs the target pte.
      
      The patch touches all architectures (adding do-nothing compatibility
      macros and inlines).  It generally mangles lots of header files and may
      break non-ia32 compiles.  I've had it in testing since 2.5.31.
      81e0a1a6
    • Andrew Morton's avatar
      [PATCH] rmap pte_chain speedup and space saving · 9dc8af80
      Andrew Morton authored
      The pte_chains presently consist of a pte pointer and a `next' link.
      So there's a 50% memory wastage here as well as potential for a lot of
      misses during walks of the singly-linked per-page list.
      
      This patch increases the pte_chain structure to occupy a full
      cacheline.  There are 7, 15 or 31 pte pointers per structure rather
      than just one.  So the wastage falls to a few percent and the number of
      misses during the walk is reduced.
      
      The patch doesn't make much difference in simple testing, because in
      those tests the pte_chain list from the previous page has good cache
      locality with the next page's list.
      
      The patch sped up Anton's "10,000 concurrently exitting shells" test by
      3x or 4x.  It gives a 10% reduction in system time for a kernel build
      on 16p NUMAQ.
      
      It saves memory and reduces the amount of work performed in the slab
      allocator.
      
      Pages which are mapped by only a single process continue to not have a
      pte_chain.  The pointer in struct page points directly at the mapping
      pte (a "PageDirect" pte pointer).  Once the page is shared a pte_chain
      is allocated and both the new and old pte pointers are moved into it.
      
      We used to collapse the pte_chain back to a PageDirect representation
      in page_remove_rmap().  That has been changed.  That collapse is now
      performed inside page reclaim, via page_referenced().  The thinking
      here is that if a page was previously shared then it may become shared
      again, so leave the pte_chain structure in place.  But if the system is
      under memory pressure then start reaping them anyway.
      9dc8af80
    • Andrew Morton's avatar
      [PATCH] buffer_head takedown for bighighmem machines · e182d612
      Andrew Morton authored
      This patch addresses the excessive consumption of ZONE_NORMAL by
      buffer_heads on highmem machines.  The algorithms which decide which
      buffers to shoot down are fairly dumb, but they only cut in on machines
      with large highmem:lowmem ratios and the code footprint is tiny.
      
      The buffer.c change implements the buffer_head accounting - it sets the
      upper limit on buffer_head memory occupancy to 10% of ZONE_NORMAL.
      
      A possible side-effect of this change is that the kernel will perform
      more calls to get_block() to map pages to disk.  This will only be
      observed when a file is being repeatadly overwritten - this is the only
      case in which the "cached get_block result" in the buffers is useful.
      
      I did quite some testing of this back in the delalloc ext2 days, and
      was not able to come up with a test in which the cached get_block
      result was measurably useful.  That's for ext2, which has a fast
      get_block().
      
      A desirable side effect of this patch is that the kernel will be able
      to cache much more blockdev pagecache in ZONE_NORMAL, so there are more
      ext2/3 indirect blocks in cache, so with some workloads, less I/O will
      be performed.
      
      In mpage_writepage(): if the number of buffer_heads is excessive then
      buffers are stripped from pages as they are submitted for writeback.
      This change is only useful for filesystems which are using the mpage
      code.  That's ext2 and ext3-writeback and JFS.  An mpage patch for
      reiserfs was floating about but seems to have got lost.
      
      There is no need to strip buffers for reads because the mpage code does
      not attach buffers for reads.
      
      These are perhaps not the most appropriate buffer_heads to toss away.
      Perhaps something smarter should be done to detect file overwriting, or
      to toss the 'oldest' buffer_heads first.
      
      In refill_inactive(): if the number of buffer_heads is excessive then
      strip buffers from pages as they move onto the inactive list.  This
      change is useful for all filesystems.  This approach is good because
      pages which are being repeatedly overwritten will remain on the active
      list and will retain their buffers, whereas pages which are not being
      overwritten will be stripped.
      e182d612
    • Andrew Morton's avatar
      [PATCH] reduce the default dirty memory thresholds · ce92adf3
      Andrew Morton authored
      Writeback parameter tuning.  Somewhat experimental, but heading in the
      right direction, I hope.
      
      - Allowing 40% of physical memory to be dirtied on massive ia32 boxes
        is unreasonable.  It pins too many buffer_heads and contribues to
        page reclaim latency.
      
        The patch changes the initial value of
        /proc/sys/vm/dirty_background_ratio, dirty_async_ratio and (the
        presently non-functional) dirty_sync_ratio so that they are reduced
        when the highmem:lowmem ratio exceeds 4:1.
      
        These ratios are scaled so that as the highmem:lowmem ratio goes
        beyond 4:1, the maximum amount of allowed dirty memory ceases to
        increase.  It is clamped at the amount of memory which a 4:1 machine
        is allowed to use.
      
      - Aggressive reduction in the dirty memory threshold at which
        background writeback cuts in.  2.4 uses 30% of ZONE_NORMAL.  2.5 uses
        40% of total memory.  This patch changes it to 10% of total memory
        (if total memory <= 4G.  Even less otherwise - see above).
      
      This means that:
      
      - Much more writeback is performed by pdflush.
      
      - When the application is generating dirty data at a moderate
        rate, background writeback cuts in much earlier, so memory is
        cleaned more promptly.
      
      - Reduces the risk of user applications getting stalled by writeback.
      
      - Will damage dbench numbers.  It turns out that the damage is
        fairly small, and dbench isn't a worthwhile workload for
        optimisation.
      
      - Moderate reduction in the dirty level at which the write(2) caller
        is forced to perform writeback (throttling).  Was 40% of total
        memory.  Is now 30% of total memory (if total memory <= 4G, less
        otherwise).
      
      This is to reduce page reclaim latency, and generally because
      allowing processes to flood the machine with dirty data is a bad
      thing in mixed workloads.
      ce92adf3
    • Andrew Morton's avatar
      [PATCH] discontigmem code cleanup #2 · e2f5e334
      Andrew Morton authored
      Patch from Martin Bligh
      
      "This mainly just rips out some magic extra structures in the boot time
       code to determine node sizes, and counts in pages instead of bytes.
       Oh, and I put the code that allocates pgdat into allocage_pgdat,
       instead of find_max_pfn_node, which seems like an incongruous home for
       it.
      
       No functionality changes, nothing touched outside i386 discontigmem ...
       just makes code cleaner and more readable.  Tested on 16-way NUMA-Q."
      e2f5e334
    • Andrew Morton's avatar
      [PATCH] discontigmem code cleanup #1 · 79a96230
      Andrew Morton authored
      Patch from Martin Bligh.
      
      "This mainly changes the PLAT_MY_MACRO_IS_ALL_CAPS() stuff to be
       normal_macro(), and takes out some unnecessary redirection of function
       names.  No functionality changes, nothing touched outside i386
       discontigmem ...  just makes code readable.  Rumour has it that the
       PLAT_* stuff came from IRIX - I don't see that as a good reason to make
       the Linux code unreadable.  Tested on 16-way NUMA-Q."
      79a96230
    • Andrew Morton's avatar
      [PATCH] exact dirty state accounting · 1f90eedd
      Andrew Morton authored
      Some adjustments to global dirty page accounting.
      
      Previously, dirty page accounting counted all dirty pages.  Even dirty
      anonymous pages.  This has potential to upset the throttling logic in
      balance_dirty_pages().  Particularly as I suspect we should decrease
      the dirty memory writeback thresholds by a lot.
      
      So this patch changes it so that we only account for dirty pagecache
      pages which have backing store.  Not anonymous pages, not swapcache,
      not in-memory filesystem pages.
      
      To support this, the `memory_backed' boolean has been added to struct
      backing_dev_info.  When an address space's backing device is marked as
      memory-backed, the core kernel knows to not include that mapping's
      pages in the dirty memory accounting.
      
      For memory-backed mappings, dirtiness is a way of pinning the page, and
      there's nothing the kernel can to do clean the page to make it freeable.
      
      driverfs, tmpfs, and ranfs have been coverted to mark their mappings as
      memory-backed.
      
      The ramdisk driver hasn't been converted.  I have a separate patch for
      ramdisk, which fails to fix the longstanding problems in there :(
      
      With this patch, /bin/sync now sends /proc/meminfo:Dirty to zero, which
      is rather comforting.
      1f90eedd
    • Andrew Morton's avatar
      [PATCH] pass the correct flags to aops->releasepage() · 6a0fb424
      Andrew Morton authored
      Restore the gfp_mask in the VM's call to a_ops->releasepage().  We can
      block in there again, and XFS (at least) can use that.
      6a0fb424
    • Andrew Morton's avatar
      [PATCH] writer throttling fix · 95b88300
      Andrew Morton authored
      The patch fixes a few problems in the writer throttling code.  Mainly
      in the situation where a single large file is being written out.
      
      That file could be parked on sb->locked_inodes due to pdflush
      writeback, and the writer throttling path coming out of
      balance_dirty_pages() forgot to look for inodes on ->locked_inodes.
      
      The net effect was that the amount of dirty memory was exceeding the
      limit set in /proc/sys/vm/dirty_async_ratio, possibly to the point
      where the system gets seriously choked.
      
      The patch removes sb->locked_inodes altogether and teaches the
      throttling code to look for inodes on sb->s_io as well as sb->s_dirty.
      
      Also, just leave unwritten dirty pages on mapping->io_pages, and
      unwritten dirty inodes on sb->s_io.  Putting them back onto
      ->dirty_pages and ->dirty_inodes was fairly pointless, given that both
      lists need to be looked at.
      95b88300
    • Ingo Molnar's avatar
      [PATCH] Re: do_syslog/__down_trylock lockup in current BK · 0d8b3b44
      Ingo Molnar authored
      This fixes the lockup.
      
      The bug happened because reparenting in the CLONE_THREAD case was done in
      a fundamentally non-atomic way, which was asking for various races to
      happen: eg. the target parent gets reparented to the currently exiting
      thread ...
      
      (the non-CLONE_THREAD case is safe because nothing reparents init.)
      
      the solution is to make all of reparenting atomic (including the
      forget_original_parent() bit) - this is possible with some reorganization
      done in signal.c and exit.c. This also made some of the loops simpler.
      0d8b3b44