- 08 Nov, 2012 10 commits
-
-
Brian Foster authored
Create a new mount workqueue and delayed_work to enable background scanning and freeing of eofblocks inodes. The scanner kicks in once speculative preallocation occurs and stops requeueing itself when no eofblocks inodes exist. The scan interval is based on the new 'speculative_prealloc_lifetime' tunable (default to 5m). The background scanner performs unfiltered, best effort scans (which skips inodes under lock contention or with a dirty cache mapping). Signed-off-by:
Brian Foster <bfoster@redhat.com> Reviewed-by:
Mark Tinguely <tinguely@sgi.com> Reviewed-by:
Dave Chinner <dchinner@redhat.com> Signed-off-by:
Ben Myers <bpm@sgi.com>
-
Brian Foster authored
Support minimum file size filtering in the eofblocks scan. The caller must set the XFS_EOF_FLAGS_MINFILESIZE flags bit and minimum file size value in bytes. Signed-off-by:
-
Brian Foster authored
Enhance the eofblocks scan code to filter based on multiply specified inode id values. When multiple inode id values are specified, only inodes that match all id values are selected. Signed-off-by:
-
Brian Foster authored
Support inode ID filtering in the eofblocks scan. The caller must set the associated XFS_EOF_FLAGS_*ID bit and ID field. Signed-off-by:
-
Brian Foster authored
The XFS_IOC_FREE_EOFBLOCKS ioctl allows users to invoke an EOFBLOCKS scan. The xfs_eofblocks structure is defined to support the command parameters (scan mode). Signed-off-by:
-
Brian Foster authored
xfs_inodes_free_eofblocks() implements scanning functionality for EOFBLOCKS inodes. It uses the AG iterator to walk the tagged inodes and free post-EOF blocks via the xfs_inode_free_eofblocks() execute function. The scan can be invoked in best-effort mode or wait (force) mode. A best-effort scan (default) handles all inodes that do not have a dirty cache and we successfully acquire the io lock via trylock. In wait mode, we continue to cycle through an AG until all inodes are handled. Signed-off-by:
-
Brian Foster authored
Turn xfs_free_eofblocks() into a non-static function, return EAGAIN to indicate trylock failure and make sure this error is not propagated in xfs_release(). Signed-off-by:
-
Brian Foster authored
This check is used in multiple places to determine whether we should check for (and potentially free) post EOF blocks on an inode. Add a helper to consolidate the check. Note that when we remove an inode from the cache (xfs_inactive()), we are required to trim post-EOF blocks even if the inode is marked preallocated or append-only to maintain correct space accounting. The 'force' parameter to xfs_can_free_eofblocks() specifies whether we should ignore the prealloc/append-only status of the inode. Signed-off-by:
-
Brian Foster authored
Genericize xfs_inode_ag_walk() to support an optional radix tree tag and args argument for the execute function. Create a new wrapper called xfs_inode_ag_iterator_tag() that performs a tag based walk of perag's and inodes. Signed-off-by:
-
Brian Foster authored
Add the XFS_ICI_EOFBLOCKS_TAG inode tag to identify inodes with speculatively preallocated blocks beyond EOF. An inode is tagged when speculative preallocation occurs and untagged either via truncate down or when post-EOF blocks are freed via release or reclaim. The tag management is intentionally not aggressive to prefer simplicity over the complexity of handling all the corner cases under which post-EOF blocks could be freed (i.e., forward truncation, fallocate, write error conditions, etc.). This means that a tagged inode may or may not have post-EOF blocks after a period of time. The tag is eventually cleared when the inode is released or reclaimed. Signed-off-by:
-
- 07 Nov, 2012 5 commits
-
-
Eric Sandeen authored
When xfs gained the projid32bit feature, it was never added to the FSGEOMETRY ioctl feature flags, so it's not queryable without this patch. Signed-off-by:
Eric Sandeen <sandeen@redhat.com> Reviewed-by:
Carlos Maiolino <cmaiolino@redhat.com> Reviewed-by:
-
Dave Chinner authored
Commit 44396476 ("xfs: reset buffer pointers before freeing them") in 3.0-rc1 introduced a regression when recovering log buffers that wrapped around the end of log. The second part of the log buffer at the start of the physical log was being read into the header buffer rather than the data buffer, and hence recovery was seeing garbage in the data buffer when it got to the region of the log buffer that was incorrectly read. Cc: <stable@vger.kernel.org> # 3.0.x, 3.2.x, 3.4.x 3.6.x Reported-by:
Torsten Kaiser <just.for.lkml@googlemail.com> Signed-off-by:
Christoph Hellwig <hch@lst.de> Reviewed-by:
-
Dave Chinner authored
When we shut down the filesystem, we have to unpin and free all the buffers currently active in the CIL. To do this we unpin and remove them in one operation as a result of a failed iclogbuf write. For buffers, we do this removal via a simultated IO completion of after marking the buffer stale. At the time we do this, we have two references to the buffer - the active LRU reference and the buf log item. The LRU reference is removed by marking the buffer stale, and the active CIL reference is by the xfs_buf_iodone() callback that is run by xfs_buf_do_callbacks() during ioend processing (via the bp->b_iodone callback). However, ioend processing requires one more reference - that of the IO that it is completing. We don't have this reference, so we free the buffer prematurely and use it after it is freed. For buffers marked with XBF_ASYNC, this leads to assert failures in xfs_buf_rele() on debug kernels because the b_hold count is zero. Fix this by making sure we take the necessary IO reference before starting IO completion processing on the stale buffer, and set the XBF_ASYNC flag to ensure that IO completion processing removes all the active references from the buffer to ensure it is fully torn down. Cc: <stable@vger.kernel.org> Signed-off-by:
-
Dave Chinner authored
Inode buffers do not need to be mapped as inodes are read or written directly from/to the pages underlying the buffer. This fixes a regression introduced by commit 611c9946 ("xfs: make XBF_MAPPED the default behaviour"). Signed-off-by:
-
Dave Chinner authored
When we free a block from the alloc btree tree, we move it to the freelist held in the AGFL and mark it busy in the busy extent tree. This typically happens when we merge btree blocks. Once the transaction is committed and checkpointed, the block can remain on the free list for an indefinite amount of time. Now, this isn't the end of the world at this point - if the free list is shortened, the buffer is invalidated in the transaction that moves it back to free space. If the buffer is allocated as metadata from the free list, then all the modifications getted logged, and we have no issues, either. And if it gets allocated as userdata direct from the freelist, it gets invalidated and so will never get written. However, during the time it sits on the free list, pressure on the log can cause the AIL to be pushed and the buffer that covers the block gets pushed for write. IOWs, we end up writing a freed metadata block to disk. Again, this isn't the end of the world because we know from the above we are only writing to free space. The problem, however, is for validation callbacks. If the block was on old btree root block, then the level of the block is going to be higher than the current tree root, and so will fail validation. There may be other inconsistencies in the block as well, and currently we don't care because the block is in free space. Shutting down the filesystem because a freed block doesn't pass write validation, OTOH, is rather unfriendly. So, make sure we always invalidate buffers as they move from the free space trees to the free list so that we guarantee they never get written to disk while on the free list. Signed-off-by:
Phil White <pwhite@sgi.com> Reviewed-by:
-
- 02 Nov, 2012 4 commits
-
-
Carlos Maiolino authored
Once inode64 is the default allocation mode now, kernel documentation should be updated to match this behaviour. Signed-off-by:
-
Carlos Maiolino authored
I found some out of date comments while studying the inode allocation code, so I believe it's worth to have these comments updated. It basically rewrites the comment regarding to "call_again" variable, which is not used anymore, but instead, callers of xfs_ialloc() decides if it needs to be called again relying only if ialloc_context is NULL or not. Also did some small changes in another comment that I thought to be pertinent to the current behaviour of these functions and some alignment on both comments. Signed-off-by:
-
Dave Chinner authored
Uninitialised variable build warning introduced by 2903ff01 ("switch simple cases of fget_light to fdget"), gcc is not smart enough to work out that the variable is not used uninitialised, and the commit removed the initialisation at declaration that the old variable had. Signed-off-by:
-
Dave Chinner authored
When updating new secondary superblocks in a growfs operation, the superblock buffer is read from the newly grown region of the underlying device. This is not guaranteed to be zero, so violates the underlying assumption that the unused parts of superblocks are zero filled. Get a new buffer for these secondary superblocks to ensure that the unused regions are zero filled correctly. Signed-off-by:
-
- 18 Oct, 2012 3 commits
-
-
Dave Chinner authored
Switching stacks are xfs_alloc_vextent can cause deadlocks when we run out of worker threads on the allocation workqueue. This can occur because xfs_bmap_btalloc can make multiple calls to xfs_alloc_vextent() and even if xfs_alloc_vextent() fails it can return with the AGF locked in the current allocation transaction. If we then need to make another allocation, and all the allocation worker contexts are exhausted because the are blocked waiting for the AGF lock, holder of the AGF cannot get it's xfs-alloc_vextent work completed to release the AGF. Hence allocation effectively deadlocks. To avoid this, move the stack switch one layer up to xfs_bmapi_allocate() so that all of the allocation attempts in a single switched stack transaction occur in a single worker context. This avoids the problem of an allocation being blocked waiting for a worker thread whilst holding the AGF. Signed-off-by:
-
Dave Chinner authored
Certain allocation paths through xfs_bmapi_write() are in situations where we have limited stack available. These are almost always in the buffered IO writeback path when convertion delayed allocation extents to real extents. The current stack switch occurs for userdata allocations, which means we also do stack switches for preallocation, direct IO and unwritten extent conversion, even those these call chains have never been implicated in a stack overrun. Hence, let's target just the single stack overun offended for stack switches. To do that, introduce a XFS_BMAPI_STACK_SWITCH flag that the caller can pass xfs_bmapi_write() to indicate it should switch stacks if it needs to do allocation. Signed-off-by:
-
Mark Tinguely authored
Zero the kernel stack space that makes up the xfs_alloc_arg structures. Signed-off-by:
-
- 17 Oct, 2012 14 commits
-
-
Dave Chinner authored
The log write code stamps each iclog with the current tail LSN in the iclog header so that recovery knows where to find the tail of thelog once it has found the head. Normally this is taken from the first item on the AIL - the log item that corresponds to the oldest active item in the log. The problem is that when the AIL is empty, the tail lsn is dervied from the the l_last_sync_lsn, which is the LSN of the last iclog to be written to the log. In most cases this doesn't happen, because the AIL is rarely empty on an active filesystem. However, when it does, it opens up an interesting case when the transaction being committed to the iclog spans multiple iclogs. That is, the first iclog is stamped with the l_last_sync_lsn, and IO is issued. Then the next iclog is setup, the changes copied into the iclog (takes some time), and then the l_last_sync_lsn is stamped into the header and IO is issued. This is still the same transaction, so the tail lsn of both iclogs must be the same for log recovery to find the entire transaction to be able to replay it. The problem arises in that the iclog buffer IO completion updates the l_last_sync_lsn with it's own LSN. Therefore, If the first iclog completes it's IO before the second iclog is filled and has the tail lsn stamped in it, it will stamp the LSN of the first iclog into it's tail lsn field. If the system fails at this point, log recovery will not see a complete transaction, so the transaction will no be replayed. The fix is simple - the l_last_sync_lsn is updated when a iclog buffer IO completes, and this is incorrect. The l_last_sync_lsn shoul dbe updated when a transaction is completed by a iclog buffer IO. That is, only iclog buffers that have transaction commit callbacks attached to them should update the l_last_sync_lsn. This means that the last_sync_lsn will only move forward when a commit record it written, not in the middle of a large transaction that is rolling through multiple iclog buffers. Signed-off-by:
-
Dave Chinner authored
The inode cache functions remaining in xfs_iget.c can be moved to xfs_icache.c along with the other inode cache functions. This removes all functionality from xfs_iget.c, so the file can simply be removed. This move results in various functions now only having the scope of a single file (e.g. xfs_inode_free()), so clean up all the definitions and exported prototypes in xfs_icache.[ch] and xfs_inode.h appropriately. Signed-off-by:
-
Dave Chinner authored
xfs_ilock() and friends really aren't related to the inode cache in any way, so move them to xfs_inode.c with all the other inode related functionality. While doing this move, move the xfs_ilock() tracepoints to *before* the lock is taken so that when a hang on a lock occurs we have events to indicate which process and what inode we were trying to lock when the hang occurred. This is much better than the current silence we get on a hang... Signed-off-by:
-
Dave Chinner authored
xfs_sync.c now only contains inode reclaim functions and inode cache iteration functions. It is not related to sync operations anymore. Rename to xfs_icache.c to reflect it's contents and prepare for consolidation with the other inode cache file that exists (xfs_iget.c). Signed-off-by:
-
Dave Chinner authored
xfs_quiesce_attr() is supposed to leave the log empty with an unmount record written. Right now it does not wait for the AIL to be emptied before writing the unmount record, not does it wait for metadata IO completion, either. Fix it to use the same method and code as xfs_log_unmount(). Signed-off-by:
-
Dave Chinner authored
Both callers of xfs_quiesce_attr() are in xfs_super.c, and there's nothing really sync-specific about this functionality so it doesn't really matter where it lives. Move it to benext to it's callers, so all the remount/sync_fs code is in the one place. Signed-off-by:
-
Dave Chinner authored
Why do we need to write the superblock to disk once we've written all the data? We don't actually - the reasons for doing this are lost in the mists of time, and go back to the way Irix used to drive VFS flushing. On linux, this code is only called from two contexts: remount and .sync_fs. In the remount case, the call is followed by a metadata sync, which unpins and writes the superblock. In the sync_fs case, we only need to force the log to disk to ensure that the superblock is correctly on disk, so we don't actually need to write it. Hence the functionality is either redundant or superfluous and thus can be removed. Seeing as xfs_quiesce_data is essentially now just a log force, remove it as well and fold the code back into the two callers. Neither of them need the log covering check, either, as that is redundant for the remount case, and unnecessary for the .sync_fs case. Signed-off-by:
-
Dave Chinner authored
With the syncd functions moved to the log and/or removed, the syncd workqueue is the only remaining bit left. It is used by the log covering/ail pushing work, as well as by the inode reclaim work. Given how cheap workqueues are these days, give the log and inode reclaim work their own work queues and kill the syncd work queue. Signed-off-by:
-
Dave Chinner authored
We don't do any data writeback from XFS any more - the VFS is completely responsible for that, including for freeze. We can replace the remaining caller with a VFS level function that achieves the same thing, but without conflicting with current writeback work. This means we can remove the flush_work and xfs_flush_inodes() - the VFS functionality completely replaces the internal flush queue for doing this writeback work in a separate context to avoid stack overruns. This does have one complication - it cannot be called with page locks held. Hence move the flushing of delalloc space when ENOSPC occurs back up into xfs_file_aio_buffered_write when we don't hold any locks that will stall writeback. Unfortunately, writeback_inodes_sb_if_idle() is not sufficient to trigger delalloc conversion fast enough to prevent spurious ENOSPC whent here are hundreds of writers, thousands of small files and GBs of free RAM. Hence we need to use sync_sb_inodes() to block callers while we wait for writeback like the previous xfs_flush_inodes implementation did. That means we have to hold the s_umount lock here, but because this call can nest inside i_mutex (the parent directory in the create case, held by the VFS), we have to use down_read_trylock() to avoid potential deadlocks. In practice, this trylock will succeed on almost every attempt as unmount/remount type operations are exceedingly rare. Note: we always need to pass a count of zero to generic_file_buffered_write() as the previously written byte count. We only do this by accident before this patch by the virtue of ret always being zero when there are no errors. Make this explicit rather than needing to specifically zero ret in the ENOSPC retry case. Signed-off-by:
-
Dave Chinner authored
When unmounting the filesystem, there are lots of operations that need to be done in a specific order, and they are spread across across a couple of functions. We have to drain the AIL before we write the unmount record, and we have to shut down the background log work before we do either of them. But this is all split haphazardly across xfs_unmountfs() and xfs_log_unmount(). Move all the AIL flushing and log manipulations to xfs_log_unmount() so that the responisbilities of each function is clear and the operations they perform obvious. Signed-off-by:
-
Dave Chinner authored
The only thing the periodic sync work does now is flush the AIL and idle the log. These are really functions of the log code, so move the work to xfs_log.c and rename it appropriately. The only wart that this leaves behind is the xfssyncd_centisecs sysctl, otherwise the xfssyncd is dead. Clean up any comments that related to xfssyncd to reflect it's passing. Signed-off-by:
-
Dave Chinner authored
If the filesystem is mounted or remounted read-only, stop the sync worker that tries to flush or cover the log if the filesystem is dirty. It's read-only, so it isn't dirty. Restart it on a remount,rw as necessary. This avoids the need for RO checks in the work. Similarly, stop the sync work when the filesystem is frozen, and start it again when the filesysetm is thawed. This avoids the need for special freeze checks in the work. Signed-off-by:
-
Dave Chinner authored
Instead of starting and stopping background work on the xfs_mount_wq all at the same time, separate them to where they really are needed to start and stop. The xfs_sync_worker, only needs to be started after all the mount processing has completed successfully, while it needs to be stopped before the log is unmounted. The xfs_reclaim_worker is started on demand, and can be stopped before the unmount process does it's own inode reclaim pass. The xfs_flush_inodes work is run on demand, and so we really only need to ensure that it has stopped running before we start processing an unmount, freeze or remount,ro. Signed-off-by:
-
Dave Chinner authored
xfs_syncd_start and xfs_syncd_stop tie a bunch of unrelated functionailty together that actually have different start and stop requirements. Kill these functions and open code the start/stop methods for each of the background functions. Subsequent patches will move the start/stop functions around to the correct places to avoid races and shutdown issues. Signed-off-by:
-
- 14 Oct, 2012 4 commits
-
-
Linus Torvalds authored
-
git://git.linux-mips.org/pub/scm/ralf/upstream-linusLinus Torvalds authored
Pull MIPS update from Ralf Baechle: "Cleanups and fixes for breakage that occured earlier during this merge phase. Also a few patches that didn't make the first pull request. Of those is the Alchemy work that merges code for many of the SOCs and evaluation boards thus among other code shrinkage, reduces the number of MIPS defconfigs by 5." * 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus: (22 commits) MIPS: SNI: Switch RM400 serial to SCCNXP driver MIPS: Remove unused empty_bad_pmd_table[] declaration. MIPS: MT: Remove kspd. MIPS: Malta: Fix section mismatch. MIPS: asm-offset.c: Delete unused irq_cpustat_t struct offsets. MIPS: Alchemy: Merge PB1100/1500 support into DB1000 code. MIPS: Alchemy: merge PB1550 support into DB1550 code MIPS: Alchemy: Single kernel for DB1200/1300/1550 MIPS: Optimize TLB refill for RI/XI configurations. MIPS: proc: Cleanup printing of ASEs. MIPS: Hardwire detection of DSP ASE Rev 2 for systems, as required. MIPS: Add detection of DSP ASE Revision 2. MIPS: Optimize pgd_init and pmd_init MIPS: perf: Add perf functionality for BMIPS5000 MIPS: perf: Split the Kconfig option CONFIG_MIPS_MT_SMP MIPS: perf: Remove unnecessary #ifdef MIPS: perf: Add cpu feature bit for PCI (performance counter interrupt) MIPS: perf: Change the "mips_perf_event" table unsupported indicator. MIPS: Align swapper_pg_dir to 64K for better TLB Refill code. vmlinux.lds.h: Allow architectures to add sections to the front of .bss ...
-
git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linuxLinus Torvalds authored
Pull module signing support from Rusty Russell: "module signing is the highlight, but it's an all-over David Howells frenzy..." Hmm "Magrathea: Glacier signing key". Somebody has been reading too much HHGTTG. * 'modules-next' of git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux: (37 commits) X.509: Fix indefinite length element skip error handling X.509: Convert some printk calls to pr_devel asymmetric keys: fix printk format warning MODSIGN: Fix 32-bit overflow in X.509 certificate validity date checking MODSIGN: Make mrproper should remove generated files. MODSIGN: Use utf8 strings in signer's name in autogenerated X.509 certs MODSIGN: Use the same digest for the autogen key sig as for the module sig MODSIGN: Sign modules during the build process MODSIGN: Provide a script for generating a key ID from an X.509 cert MODSIGN: Implement module signature checking MODSIGN: Provide module signing public keys to the kernel MODSIGN: Automatically generate module signing keys if missing MODSIGN: Provide Kconfig options MODSIGN: Provide gitignore and make clean rules for extra files MODSIGN: Add FIPS policy module: signature checking hook X.509: Add a crypto key parser for binary (DER) X.509 certificates MPILIB: Provide a function to read raw data into an MPI X.509: Add an ASN.1 decoder X.509: Add simple ASN.1 grammar compiler ...
-
Matt Fleming authored
The hostprogs need access to the CONFIG_* symbols found in include/generated/autoconf.h. But commit abbf1590 ("UAPI: Partition the header include path sets and add uapi/ header directories") replaced $(LINUXINCLUDE) with $(USERINCLUDE) which doesn't contain the necessary include paths. This has the undesirable effect of breaking the EFI boot stub because the #ifdef CONFIG_EFI_STUB code in arch/x86/boot/tools/build.c is never compiled. It should also be noted that because $(USERINCLUDE) isn't exported by the top-level Makefile it's actually empty in arch/x86/boot/Makefile. Cc: H. Peter Anvin <hpa@zytor.com> Cc: Ingo Molnar <mingo@kernel.org> Acked-by:
David Howells <dhowells@redhat.com> Signed-off-by:
Matt Fleming <matt.fleming@intel.com> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
