- 21 Dec, 2012 25 commits
-
-
Mikulas Patocka authored
Introduce a field per_bio_data_size in struct dm_target. Targets can set this field in the constructor. If a target sets this field to a non-zero value, "per_bio_data_size" bytes of auxiliary data are allocated for each bio submitted to the target. These data can be used for any purpose by the target and help us improve performance by removing some per-target mempools. Per-bio data is accessed with dm_per_bio_data. The argument data_size must be the same as the value per_bio_data_size in dm_target. If the target has a pointer to per_bio_data, it can get a pointer to the bio with dm_bio_from_per_bio_data() function (data_size must be the same as the value passed to dm_per_bio_data). Signed-off-by:
Mikulas Patocka <mpatocka@redhat.com> Signed-off-by:
Alasdair G Kergon <agk@redhat.com>
-
Mike Snitzer authored
Add WRITE SAME support to dm-io and make it accessible to dm_kcopyd_zero(). dm_kcopyd_zero() provides an asynchronous interface whereas the blkdev_issue_write_same() interface is synchronous. WRITE SAME is a SCSI command that can be leveraged for more efficient zeroing of a specified logical extent of a device which supports it. Only a single zeroed logical block is transfered to the target for each WRITE SAME and the target then writes that same block across the specified extent. The dm thin target uses this. Signed-off-by:
Mike Snitzer <snitzer@redhat.com> Signed-off-by:
-
Mike Snitzer authored
The linear target can already support WRITE SAME requests so signal this by setting num_write_same_requests to 1. Signed-off-by:
-
Mike Snitzer authored
WRITE SAME bios have a payload that contain a single page. When cloning WRITE SAME bios DM has no need to modify the bi_io_vec attributes (and doing so would be detrimental). DM need only alter the start and end of the WRITE SAME bio accordingly. Rather than duplicate __clone_and_map_discard, factor out a common function that is also used by __clone_and_map_write_same. Signed-off-by:
-
Mike Snitzer authored
Allow targets to opt in to WRITE SAME support by setting 'num_write_same_requests' in the dm_target structure. A dm device will only advertise WRITE SAME support if all its targets and all its underlying devices support it. Signed-off-by:
-
Mikulas Patocka authored
If the parameter buffer is small enough, try to allocate it with kmalloc() rather than vmalloc(). vmalloc is noticeably slower than kmalloc because it has to manipulate page tables. In my tests, on PA-RISC this patch speeds up activation 13 times. On Opteron this patch speeds up activation by 5%. This patch introduces a new function free_params() to free the parameters and this uses new flags that record whether or not vmalloc() was used and whether or not the input buffer must be wiped after use. Signed-off-by:
-
Mikulas Patocka authored
When allocating memory for the userspace ioctl data, set some appropriate GPF flags directly instead of using PF_MEMALLOC. Signed-off-by:
-
Joe Thornber authored
Improve space map error message when unable to allocate a new metadata block. Signed-off-by:
Joe Thornber <ejt@redhat.com> Signed-off-by:
-
Mike Snitzer authored
Throttle all errors logged from the IO path by dm thin. Signed-off-by:
-
Mike Snitzer authored
Nearly all of persistent-data is in the IO path so throttle error messages with DMERR_LIMIT to limit the amount logged when something has gone wrong. Signed-off-by:
-
Mike Snitzer authored
Reinstate a useful error message when the block manager buffer validator fails. This was mistakenly eliminated when the block manager was converted to use dm-bufio. Signed-off-by:
-
Jonathan Brassow authored
If the user does not supply a bitmap region_size to the dm raid target, a reasonable size is computed automatically. If this is not a power of 2, the md code will report an error later. This patch catches the problem early and rounds the region_size to the next power of two. Signed-off-by:
Jonathan Brassow <jbrassow@redhat.com> Signed-off-by:
-
Joe Thornber authored
Remove unused @data_block parameter from cell_defer. Change thin_bio_map to use many returns rather than setting a variable. Signed-off-by:
-
Joe Thornber authored
Rename cell_defer_except() to cell_defer_no_holder() which describes its function more clearly. Signed-off-by:
-
Mikulas Patocka authored
track_chunk is always called with interrupts enabled. Consequently, we do not need to save and restore interrupt state in "flags" variable. This patch changes spin_lock_irqsave to spin_lock_irq and spin_unlock_irqrestore to spin_unlock_irq. Signed-off-by:
-
Mikulas Patocka authored
Use a defined macro DM_ENDIO_INCOMPLETE instead of a numeric constant. Signed-off-by:
-
Mikulas Patocka authored
mempool_alloc can't fail if __GFP_WAIT is specified, so the condition that tests if read_record is non-NULL is always true. Signed-off-by:
-
Mike Snitzer authored
If "ignore_discard" is specified when creating the thin pool device then discard support is disabled for that device. The pool device's status should reflect this fact rather than stating "no_discard_passdown" (which implies discards are enabled but passdown is disabled). Reported-by:
Zdenek Kabelac <zkabelac@redhat.com> Signed-off-by:
-
Joe Thornber authored
When deleting nested btrees, the code forgets to delete the innermost btree. The thin-metadata code serendipitously compensates for this by claiming there is one extra layer in the tree. This patch corrects both problems. Signed-off-by:
-
Joe Thornber authored
When discards are prepared it is best to directly wake the worker that will process them. The worker will be woken anyway, via periodic commit, but there is no reason to not wake_worker here. Signed-off-by:
-
Joe Thornber authored
There is a race when discard bios and non-discard bios are issued simultaneously to the same block. Discard support is expensive for all thin devices precisely because you have to be careful to quiesce the area you're discarding. DM thin must handle this conflicting IO pattern (simultaneous non-discard vs discard) even though a sane application shouldn't be issuing such IO. The race manifests as follows: 1. A non-discard bio is mapped in thin_bio_map. This doesn't lock out parallel activity to the same block. 2. A discard bio is issued to the same block as the non-discard bio. 3. The discard bio is locked in a dm_bio_prison_cell in process_discard to lock out parallel activity against the same block. 4. The non-discard bio's mapping continues and its all_io_entry is incremented so the bio is accounted for in the thin pool's all_io_ds which is a dm_deferred_set used to track time locality of non-discard IO. 5. The non-discard bio is finally locked in a dm_bio_prison_cell in process_bio. The race can result in deadlock, leaving the block layer hanging waiting for completion of a discard bio that never completes, e.g.: INFO: task ruby:15354 blocked for more than 120 seconds. "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. ruby D ffffffff8160f0e0 0 15354 15314 0x00000000 ffff8802fb08bc58 0000000000000082 ffff8802fb08bfd8 0000000000012900 ffff8802fb08a010 0000000000012900 0000000000012900 0000000000012900 ffff8802fb08bfd8 0000000000012900 ffff8803324b9480 ffff88032c6f14c0 Call Trace: [<ffffffff814e5a19>] schedule+0x29/0x70 [<ffffffff814e3d85>] schedule_timeout+0x195/0x220 [<ffffffffa06b9bc1>] ? _dm_request+0x111/0x160 [dm_mod] [<ffffffff814e589e>] wait_for_common+0x11e/0x190 [<ffffffff8107a170>] ? try_to_wake_up+0x2b0/0x2b0 [<ffffffff814e59ed>] wait_for_completion+0x1d/0x20 [<ffffffff81233289>] blkdev_issue_discard+0x219/0x260 [<ffffffff81233e79>] blkdev_ioctl+0x6e9/0x7b0 [<ffffffff8119a65c>] block_ioctl+0x3c/0x40 [<ffffffff8117539c>] do_vfs_ioctl+0x8c/0x340 [<ffffffff8119a547>] ? block_llseek+0x67/0xb0 [<ffffffff811756f1>] sys_ioctl+0xa1/0xb0 [<ffffffff810561f6>] ? sys_rt_sigprocmask+0x86/0xd0 [<ffffffff814ef099>] system_call_fastpath+0x16/0x1b The thinp-test-suite's test_discard_random_sectors reliably hits this deadlock on fast SSD storage. The fix for this race is that the all_io_entry for a bio must be incremented whilst the dm_bio_prison_cell is held for the bio's associated virtual and physical blocks. That cell locking wasn't occurring early enough in thin_bio_map. This patch fixes this. Care is taken to always call the new function inc_all_io_entry() with the relevant cells locked, but they are generally unlocked before calling issue() to try to avoid holding the cells locked across generic_submit_request. Also, now that thin_bio_map may lock bios in a cell, process_bio() is no longer the only thread that will do so. Because of this we must be sure to use cell_defer_except() to release all non-holder entries, that were added by the other thread, because they must be deferred. This patch depends on "dm thin: replace dm_cell_release_singleton with cell_defer_except". Signed-off-by:
-
Joe Thornber authored
Change existing users of the function dm_cell_release_singleton to share cell_defer_except instead, and then remove the now-unused function. Everywhere that calls dm_cell_release_singleton, the bio in question is the holder of the cell. If there are no non-holder entries in the cell then cell_defer_except behaves exactly like dm_cell_release_singleton. Conversely, if there *are* non-holder entries then dm_cell_release_singleton must not be used because those entries would need to be deferred. Consequently, it is safe to replace use of dm_cell_release_singleton with cell_defer_except. This patch is a pre-requisite for "dm thin: fix race between simultaneous io and discards to same block". Signed-off-by:
-
Mike Snitzer authored
WRITE SAME bios are not yet handled correctly by device-mapper so disable their use on device-mapper devices by setting max_write_same_sectors to zero. As an example, a ciphertext device is incompatible because the data gets changed according to the location at which it written and so the dm crypt target cannot support it. Signed-off-by:
-
Alasdair G Kergon authored
Abort dm ioctl processing if userspace changes the data_size parameter after we validated it but before we finished copying the data buffer from userspace. The dm ioctl parameters are processed in the following sequence: 1. ctl_ioctl() calls copy_params(); 2. copy_params() makes a first copy of the fixed-sized portion of the userspace parameters into the local variable "tmp"; 3. copy_params() then validates tmp.data_size and allocates a new structure big enough to hold the complete data and copies the whole userspace buffer there; 4. ctl_ioctl() reads userspace data the second time and copies the whole buffer into the pointer "param"; 5. ctl_ioctl() reads param->data_size without any validation and stores it in the variable "input_param_size"; 6. "input_param_size" is further used as the authoritative size of the kernel buffer. The problem is that userspace code could change the contents of user memory between steps 2 and 4. In particular, the data_size parameter can be changed to an invalid value after the kernel has validated it. This lets userspace force the kernel to access invalid kernel memory. The fix is to ensure that the size has not changed at step 4. This patch shouldn't have a security impact because CAP_SYS_ADMIN is required to run this code, but it should be fixed anyway. Reported-by: -
Mikulas Patocka authored
This patch fixes a compilation failure on sparc32 by renaming struct node. struct node is already defined in include/linux/node.h. On sparc32, it happens to be included through other dependencies and persistent-data doesn't compile because of conflicting declarations. Signed-off-by:
-
- 11 Dec, 2012 3 commits
-
-
Linus Torvalds authored
-
Florian Fainelli authored
The matrix-keymap module is currently lacking a proper module license, add one so we don't have this module tainting the entire kernel. This issue has been present since commit 1932811f ("Input: matrix-keymap - uninline and prepare for device tree support") Signed-off-by:
Florian Fainelli <florian@openwrt.org> CC: stable@vger.kernel.org # v3.5+ Signed-off-by:
Dmitry Torokhov <dmitry.torokhov@gmail.com> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
git://git.kernel.org/pub/scm/linux/kernel/git/davem/netLinus Torvalds authored
Pull networking fixes from David Miller: 1) Netlink socket dumping had several missing verifications and checks. In particular, address comparisons in the request byte code interpreter could access past the end of the address in the inet_request_sock. Also, address family and address prefix lengths were not validated properly at all. This means arbitrary applications can read past the end of certain kernel data structures. Fixes from Neal Cardwell. 2) ip_check_defrag() operates in contexts where we're in the process of, or about to, input the packet into the real protocols (specifically macvlan and AF_PACKET snooping). Unfortunately, it does a pskb_may_pull() which can modify the backing packet data which is not legal if the SKB is shared. It very much can be shared in this context. Deal with the possibility that the SKB is segmented by using skb_copy_bits(). Fix from Johannes Berg based upon a report by Eric Leblond. * git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: ipv4: ip_check_defrag must not modify skb before unsharing inet_diag: validate port comparison byte code to prevent unsafe reads inet_diag: avoid unsafe and nonsensical prefix matches in inet_diag_bc_run() inet_diag: validate byte code to prevent oops in inet_diag_bc_run() inet_diag: fix oops for IPv4 AF_INET6 TCP SYN-RECV state
-
- 10 Dec, 2012 4 commits
-
-
Linus Torvalds authored
This reverts commits a5091539 and d7c3b937. This is a revert of a revert of a revert. In addition, it reverts the even older i915 change to stop using the __GFP_NO_KSWAPD flag due to the original commits in linux-next. It turns out that the original patch really was bogus, and that the original revert was the correct thing to do after all. We thought we had fixed the problem, and then reverted the revert, but the problem really is fundamental: waking up kswapd simply isn't the right thing to do, and direct reclaim sometimes simply _is_ the right thing to do. When certain allocations fail, we simply should try some direct reclaim, and if that fails, fail the allocation. That's the right thing to do for THP allocations, which can easily fail, and the GPU allocations want to do that too. So starting kswapd is sometimes simply wrong, and removing the flag that said "don't start kswapd" was a mistake. Let's hope we never revisit this mistake again - and certainly not this many times ;) Acked-by:
Mel Gorman <mgorman@suse.de> Acked-by:
Johannes Weiner <hannes@cmpxchg.org> Cc: Rik van Riel <riel@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
-
Johannes Berg authored
ip_check_defrag() might be called from af_packet within the RX path where shared SKBs are used, so it must not modify the input SKB before it has unshared it for defragmentation. Use skb_copy_bits() to get the IP header and only pull in everything later. The same is true for the other caller in macvlan as it is called from dev->rx_handler which can also get a shared SKB. Reported-by:
Eric Leblond <eric@regit.org> Cc: stable@vger.kernel.org Signed-off-by:
Johannes Berg <johannes.berg@intel.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Linus Torvalds authored
This reverts commit 782fd304. We are going to reinstate the __GFP_NO_KSWAPD flag that has been removed, the removal reverted, and then removed again. Making this commit a pointless fixup for a problem that was caused by the removal of __GFP_NO_KSWAPD flag. The thing is, we really don't want to wake up kswapd for THP allocations (because they fail quite commonly under any kind of memory pressure, including when there is tons of memory free), and these patches were just trying to fix up the underlying bug: the original removal of __GFP_NO_KSWAPD in commit c6543459 ("mm: remove __GFP_NO_KSWAPD") was simply bogus. Signed-off-by:
-
Neal Cardwell authored
Add logic to verify that a port comparison byte code operation actually has the second inet_diag_bc_op from which we read the port for such operations. Previously the code blindly referenced op[1] without first checking whether a second inet_diag_bc_op struct could fit there. So a malicious user could make the kernel read 4 bytes beyond the end of the bytecode array by claiming to have a whole port comparison byte code (2 inet_diag_bc_op structs) when in fact the bytecode was not long enough to hold both. Signed-off-by:
Neal Cardwell <ncardwell@google.com> Signed-off-by:
-
- 09 Dec, 2012 3 commits
-
-
Neal Cardwell authored
Add logic to check the address family of the user-supplied conditional and the address family of the connection entry. We now do not do prefix matching of addresses from different address families (AF_INET vs AF_INET6), except for the previously existing support for having an IPv4 prefix match an IPv4-mapped IPv6 address (which this commit maintains as-is). This change is needed for two reasons: (1) The addresses are different lengths, so comparing a 128-bit IPv6 prefix match condition to a 32-bit IPv4 connection address can cause us to unwittingly walk off the end of the IPv4 address and read garbage or oops. (2) The IPv4 and IPv6 address spaces are semantically distinct, so a simple bit-wise comparison of the prefixes is not meaningful, and would lead to bogus results (except for the IPv4-mapped IPv6 case, which this commit maintains). Signed-off-by:
-
Neal Cardwell authored
Add logic to validate INET_DIAG_BC_S_COND and INET_DIAG_BC_D_COND operations. Previously we did not validate the inet_diag_hostcond, address family, address length, and prefix length. So a malicious user could make the kernel read beyond the end of the bytecode array by claiming to have a whole inet_diag_hostcond when the bytecode was not long enough to contain a whole inet_diag_hostcond of the given address family. Or they could make the kernel read up to about 27 bytes beyond the end of a connection address by passing a prefix length that exceeded the length of addresses of the given family. Signed-off-by:
-
Neal Cardwell authored
Fix inet_diag to be aware of the fact that AF_INET6 TCP connections instantiated for IPv4 traffic and in the SYN-RECV state were actually created with inet_reqsk_alloc(), instead of inet6_reqsk_alloc(). This means that for such connections inet6_rsk(req) returns a pointer to a random spot in memory up to roughly 64KB beyond the end of the request_sock. With this bug, for a server using AF_INET6 TCP sockets and serving IPv4 traffic, an inet_diag user like `ss state SYN-RECV` would lead to inet_diag_fill_req() causing an oops or the export to user space of 16 bytes of kernel memory as a garbage IPv6 address, depending on where the garbage inet6_rsk(req) pointed. Signed-off-by:
-
- 08 Dec, 2012 3 commits
-
-
Johannes Weiner authored
commit c702418f ("mm: vmscan: do not keep kswapd looping forever due to individual uncompactable zones") removed zone watermark checks from the compaction code in kswapd but left in the zone congestion clearing, which now happens unconditionally on higher order reclaim. This messes up the reclaim throttling logic for zones with dirty/writeback pages, where zones should only lose their congestion status when their watermarks have been restored. Remove the clearing from the zone compaction section entirely. The preliminary zone check and the reclaim loop in kswapd will clear it if the zone is considered balanced. Signed-off-by:
Rik van Riel <riel@redhat.com> Signed-off-by:
-
Linus Torvalds authored
The direct-IO write path already had the i_size checks in mm/filemap.c, but it turns out the read path did not, and removing the block size checks in fs/block_dev.c (commit bbec0270: "blkdev_max_block: make private to fs/buffer.c") removed the magic "shrink IO to past the end of the device" code there. Fix it by truncating the IO to the size of the block device, like the write path already does. NOTE! I suspect the write path would be *much* better off doing it this way in fs/block_dev.c, rather than hidden deep in mm/filemap.c. The mm/filemap.c code is extremely hard to follow, and has various conditionals on the target being a block device (ie the flag passed in to 'generic_write_checks()', along with a conditional update of the inode timestamp etc). It is also quite possible that we should treat this whole block device size as a "s_maxbytes" issue, and try to make the logic even more generic. However, in the meantime this is the fairly minimal targeted fix. Noted by Milan Broz thanks to a regression test for the cryptsetup reencrypt tool. Reported-and-tested-by:
Milan Broz <mbroz@redhat.com> Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/davem/netLinus Torvalds authored
Pull networking fixes from David Miller: "Two stragglers: 1) The new code that adds new flushing semantics to GRO can cause SKB pointer list corruption, manage the lists differently to avoid the OOPS. Fix from Eric Dumazet. 2) When TCP fast open does a retransmit of data in a SYN-ACK or similar, we update retransmit state that we shouldn't triggering a WARN_ON later. Fix from Yuchung Cheng." * git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: net: gro: fix possible panic in skb_gro_receive() tcp: bug fix Fast Open client retransmission
-
- 07 Dec, 2012 2 commits
-
-
Eric Dumazet authored
commit 2e71a6f8 (net: gro: selective flush of packets) added a bug for skbs using frag_list. This part of the GRO stack is rarely used, as it needs skb not using a page fragment for their skb->head. Most drivers do use a page fragment, but some of them use GFP_KERNEL allocations for the initial fill of their RX ring buffer. napi_gro_flush() overwrite skb->prev that was used for these skb to point to the last skb in frag_list. Fix this using a separate field in struct napi_gro_cb to point to the last fragment. Signed-off-by:
Eric Dumazet <edumazet@google.com> Signed-off-by:
-
Yuchung Cheng authored
If SYN-ACK partially acks SYN-data, the client retransmits the remaining data by tcp_retransmit_skb(). This increments lost recovery state variables like tp->retrans_out in Open state. If loss recovery happens before the retransmission is acked, it triggers the WARN_ON check in tcp_fastretrans_alert(). For example: the client sends SYN-data, gets SYN-ACK acking only ISN, retransmits data, sends another 4 data packets and get 3 dupacks. Since the retransmission is not caused by network drop it should not update the recovery state variables. Further the server may return a smaller MSS than the cached MSS used for SYN-data, so the retranmission needs a loop. Otherwise some data will not be retransmitted until timeout or other loss recovery events. Signed-off-by:
Yuchung Cheng <ycheng@google.com> Acked-by:
-
