- 11 Apr, 2014 14 commits
-
-
Roland Dreier authored
-
Steve Wise authored
Need to get the endpoint reference before calling rdma_fini(), which might fail causing us to not get the reference. Signed-off-by:
Steve Wise <swise@opengridcomputing.com> Signed-off-by:
Roland Dreier <roland@purestorage.com>
-
Steve Wise authored
Signed-off-by:
-
Steve Wise authored
Signed-off-by:
-
Steve Wise authored
Signed-off-by:
-
Hariprasad Shenai authored
Signed-off-by:
Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by:
-
Steve Wise authored
The max depth of a fastreg mr depends on whether the device supports DSGL or not. So compute it dynamically based on the device support and the module use_dsgl option. Signed-off-by:
-
Steve Wise authored
There is a race when moving a QP from RTS->CLOSING where a SQ work request could be posted after the FW receives the RDMA_RI/FINI WR. The SQ work request will never get processed, and should be completed with FLUSHED status. Function c4iw_flush_sq(), however was dropping the oldest SQ work request when in CLOSING or IDLE states, instead of completing the pending work request. If that oldest pending work request was actually complete and has a CQE in the CQ, then when that CQE is proceessed in poll_cq, we'll BUG_ON() due to the inconsistent SQ/CQ state. This is a very small timing hole and has only been hit once so far. The fix is two-fold: 1) c4iw_flush_sq() MUST always flush all non-completed WRs with FLUSHED status regardless of the QP state. 2) In c4iw_modify_rc_qp(), always set the "in error" bit on the queue before moving the state out of RTS. This ensures that the state transition will not happen while another thread is in post_rc_send(), because set_state() and post_rc_send() both aquire the qp spinlock. Also, once we transition the state out of RTS, subsequent calls to post_rc_send() will fail because the "in error" bit is set. I don't think this fully closes the race where the FW can get a FINI followed a SQ work request being posted (because they are posted to differente EQs), but the #1 fix will handle the issue by flushing the SQ work request. Signed-off-by:
-
Steve Wise authored
Some HW platforms can reorder read operations, so we must rmb() after we see a valid gen bit in a CQE but before we read any other fields from the CQE. Signed-off-by:
-
Steve Wise authored
1) timedout endpoint processing can be starved. If there are continual CPL messages flowing into the driver, the endpoint timeout processing can be starved. This condition exposed the other bugs below. Solution: In process_work(), call process_timedout_eps() after each CPL is processed. 2) Connection events can be processed even though the endpoint is on the timeout list. If the endpoint is scheduled for timeout processing, then we must ignore MPA Start Requests and Replies. Solution: Change stop_ep_timer() to return 1 if the ep has already been queued for timeout processing. All the callers of stop_ep_timer() need to check this and act accordingly. There are just a few cases where the caller needs to do something different if stop_ep_timer() returns 1: 1) in process_mpa_reply(), ignore the reply and process_timeout() will abort the connection. 2) in process_mpa_request, ignore the request and process_timeout() will abort the connection. It is ok for callers of stop_ep_timer() to abort the connection since that will leave the state in ABORTING or DEAD, and process_timeout() now ignores timeouts when the ep is in these states. 3) Double insertion on the timeout list. Since the endpoint timers are used for connection setup and teardown, we need to guard against the possibility that an endpoint is already on the timeout list. This is a rare condition and only seen under heavy load and in the presense of the above 2 bugs. Solution: In ep_timeout(), don't queue the endpoint if it is already on the queue. Signed-off-by:
-
Steve Wise authored
Signed-off-by:
-
Eli Cohen authored
Add support for the block multicast loopback QP creation flag along the proper firmware API for that. Signed-off-by:
Eli Cohen <eli@mellanox.com> Signed-off-by:
Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by:
-
Alexander Gordeev authored
As result of the deprecation of the MSI-X/MSI enablement functions pci_enable_msix() and pci_enable_msi_block(), all drivers using these two interfaces need to be updated to use the new pci_enable_msi_range() or pci_enable_msi_exact() and pci_enable_msix_range() or pci_enable_msix_exact() interfaces. Signed-off-by:
Alexander Gordeev <agordeev@redhat.com> Signed-off-by:
-
Alexander Gordeev authored
As result of the deprecation of the MSI-X/MSI enablement functions pci_enable_msix() and pci_enable_msi_block(), all drivers using these two interfaces need to be updated to use the new pci_enable_msi_range() and pci_enable_msix_range() interfaces. Signed-off-by:
Mike Marciniszyn <mike.marciniszyn@intel.com> Signed-off-by:
-
- 03 Apr, 2014 26 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/roland/infinibandLinus Torvalds authored
Pull infiniband updates from Roland Dreier: "Main batch of InfiniBand/RDMA changes for 3.15: - The biggest change is core API extensions and mlx5 low-level driver support for handling DIF/DIX-style protection information, and the addition of PI support to the iSER initiator. Target support will be arriving shortly through the SCSI target tree. - A nice simplification to the "umem" memory pinning library now that we have chained sg lists. Kudos to Yishai Hadas for realizing our code didn't have to be so crazy. - Another nice simplification to the sg wrappers used by qib, ipath and ehca to handle their mapping of memory to adapter. - The usual batch of fixes to bugs found by static checkers etc. from intrepid people like Dan Carpenter and Yann Droneaud. - A large batch of cxgb4, ocrdma, qib driver updates" * tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (102 commits) RDMA/ocrdma: Unregister inet notifier when unloading ocrdma RDMA/ocrdma: Fix warnings about pointer <-> integer casts RDMA/ocrdma: Code clean-up RDMA/ocrdma: Display FW version RDMA/ocrdma: Query controller information RDMA/ocrdma: Support non-embedded mailbox commands RDMA/ocrdma: Handle CQ overrun error RDMA/ocrdma: Display proper value for max_mw RDMA/ocrdma: Use non-zero tag in SRQ posting RDMA/ocrdma: Memory leak fix in ocrdma_dereg_mr() RDMA/ocrdma: Increment abi version count RDMA/ocrdma: Update version string be2net: Add abi version between be2net and ocrdma RDMA/ocrdma: ABI versioning between ocrdma and be2net RDMA/ocrdma: Allow DPP QP creation RDMA/ocrdma: Read ASIC_ID register to select asic_gen RDMA/ocrdma: SQ and RQ doorbell offset clean up RDMA/ocrdma: EQ full catastrophe avoidance RDMA/cxgb4: Disable DSGL use by default RDMA/cxgb4: rx_data() needs to hold the ep mutex ... -
git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-stagingLinus Torvalds authored
Pull second round of hwmon updates from Guenter Roeck: "Add support for AMD F16 M30h processor to k10temp driver. This adds one more patch which had secondary dependencies. The branch point is arbitrary, but I did run a full set of build and qemu tests on it. While there are some new build failures (6 out of 122 in my builds), none are due to this commit" * tag 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging: hwmon: (k10temp) Add support for AMD F16 M30h processor
-
git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpioLinus Torvalds authored
Pull bulk of gpio updates from Linus Walleij: "A pretty big chunk of changes this time, but it has all been on rotation in linux-next and had some testing. Of course there will be some amount of fixes on top... - Merged in a branch of irqchip changes from Thomas Gleixner: we need to have new callbacks from the irqchip to determine if the GPIO line will be eligible for IRQs, and this callback must be able to say "no". After some thinking I got the branch from tglx and have switched all current users over to use this. - Based on tglx patches, we have added some generic irqchip helpers in the gpiolib core. These will help centralize code when GPIO drivers have simple chained/cascaded IRQs. Drivers will still define their irqchip vtables, but the gpiolib core will take care of irqdomain set-up, mapping from local offsets to Linux irqs, and reserve resources by marking the GPIO lines for IRQs. - Initially the PL061 and Nomadik GPIO/pin control drivers have been switched over to use the new gpiochip-to-irqchip infrastructure with more drivers expected for the next kernel cycle. The factoring of just two drivers still makes it worth it so it is already a win. - A new driver for the Synopsys DesignWare APB GPIO block. - Modify the DaVinci GPIO driver to be reusable also for the new TI Keystone architecture. - A new driver for the LSI ZEVIO SoCs. - Delete the obsolte tnetv107x driver. - Some incremental work on GPIO descriptors: have gpiod_direction_output() use a logical level, respecting assertion polarity through ACTIVE_LOW flags, adding gpiod_direction_output_raw() for the case where you want to set that very value. Add gpiochip_get_desc() to fetch a GPIO descriptor from a specific offset on a certain chip inside driver code. - Switch ACPI GPIO code over to using gpiochip_get_desc() and get rid of gpio_to_desc(). - The ACPI GPIO event handling code has been reworked after encountering an actual real life implementation. - Support for ACPI GPIO operation regions. - Generic GPIO chips can now be assigned labels/names from platform data. - We now clamp values returned from GPIO drivers to the boolean [0,1] range. - Some improved documentation on how to use the polarity flag was added. - a large slew of incremental driver updates and non-critical fixes. Some targeted for stable" * tag 'gpio-v3.15-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio: (80 commits) gpio: rcar: Add helper variable dev = &pdev->dev gpio-lynxpoint: force gpio_get() to return "1" and "0" only gpio: unmap gpio irqs properly pch_gpio: set value before enabling output direction gpio: moxart: Actually set output state in moxart_gpio_direction_output() gpio: moxart: Avoid forward declaration gpio: mxs: Allow for recursive enable_irq_wake() call gpio: samsung: Add missing "break" statement gpio: twl4030: Remove redundant assignment gpio: dwapb: correct gpio-cells in binding document gpio: iop: fix devm_ioremap_resource() return value checking pinctrl: coh901: convert driver to use gpiolib irqchip pinctrl: nomadik: convert driver to use gpiolib irqchip gpio: pl061: convert driver to use gpiolib irqchip gpio: add IRQ chip helpers in gpiolib pinctrl: nomadik: factor in platform data container pinctrl: nomadik: rename secondary to latent gpio: Driver for SYSCON-based GPIOs gpio: generic: Use platform_device_id->driver_data field for driver flags pinctrl: coh901: move irq line locking to resource callbacks ... -
Linus Torvalds authored
Merge first patch-bomb from Andrew Morton: - Various misc bits - kmemleak fixes - small befs, codafs, cifs, efs, freexxfs, hfsplus, minixfs, reiserfs things - fanotify - I appear to have become SuperH maintainer - ocfs2 updates - direct-io tweaks - a bit of the MM queue - printk updates - MAINTAINERS maintenance - some backlight things - lib/ updates - checkpatch updates - the rtc queue - nilfs2 updates - Small Documentation/ updates * emailed patches from Andrew Morton <akpm@linux-foundation.org>: (237 commits) Documentation/SubmittingPatches: remove references to patch-scripts Documentation/SubmittingPatches: update some dead URLs Documentation/filesystems/ntfs.txt: remove changelog reference Documentation/kmemleak.txt: updates fs/reiserfs/super.c: add __init to init_inodecache fs/reiserfs: move prototype declaration to header file fs/hfsplus/attributes.c: add __init to hfsplus_create_attr_tree_cache() fs/hfsplus/extents.c: fix concurrent acess of alloc_blocks fs/hfsplus/extents.c: remove unused variable in hfsplus_get_block nilfs2: update project's web site in nilfs2.txt nilfs2: update MAINTAINERS file entries fix nilfs2: verify metadata sizes read from disk nilfs2: add FITRIM ioctl support for nilfs2 nilfs2: add nilfs_sufile_trim_fs to trim clean segs nilfs2: implementation of NILFS_IOCTL_SET_SUINFO ioctl nilfs2: add nilfs_sufile_set_suinfo to update segment usage nilfs2: add struct nilfs_suinfo_update and flags nilfs2: update MAINTAINERS file entries fs/coda/inode.c: add __init to init_inodecache() BEFS: logging cleanup ...
-
Mitchel Humpherys authored
The link to the tarball for Andrew Morton's patch scripts is dead. These scripts don't seem to be used for kernel development these days anyways so just rip out all references to them. Signed-off-by:
Mitchel Humpherys <mitchelh@codeaurora.org> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Mitchel Humpherys authored
The links to "The perfect patch" and "NO!!!! No more huge patch bombs..." have gone stale. Update them to some working locations. Signed-off-by:
-
Fabian Frederick authored
File was removed in commit 7c821a17 ("Remove fs/ntfs/ChangeLog"). Signed-off-by:
Fabian Frederick <fabf@skynet.be> Acked-by:
Anton Altaparmakov <anton@tuxera.com> Signed-off-by:
-
Wang YanQing authored
Update Documentatin/kmemleak.txt to reflect the following changes: Commit b69ec42b ("Kconfig: clean up the long arch list for the DEBUG_KMEMLEAK config option") made it so that we can't check supported architectures by read Kconfig.debug. Commit 85d3a316 ("kmemleak: use rbtree instead of prio tree") converted kmemleak to use rbtree instead of prio tree. Signed-off-by:
Wang YanQing <udknight@gmail.com> Cc: Catalin Marinas <catalin.marinas@arm.com> Signed-off-by:
-
Fabian Frederick authored
init_inodecache is only called by __init init_reiserfs_fs. Signed-off-by:
Jeff Mahoney <jeffm@suse.com> Signed-off-by:
-
Rashika Kheria authored
Move prototype declaration to header file reiserfs/reiserfs.h from reiserfs/super.c because they are used by more than one file. This eliminates the following warning in reiserfs/bitmap.c: fs/reiserfs/bitmap.c:647:6: warning: no previous prototype for `show_alloc_options' [-Wmissing-prototypes] Signed-off-by:
Rashika Kheria <rashika.kheria@gmail.com> Reviewed-by:
Josh Triplett <josh@joshtriplett.org> Acked-by:
-
Fabian Frederick authored
hfsplus_create_attr_tree_cache is only called by __init init_hfsplus_fs Signed-off-by:
Vyacheslav Dubeyko <slava@dubeyko.com> Signed-off-by:
-
Sougata Santra authored
Concurrent access to alloc_blocks in hfsplus_inode_info() is protected by extents_lock mutex. This patch fixes two instances where alloc_blocks modification was not protected with this lock. This fixes possible allocation bitmap corruption in race conditions while extending and truncating files. [akpm@linux-foundation.org: take extents_lock before taking a copy of ->alloc_blocks] [akpm@linux-foundation.org: remove now-unused label `out'] Signed-off-by:
Sougata Santra <sougata@tuxera.com> Reviewed-by:
Christoph Hellwig <hch@lst.de> Cc: Vyacheslav Dubeyko <slava@dubeyko.com> Cc: Alexey Khoroshilov <khoroshilov@ispras.ru> Signed-off-by:
-
Sougata Santra authored
The variable is defined but not used. Generally it compiles away with -O2 optimization hence it does not show a warning. Signed-off-by:
-
Ryusuke Konishi authored
Project's web site was moved to nilfs.sourceforge.net from www.nilfs.org. This updates the site information in Documentation/filesystems/nilfs2.txt with the new location. Signed-off-by:
Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by:
-
Ryusuke Konishi authored
Also, web-page entry is updated according to relocation of project's web site. Signed-off-by:
-
Ryusuke Konishi authored
Add code to check sizes of on-disk data of metadata files such as inode size, segment usage size, DAT entry size, and checkpoint size. Although these sizes are read from disk, the current implementation doesn't check them. If these sizes are not sane on disk, it can cause out-of-range access to metadata or memory access overrun on metadata block buffers due to overflow in sundry calculations. Both lower limit and upper limit of metadata sizes are verified to prevent these issues. Signed-off-by:
-
Andreas Rohner authored
Add support for the FITRIM ioctl, which enables user space tools to issue TRIM/DISCARD requests to the underlying device. Every clean segment within the specified range will be discarded. Signed-off-by:
Andreas Rohner <andreas.rohner@gmx.net> Signed-off-by:
-
Andreas Rohner authored
Add nilfs_sufile_trim_fs(), which takes an fstrim_range structure and calls blkdev_issue_discard for every clean segment in the specified range. The range is truncated to file system block boundaries. Signed-off-by:
-
Andreas Rohner authored
With this ioctl the segment usage entries in the SUFILE can be updated from userspace. This is useful, because it allows the userspace GC to modify and update segment usage entries for specific segments, which enables it to avoid unnecessary write operations. If a segment needs to be cleaned, but there is no or very little reclaimable space in it, the cleaning operation basically degrades to a useless moving operation. In the end the only thing that changes is the location of the data and a timestamp in the segment usage information. With this ioctl the GC can skip the cleaning and update the segment usage entries directly instead. This is basically a shortcut to cleaning the segment. It is still necessary to read the segment summary information, but the writing of the live blocks can be skipped if it's not worth it. [konishi.ryusuke@lab.ntt.co.jp: add description of NILFS_IOCTL_SET_SUINFO ioctl] Signed-off-by:
-
Andreas Rohner authored
Introduce nilfs_sufile_set_suinfo(), which expects an array of nilfs_suinfo_update structures and updates the segment usage information accordingly. This is basically a helper function for the newly introduced NILFS_IOCTL_SET_SUINFO ioctl. [konishi.ryusuke@lab.ntt.co.jp: use put_bh() instead of brelse() because we know bh != NULL] Signed-off-by:
-
Andreas Rohner authored
Add the nilfs_suinfo_update structure, which contains the information needed to update one segment usage entry. The flags specify, which fields need to be updated. Signed-off-by:
-
Ryusuke Konishi authored
Update git repository entry of nilfs2 file system and maintainer's email description. Signed-off-by:
-
Fabian Frederick authored
init_inodecache is only called by __init init_coda Signed-off-by:
-
Fabian Frederick authored
Summary: - all printk(KERN_foo converted to pr_foo() - add pr_fmt and remove redundant prefixes - convert befs_() to va_format (based on patch by Joe Perches) - remove non standard %Lu - use __func__ for all debugging [akpm@linux-foundation.org: fix printk warnings, reported by Fengguang] Signed-off-by:
-
Fabian Frederick authored
init_inodecache is only called by __init init_befs_fs. Signed-off-by:
-
Fabian Frederick authored
Use kzalloc for clean fs_info allocation like other filesystems. Signed-off-by:
-
