- 14 May, 2017 25 commits
-
-
Krzysztof Kozlowski authored
commit 68bd6fc3 upstream. Returning from for_each_available_child_of_node() loop requires cleaning up node refcount. Error paths lacked it so for example in case of deferred probe, the refcount of phy node was left increased. Fixes: 6d40500a ("usb: ehci/ohci-exynos: Fix of_node_put() for child when getting PHYs") Signed-off-by:
Krzysztof Kozlowski <krzk@kernel.org> Acked-by:
Alan Stern <stern@rowland.harvard.edu> Reviewed-by:
Javier Martinez Canillas <javier@osg.samsung.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Krzysztof Kozlowski authored
commit 3f6026b1 upstream. Returning from for_each_available_child_of_node() loop requires cleaning up node refcount. Error paths lacked it so for example in case of deferred probe, the refcount of phy node was left increased. Fixes: 6d40500a ("usb: ehci/ohci-exynos: Fix of_node_put() for child when getting PHYs") Signed-off-by:
-
Ladi Prosek authored
commit ab007cc9 upstream. The PML feature is not exposed to guests so we should not be forwarding the vmexit either. This commit fixes BSOD 0x20001 (HYPERVISOR_ERROR) when running Hyper-V enabled Windows Server 2016 in L1 on hardware that supports PML. Fixes: 843e4330 ("KVM: VMX: Add PML support in VMX") Signed-off-by:
Ladi Prosek <lprosek@redhat.com> Reviewed-by:
David Hildenbrand <david@redhat.com> Signed-off-by:
Radim Krčmář <rkrcmar@redhat.com> Signed-off-by:
-
Ladi Prosek authored
commit 1fb883bb upstream. L2 was running with uninitialized PML fields which led to incomplete dirty bitmap logging. This manifested as all kinds of subtle erratic behavior of the nested guest. Fixes: 843e4330 ("KVM: VMX: Add PML support in VMX") Signed-off-by:
-
Jim Mattson authored
commit 0b4c208d upstream. This reverts commit bc613494. A CPUID instruction executed in VMX non-root mode always causes a VM-exit, regardless of the leaf being queried. Fixes: bc613494 ("KVM: nested VMX: disable perf cpuid reporting") Signed-off-by:
Jim Mattson <jmattson@google.com> [The issue solved by bc613494 has been resolved with ff651cb6 ("KVM: nVMX: Add nested msr load/restore algorithm").] Signed-off-by:
-
Andy Shevchenko authored
commit 80354c29 upstream. The interrupt line used for the watchdog is 12, according to the official Intel Edison BSP code. And indeed after fixing it we start getting an interrupt and thus the watchdog starts working again: [ 191.699951] Kernel panic - not syncing: Kernel Watchdog Signed-off-by:
Andy Shevchenko <andriy.shevchenko@linux.intel.com> Cc: Borislav Petkov <bp@alien8.de> Cc: David Cohen <david.a.cohen@linux.intel.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: 78a3bb9e ("x86: intel-mid: add watchdog platform code for Merrifield") Link: http://lkml.kernel.org/r/20170312150744.45493-1-andriy.shevchenko@linux.intel.comSigned-off-by:
Ingo Molnar <mingo@kernel.org> Signed-off-by:
-
Masami Hiramatsu authored
commit 75013fb1 upstream. Fix to the exception table entry check by using probed address instead of the address of copied instruction. This bug may cause unexpected kernel panic if user probe an address where an exception can happen which should be fixup by __ex_table (e.g. copy_from_user.) Unless user puts a kprobe on such address, this doesn't cause any problem. This bug has been introduced years ago, by commit: 46484688 ("x86/kprobes: Fix a bug which can modify kernel code permanently"). Signed-off-by:
Masami Hiramatsu <mhiramat@kernel.org> Cc: Borislav Petkov <bp@alien8.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: 46484688 ("x86/kprobes: Fix a bug which can modify kernel code permanently") Link: http://lkml.kernel.org/r/148829899399.28855.12581062400757221722.stgit@devboxSigned-off-by:
-
Pierre-Louis Bossart authored
commit f35b6542 upstream. Fix Makefile for x86 support, dependency on CONFIG_COMMON_CLK was not explicit Fixes: 701190fd ('clk: x86: add support for Lynxpoint LPSS clocks') Signed-off-by:
Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Acked-by:
Stephen Boyd <sboyd@codeaurora.org> Signed-off-by:
-
Nikola Pajkovsky authored
commit 68dee8e2 upstream. commit 8fd524b3 ("x86: Kill bad_dma_address variable") has killed bad_dma_address variable and used instead of macro DMA_ERROR_CODE which is always zero. Since dma_addr is unsigned, the statement dma_addr >= DMA_ERROR_CODE is always true, and not needed. arch/x86/kernel/pci-calgary_64.c: In function ‘iommu_free’: arch/x86/kernel/pci-calgary_64.c:299:2: warning: comparison of unsigned expression >= 0 is always true [-Wtype-limits] if (unlikely((dma_addr >= DMA_ERROR_CODE) && (dma_addr < badend))) { Fixes: 8fd524b3 ("x86: Kill bad_dma_address variable") Signed-off-by:
Nikola Pajkovsky <npajkovsky@suse.cz> Cc: iommu@lists.linux-foundation.org Cc: Jon Mason <jdmason@kudzu.us> Cc: Muli Ben-Yehuda <mulix@mulix.org> Link: http://lkml.kernel.org/r/7612c0f9dd7c1290407dbf8e809def922006920b.1479161177.git.npajkovsky@suse.czSigned-off-by:
Thomas Gleixner <tglx@linutronix.de> Signed-off-by:
-
Ruslan Ruslichenko authored
commit a9b4f087 upstream. commit d32932d0 removed the irq_retrigger callback from the IO-APIC chip and did not add it to the new IO-APIC-IR irq chip. There is no harm because the interrupts are resent in software when the retrigger callback is NULL, but it's less efficient. So restore them. [ tglx: Massaged changelog ] Fixes: d32932d0 ("x86/irq: Convert IOAPIC to use hierarchical irqdomain interfaces") Signed-off-by:
Ruslan Ruslichenko <rruslich@cisco.com> Cc: xe-linux-external@cisco.com Link: http://lkml.kernel.org/r/1484662432-13580-1-git-send-email-rruslich@cisco.comSigned-off-by:
-
Ganapathi Bhat authored
commit a5b60de6 upstream. This patch fixes the issue specific to AP. AP is started with WEP security and external station is connected to it. Data path works in this case. Now if AP is restarted with WPA/WPA2 security, station is able to connect but ping fails. Driver skips the deletion of WEP keys if interface type is AP. Removing that redundant check resolves the issue. Fixes: e57f1734 ("mwifiex: add key material v2 support") Signed-off-by:
Ganapathi Bhat <gbhat@marvell.com> Signed-off-by:
Amitkumar Karwar <akarwar@marvell.com> Signed-off-by:
Kalle Valo <kvalo@codeaurora.org> Signed-off-by:
-
Xinming Hu authored
commit 5f0a221f upstream. We already ensure 64 bytes alignment and add padding if required during skb_aggr allocation. Alignment and padding in mwifiex_11n_form_amsdu_txpd() is redundant. We may end up accessing more data than allocated size with this. This patch fixes following issue by removing redundant padding. [ 370.241338] skbuff: skb_over_panic: text:ffffffffc046946a len:3550 put:72 head:ffff880000110000 data:ffff8800001100e4 tail:0xec2 end:0xec0 dev:<NULL> [ 370.241374] ------------[ cut here ]------------ [ 370.241382] kernel BUG at net/core/skbuff.c:104! 370.244032] Call Trace: [ 370.244041] [<ffffffff8c3df5ec>] skb_put+0x44/0x45 [ 370.244055] [<ffffffffc046946a>] mwifiex_11n_aggregate_pkt+0x1e9/0xa50 [mwifiex] [ 370.244067] [<ffffffffc0467c16>] mwifiex_wmm_process_tx+0x44a/0x6b7 [mwifiex] [ 370.244074] [<ffffffffc0411eb8>] ? 0xffffffffc0411eb8 [ 370.244084] [<ffffffffc046116b>] mwifiex_main_process+0x476/0x5a5 [mwifiex] [ 370.244098] [<ffffffffc0461298>] mwifiex_main_process+0x5a3/0x5a5 [mwifiex] [ 370.244113] [<ffffffff8be7e9ff>] process_one_work+0x1a4/0x309 [ 370.244123] [<ffffffff8be7f4ca>] worker_thread+0x20c/0x2ee [ 370.244130] [<ffffffff8be7f2be>] ? rescuer_thread+0x383/0x383 [ 370.244136] [<ffffffff8be7f2be>] ? rescuer_thread+0x383/0x383 [ 370.244143] [<ffffffff8be83742>] kthread+0x11c/0x124 [ 370.244150] [<ffffffff8be83626>] ? kthread_parkme+0x24/0x24 [ 370.244157] [<ffffffff8c4da1ef>] ret_from_fork+0x3f/0x70 [ 370.244168] [<ffffffff8be83626>] ? kthread_parkme+0x24/0x24 Fixes: 84b313b3 ("mwifiex: make tx packet 64 byte DMA aligned") Signed-off-by:
Xinming Hu <huxm@marvell.com> Signed-off-by:
-
Brian Norris authored
commit 6183468a upstream. Similar to commit fcd2042e ("mwifiex: printk() overflow with 32-byte SSIDs"), we failed to account for the existence of 32-char SSIDs in our debugfs code. Unlike in that case though, we zeroed out the containing struct first, and I'm pretty sure we're guaranteed to have some padding after the 'ssid.ssid' and 'ssid.ssid_len' fields (the struct is 33 bytes long). So, this is the difference between: # cat /sys/kernel/debug/mwifiex/mlan0/info ... essid="0123456789abcdef0123456789abcdef " ... and the correct output: # cat /sys/kernel/debug/mwifiex/mlan0/info ... essid="0123456789abcdef0123456789abcdef" ... Fixes: 5e6e3a92 ("wireless: mwifiex: initial commit for Marvell mwifiex driver") Signed-off-by:
Brian Norris <briannorris@chromium.org> Signed-off-by:
-
Matthijs van Duin authored
commit 448c077e upstream. 'adr' yields a data-pointer, not a function-pointer. Fixes: 999f934d ("ARM: omap5/dra7xx: Enable booting secondary CPU in HYP mode") Signed-off-by:
Matthijs van Duin <matthijsvanduin@gmail.com> Signed-off-by:
Tony Lindgren <tony@atomide.com> Signed-off-by:
-
Arnd Bergmann authored
commit cbe99c53 upstream. gcc gets confused about the control flow in ktd2692_parse_dt(), causing it to warn about what seems like a potential bug: drivers/leds/leds-ktd2692.c: In function 'ktd2692_probe': drivers/leds/leds-ktd2692.c:244:15: error: '*((void *)&led_cfg+8)' may be used uninitialized in this function [-Werror=maybe-uninitialized] drivers/leds/leds-ktd2692.c:225:7: error: 'led_cfg.flash_max_microamp' may be used uninitialized in this function [-Werror=maybe-uninitialized] drivers/leds/leds-ktd2692.c:232:3: error: 'led_cfg.movie_max_microamp' may be used uninitialized in this function [-Werror=maybe-uninitialized] The code is fine, and slightly reworking it in an equivalent way lets gcc figure that out too, which gets rid of the warning. Fixes: 77e7915b ("leds: ktd2692: Add missing of_node_put") Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Acked-by:
Pavel Machek <pavel@ucw.cz> Signed-off-by:
Jacek Anaszewski <jacek.anaszewski@gmail.com> Signed-off-by:
-
Liam Breck authored
commit ba52e757 upstream. Reading both fault and status registers and logging any fault should take priority over handling status register update. Fix by moving the status handling to later in interrupt routine. Fixes: d7bf353f ("bq24190_charger: Add support for TI BQ24190 Battery Charger") Signed-off-by:
Liam Breck <kernel@networkimprov.net> Acked-by:
Mark Greer <mgreer@animalcreek.com> Acked-by:
Sebastian Reichel <sre@kernel.org> Signed-off-by:
-
Liam Breck authored
commit 68abfb80 upstream. Caching the fault register after a single I2C read may not keep an accurate value. Fix by doing two reads in irq_handle_thread() and using the cached value elsewhere. If a safety timer fault later clears itself, we apparently don't get an interrupt (INT), however other interrupts would refresh the register cache. From the data sheet: "When a fault occurs, the charger device sends out INT and keeps the fault state in REG09 until the host reads the fault register. Before the host reads REG09 and all the faults are cleared, the charger device would not send any INT upon new faults. In order to read the current fault status, the host has to read REG09 two times consecutively. The 1st reads fault register status from the last read [1] and the 2nd reads the current fault register status." [1] presumably a typo; should be "last fault" Fixes: d7bf353f ("bq24190_charger: Add support for TI BQ24190 Battery Charger") Signed-off-by:
-
Liam Breck authored
commit 2d9fee6a upstream. We wrongly get uevents for bq24190-charger and bq24190-battery on every register change. Fix by checking the association with charger and battery before emitting uevent(s). Fixes: d7bf353f ("bq24190_charger: Add support for TI BQ24190 Battery Charger") Signed-off-by:
-
Liam Breck authored
commit d62acc5e upstream. The device specific data is not fully initialized on request_threaded_irq(). This may cause a crash when the IRQ handler tries to reference them. Fix the issue by installing IRQ handler at the end of the probe. Fixes: d7bf353f ("bq24190_charger: Add support for TI BQ24190 Battery Charger") Signed-off-by:
-
Liam Breck authored
commit e05ad7e0 upstream. pm_resume() does a register_reset() which clears charger host mode. Fix by calling set_mode_host() after the reset. Fixes: d7bf353f ("bq24190_charger: Add support for TI BQ24190 Battery Charger") Signed-off-by:
-
Liam Breck authored
commit 767eee36 upstream. The interrupt signal is TRIGGER_FALLING. This is is specified in the data sheet PIN FUNCTIONS: "The INT pin sends active low, 256us pulse to host to report charger device status and fault." Also the direction can be seen in the data sheet Figure 37 "BQ24190 with D+/D- Detection and USB On-The-Go (OTG)" which shows a 10k pull-up resistor installed for the sample configurations. Fixes: d7bf353f ("bq24190_charger: Add support for TI BQ24190 Battery Charger") Signed-off-by:
-
Michael Ellerman authored
commit a7e0fb6c upstream. Currently the opal_exit tracepoint usually shows the opcode as 0: <idle>-0 [047] d.h. 635.654292: opal_entry: opcode=63 <idle>-0 [047] d.h. 635.654296: opal_exit: opcode=0 retval=0 kopald-1209 [019] d... 636.420943: opal_entry: opcode=10 kopald-1209 [019] d... 636.420959: opal_exit: opcode=0 retval=0 This is because we incorrectly load the opcode into r0 before calling __trace_opal_exit(), whereas it expects the opcode in r3 (first function parameter). In fact we are leaving the retval in r3, so opcode and retval will always show the same value. Instead load the opcode into r3, resulting in: <idle>-0 [040] d.h. 636.618625: opal_entry: opcode=63 <idle>-0 [040] d.h. 636.618627: opal_exit: opcode=63 retval=0 Fixes: c49f6353 ("powernv: Add OPAL tracepoints") Signed-off-by:
Michael Ellerman <mpe@ellerman.id.au> Signed-off-by:
-
Ben Hutchings authored
commit 4cca0457 upstream. The switch that conditionally sets CPUPOWER_CAP_HAS_TURBO_RATIO and CPUPOWER_CAP_IS_SNB flags is missing a break, so all cores get both flags set and an assumed base clock of 100 MHz for turbo values. Reported-by:
GSR <gsr.bugs@infernal-iceberg.com> Tested-by:
Ben Hutchings <ben@decadent.org.uk> Signed-off-by:
Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by:
-
Ard Biesheuvel authored
commit 5008efc8 upstream. The PJ4 inline asm sequence to write to cp15 cannot be built in Thumb-2 mode, due to the way it performs arithmetic on the program counter, so it is built in ARM mode instead. However, building C files in ARM mode under CONFIG_THUMB2_KERNEL is problematic, since the instrumentation performed by subsystems like ftrace does not expect having to deal with interworking branches. Since the sequence in question is simply a poor man's ISB instruction, let's use a straight 'isb' instead when building in Thumb2 mode. Thumb2 implies V7, so 'isb' should always be supported in that case. Acked-by:
Nicolas Pitre <nico@linaro.org> Signed-off-by:
Ard Biesheuvel <ard.biesheuvel@linaro.org> Signed-off-by:
Russell King <rmk+kernel@arm.linux.org.uk> Signed-off-by:
-
Cong Wang authored
commit b5c66bab upstream. posix_acl_update_mode() could possibly clear 'acl', if so we leak the memory pointed by 'acl'. Save this pointer before calling posix_acl_update_mode() and release the memory if 'acl' really gets cleared. Link: http://lkml.kernel.org/r/1486678332-2430-1-git-send-email-xiyou.wangcong@gmail.comSigned-off-by:
Cong Wang <xiyou.wangcong@gmail.com> Reported-by:
Mark Salyzyn <salyzyn@android.com> Reviewed-by:
Jan Kara <jack@suse.cz> Reviewed-by:
Greg Kurz <groug@kaod.org> Cc: Eric Van Hensbergen <ericvh@gmail.com> Cc: Ron Minnich <rminnich@sandia.gov> Cc: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by:
-
- 08 May, 2017 15 commits
-
-
Greg Kroah-Hartman authored
-
Adrian Salido authored
commit 4617f564 upstream. When calling a dm ioctl that doesn't process any data (IOCTL_FLAGS_NO_PARAMS), the contents of the data field in struct dm_ioctl are left initialized. Current code is incorrectly extending the size of data copied back to user, causing the contents of kernel stack to be leaked to user. Fix by only copying contents before data and allow the functions processing the ioctl to override. Signed-off-by:
Adrian Salido <salidoa@google.com> Reviewed-by:
Alasdair G Kergon <agk@redhat.com> Signed-off-by:
Mike Snitzer <snitzer@redhat.com> Signed-off-by:
-
J. Bruce Fields authored
commit 13bf9fbf upstream. The NFSv2/v3 code does not systematically check whether we decode past the end of the buffer. This generally appears to be harmless, but there are a few places where we do arithmetic on the pointers involved and don't account for the possibility that a length could be negative. Add checks to catch these. Reported-by:
Tuomas Haanpää <thaan@synopsys.com> Reported-by:
Ari Kauppi <ari@synopsys.com> Reviewed-by:
NeilBrown <neilb@suse.com> Signed-off-by:
J. Bruce Fields <bfields@redhat.com> Signed-off-by:
-
J. Bruce Fields authored
commit db44bac4 upstream. Use a couple shortcuts that will simplify a following bugfix. (Minor backporting required to account for a change from f34b9568 "The NFSv2/NFSv3 server does not handle zero length WRITE requests correctly".) Signed-off-by:
-
Jaegeuk Kim authored
commit 03a8bb0e upstream. As Al pointed, d_revalidate should return RCU lookup before using d_inode. This was originally introduced by: commit 34286d66 ("fs: rcu-walk aware d_revalidate method"). Reported-by:
Al Viro <viro@zeniv.linux.org.uk> Signed-off-by:
Jaegeuk Kim <jaegeuk@kernel.org> Cc: Theodore Ts'o <tytso@mit.edu> Signed-off-by:
Eric Biggers <ebiggers@google.com> Signed-off-by:
-
Theodore Ts'o authored
commit 3d43bcfe upstream. This avoids potential problems caused by a race where the inode gets renamed out from its parent directory and the parent directory is deleted while ext4_d_revalidate() is running. Fixes: 28b4c263Reported-by:
Theodore Ts'o <tytso@mit.edu> Signed-off-by:
-
Theodore Ts'o authored
commit 28b4c263 upstream. Add a validation check for dentries for encrypted directory to make sure we're not caching stale data after a key has been added or removed. Also check to make sure that status of the encryption key is updated when readdir(2) is executed. Signed-off-by:
-
Richard Weinberger authored
commit 9a200d07 upstream. ...otherwise an user can enable encryption for certain files even when the filesystem is unable to support it. Such a case would be a filesystem created by mkfs.ext4's default settings, 1KiB block size. Ext4 supports encyption only when block size is equal to PAGE_SIZE. But this constraint is only checked when the encryption feature flag is set. Signed-off-by:
Richard Weinberger <richard@nod.at> Signed-off-by:
-
Arnd Bergmann authored
The driver causes two warnings about possibly uninitialized variables: drivers/infiniband/hw/ehca/ehca_mrmw.c: In function 'ehca_set_pagebuf': drivers/infiniband/hw/ehca/ehca_mrmw.c:1908:4: warning: 'prev_pgaddr' may be used uninitialized in this function [-Wmaybe-uninitialized] drivers/infiniband/hw/ehca/ehca_mrmw.c:1924:14: note: 'prev_pgaddr' was declared here drivers/infiniband/hw/ehca/ehca_mrmw.c: In function 'ehca_reg_mr': drivers/infiniband/hw/ehca/ehca_mrmw.c:2430:5: warning: 'hret' may be used uninitialized in this function [-Wmaybe-uninitialized] The first one is definitely a false positive, the second one may or may not be one. In both cases, adding an intialization is the safe and easy workaround. The driver was removed in mainline in commit e581d111 ("staging/rdma: remove deprecated ehca driver"), in linux-4.6. In 4.4, the file is located in drivers/staging/rdma/ehca/ehca_mrmw.c, and the fix still applies. Signed-off-by:
-
Arnd Bergmann authored
We get this build warning on arm64 drivers/infiniband/hw/qib/qib_qp.c:44:0: error: "BITS_PER_PAGE" redefined [-Werror] #define BITS_PER_PAGE (PAGE_SIZE*BITS_PER_BYTE) This is fixed upstream in commit 898fa52b ("IB/qib: Remove qpn, qp tables and related variables from qib"), which does a lot of other things as well. Instead, I just backport the rename of the local BITS_PER_PAGE definition to RVT_BITS_PER_PAGE. The driver first showed up in linux-2.6.35, and the fixup should still apply to that. The upstream fix went into v4.6, so we could apply this workaround to both 3.18 and 4.4. Fixes: f931551b ("IB/qib: Add new qib driver for QLogic PCIe InfiniBand adapters") Signed-off-by:
-
Ross Lagerwall authored
The backport of d35c99ff ("netlink: do not enter direct reclaim from netlink_dump()") to the 4.4 branch (first in 4.4.32) mistakenly removed direct claim from the initial large allocation _and_ the fallback allocation which means that allocations can spuriously fail. Fix the issue by adding back the direct reclaim flag to the fallback allocation. Fixes: 6d123f1d ("netlink: do not enter direct reclaim from netlink_dump()") Signed-off-by:
Ross Lagerwall <ross.lagerwall@citrix.com> Signed-off-by:
-
Gabriel Krisman Bertazi authored
commit c130b666 upstream. Commit f209fa03 ("serial: 8250_pci: Detach low-level driver during PCI error recovery") introduces a potential use-after-free in case the pciserial_init_ports call in serial8250_io_resume fails, which may happen if a memory allocation fails or if the .init quirk failed for whatever reason). If this happen, further pci_get_drvdata will return a pointer to freed memory. This patch reworks the PCI recovery resume hook to restore the old priv structure in this case, which should be ok, since the ports were already detached. Such error during recovery causes us to give up on the recovery. Fixes: f209fa03 ("serial: 8250_pci: Detach low-level driver during PCI error recovery") Reported-by:
Michal Suchanek <msuchanek@suse.com> Signed-off-by:
Gabriel Krisman Bertazi <krisman@linux.vnet.ibm.com> Signed-off-by:
Guilherme G. Piccoli <gpiccoli@linux.vnet.ibm.com> Cc: Sumit Semwal <sumit.semwal@linaro.org> Signed-off-by:
-
Matthew R. Ochs authored
commit 05dab432 upstream. When an EEH occurs during device initialization, the port timeout logic can cause excessive delays as MMIO reads will fail. Depending on where they are experienced, these delays can lead to a prolonged reset, causing an unnecessary triggering of other timeout logic in the SCSI stack or user applications. To expedite recovery, the port timeout logic is updated to decay the timeout at a much faster rate when in the presence of a likely EEH frozen event. Signed-off-by:
Matthew R. Ochs <mrochs@linux.vnet.ibm.com> Acked-by:
Uma Krishnan <ukrishn@linux.vnet.ibm.com> Signed-off-by:
Martin K. Petersen <martin.petersen@oracle.com> Cc: Sumit Semwal <sumit.semwal@linaro.org> Signed-off-by:
-
Matthew R. Ochs authored
commit 1d3324c3 upstream. The EEH reset handler is ignorant to the current state of the driver when processing a frozen event and initiating a device reset. This can be an issue if an EEH event occurs while a user or stack initiated reset is executing. More specifically, if an EEH occurs while the SCSI host reset handler is active, the reset initiated by the EEH thread will likely collide with the host reset thread. This can leave the device in an inconsistent state, or worse, cause a system crash. As a remedy, the EEH handler is updated to evaluate the device state and take appropriate action (proceed, wait, or disconnect host). The host reset handler is also updated to handle situations where an EEH occurred during a host reset. In such situations, the host reset handler will delay reporting back a success to give the EEH reset an opportunity to complete. Signed-off-by:
-
Uma Krishnan authored
commit bbbfae96 upstream. When a port link is established, the AFU sends a 'link up' interrupt. After the link is up, corresponding initialization steps are performed on the card. Following that, when the card is ready for I/O, the AFU sends 'login succeeded' interrupt. Today, cxlflash invokes scsi_scan_host() upon receipt of both interrupts. SCSI commands sent to the port prior to the 'login succeeded' interrupt will fail with 'port not available' error. This is not desirable. Moreover, when async_scan is active for the host, subsequent scan calls are terminated with error. Due to this, the scsi_scan_host() call performed after 'login succeeded' interrupt could portentially return error and the devices may not be scanned properly. To avoid this problem, scsi_scan_host() should be called only after the 'login succeeded' interrupt. Signed-off-by:
-
