- 31 Oct, 2016 40 commits
-
-
Geert Uytterhoeven authored
commit 2cf9a578 upstream. clk-divider uses clk_readl()/clk_writel() everywhere, except in clk_divider_round_rate(), where plain readl() is used. Change this to clk_readl(), as it makes a difference on powerpc. Fixes: e6d5e7d9 ("clk-divider: Fix READ_ONLY when divider > 1") Signed-off-by:
Geert Uytterhoeven <geert+renesas@glider.be> Reviewed-by:
James Hogan <james.hogan@imgtec.com> Signed-off-by:
Stephen Boyd <sboyd@codeaurora.org> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Jean-Francois Moine authored
commit 3174b0c9 upstream. This patch reverts commit 023bd716 ("clk: skip unnecessary set_phase if nothing to do"), fixing two problems: * in some SoCs, the hardware phase delay depends on the rate ratio of the clock and its parent. So, changing this ratio may imply to set new hardware values, even if the logical delay is the same. * when the delay was the same as previously, an error was returned. Signed-off-by:
Jean-Francois Moine <moinejf@free.fr> Fixes: 023bd716 ("clk: skip unnecessary set_phase if nothing to do") Signed-off-by:
-
Stephen Boyd authored
commit f155d15b upstream. Before commit 0861e5b8 (clk: Add clk_hw OF clk providers, 2016-02-05) __of_clk_get_from_provider() would return an error pointer of the provider's choosing if there was a provider registered and EPROBE_DEFER otherwise. After that commit, it would return EPROBE_DEFER regardless of whether or not the provider returned an error. This is odd and can lead to behavior where clk consumers keep probe deferring when they should be seeing some other error. Let's restore the previous behavior where we only return EPROBE_DEFER when there isn't a provider in our of_clk_providers list. Otherwise, return the error from the last provider we find that matches the node. Reported-by:
Masahiro Yamada <yamada.masahiro@socionext.com> Fixes: 0861e5b8 ("clk: Add clk_hw OF clk providers") Signed-off-by:
-
Tang Yuantian authored
commit 8964193f upstream. The offset of Core Cluster clock control/status register on cluster group V3 version is different from others, and should be plus 0x70000. Signed-off-by:
Tang Yuantian <yuantian.tang@nxp.com> Reviewed-by:
Scott Wood <oss@buserror.net> Fixes: 9e19ca2f ("clk: qoriq: Add ls2080a support.") Signed-off-by:
-
Srinivas Kandagatla authored
commit 6d91f2c0 upstream. This patch selects QCOM_GDSC Kconfig for msm8996 GCC and MMCC clock controllers, as these provide some of the gdscs on the SOC. Also selecting this config will make it align with other drivers which do the same. Signed-off-by:
Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Fixes: 52111672 ("clk: qcom: gdsc: Add GDSCs in msm8996 GCC") Fixes: 7e824d50 ("clk: qcom: gdsc: Add mmcc gdscs for msm8996 family") Signed-off-by:
-
Srinivas Kandagatla authored
commit ce61966c upstream. This patch corrects the register offset for pcie2 pipe clock. Offset according to datasheet is 0x6e018 instead of 0x6e108. Signed-off-by:
-
Eric Anholt authored
commit 67615c58 upstream. If the firmware had set up a clock to source from PLLC, go along with it. But if we're looking for a new parent, we don't want to switch it to PLLC because the firmware will force PLLC (and thus the AXI bus clock) to different frequencies during over-temp/under-voltage, without notification to Linux. On my system, this moves the Linux-enabled HDMI state machine and DSI1 escape clock over to plld_per from pllc_per. EMMC still ends up on pllc_per, because the firmware had set it up to use that. Signed-off-by:
Eric Anholt <eric@anholt.net> Fixes: 41691b88 ("clk: bcm2835: Add support for programming the audio domain clocks") Acked-by:
Martin Sperl <kernel@martin.sperl.org> Signed-off-by:
-
Sascha Silbe authored
commit 6cd997db upstream. con3270 contains an optimisation that reduces the amount of data to be transmitted to the 3270 terminal by putting a Repeat to Address (RA) order into the data stream. The RA order itself takes up space, so con3270 only uses it if there's enough space left in the line buffer. Otherwise it just pads out the line manually. For lines that were _just_ short enough that the RA order still fit in the line buffer, the line was instead padded with an insufficient amount of spaces. This was caused by examining the size of the allocated line buffer rather than the length of the string to be displayed. For con3270_cline_end(), we just compare against the line length. For con3270_update_string() however that isn't available anymore, so we check whether the Repeat to Address order is present. Fixes: f51320a5 ("[PATCH] s390: new 3270 driver.") (tglx/history.git) Tested-by:
Jing Liu <liujbjl@linux.vnet.ibm.com> Tested-by:
Yang Chen <bjcyang@linux.vnet.ibm.com> Signed-off-by:
Sascha Silbe <silbe@linux.vnet.ibm.com> Signed-off-by:
Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by:
-
Sascha Silbe authored
commit c14f2aac upstream. con3270 contains an optimisation that reduces the amount of data to be transmitted to the 3270 terminal by putting a Repeat to Address (RA) order into the data stream. The RA order itself takes up space, so con3270 only uses it if there's enough space left in the line buffer. Otherwise it just pads out the line manually. For lines too long to include the RA order, one byte was left uninitialised. This was caused by an off-by-one bug in the loop that pads out the line. Since the buffer is allocated from a common pool, the single byte left uninitialised contained some previous buffer content. Usually this was just a space or some character (which can result in clutter but is otherwise harmless). Sometimes, however, it was a Repeat to Address order, messing up the entire screen layout and causing the display to send the entire buffer content on every keystroke. Fixes: f51320a5 ("[PATCH] s390: new 3270 driver.") (tglx/history.git) Reported-by:
-
Sebastian Ott authored
commit d53c51f2 upstream. Since commit 9f3d6d7a chsc_get_channel_measurement_chars is called with interrupts disabled during resume from hibernate. Since this function used spin_unlock_irq, interrupts have been enabled accidentally. Fix this by using the irqsave variant. Since we can't guarantee the IRQ-enablement state for all (future/ external) callers, change the locking in related functions to prevent similar bugs in the future. Fixes: 9f3d6d7a ("s390/cio: update measurement characteristics") Signed-off-by:
Sebastian Ott <sebott@linux.vnet.ibm.com> Reviewed-by:
Peter Oberparleiter <oberpar@linux.vnet.ibm.com> Signed-off-by:
-
Thomas Hellstrom authored
commit a1944030 upstream. When a view destruction command was present in the command stream, the view was validated to avoid a device error. That caused excessive and unnecessary validations of views, surfaces and mobs on view destruction. Replace this with a new relocation type that patches the view destruction command to a NOP if the view is not present in the device after the execbuf validation sequence. Also add checks for the member size of the vmw_res_relocation struct. Fixes sporadic command submission errors on google-earth exit. Reported-by:
Brian Paul <brianp@vmware.com> Signed-off-by:
Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by:
Sinclair Yeh <syeh@vmware.com> Signed-off-by:
-
Gerald Schaefer authored
commit eb03aa00 upstream. For every pfn aligned to minimum_order, dissolve_free_huge_pages() will call dissolve_free_huge_page() which takes the hugetlb spinlock, even if the page is not huge at all or a hugepage that is in-use. Improve this by doing the PageHuge() and page_count() checks already in dissolve_free_huge_pages() before calling dissolve_free_huge_page(). In dissolve_free_huge_page(), when holding the spinlock, those checks need to be revalidated. Link: http://lkml.kernel.org/r/20160926172811.94033-4-gerald.schaefer@de.ibm.comSigned-off-by:
Gerald Schaefer <gerald.schaefer@de.ibm.com> Acked-by:
Michal Hocko <mhocko@suse.com> Acked-by:
Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Cc: "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Mike Kravetz <mike.kravetz@oracle.com> Cc: "Aneesh Kumar K . V" <aneesh.kumar@linux.vnet.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Rui Teng <rui.teng@linux.vnet.ibm.com> Cc: Dave Hansen <dave.hansen@linux.intel.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by:
-
Gerald Schaefer authored
commit 082d5b6b upstream. In dissolve_free_huge_pages(), free hugepages will be dissolved without making sure that there are enough of them left to satisfy hugepage reservations. Fix this by adding a return value to dissolve_free_huge_pages() and checking h->free_huge_pages vs. h->resv_huge_pages. Note that this may lead to the situation where dissolve_free_huge_page() returns an error and all free hugepages that were dissolved before that error are lost, while the memory block still cannot be set offline. Fixes: c8721bbb ("mm: memory-hotplug: enable memory hotplug to handle hugepage") Link: http://lkml.kernel.org/r/20160926172811.94033-3-gerald.schaefer@de.ibm.comSigned-off-by:
-
Arnd Bergmann authored
commit 86c7e683 upstream. A workaround for a warning introduced a use of the NO_IRQ macro that should have been gone for a long time. It is clear from the code that the value cannot actually be used, but apparently there was a configuration at some point that caused a warning, so instead of just reverting that patch, this rearranges the code in a way that the warning cannot reappear. Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Fixes: 6ef41cf6 ("dmaengine :ipu: change ipu_irq_handler() to remove compile warning") Signed-off-by:
Vinod Koul <vinod.koul@intel.com> Signed-off-by:
-
Georges Savoundararadj authored
commit 06107359 upstream. bq->charger is initialized in bq24257_power_supply_init. Therefore, bq24257_power_supply_init should be called before the registration of the IRQ handler bq24257_irq_handler_thread that calls power_supply_changed(bq->charger). Signed-off-by:
Georges Savoundararadj <savoundg@gmail.com> Cc: Aurelien Chanot <chanot.a@gmail.com> Cc: Andreas Dannenberg <dannenberg@ti.com> Cc: Sebastian Reichel <sre@kernel.org> Cc: David Woodhouse <dwmw2@infradead.org> Fixes: 2219a935 ("power_supply: Add TI BQ24257 charger driver") Signed-off-by:
Sebastian Reichel <sre@kernel.org> Signed-off-by:
-
Maor Gottlieb authored
commit 7055a294 upstream. Fix multicast flow rule leak on adding unicast rule failure. Fixes: 038d2ef8 ('IB/mlx5: Add flow steering support') Signed-off-by:
Maor Gottlieb <maorg@mellanox.com> Signed-off-by:
Leon Romanovsky <leon@kernel.org> Signed-off-by:
Doug Ledford <dledford@redhat.com> Signed-off-by:
-
Dennis Dalessandro authored
commit 84b3adc2 upstream. There is no need to have a global qpt_mask as that does not support the multiple chip model which qib has. Instead rely on the value which exists already in the device data (dd). Fixes: 898fa52b "IB/qib: Remove qpn, qp tables and related variables from qib" Reviewed-by:
Mike Marciniszyn <mike.marciniszyn@intel.com> Signed-off-by:
Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by:
-
Mike Marciniszyn authored
commit eefa1d89 upstream. The __must_hold() is sufficent to correct the sparse context imbalance inside a function. Per Documentation/sparse.txt: __must_hold - The specified lock is held on function entry and exit. Fixes: Commit c0a67f6b ("IB/rdmavt: Annotate rvt_reset_qp()") Reviewed-by:
-
Mike Marciniszyn authored
commit 5a648dfa upstream. The call is misplaced in the reset calldown function and causes issues with lockdep assertions that are to be added. Fixes: Commit a2c2d608 ("staging/rdma/hfi1: Remove create_qp functionality") Reviewed-by:
Ira Weiny <ira.weiny@intel.com> Reviewed-by:
-
Nicholas Mc Guire authored
commit 9afe11e9 upstream. Private functions in ks_hostif.c can be declared static. Fixes: 13a9930d ("staging: ks7010: add driver from Nanonote extra-repository") Signed-off-by:
Nicholas Mc Guire <hofrat@osadl.org> Reviewed-by:
Wolfram Sang <wsa@the-dreams.de> Signed-off-by:
-
Nicholas Mc Guire authored
commit 9d29f14d upstream. wait_for_completion_interruptible_timeout return 0 on timeout and -ERESTARTSYS if interrupted. The check for !wait_for_completion_interruptible_timeout() would report an interrupt as timeout. Further, while HZ/50 will work most of the time it could fail for HZ < 50, so this is switched to msecs_to_jiffies(20). Fixes: 13a9930d ("staging: ks7010: add driver from Nanonote extra-repository") Signed-off-by:
-
Larry Finger authored
commit 1335a951 upstream. Commit fadbe0cd ("staging: rtl8188eu: Remove rtw_zmalloc(), wrapper for kzalloc()") changed all allocation calls to be GFP_KERNEL even though the original wrapper was testing to determine if the caller was in atomic mode. Most of the mistakes were corrected with commit 33dc85c3 ("staging: r8188eu: Fix scheduling while atomic error introduced in commit fadbe0cd"); however, two kzalloc calls were missed as the call only happens when the driver is shutting down. Fixes: fadbe0cd ("staging: rtl8188eu: Remove rtw_zmalloc(), wrapper for kzalloc()") Signed-off-by:
Larry Finger <Larry.Finger@lwfinger.net> Cc: navin patidar <navin.patidar@gmail.com> Signed-off-by:
-
Phil Turnbull authored
commit 992f9614 upstream. Commit 6fba39cf ("staging: sm750fb: use BIT macro for PANEL_DISPLAY_CTRL single-bit fields") accidentally changed the CLOCK_PHASE logic from '|=' to '=' which clears all the previously set bits. Fixes: 6fba39cf ("staging: sm750fb: use BIT macro for PANEL_DISPLAY_CTRL single-bit fields") Signed-off-by:
Phil Turnbull <phil.turnbull@oracle.com> Signed-off-by:
-
Arnd Bergmann authored
commit 4a236d01 upstream. The newly added Hi6220 Ion code fails to build when the ION_OF helpers are not present: drivers/staging/android/ion/hisilicon/hi6220_ion.o: In function `hi6220_ion_remove': hi6220_ion.c:(.text.hi6220_ion_remove+0x4c): undefined reference to `ion_destroy_platform_data' drivers/staging/android/ion/hisilicon/hi6220_ion.o: In function `hi6220_ion_probe': hi6220_ion.c:(.text.hi6220_ion_probe+0x5c): undefined reference to `ion_parse_dt' hi6220_ion.c:(.text.hi6220_ion_probe+0xf8): undefined reference to `ion_destroy_platform_data' This selects the symbol when needed. Fixes: 2b40182a ("staging: android: ion: Add ion driver for Hi6220 SoC platform") Signed-off-by:
Laura Abbott <labbott@redhat.com> Reviewed-by:
Sumit Semwal <sumit.semwal@linaro.org> Signed-off-by:
-
Christophe JAILLET authored
commit 281269f8 upstream. It is likely that checking the result of 'pm_runtime_set_active' is expected here. Fixes: f0e5f57d ("iio: light: us8152d: Add power management support") Signed-off-by:
Christophe JAILLET <christophe.jaillet@wanadoo.fr> Signed-off-by:
Jonathan Cameron <jic23@kernel.org> Signed-off-by:
-
Colin Ian King authored
commit 9d47964b upstream. The comparison for devnr limits is off-by-one, the current check allows 0 to AD5755_NUM_CHANNELS and the limit should be in fact 0 to AD5755_NUM_CHANNELS - 1. This can lead to an out of bounds write to pdata->dac[devnr]. Fix this by replacing > with >= on the comparison. Signed-off-by:
Colin Ian King <colin.king@canonical.com> Fixes: c9474599 ("iio: ad5755: add support for dt bindings") Signed-off-by:
-
Takashi Sakamoto authored
commit 43d443dc upstream. Sparse reports below warnings. bxt_da7219_max98357a.c:250:9: warning: obsolete array initializer, use C99 syntax bxt_da7219_max98357a.c:275:9: warning: obsolete array initializer, use C99 syntax bxt_da7219_max98357a.c:290:9: warning: obsolete array initializer, use C99 syntax bxt_da7219_max98357a.c:304:9: warning: obsolete array initializer, use C99 syntax bxt_da7219_max98357a.c:317:9: warning: obsolete array initializer, use C99 syntax There's no need to use obsoleted way. This commit fixes it. Fixes: 723bad3f (ASoC: Intel: Add Broxton-P Dialog Maxim machine driver) Signed-off-by:
Takashi Sakamoto <o-takashi@sakamocchi.jp> Acked-by:
Mark Brown <broonie@kernel.org> Signed-off-by:
-
Takashi Sakamoto authored
commit 5919a389 upstream. Sparse reports below warnings. bxt_rt298.c:275:9: warning: obsolete array initializer, use C99 syntax bxt_rt298.c:290:9: warning: obsolete array initializer, use C99 syntax bxt_rt298.c:304:9: warning: obsolete array initializer, use C99 syntax bxt_rt298.c:317:9: warning: obsolete array initializer, use C99 syntax bxt_rt298.c:331:9: warning: obsolete array initializer, use C99 syntax bxt_rt298.c:344:9: warning: obsolete array initializer, use C99 syntax bxt_rt298.c:357:9: warning: obsolete array initializer, use C99 syntax There's no need to use obsoleted way. This commit fixes it. Fixes: 76016322 (ASoC: Intel: Add Broxton-P machine driver) Signed-off-by:
-
Chen-Yu Tsai authored
commit a3930ed0 upstream. Commit d88429a6 ("ASoC: dapm: Add output driver widget") added the snd_soc_dapm_out_drv ID for the output driver widget, which is the same as the PGA widget, with a later power sequence number. Commit 19a2557b ("ASoC: dapm: Add kcontrol support for PGAs") then added kcontrol support for PGA widgets, but failed to account for output driver widgets. Attempts to use kcontrols with output driver widgets result in silent failures, with the developer having little idea about what went on. Add snd_soc_dapm_out_drv to the switch/case block under snd_soc_dapm_pga in dapm_create_or_share_kcontrol, since they are essentially the same. Fixes: 19a2557b (ASoC: dapm: Add kcontrol support for PGAs) Signed-off-by:
Chen-Yu Tsai <wens@csie.org> Signed-off-by:
-
Chen-Yu Tsai authored
commit 071133a2 upstream. The value for the second channel in _ENUM_DOUBLE (double channel) MUXs is not correctly updated, due to using the wrong bit shift. Use the correct bit shift, so both channels toggle together. Fixes: 3727b496 (ASoC: dapm: Consolidate MUXs and value MUXs) Signed-off-by:
Charles Keepax <ckeepax@opensource.wolfsonmicro.com> Signed-off-by:
-
Chen-Yu Tsai authored
commit 01ad5e7d upstream. If soc_dapm_read() fails, val will be uninitialized, and bogus values will be written later: ret = soc_dapm_read(dapm, reg, &val); val = (val >> shift) & mask; However, the compiler does not give a warning. Return on error before val is really used to avoid this. This is similar to the commit 69128316 ("ASoC: dapm: Fix uninitialized variable in snd_soc_dapm_get_enum_double()") Fixes: ce0fc93a (ASoC: Add DAPM support at the component level) Signed-off-by:
-
Wei Yongjun authored
commit 8ae3ea48 upstream. Fix to return error code -ENOMEM instead of 0 when failed to create widget, as done elsewhere in this function. Fixes: 8a978234 ("ASoC: topology: Add topology core") Signed-off-by:
Wei Yongjun <weiyj.lk@gmail.com> Signed-off-by:
-
Dave Gerlach authored
commit ad8529fd upstream. Currently omap-rng checks the return value of pm_runtime_get_sync and reports failure if anything is returned, however it should be checking if ret < 0 as pm_runtime_get_sync return 0 on success but also can return 1 if the device was already active which is not a failure case. Only values < 0 are actual failures. Fixes: 61dc0a44 ("hwrng: omap - Fix assumption that runtime_get_sync will always succeed") Signed-off-by:
Dave Gerlach <d-gerlach@ti.com> Signed-off-by:
Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by:
-
Wei Yongjun authored
commit 7514e368 upstream. Fix the retrn value check which testing the wrong variable in ccp_dmaengine_register(). Fixes: 58ea8abf ("crypto: ccp - Register the CCP as a DMA resource") Signed-off-by:
Wei Yongjun <weiyongjun1@huawei.com> Acked-by:
Gary R Hook <gary.hook@amd.com> Signed-off-by:
-
Ard Biesheuvel authored
commit ed4767d6 upstream. Since commit 8996eafd ("crypto: ahash - ensure statesize is non-zero"), all ahash drivers are required to implement import()/export(), and must have a non-zero statesize. Fix this for the ARM Crypto Extensions GHASH implementation. Fixes: 8996eafd ("crypto: ahash - ensure statesize is non-zero") Signed-off-by:
Ard Biesheuvel <ard.biesheuvel@linaro.org> Signed-off-by:
-
Romain Perier authored
commit 09951d83 upstream. So far, sub part of mv_cesa_int was responsible of dequeuing complete requests, then call the 'cleanup' operation on these reqs and call the crypto api callback 'complete'. The problem is that the transformation context 'ctx' is retrieved only once before the while loop. Which means that the wrong 'cleanup' operation might be called on the wrong type of cesa requests, it can lead to memory corruptions with this message: marvell-cesa f1090000.crypto: dma_pool_free cesa_padding, 5a5a5a5a/5a5a5a5a (bad dma) This commit fixes the issue, by updating the transformation context for each dequeued cesa request. Fixes: commit 85030c51 ("crypto: marvell - Add support for chai...") Signed-off-by:
Romain Perier <romain.perier@free-electrons.com> Signed-off-by:
-
Ondrej Mosnáček authored
commit 50d2e6dc upstream. The cipher block size for GCM is 16 bytes, and thus the CTR transform used in crypto_gcm_setkey() will also expect a 16-byte IV. However, the code currently reserves only 8 bytes for the IV, causing an out-of-bounds access in the CTR transform. This patch fixes the issue by setting the size of the IV buffer to 16 bytes. Fixes: 84c91152 ("[CRYPTO] gcm: Add support for async ciphers") Signed-off-by:
Ondrej Mosnacek <omosnacek@gmail.com> Signed-off-by:
-
Romain Perier authored
commit 57cfda1a upstream. Currently, in mv_cesa_{md5,sha1,sha256}_init creq->state is initialized before the call to mv_cesa_ahash_init. This is wrong because this function fills creq with zero by using memset, so its 'state' that contains the default DIGEST is overwritten. This commit fixes the issue by initializing creq->state just after the call to mv_cesa_ahash_init. Fixes: commit b0ef5106 ("crypto: marvell/cesa - initialize hash...") Signed-off-by:
-
Rajkumar Manoharan authored
commit 18f53fe0 upstream. commit 7a0adc83 ("ath10k: improve tx scheduling") is causing severe throughput drop in multi client mode. This issue is originally reported in veriwave setup with 50 clients with TCP downlink traffic. While increasing number of clients, the average throughput drops gradually. With 50 clients, the combined peak throughput is decreased to 98 Mbps whereas reverting given commit restored it to 550 Mbps. Processing txqs for every tx completion is causing overhead. Ideally for management frame tx completion, pending txqs processing can be avoided. The change partly reverts the commit "ath10k: improve tx scheduling". Processing pending txqs after all skbs tx completion will yeild enough room to burst tx frames. Fixes: 7a0adc83 ("ath10k: improve tx scheduling") Signed-off-by:
Rajkumar Manoharan <rmanohar@qti.qualcomm.com> Signed-off-by:
Kalle Valo <kvalo@qca.qualcomm.com> Signed-off-by:
-
Ashok Raj Nagarajan authored
commit 77eb3d69 upstream. When user requests for survey dump data, driver is providing wrong survey information. This information we sent is the survey data that we have collected during previous user request. This issue occurs because we request survey dump for wrong channel. With this change, we correctly display the correct and current survey information to userspace. Fixes: fa7937e3 ("ath10k: update bss channel survey information") Signed-off-by:
Ashok Raj Nagarajan <arnagara@qti.qualcomm.com> Signed-off-by:
-
