aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2019-06-05arm64: smp: Moved cpu_logical_map[] to smp.hFlorian Fainelli2-5/+6
asm/smp.h is included by linux/smp.h and some drivers, in particular irqchip drivers can access cpu_logical_map[] in order to perform SMP affinity tasks. Make arm64 consistent with other architectures here. Signed-off-by: Florian Fainelli <[email protected]> Signed-off-by: Will Deacon <[email protected]>
2019-06-05arm64: cpufeature: Fix missing ZFR0 in __read_sysreg_by_encoding()Dave Martin1-0/+1
In commit 06a916feca2b ("arm64: Expose SVE2 features for userspace"), new hwcaps are added that are detected via fields in the SVE-specific ID register ID_AA64ZFR0_EL1. In order to check compatibility of secondary cpus with the hwcaps established at boot, the cpufeatures code uses __read_sysreg_by_encoding() to read this ID register based on the sys_reg field of the arm64_elf_hwcaps[] table. This leads to a kernel splat if an hwcap uses an ID register that __read_sysreg_by_encoding() doesn't explicitly handle, as now happens when exercising cpu hotplug on an SVE2-capable platform. So fix it by adding the required case in there. Fixes: 06a916feca2b ("arm64: Expose SVE2 features for userspace") Signed-off-by: Dave Martin <[email protected]> Signed-off-by: Will Deacon <[email protected]>
2019-06-05usbip: Replace unused kvec array with single variable in vhci_send_cmd_unlink()Suwan Kim1-7/+5
vhci_send_cmd_unlink() declears kvec array of size 3 but it actually uses just one element of the array. So, remove kvec array and replace it with single kvec variable. Signed-off-by: Suwan Kim <[email protected]> Acked-by: Shuah Khan <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]>
2019-06-05usb: renesas_usbhs: remove controlling PWEN/EXTLP supportYoshihiro Shimoda2-5/+0
Controlling PWMEN/EXTLP (named as "has_otg") was supported in v3.2, but the last user (kzm9g) was removed by the commit 30f8925a57d8ad49 ("ARM: shmobile: Remove legacy board code for KZM-A9-GT"). So, this patch remove it. Signed-off-by: Yoshihiro Shimoda <[email protected]> Reviewed-by: Geert Uytterhoeven <[email protected]> Reviewed-by: Simon Horman <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]>
2019-06-05usb: renesas_usbhs: remove sudmac supportYoshihiro Shimoda2-6/+1
SUDMAC feature was supported in v3.10, but was never used by any platform. So, this patch removes it. Signed-off-by: Yoshihiro Shimoda <[email protected]> Reviewed-by: Geert Uytterhoeven <[email protected]> Reviewed-by: Simon Horman <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]>
2019-06-05usb: xhci: Add Clear_TT_BufferJim Lin3-1/+52
USB 2.0 specification chapter 11.17.5 says "as part of endpoint halt processing for full-/low-speed endpoints connected via a TT, the host software must use the Clear_TT_Buffer request to the TT to ensure that the buffer is not in the busy state". In our case, a full-speed speaker (ConferenceCam) is behind a high- speed hub (ConferenceCam Connect), sometimes once we get STALL on a request we may continue to get STALL with the folllowing requests, like Set_Interface. Here we invoke usb_hub_clear_tt_buffer() to send Clear_TT_Buffer request to the hub of the device for the following Set_Interface requests to the device to get ACK successfully. Signed-off-by: Jim Lin <[email protected]> Acked-by: Mathias Nyman <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]>
2019-06-05usb: Add devaddr in struct usb_deviceJim Lin3-1/+7
The Clear_TT_Buffer request sent to the hub includes the address of the LS/FS child device in wValue field. usb_hub_clear_tt_buffer() uses udev->devnum to set the address wValue. This won't work for devices connected to xHC. For other host controllers udev->devnum is the same as the address of the usb device, chosen and set by usb core. With xHC the controller hardware assigns the address, and won't be the same as devnum. Here we add devaddr in "struct usb_device" for usb_hub_clear_tt_buffer() to use. Signed-off-by: Jim Lin <[email protected]> Acked-by: Alan Stern <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]>
2019-06-05USB: Fix chipmunk-like voice when using Logitech C270 for recording audio.Marco Zatta1-0/+3
This patch fixes the chipmunk-like voice that manifets randomly when using the integrated mic of the Logitech Webcam HD C270. The issue was solved initially for this device by commit 2394d67e446b ("USB: add RESET_RESUME for webcams shown to be quirky") but it was then reintroduced by e387ef5c47dd ("usb: Add USB_QUIRK_RESET_RESUME for all Logitech UVC webcams"). This patch is to have the fix back. Signed-off-by: Marco Zatta <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]> Cc: stable <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]>
2019-06-05USB: usb-storage: Add new ID to ums-realtekKai-Heng Feng1-0/+5
There is one more Realtek card reader requires ums-realtek to work correctly. Add the device ID to support it. Signed-off-by: Kai-Heng Feng <[email protected]> Cc: stable <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]>
2019-06-05Merge tag 'gvt-fixes-2019-06-05' of https://github.com/intel/gvt-linux into ↵Joonas Lahtinen5-7/+46
drm-intel-fixes gvt-fixes-2019-06-05 - Fix i915 guest debug build for register command access (Weinan) - Fix guest ring state after execution for hangcheck (Xiaolin) - klocwork static check fixes (Alek) Signed-off-by: Joonas Lahtinen <[email protected]> From: Zhenyu Wang <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2019-06-05udmabuf: actually unmap the scatterlistLucas Stach1-0/+1
unmap_udmabuf fails to actually unmap the scatterlist, leaving dangling mappings around. Fixes: fbb0de795078 ("Add udmabuf misc device") Signed-off-by: Lucas Stach <[email protected]> Link: http://patchwork.freedesktop.org/patch/msgid/[email protected] Signed-off-by: Gerd Hoffmann <[email protected]>
2019-06-04net: fix indirect calls helpers for ptype list hooks.Paolo Abeni1-3/+3
As Eric noted, the current wrapper for ptype func hook inside __netif_receive_skb_list_ptype() has no chance of avoiding the indirect call: we enter such code path only for protocols other than ipv4 and ipv6. Instead we can wrap the list_func invocation. v1 -> v2: - use the correct fix tag Fixes: f5737cbadb7d ("net: use indirect calls helpers for ptype hook") Suggested-by: Eric Dumazet <[email protected]> Signed-off-by: Paolo Abeni <[email protected]> Acked-by: Edward Cree <[email protected]> Reviewed-by: Eric Dumazet <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2019-06-04net: ipvlan: Fix ipvlan device tso disabled while NETIF_F_IP_CSUM is setMiaohe Lin1-1/+1
There's some NICs, such as hinic, with NETIF_F_IP_CSUM and NETIF_F_TSO on but NETIF_F_HW_CSUM off. And ipvlan device features will be NETIF_F_TSO on with NETIF_F_IP_CSUM and NETIF_F_IP_CSUM both off as IPVLAN_FEATURES only care about NETIF_F_HW_CSUM. So TSO will be disabled in netdev_fix_features. For example: Features for enp129s0f0: rx-checksumming: on tx-checksumming: on tx-checksum-ipv4: on tx-checksum-ip-generic: off [fixed] tx-checksum-ipv6: on Fixes: a188222b6ed2 ("net: Rename NETIF_F_ALL_CSUM to NETIF_F_CSUM_MASK") Signed-off-by: Miaohe Lin <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2019-06-04scsi: smartpqi: unlock on error in pqi_submit_raid_request_synchronous()Dan Carpenter1-2/+4
We need to drop the "ctrl_info->sync_request_sem" lock before returning. Fixes: 6c223761eb54 ("smartpqi: initial commit of Microsemi smartpqi driver") Signed-off-by: Dan Carpenter <[email protected]> Acked-by: Don Brace <[email protected]> Signed-off-by: Martin K. Petersen <[email protected]>
2019-06-04scsi: ufs: Check that space was properly alloced in copy_query_responseAvri Altman1-1/+2
struct ufs_dev_cmd is the main container that supports device management commands. In the case of a read descriptor request, we assume that the proper space was allocated in dev_cmd to hold the returning descriptor. This is no longer true, as there are flows that doesn't use dev_cmd for device management requests, and was wrong in the first place. Fixes: d44a5f98bb49 (ufs: query descriptor API) Signed-off-by: Avri Altman <[email protected]> Reviewed-by: Alim Akhtar <[email protected]> Acked-by: Bean Huo <[email protected]> Signed-off-by: Martin K. Petersen <[email protected]>
2019-06-04udp: only choose unbound UDP socket for multicast when not in a VRFTim Beale1-2/+1
By default, packets received in another VRF should not be passed to an unbound socket in the default VRF. This patch updates the IPv4 UDP multicast logic to match the unicast VRF logic (in compute_score()), as well as the IPv6 mcast logic (in __udp_v6_is_mcast_sock()). The particular case I noticed was DHCP discover packets going to the 255.255.255.255 address, which are handled by __udp4_lib_mcast_deliver(). The previous code meant that running multiple different DHCP server or relay agent instances across VRFs did not work correctly - any server/relay agent in the default VRF received DHCP discover packets for all other VRFs. Fixes: 6da5b0f027a8 ("net: ensure unbound datagram socket to be chosen when not in a VRF") Signed-off-by: Tim Beale <[email protected]> Reviewed-by: David Ahern <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2019-06-04Merge branch 'net-tls-redo-the-RX-resync-locking'David S. Miller2-8/+22
Jakub Kicinski says: ==================== net/tls: redo the RX resync locking Take two of making sure we don't use a NULL netdev pointer for RX resync. This time using a bit and an open coded wait loop. v2: - fix build warning (DaveM). ==================== Signed-off-by: David S. Miller <[email protected]>
2019-06-04net/tls: replace the sleeping lock around RX resync with a bit lockJakub Kicinski2-6/+25
Commit 38030d7cb779 ("net/tls: avoid NULL-deref on resync during device removal") tried to fix a potential NULL-dereference by taking the context rwsem. Unfortunately the RX resync may get called from soft IRQ, so we can't use the rwsem to protect from the device disappearing. Because we are guaranteed there can be only one resync at a time (it's called from strparser) use a bit to indicate resync is busy and make device removal wait for the bit to get cleared. Note that there is a leftover "flags" field in struct tls_context already. Fixes: 4799ac81e52a ("tls: Add rx inline crypto offload") Signed-off-by: Jakub Kicinski <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2019-06-04Revert "net/tls: avoid NULL-deref on resync during device removal"Jakub Kicinski1-10/+5
This reverts commit 38030d7cb77963ba84cdbe034806e2b81245339f. Unfortunately the RX resync may get called from soft IRQ, so we can't take the rwsem to protect from the device disappearing. Signed-off-by: Jakub Kicinski <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2019-06-04block: aoe: no need to check return value of debugfs_create functionsGreg Kroah-Hartman1-14/+2
When calling debugfs functions, there is no need to ever check the return value. The function can work or not, but the code logic should never do something different based on this. Cc: "Ed L. Cashin" <[email protected]> Cc: [email protected] Cc: Omar Sandoval <[email protected]> Acked-by: Justin Sanders <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]> Signed-off-by: Jens Axboe <[email protected]>
2019-06-04net: dsa: sja1105: Fix link speed not working at 100 Mbps and belowVladimir Oltean1-16/+16
The hardware values for link speed are held in the sja1105_speed_t enum. However they do not increase in the order that sja1105_get_speed_cfg was iterating over them (basically from SJA1105_SPEED_AUTO - 0 - to SJA1105_SPEED_1000MBPS - 1 - skipping the other two). Another bug is that the code in sja1105_adjust_port_config relies on the fact that an invalid link speed is detected by sja1105_get_speed_cfg and returned as -EINVAL. However storing this into an enum that only has positive members will cast it into an unsigned value, and it will miss the negative check. So take the simplest approach and remove the sja1105_get_speed_cfg function and replace it with a simple switch-case statement. Fixes: 8aa9ebccae87 ("net: dsa: Introduce driver for NXP SJA1105 5-port L2 switch") Signed-off-by: Vladimir Oltean <[email protected]> Suggested-by: Andrew Lunn <[email protected]> Reviewed-by: Andrew Lunn <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2019-06-04net: phylink: avoid reducing support maskRussell King1-4/+9
Avoid reducing the support mask as a result of the interface type selected for SFP modules, or when setting the link settings through ethtool - this should only change when the supported link modes of the hardware combination change. Signed-off-by: Russell King <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2019-06-05scripts/checkstack.pl: Fix arm64 wrong or unknown architectureGeorge G. Davis1-1/+1
The following error occurs for the `make ARCH=arm64 checkstack` case: aarch64-linux-gnu-objdump -d vmlinux $(find . -name '*.ko') | \ perl ./scripts/checkstack.pl arm64 wrong or unknown architecture "arm64" As suggested by Masahiro Yamada, fix the above error using regular expressions in the same way it was fixed for the `ARCH=x86` case via commit fda9f9903be6 ("scripts/checkstack.pl: automatically handle 32-bit and 64-bit mode for ARCH=x86"). Suggested-by: Masahiro Yamada <[email protected]> Signed-off-by: George G. Davis <[email protected]> Signed-off-by: Masahiro Yamada <[email protected]>
2019-06-05kbuild: tar-pkg: enable communication with jobserverTrevor Bourget1-1/+1
The buildtar script might want to invoke a make, so tell the parent make to pass the jobserver token pipe to the subcommand by prefixing the command with a +. This addresses the issue seen here: /bin/sh ../scripts/package/buildtar tar-pkg make[3]: warning: jobserver unavailable: using -j1. Add '+' to parent make rule. See https://www.gnu.org/software/make/manual/html_node/Job-Slots.html for more information. Signed-off-by: Trevor Bourget <[email protected]> Signed-off-by: Masahiro Yamada <[email protected]>
2019-06-05kconfig: tests: fix recursive inclusion unit testMasahiro Yamada1-3/+3
Adding SPDX license identifier is pretty safe; however, here is one exception. Since commit ec8f24b7faaf ("treewide: Add SPDX license identifier - Makefile/Kconfig"), "make testconfig" would not pass. When Kconfig detects a circular file inclusion, it displays error messages with a file name and a line number prefixed to each line. The unit test checks if Kconfig emits the error messages correctly (this also checks the line number correctness). Now that the test input has the SPDX license identifier at the very top, the line numbers in the expected stderr should be incremented by 1. Fixes: ec8f24b7faaf ("treewide: Add SPDX license identifier - Makefile/Kconfig") Signed-off-by: Masahiro Yamada <[email protected]>
2019-06-05kbuild: teach kselftest-merge to find nested config filesDan Rue1-3/+2
Current implementation of kselftest-merge only finds config files that are one level deep using `$(srctree)/tools/testing/selftests/*/config`. Often, config files are added in nested directories, and do not get picked up by kselftest-merge. Use `find` to catch all config files under `$(srctree)/tools/testing/selftests` instead. Signed-off-by: Dan Rue <[email protected]> Signed-off-by: Masahiro Yamada <[email protected]>
2019-06-04nvmet: fix data_len to 0 for bdev-backed write_zeroesMinwoo Im1-0/+1
The WRITE ZEROES command has no data transfer so that we need to initialize the struct (nvmet_req *req)->data_len to 0x0. While (nvmet_req *req)->transfer_len is initialized in nvmet_req_init(), data_len will be initialized by nowhere which might cause the failure with status code NVME_SC_SGL_INVALID_DATA | NVME_SC_DNR randomly. It's because nvmet_req_execute() checks like: if (unlikely(req->data_len != req->transfer_len)) { req->error_loc = offsetof(struct nvme_common_command, dptr); nvmet_req_complete(req, NVME_SC_SGL_INVALID_DATA | NVME_SC_DNR); } else req->execute(req); This patch fixes req->data_len not to be a randomly assigned by initializing it to 0x0 when preparing the command in nvmet_bdev_parse_io_cmd(). nvmet_file_parse_io_cmd() which is for file-backed I/O has already initialized the data_len field to 0x0, though. Cc: Christoph Hellwig <[email protected]> Cc: Sagi Grimberg <[email protected]> Cc: Chaitanya Kulkarni <[email protected]> Signed-off-by: Minwoo Im <[email protected]> Reviewed-by: Chaitanya Kulkarni <[email protected]> Reviewed-by: Christoph Hellwig <[email protected]> Signed-off-by: Sagi Grimberg <[email protected]>
2019-06-04MAINTAINERS: Hand over skd maintainershipBart Van Assche1-1/+1
Since I do no longer have access to any STEC SSDs, hand over maintainership of the skd driver to Damien who still has access to STEC SSDs. Cc: Bart Van Assche <[email protected]> Cc: Christoph Hellwig <[email protected]> Cc: Hannes Reinecke <[email protected]> Acked-by: Damien Le Moal <[email protected]> Signed-off-by: Bart Van Assche <[email protected]> Signed-off-by: Jens Axboe <[email protected]>
2019-06-04ASoC: sun4i-i2s: Add offset to RX channel selectMarcus Cooper1-0/+4
Whilst testing the capture functionality of the i2s on the newer SoCs it was noticed that the recording was somewhat distorted. This was due to the offset not being set correctly on the receiver side. Signed-off-by: Marcus Cooper <[email protected]> Acked-by: Maxime Ripard <[email protected]> Acked-by: Chen-Yu Tsai <[email protected]> Signed-off-by: Mark Brown <[email protected]>
2019-06-04ASoC: sun4i-i2s: Fix sun8i tx channel offset maskMarcus Cooper1-1/+1
Although not causing any noticeable issues, the mask for the channel offset is covering too many bits. Signed-off-by: Marcus Cooper <[email protected]> Acked-by: Maxime Ripard <[email protected]> Acked-by: Chen-Yu Tsai <[email protected]> Signed-off-by: Mark Brown <[email protected]>
2019-06-04ASoC: max98090: remove 24-bit format support if RJ is 0Yu-Hsuan Hsu1-0/+16
The supported formats are S16_LE and S24_LE now. However, by datasheet of max98090, S24_LE is only supported when it is in the right justified mode. We should remove 24-bit format if it is not in that mode to avoid triggering error. Signed-off-by: Yu-Hsuan Hsu <[email protected]> Signed-off-by: Mark Brown <[email protected]>
2019-06-04ASoC: da7219: Fix build error without CONFIG_I2CYueHaibing1-1/+1
Fix gcc build error while CONFIG_I2C is not set sound/soc/codecs/da7219.c:2640:1: warning: data definition has no type or storage class module_i2c_driver(da7219_i2c_driver); ^~~~~~~~~~~~~~~~~ sound/soc/codecs/da7219.c:2640:1: error: type defaults to int in declaration of module_i2c_driver [-Werror=implicit-int] sound/soc/codecs/da7219.c:2640:1: warning: parameter names (without types) in function declaration sound/soc/codecs/da7219.c:2629:26: warning: da7219_i2c_driver defined but not used [-Wunused-variable] Reported-by: Hulk Robot <[email protected]> Fixes: 6d817c0e9fd7 ("ASoC: codecs: Add da7219 codec driver") Signed-off-by: YueHaibing <[email protected]> Signed-off-by: Mark Brown <[email protected]>
2019-06-04ASoC: SOF: Intel: hda: Fix COMPILE_TEST build errorYueHaibing1-0/+4
while building without PCI: sound/soc/sof/intel/hda.o: In function `hda_dsp_probe': hda.c:(.text+0x79c): undefined reference to `pci_ioremap_bar' hda.c:(.text+0x79c): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `pci_ioremap_bar' hda.c:(.text+0x7c4): undefined reference to `pci_ioremap_bar' hda.c:(.text+0x7c4): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `pci_ioremap_bar' Reported-by: Hulk Robot <[email protected]> Fixes: e13ef82a9ab8 ("ASoC: SOF: add COMPILE_TEST for PCI options") Signed-off-by: YueHaibing <[email protected]> Acked-by: Pierre-Louis Bossart <[email protected]> Signed-off-by: Mark Brown <[email protected]>
2019-06-04drm/arm/hdlcd: Allow a bit of clock toleranceRobin Murphy1-1/+2
On the Arm Juno platform, the HDLCD pixel clock is constrained to 250KHz resolution in order to avoid the tiny System Control Processor spending aeons trying to calculate exact PLL coefficients. This means that modes like my oddball 1600x1200 with 130.89MHz clock get rejected since the rate cannot be matched exactly. In practice, though, this mode works quite happily with the clock at 131MHz, so let's relax the check to allow a little bit of slop. Signed-off-by: Robin Murphy <[email protected]> Signed-off-by: Liviu Dudau <[email protected]>
2019-06-04drm/arm/hdlcd: Actually validate CRTC modesRobin Murphy1-6/+5
Rather than allowing any old mode through, then subsequently refusing unmatchable clock rates in atomic_check when it's too late to back out and pick a different mode, let's do that validation up-front where it will cause unsupported modes to be correctly pruned in the first place. This also eliminates an issue whereby a perceived clock rate of 0 would cause atomic disable to fail and prevent the module from being unloaded. Signed-off-by: Robin Murphy <[email protected]> Signed-off-by: Liviu Dudau <[email protected]>
2019-06-04drm/arm/mali-dp: Add a loop around the second set CVAL and try 5 timesWen He1-1/+12
This patch trying to fix monitor freeze issue caused by drm error 'flip_done timed out' on LS1028A platform. this set try is make a loop around the second setting CVAL and try like 5 times before giveing up. Signed-off-by: Wen He <[email protected]> Signed-off-by: Liviu Dudau <[email protected]>
2019-06-04drm/komeda: fixing of DMA mapping sg segment warningLowry Li (Arm Technology China)2-0/+6
Fixing the DMA mapping sg segment warning, which shows "DMA-API: mapping sg segment longer than device claims to support [len=921600] [max=65536]". Fixed by setting the max segment size at Komeda driver. This patch depends on: - https://patchwork.freedesktop.org/series/54448/ - https://patchwork.freedesktop.org/series/54449/ - https://patchwork.freedesktop.org/series/54450/ - https://patchwork.freedesktop.org/series/58976/ Changes since v1: - Adds member description - Adds patch denpendency in the comment Signed-off-by: Lowry Li (Arm Technology China) <[email protected]> Reviewed-by: James Qian Wang (Arm Technology China) <[email protected]> Reviewed-by: Ayan Kumar Halder <[email protected]> Signed-off-by: Liviu Dudau <[email protected]>
2019-06-04habanalabs: Read upper bits of trace buffer from RWPHITomer Tayar1-2/+12
The trace buffer address is 40 bits wide. The end of the buffer is set in the RWP register (lower 32 bits), and in the RWPHI register (upper 8 bits). Currently only the lower 32 bits are read, and this patch fixes it and concatenates the upper 8 bits to the output address. Signed-off-by: Tomer Tayar <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2019-06-04drm: don't block fb changes for async plane updatesHelen Koike2-10/+20
In the case of a normal sync update, the preparation of framebuffers (be it calling drm_atomic_helper_prepare_planes() or doing setups with drm_framebuffer_get()) are performed in the new_state and the respective cleanups are performed in the old_state. In the case of async updates, the preparation is also done in the new_state but the cleanups are done in the new_state (because updates are performed in place, i.e. in the current state). The current code blocks async udpates when the fb is changed, turning async updates into sync updates, slowing down cursor updates and introducing regressions in igt tests with errors of type: "CRITICAL: completed 97 cursor updated in a period of 30 flips, we expect to complete approximately 15360 updates, with the threshold set at 7680" Fb changes in async updates were prevented to avoid the following scenario: - Async update, oldfb = NULL, newfb = fb1, prepare fb1, cleanup fb1 - Async update, oldfb = fb1, newfb = fb2, prepare fb2, cleanup fb2 - Non-async commit, oldfb = fb2, newfb = fb1, prepare fb1, cleanup fb2 (wrong) Where we have a single call to prepare fb2 but double cleanup call to fb2. To solve the above problems, instead of blocking async fb changes, we place the old framebuffer in the new_state object, so when the code performs cleanups in the new_state it will cleanup the old_fb and we will have the following scenario instead: - Async update, oldfb = NULL, newfb = fb1, prepare fb1, no cleanup - Async update, oldfb = fb1, newfb = fb2, prepare fb2, cleanup fb1 - Non-async commit, oldfb = fb2, newfb = fb1, prepare fb1, cleanup fb2 Where calls to prepare/cleanup are balanced. Cc: <[email protected]> # v4.14+ Fixes: 25dc194b34dd ("drm: Block fb changes for async plane updates") Suggested-by: Boris Brezillon <[email protected]> Signed-off-by: Helen Koike <[email protected]> Reviewed-by: Boris Brezillon <[email protected]> Reviewed-by: Nicholas Kazlauskas <[email protected]> Signed-off-by: Boris Brezillon <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2019-06-04drm/vc4: fix fb references in async updateHelen Koike1-1/+1
Async update callbacks are expected to set the old_fb in the new_state so prepare/cleanup framebuffers are balanced. Calling drm_atomic_set_fb_for_plane() (which gets a reference of the new fb and put the old fb) is not required, as it's taken care by drm_mode_cursor_universal() when calling drm_atomic_helper_update_plane(). Cc: <[email protected]> # v4.19+ Fixes: 539c320bfa97 ("drm/vc4: update cursors asynchronously through atomic") Suggested-by: Boris Brezillon <[email protected]> Signed-off-by: Helen Koike <[email protected]> Reviewed-by: Boris Brezillon <[email protected]> Signed-off-by: Boris Brezillon <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2019-06-04drm/msm: fix fb references in async updateHelen Koike1-0/+4
Async update callbacks are expected to set the old_fb in the new_state so prepare/cleanup framebuffers are balanced. Cc: <[email protected]> # v4.14+ Fixes: 224a4c970987 ("drm/msm: update cursors asynchronously through atomic") Suggested-by: Boris Brezillon <[email protected]> Signed-off-by: Helen Koike <[email protected]> Acked-by: Rob Clark <[email protected]> Signed-off-by: Boris Brezillon <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2019-06-04drm/amd: fix fb references in async updateHelen Koike1-2/+1
Async update callbacks are expected to set the old_fb in the new_state so prepare/cleanup framebuffers are balanced. Calling drm_atomic_set_fb_for_plane() (which gets a reference of the new fb and put the old fb) is not required, as it's taken care by drm_mode_cursor_universal() when calling drm_atomic_helper_update_plane(). Cc: <[email protected]> # v4.20+ Fixes: 674e78acae0d ("drm/amd/display: Add fast path for cursor plane updates") Suggested-by: Boris Brezillon <[email protected]> Signed-off-by: Helen Koike <[email protected]> Reviewed-by: Nicholas Kazlauskas <[email protected]> Signed-off-by: Boris Brezillon <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2019-06-04drm/rockchip: fix fb references in async updateHelen Koike1-25/+26
In the case of async update, modifications are done in place, i.e. in the current plane state, so the new_state is prepared and the new_state is cleaned up (instead of the old_state, unlike what happens in a normal sync update). To cleanup the old_fb properly, it needs to be placed in the new_state in the end of async_update, so cleanup call will unreference the old_fb correctly. Also, the previous code had a: plane_state = plane->funcs->atomic_duplicate_state(plane); ... swap(plane_state, plane->state); if (plane->state->fb && plane->state->fb != new_state->fb) { ... } Which was wrong, as the fb were just assigned to be equal, so this if statement nevers evaluates to true. Another details is that the function drm_crtc_vblank_get() can only be called when vop->is_enabled is true, otherwise it has no effect and trows a WARN_ON(). Calling drm_atomic_set_fb_for_plane() (which get a referent of the new fb and pus the old fb) is not required, as it is taken care by drm_mode_cursor_universal() when calling drm_atomic_helper_update_plane(). Fixes: 15609559a834 ("drm/rockchip: update cursors asynchronously through atomic.") Cc: <[email protected]> # v4.20+ Signed-off-by: Helen Koike <[email protected]> Signed-off-by: Boris Brezillon <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2019-06-03xen-blkfront: switch kcalloc to kvcalloc for large array allocationRoger Pau Monne1-19/+19
There's no reason to request physically contiguous memory for those allocations. [boris: added CC to stable] Cc: [email protected] Reported-by: Ian Jackson <[email protected]> Signed-off-by: Roger Pau Monné <[email protected]> Reviewed-by: Juergen Gross <[email protected]> Acked-by: Konrad Rzeszutek Wilk <[email protected]> Signed-off-by: Boris Ostrovsky <[email protected]>
2019-06-04drm/mediatek: call mtk_dsi_stop() after mtk_drm_crtc_atomic_disable()Hsin-Yi Wang1-1/+9
mtk_dsi_stop() should be called after mtk_drm_crtc_atomic_disable(), which needs ovl irq for drm_crtc_wait_one_vblank(), since after mtk_dsi_stop() is called, ovl irq will be disabled. If drm_crtc_wait_one_vblank() is called after last irq, it will timeout with this message: "vblank wait timed out on crtc 0". This happens sometimes when turning off the screen. In drm_atomic_helper.c#disable_outputs(), the calling sequence when turning off the screen is: 1. mtk_dsi_encoder_disable() --> mtk_output_dsi_disable() --> mtk_dsi_stop(); /* sometimes make vblank timeout in atomic_disable */ --> mtk_dsi_poweroff(); 2. mtk_drm_crtc_atomic_disable() --> drm_crtc_wait_one_vblank(); ... --> mtk_dsi_ddp_stop() --> mtk_dsi_poweroff(); mtk_dsi_poweroff() has reference count design, change to make mtk_dsi_stop() called in mtk_dsi_poweroff() when refcount is 0. Fixes: 0707632b5bac ("drm/mediatek: update DSI sub driver flow for sending commands to panel") Signed-off-by: Hsin-Yi Wang <[email protected]> Signed-off-by: CK Hu <[email protected]>
2019-06-04drm/mediatek: clear num_pipes when unbind driverHsin-Yi Wang1-0/+1
num_pipes is used for mutex created in mtk_drm_crtc_create(). If we don't clear num_pipes count, when rebinding driver, the count will be accumulated. From mtk_disp_mutex_get(), there can only be at most 10 mutex id. Clear this number so it starts from 0 in every rebind. Fixes: 119f5173628a ("drm/mediatek: Add DRM Driver for Mediatek SoC MT8173.") Signed-off-by: Hsin-Yi Wang <[email protected]> Signed-off-by: CK Hu <[email protected]>
2019-06-04drm/mediatek: call drm_atomic_helper_shutdown() when unbinding driverHsin-Yi Wang1-0/+1
shutdown all CRTC when unbinding drm driver. Fixes: 119f5173628a ("drm/mediatek: Add DRM Driver for Mediatek SoC MT8173.") Signed-off-by: Hsin-Yi Wang <[email protected]> Signed-off-by: CK Hu <[email protected]>
2019-06-04drm/mediatek: unbind components in mtk_drm_unbind()Hsin-Yi Wang1-5/+1
Unbinding components (i.e. mtk_dsi and mtk_disp_ovl/rdma/color) will trigger master(mtk_drm)'s .unbind(), and currently mtk_drm's unbind won't actually unbind components. During the next bind, mtk_drm_kms_init() is called, and the components are added back. .unbind() should call mtk_drm_kms_deinit() to unbind components. And since component_master_del() in .remove() will trigger .unbind(), which will also unregister device, it's fine to remove original functions called here. Fixes: 119f5173628a ("drm/mediatek: Add DRM Driver for Mediatek SoC MT8173.") Signed-off-by: Hsin-Yi Wang <[email protected]> Signed-off-by: CK Hu <[email protected]>
2019-06-04drm/mediatek: fix unbind functionsHsin-Yi Wang1-0/+2
detatch panel in mtk_dsi_destroy_conn_enc(), since .bind will try to attach it again. Fixes: 2e54c14e310f ("drm/mediatek: Add DSI sub driver") Signed-off-by: Hsin-Yi Wang <[email protected]> Signed-off-by: CK Hu <[email protected]>
2019-06-03net: sfp: read eeprom in maximum 16 byte incrementsRussell King1-4/+20
Some SFP modules do not like reads longer than 16 bytes, so read the EEPROM in chunks of 16 bytes at a time. This behaviour is not specified in the SFP MSAs, which specifies: "The serial interface uses the 2-wire serial CMOS E2PROM protocol defined for the ATMEL AT24C01A/02/04 family of components." and "As long as the SFP+ receives an acknowledge, it shall serially clock out sequential data words. The sequence is terminated when the host responds with a NACK and a STOP instead of an acknowledge." We must avoid breaking a read across a 16-bit quantity in the diagnostic page, thankfully all 16-bit quantities in that page are naturally aligned. Signed-off-by: Russell King <[email protected]> Reviewed-by: Andrew Lunn <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.