hyesoo.yu [Mon, 24 Dec 2018 06:32:06 +0000 (15:32 +0900)]
[RAMEN9610-10376] g2d: add blocking interface by module param
If the jobs are queued larger than max_queued set by
module param or there is not free task on list,
the user request of task could be returned with -EBUSY
immediately, or wait until the number of queued job
is less than max_queued and free list is not empty.
Change-Id: I71dde2620830ea7e4426c86d70b7cd8eba9af952
Signed-off-by: hyesoo.yu <hyesoo.yu@samsung.com>
Shinkyu Park [Wed, 2 Jan 2019 11:59:04 +0000 (20:59 +0900)]
[RAMEN9610-10369][ERD][9610]dtsi: disable USI in SHUB block
[Issue#]
There is an issue when accessing USI in SHUB block when SHUB is
in Low-Power Mode.
[Description]
Disable USI control in SHUB block.
Platform Development Team
Shinkyu Park (shinkyu.park@samsung.com)
Change-Id: I3f5c4b28ebd2d1f2197288096ec411692c068d82
Youngmin Nam [Mon, 19 Nov 2018 05:41:06 +0000 (14:41 +0900)]
[RAMEN9610-10369][COMMON] pinctrl: samsung: Disable SHUB interrupt init, suspend/resume
Change-Id: I83502ec7abced0beaa79876d033287d0ae3bf044
Signed-off-by: Youngmin Nam <youngmin.nam@samsung.com>
JaeHun Jung [Thu, 3 Jan 2019 05:02:25 +0000 (14:02 +0900)]
[RAMEN9610-10347] Subject: [PATCH] ANDROID: sdcardfs: Don't use OVERRIDE_CRED macro
The macro hides some control flow, making it easier
to run into bugs.
bug:
111642636
Change-Id: I37ec207c277d97c4e7f1e8381bc9ae743ad78435
Reported-by: Jann Horn <jannh@google.com>
Signed-off-by: Daniel Rosenberg <drosen@google.com>
JaeHun Jung [Thu, 3 Jan 2019 05:27:49 +0000 (14:27 +0900)]
[RAMEN9610-10347] Subject: [PATCH] ANDROID: sdcardfs: Change current->fs under lock
bug:
111641492
Change-Id: I79e9894f94880048edaf0f7cfa2d180f65cbcf3b
Reported-by: Jann Horn <jannh@google.com>
Signed-off-by: Daniel Rosenberg <drosen@google.com>
Taekki Kim [Thu, 3 Jan 2019 04:20:43 +0000 (13:20 +0900)]
[RAMEN9610-10302][Common][9610] bts: remove pm_qos for int freq
The INT domain of Exynos9610 does not have a shared bus.
So, the bts driver does not need to control the INT frequency.
Change-Id: I38f6a80c9287ecf25c33f795a74fe1f0c0d89d5d
Signed-off-by: Taekki Kim <taekki.kim@samsung.com>
Hyeonseong Gil [Mon, 13 Aug 2018 05:16:30 +0000 (14:16 +0900)]
[RAMEN9610-10216][COMMON] thermal: gpu_cooling: Modify GPU Frequency unit to KHz
GPU API was modified to use KHz frequency.
Change-Id: I0870cccf51cffd3f7b3b42f94ce7a245e2f6cffe
Signed-off-by: Hyeonseong Gil <hs.gil@samsung.com>
Jiyoung Jeong [Wed, 2 Jan 2019 07:34:20 +0000 (16:34 +0900)]
[RAMEN9610-10259][COMMON][MODAP-13581] drivers: modem_if: change QoS setting order for bus bandwidth
before: CPU->MIF->INT
after: MIF->INT->CPU
Change-Id: I04fabaf9d7c9b04f73ab28a7bb1397d1283005be
Signed-off-by: Jiyoung Jeong <ji_0.jeong@samsung.com>
ChiHun Won [Mon, 31 Dec 2018 02:22:07 +0000 (11:22 +0900)]
[RAMEN9610-10169] fbdev: dpu20: added dqe hsc full pixel num
Change-Id: I89c38cfa93a2247cfdfa844438cd129c70b8175c
Signed-off-by: ChiHun Won <chihun.won@samsung.com>
Hyunwoong Kim [Wed, 7 Mar 2018 08:34:34 +0000 (17:34 +0900)]
[RAMEN9610-10167][COMMON] ASoC: abox: Save abox DRAM memory
ABOX reserved memory for saving DRAM was removed.
But dumped DRAM data is important for working out a problem.
This patch allocates actual memory region for saving abox DRAM memory.
Change-Id: Id33aa8c4ce77916e63589b29e8fd7384d07d9432
Signed-off-by: Hyunwoong Kim <khw0178.kim@samsung.com>
Hyunwoong Kim [Wed, 12 Dec 2018 01:10:09 +0000 (10:10 +0900)]
[RAMEN9610-10167][COMMON] ASoC: abox: remove unused SRAM feature
Change-Id: I8021188df50533aeaae13213f8ec9c2e6ac7a921
Signed-off-by: Hyunwoong Kim <khw0178.kim@samsung.com>
kiwoo Kwon [Fri, 28 Dec 2018 03:58:35 +0000 (12:58 +0900)]
[RAMEN9610-10136][MODAP-13272][NFC] dts: add nfc pd pin config
Change-Id: I4835105e52fc2443612e06de6f5cabec24da25e3
Signed-off-by: kiwoo Kwon <kiwo.kwon@samsung.com>
ChiHun Won [Thu, 27 Dec 2018 01:32:59 +0000 (10:32 +0900)]
[RAMEN9610-10033] fbdev: dpu20: added exception for checking diagonal matrix
Change-Id: If3dceec7047aaf537677656ec532d1e23f2c5483
Signed-off-by: ChiHun Won <chihun.won@samsung.com>
Ayoung Sim [Tue, 18 Dec 2018 04:48:15 +0000 (13:48 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: fix the print type
Change-Id: I924647684da96759cfb94d35e7e5ab5293bdc9e6
Signed-off-by: Ayoung Sim <a.sim@samsung.com>
Sunyoung Kang [Tue, 11 Dec 2018 00:21:19 +0000 (09:21 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: fix the 10bit memtype for DRM
When DRM playback, the 8+2 format is only used for 10bit.
Change-Id: I6613ba128db77178fa4b302c3ec78fbc17fe237b
Signed-off-by: Sunyoung Kang <sy0816.kang@samsung.com>
Sunyoung Kang [Thu, 6 Dec 2018 06:03:11 +0000 (15:03 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: use smc call for MMCache flush
The MFC driver can flush only normal data. But the secure OS
can flush the all normal and secure data in the MMCache.
This changes to use smc call to flush MMCache.
Change-Id: I92f6905d83e6a48b780b6eac8a42387daf17b01b
Signed-off-by: Sunyoung Kang <sy0816.kang@samsung.com>
Sunyoung Kang [Thu, 6 Dec 2018 06:45:03 +0000 (15:45 +0900)]
[RAMEN9610-10029][COMMON] smc: add the MMCACHE_OPERATION command
This adds MMCACHE_OPERATION definition for MMCache flush.
Change-Id: Ic09c11f7ff1ff5157a5006e7330488fe53addba6
Signed-off-by: Sunyoung Kang <sy0816.kang@samsung.com>
Ayoung Sim [Mon, 3 Dec 2018 07:26:56 +0000 (16:26 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: fix information leak
This patch doesn't print the pointers and use %pK
instead %p, if it is necessary to show.
Change-Id: I6bb2f272129d707c585bb1a98596e00442418999
Signed-off-by: Ayoung Sim <a.sim@samsung.com>
Ayoung Sim [Mon, 3 Dec 2018 06:50:43 +0000 (15:50 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: fix the prevent defect
33594, 33146: Other violation
24990, 24406: Dereference null return value
23952, 23923: Unsigned compared against 0
Change-Id: Ic5251e490d4ff5d55b027892d2973492076e5249
Signed-off-by: Ayoung Sim <a.sim@samsung.com>
Ayoung Sim [Thu, 29 Nov 2018 07:04:08 +0000 (16:04 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: set the 10bit memtype when INIT_BUF also
10bit format can be changed after SEQ_DONE from P010(P210) to 8+2 type.
However, original F/W couldn't get the memtype setting when INIT_BUF.
It was valid when only SEQ_START.
Because new F/W supports this when only SEQ_START but also INIT_BUF,
driver set the 10bit memtype when INIT_BUF time also.
Change-Id: I2ec92c0169950b28b23ae13fb3686f9207677203
Signed-off-by: Ayoung Sim <a.sim@samsung.com>
Sunyoung Kang [Tue, 20 Nov 2018 05:10:58 +0000 (14:10 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: fix the clock on sequence
When the clock on with drm instance, the sequence is changed as following:
cache flush (cmd 12) -> clock off -> set DRM base address
-> IP protection enable -> clock on
The base addr and protection control should be applied before clock enable
because the MFC core can run immediately when the clock enable.
Change-Id: Ib371a7b3079eb76515a38e977c76b78cc37f4c78
Signed-off-by: Sunyoung Kang <sy0816.kang@samsung.com>
Ayoung Sim [Tue, 20 Nov 2018 06:04:31 +0000 (15:04 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: handle MFC QoS atomically
MFC QoS should be controled atomically
because various thread can request qos_on.
Overlapped QoS calls(pm_qos_xxx_request())
cause of countless warning log messages.
Change-Id: Ifd74cf009fd263cd6732646ae92d4d46fb1d4ca0
Signed-off-by: Ayoung Sim <a.sim@samsung.com>
Ayoung Sim [Tue, 30 Oct 2018 01:07:40 +0000 (10:07 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: use the real 10bit mem type
Even if pdata->P010_decoding is set, user can use the 8+2 format
using with s_fmt(ex: V4L2_PIX_FMT_NV12M_S10B).
Driver set the options according to real 10bit format
not project default 10bit format.
Change-Id: I52dc35ff02982d19978399f67bfb84dd2514ff5c
Signed-off-by: Ayoung Sim <a.sim@samsung.com>
Sunyoung Kang [Thu, 15 Nov 2018 00:58:40 +0000 (09:58 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: add memset only for firmware except code region
This adds memset '0' to the firmware memory except fw code region.
And also this change to do not use the memory clear by ION allocator
because the firmware performs the memory clear. Since the firmware can not
clear the stack area by itself, it is only cleared by the driver.
Change-Id: Ifea58165d562280b5459c39852f21f34501c4370
Signed-off-by: Sunyoung Kang <sy0816.kang@samsung.com>
hyesoo.yu [Thu, 15 Nov 2018 05:49:45 +0000 (14:49 +0900)]
[RAMEN9610-10029]ion: add ION_FLAG_NOZEROED
Change-Id: I27684371954c634b2c061d3d516e54d9077b14b2
Signed-off-by: hyesoo.yu <hyesoo.yu@samsung.com>
Sunyoung Kang [Tue, 23 Oct 2018 09:23:23 +0000 (18:23 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: fix the WAIT state for decoder DRC
In case of UTC, it calls g_fmt after streamoff so can't out from
WAIT_DPB_FLUSH state. This modified that it can wait both g_fmt and
streamoff regardless of the order.
When DRC is detected the wait_state set to (WAIT_G_FMT | WAIT_STOP).
If the g_fmt is called WAIT_G_FMT is cleared and
if the streamoff is called WAIT_STOP is cleared.
Change-Id: I90326d191de2ba164060673e8b124227fda3a397
Signed-off-by: Sunyoung Kang <sy0816.kang@samsung.com>
Ayoung Sim [Fri, 26 Oct 2018 00:12:40 +0000 (09:12 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: add the MFCINST_FINISHED status
MFCINST_FINISHED status has following meaning.
Instance has already been opened but can't do anything.
When new format settings come, encoding can be re-started
begin with SEQ_START.
Change-Id: I679f77541942db75db9ba76e936b88668ba25e45
Signed-off-by: Ayoung Sim <a.sim@samsung.com>
Sunyoung Kang [Thu, 25 Oct 2018 06:32:10 +0000 (15:32 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: add to check clock before register dump
This adds to check the clock and power state before register dump.
Change-Id: I315b6346c6175232a87453f603292fe8a547ff83
Signed-off-by: Sunyoung Kang <sy0816.kang@samsung.com>
Sunyoung Kang [Thu, 25 Oct 2018 00:05:38 +0000 (09:05 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: fix the potential memory leak
If user calls repeatedly s_fmt(), the roi_buffer will be re-allocated and
it can be memory leak. So this add to check if it is already allocated.
Change-Id: I0eea5f745c13630ce951f4c45109a01e92ddb7a9
Signed-off-by: Sunyoung Kang <sy0816.kang@samsung.com>
Sunyoung Kang [Wed, 24 Oct 2018 08:04:59 +0000 (17:04 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: fix Out-of-Bound defect
This adds to check the size of copy_from_user().
Change-Id: Icb869f2906881889305beea6b4fb3bbd9ef14f08
Signed-off-by: Sunyoung Kang <sy0816.kang@samsung.com>
Ayoung Sim [Mon, 1 Oct 2018 07:35:26 +0000 (16:35 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: fix the 10bit options
We do not need INIT_BUF_OPT_STRIDE_SIZE_ALIGN option when P010 format,
because this option should be only 8+2 format for G2D.
Dithering option is also needed only for 8+2 format.
Dithering option are intended to be applied
when the display device doesn't support 10bit.
Change-Id: I132bf84d569c4488bda5386e3e817cd3ad96d4d8
Signed-off-by: Ayoung Sim <a.sim@samsung.com>
Sunyoung Kang [Thu, 13 Sep 2018 07:17:20 +0000 (16:17 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: fix interlace processing when NAL_Q mode
The H264/VC1/MPEG2/MPEG4 codec can have interlace type.
But the only MPEG4 can continue to use NAL_Q mode
because it doesn't handle field unit.
The other codecs should stop NAL_Q mode and use NAL_START.
Change-Id: Ied5972b8f567d91290ca937efd25ae14f9ece19e
Signed-off-by: Sunyoung Kang <sy0816.kang@samsung.com>
Ayoung Sim [Mon, 17 Sep 2018 06:46:44 +0000 (15:46 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: fix wrong mem type check
When an error occures during header parsing,
the driver dumps the header stream.
In order to print the header stream, driver should get the KVA
when only the memory type of source buffer is DMA_BUF type.
This patch fixs this type condition.
Change-Id: I557bef77b54aae266e6d75e76814c506239d4260
Signed-off-by: Ayoung Sim <a.sim@samsung.com>
Ayoung Sim [Tue, 4 Sep 2018 06:40:27 +0000 (15:40 +0900)]
[RAMEN9610-10029][COMMON] media: mfc: support decoded order decoding
If the DECODING_ORDER_ENABLE is set, F/W returns
decoded information in same register.
frame type, status and addr are same whether decoded or display.
Change-Id: I92b1c8a21205258bce2a13262d59b6663476ff93
Signed-off-by: Ayoung Sim <a.sim@samsung.com>
Jang JeongHoon [Fri, 2 Mar 2018 07:14:47 +0000 (16:14 +0900)]
[RAMEN9610-10009][COMMON] soc: samsung: cal-if: Changed DLL freq from 360 to 104Mhz
Change-Id: I2954fc93ac52aecc837f64b94dc02345c1bff1b6
Signed-off-by: Jang JeongHoon <jnghn.jang@samsung.com>
Hosung Kim [Wed, 26 Dec 2018 11:45:38 +0000 (20:45 +0900)]
[RAMEN9610-9963][COMMON] arm64: dts: add dss reserved memory for user/eng
Change-Id: I4692d2d5d69a434b306c80ea57f5ccd33189f00d
Signed-off-by: Hosung Kim <hosung0.kim@samsung.com>
Hosung Kim [Wed, 26 Dec 2018 11:44:56 +0000 (20:44 +0900)]
[RAMEN9610-9963][COMMON] lib: dss: split to define reserve memory for user/eng
Change-Id: I66fb5689f214391f56e54a4fc1e62370d9f9aaee
Signed-off-by: Hosung Kim <hosung0.kim@samsung.com>
Hosung Kim [Wed, 26 Dec 2018 10:57:17 +0000 (19:57 +0900)]
[RAMEN9610-9963][COMMON] lib: dss: modify ramoops size in reserved mem
this commit modified ramoops size depend on build mode
eng mode and user, userdebug mode.
Change-Id: I790c1efc9003a18a5dcb93c41c28a682add5f077
Signed-off-by: Hosung Kim <hosung0.kim@samsung.com>
Hosung Kim [Wed, 26 Dec 2018 09:29:35 +0000 (18:29 +0900)]
[RAMEN9610-9963][COMMON] lib: Kconfig: fix to define DEBUG_SNAPSHOT_MINIMIZED_MODE
This define is depend on DEBUG_SNAPSHOT_USER_MODE.
If DEBUG_SNAPSHOT_USER_MODE is enabled, MINIMIZED_MODE should be
activated. This patch fixes it.
Change-Id: Iaa47922cd053a6f6ae4f1c31afbe4a2abb37804e
Signed-off-by: Hosung Kim <hosung0.kim@samsung.com>
Changki Kim [Thu, 22 Nov 2018 06:41:37 +0000 (15:41 +0900)]
[RAMEN9610-9963][COMMON] lib: dss: modified task state flag
Change-Id: Ic04847e5669a8a816c987bc0b2c654830c05ec6a
Signed-off-by: Changki Kim <changki.kim@samsung.com>
Hosung Kim [Fri, 14 Sep 2018 11:31:04 +0000 (20:31 +0900)]
[RAMEN9610-9963][COMMON] arm64: exynos9610_userdebug: replace to DEBUG_SNAPSHOT_USER_MODE
CONFIG_DEBUG_SNAPSHOT_MINIMIZED_MODE=y is replaced to
CONFIG_DEBUG_SNAPSHOT_USER_MODE
Change-Id: I13573697205d303d599593bfdefe78993bcf5435
Signed-off-by: Hosung Kim <hosung0.kim@samsung.com>
Changki Kim [Fri, 17 Aug 2018 06:43:00 +0000 (15:43 +0900)]
[RAMEN9610-9963][COMMON] lib: dss: move definition of DEBUG_LEVEL.
Change-Id: Ice038a6acc61c16f17a0ec5750be97d1ffec4cc4
Signed-off-by: Changki Kim <changki.kim@samsung.com>
Hyunki Koo [Thu, 16 Aug 2018 23:38:27 +0000 (08:38 +0900)]
[RAMEN9610-9963][COMMON] lib: dss: remove unused variable
Change-Id: I72488b43f5b54ce82b69ea3c59a1f9f3c06fd507
Signed-off-by: Hyunki Koo <hyunki00.koo@samsung.com>
Hosung Kim [Tue, 14 Aug 2018 12:10:08 +0000 (21:10 +0900)]
[RAMEN9610-9963][COMMON] lib: dss: fix to set debug level
This commit modified to set debug level more simple.
The debug level is able to set by ENG/USER image or
environment value of u-boot. No more device-tree
supported.
Change-Id: Ie3d1201591213321f2d864de9bec69d93cb36594
Signed-off-by: Hosung Kim <hosung0.kim@samsung.com>
Hosung Kim [Tue, 17 Jul 2018 04:13:24 +0000 (13:13 +0900)]
[RAMEN9610-9963][COMMON] lib: dss: add DEBUG_SNAPSHOT_USER_MODE and debug level
This commit adds DEBUG_SNAPSHOT_USER_MODE and debug level.
If DEBUG_SNAPSHOT_USER_MODE is enabled, DEBUG_SNAPSHOT_MINIMIZED_MODE
is enabled too and debug_level is set by low.
It means that DEBUG_SNAPSHOT_USER_MODE is static level(build time)
debugging, debug level is runtime debugging.
debug level is able to conrol in bootloader.
Change-Id: Idedb764b82d7bfcfd313774827b71517d92e6fae
Signed-off-by: Hosung Kim <hosung0.kim@samsung.com>
sanghwan park [Wed, 26 Dec 2018 01:49:37 +0000 (10:49 +0900)]
[RAMEN9610-9962][9610] vipx: reduce reserved memory for dram binary
Change-Id: I8e86a301f4e880050814eef209e3de996e2056a4
Signed-off-by: sanghwan park <senius.park@samsung.com>
Taekki Kim [Wed, 26 Dec 2018 10:42:19 +0000 (19:42 +0900)]
[RAMEN9610-9961][Common] soc: samsung: debug: fix build error
This patch fixes build error when BCMDBG disabled.
Change-Id: I4511d479142b37dc926a27acf470bbc8dcd8f21c
Signed-off-by: Taekki Kim <taekki.kim@samsung.com>
Taekki Kim [Wed, 26 Dec 2018 10:38:44 +0000 (19:38 +0900)]
[RAMEN9610-9961][Common][9610] arm64: configs: disable BCMDBG for user & userdebug
Change-Id: I118102481ab6e14eaa4a05aabbae2f4a4b897f5a
Signed-off-by: Taekki Kim <taekki.kim@samsung.com>
ShinHyung [Wed, 26 Dec 2018 10:06:42 +0000 (19:06 +0900)]
[RAMEN9610-9953][COMMON] arm64: dts: Decrease ABOX reserved memory
From 40MB to 4MB
Change-Id: Ic0a9607413e1c969d86d0f1875b829e695e4e0fd
Signed-off-by: ShinHyung <s47.kang@samsung.com>
Cho KyongHo [Wed, 26 Dec 2018 06:17:13 +0000 (15:17 +0900)]
[RAMEN9610-9924] android: ion: add limit to page pool of systemheap
Users may want to restrict the number of free pages in the page pool
because it causes memory profiling and tuning hard.
Change-Id: I154ffa5eb176b0dc2bad0b3eb5c5747cde3a0bd5
Signed-off-by: Cho KyongHo <pullip.cho@samsung.com>
Cho KyongHo [Mon, 24 Dec 2018 07:00:16 +0000 (16:00 +0900)]
[RAMEN9610-9921] arm64: dts: shrink swiotlb to 1MB in Exynos9610
Change-Id: Iddf1d53bdf515999a53aabca7bffd512a25e810e
Signed-off-by: Cho KyongHo <pullip.cho@samsung.com>
Yoojin Park [Mon, 24 Dec 2018 05:03:54 +0000 (14:03 +0900)]
[RAMEN9610-9865][9609] driver: gpu: Update Robusta2 Mali-GPU DDK from r10p0 to r16p0
Update DDK for new features in Robosta2
Change-Id: I1f7ad7454ec96994fa90b6af439e346654d512eb
Signed-off-by: Yoojin Park <yoojin1.park@samsung.com>
JaeHun Jung [Fri, 14 Dec 2018 11:06:43 +0000 (19:06 +0800)]
[RAMEN9610-9575][9610: UFS] Fix mainline bug fix
Change-Id: I35be24402d6762c9c56afa715a50e3be502e5086
Signed-off-by: JaeHun Jung <jh0801.jung@samsung.com>
ChiHun Won [Wed, 19 Dec 2018 02:17:43 +0000 (11:17 +0900)]
[RAMEN9610-9644] fbdev: dpu20: updated xml show for supporting various panel ddi
Change-Id: I3837e676ce1b771f13dbca487f8d6b25f977f515
Signed-off-by: ChiHun Won <chihun.won@samsung.com>
ChiHun Won [Wed, 19 Dec 2018 02:16:51 +0000 (11:16 +0900)]
[RAMEN9610-9644] fbdev: dpu20: added dsim ddi device type
Change-Id: I54c7e88dda5a9dc2cf988cc90f4efbeedb3e4b69
Signed-off-by: ChiHun Won <chihun.won@samsung.com>
Yoojin Park [Tue, 18 Dec 2018 04:55:51 +0000 (13:55 +0900)]
[RAMEN9610-9616][MODAP-11585] drivers: gpu: Fix ui not smooth
MIDCET-1863 : Don't evict NEXT atom on soft stop if it's from a different context
This commit changes the behaviour of the job manager on a soft stop
request. Before this commit, if the atom to stop was on the HEAD of a
job slot and an atom from a different context was on the NEXT slot, both
atoms would be removed. This was once required due to ordering
restrictions in an older job scheduler implementation.
It has the side-effect that if these were the only two contexts
attempting to schedule on this slot, then there was a chance that they
would be removed from the job manager and re-submitted in exactly the
same order. If the soft stop was due to the scheduler attempting to
schedule out the HEAD atom so that the NEXT atom could run, then this
would be entirely counter productive.
Change-Id: I9b85bf10b3d720f248f5976f24a4ecac9d5203bb
Signed-off-by: Yoojin Park <yoojin1.park@samsung.com>
hwangjae lee [Tue, 18 Dec 2018 04:29:07 +0000 (13:29 +0900)]
[RAMEN9610-9592][COMMON] fbdev: dpu: color format is modified in pan_display
In pan_display, 32bit color format is modified BGRA8888 to ABGR8888.
Because of Android recovery UI color format is ABGR8888.
Change-Id: I5741884bbe5df1d2fccb7988fb50d752f8f50e45
Signed-off-by: hwangjae lee <hj-yo.lee@samsung.com>
ChiHun Won [Mon, 17 Dec 2018 08:15:43 +0000 (17:15 +0900)]
[RAMEN9610-9571] fbdev: dpu20: update dqe drv for supporting tune result
Change-Id: Iebbe4a3043fd23e1dbfd854e4d826bb64506b9c0
Signed-off-by: ChiHun Won <chihun.won@samsung.com>
Kisang Lee [Sat, 15 Dec 2018 03:52:59 +0000 (12:52 +0900)]
[RAMEN9610-9517][COMMON] usb: host: re-init Runtime-PM values to fit PM balance
Change-Id: I5fdca72c85daebc717c96398fad1ea9aac39aef6
Signed-off-by: Kisang Lee <kisang80.lee@samsung.com>
Sangkyu Kim [Mon, 17 Dec 2018 06:35:22 +0000 (15:35 +0900)]
[RAMEN9610-9534][COMMON] cpufreq: modified parameter type for CFI
Change-Id: I277d2e3b12f9ce9bab15ca5ab3e563e6e6ea05dc
Signed-off-by: Sangkyu Kim <skwith.kim@samsung.com>
Soren Andersen [Thu, 29 Nov 2018 12:45:06 +0000 (13:45 +0100)]
[RAMEN9610-9516][9610] wlbt: Add memory barrier in BT IRQ handler
Added a memory barrier in the IRQ handler for BT,
to ensure that the irq bit is cleared before starting
to read the shared mailbox indexes.
Change-Id: Ic8c4f2bb9ab675d692ff818800012cc0733c632f
SCSC-Bug-Id: SSB-46925
Signed-off-by: Soren Andersen <s.andersen@samsung.com>
hwangjae lee [Wed, 12 Dec 2018 01:02:02 +0000 (10:02 +0900)]
[RAMEN9610-9498][COMMON] fbdev: dpu: dphy timing parameter is modified
Change-Id: I3752458a1464139754ee564be6d9b73b1c15859e
Signed-off-by: hwangjae lee <hj-yo.lee@samsung.com>
Wooyeon Kim [Thu, 13 Dec 2018 10:17:28 +0000 (19:17 +0900)]
[RAMEN9610-9493][HACK][9610] fimc-is2: c2 disable when camera operated
PR JIRA ID: CPR-501
Change-Id: I2f57ed76f3aeaecc0a04be853d429a7883d40c1f
Signed-off-by: Wooyeon Kim <wooy88.kim@samsung.com>
Hyeyeon Chung [Wed, 5 Dec 2018 10:59:27 +0000 (19:59 +0900)]
[RAMEN9610-9484][COMMON] scsi: rpmb: print information log when rpmb result is not zero
- Print result code in rpmb packet for debug information
- Remove code checking rpmb result and returning ufs status for consistency
Change-Id: Ib201e34958684a8b7f17a94c718f699e2745e9ae
Signed-off-by: Hyeyeon Chung <hyeon.chung@samsung.com>
Hyeyeon Chung [Fri, 7 Dec 2018 03:37:36 +0000 (12:37 +0900)]
[RAMEN9610-9484][COMMON] scsi: rpmb: fix bug of data lengh in memcopy
Change-Id: If1849061ab3fa3adc14d3a2bb3a43121721b362c
Signed-off-by: Hyeyeon Chung <hyeon.chung@samsung.com>
Nicolai Stange [Mon, 30 Oct 2017 23:15:54 +0000 (00:15 +0100)]
[RAMEN9610-9428] debugfs: defer debugfs_fsdata allocation to first usage
Currently, __debugfs_create_file allocates one struct debugfs_fsdata
instance for every file created. However, there are potentially many
debugfs file around, most of which are never touched by userspace.
Thus, defer the allocations to the first usage, i.e. to the first
debugfs_file_get().
A dentry's ->d_fsdata starts out to point to the "real", user provided
fops. After a debugfs_fsdata instance has been allocated (and the real
fops pointer has been moved over into its ->real_fops member),
->d_fsdata is changed to point to it from then on. The two cases are
distinguished by setting BIT(0) for the real fops case.
struct debugfs_fsdata's foremost purpose is to track active users and to
make debugfs_remove() block until they are done. Since no debugfs_fsdata
instance means no active users, make debugfs_remove() return immediately
in this case.
Take care of possible races between debugfs_file_get() and
debugfs_remove(): either debugfs_remove() must see a debugfs_fsdata
instance and thus wait for possible active users or debugfs_file_get() must
see a dead dentry and return immediately.
Make a dentry's ->d_release(), i.e. debugfs_release_dentry(), check whether
->d_fsdata is actually a debugfs_fsdata instance before kfree()ing it.
Similarly, make debugfs_real_fops() check whether ->d_fsdata is actually
a debugfs_fsdata instance before returning it, otherwise emit a warning.
The set of possible error codes returned from debugfs_file_get() has grown
from -EIO to -EIO and -ENOMEM. Make open_proxy_open() and full_proxy_open()
pass the -ENOMEM onwards to their callers.
Signed-off-by: Nicolai Stange <nicstange@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Nicolai Stange [Mon, 30 Oct 2017 23:15:53 +0000 (00:15 +0100)]
[RAMEN9610-9428] debugfs: call debugfs_real_fops() only after debugfs_file_get()
The current implementation of debugfs_real_fops() relies on a
debugfs_fsdata instance to be installed at ->d_fsdata.
With future patches introducing lazy allocation of these, this requirement
will be guaranteed to be fullfilled only inbetween a
debugfs_file_get()/debugfs_file_put() pair.
The full proxies' fops implemented by debugfs happen to be the only
offenders. Fix them up by moving their debugfs_real_fops() calls past those
to debugfs_file_get().
full_proxy_release() is special as it doesn't invoke debugfs_file_get() at
all. Leave it alone for now.
Signed-off-by: Nicolai Stange <nicstange@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Nicolai Stange [Mon, 30 Oct 2017 23:15:52 +0000 (00:15 +0100)]
[RAMEN9610-9428] debugfs: purge obsolete SRCU based removal protection
Purge the SRCU based file removal race protection in favour of the new,
refcount based debugfs_file_get()/debugfs_file_put() API.
Fixes:
49d200deaa68 ("debugfs: prevent access to removed files' private data")
Signed-off-by: Nicolai Stange <nicstange@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Nicolai Stange [Mon, 30 Oct 2017 23:15:50 +0000 (00:15 +0100)]
[RAMEN9610-9428] debugfs: convert to debugfs_file_get() and -put()
Convert all calls to the now obsolete debugfs_use_file_start() and
debugfs_use_file_finish() from the debugfs core itself to the new
debugfs_file_get() and debugfs_file_put() API.
Fixes:
49d200deaa68 ("debugfs: prevent access to removed files' private data")
Signed-off-by: Nicolai Stange <nicstange@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Nicolai Stange [Mon, 30 Oct 2017 23:15:49 +0000 (00:15 +0100)]
[RAMEN9610-9428] debugfs: debugfs_real_fops(): drop __must_hold sparse annotation
Currently, debugfs_real_fops() is annotated with a
__must_hold(&debugfs_srcu) sparse annotation.
With the conversion of the SRCU based protection of users against
concurrent file removals to a per-file refcount based scheme, this becomes
wrong.
Drop this annotation.
Signed-off-by: Nicolai Stange <nicstange@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Nicolai Stange [Mon, 30 Oct 2017 23:15:48 +0000 (00:15 +0100)]
[RAMEN9610-9428] debugfs: implement per-file removal protection
Since commit
49d200deaa68 ("debugfs: prevent access to removed files'
private data"), accesses to a file's private data are protected from
concurrent removal by covering all file_operations with a SRCU read section
and sychronizing with those before returning from debugfs_remove() by means
of synchronize_srcu().
As pointed out by Johannes Berg, there are debugfs files with forever
blocking file_operations. Their corresponding SRCU read side sections would
block any debugfs_remove() forever as well, even unrelated ones. This
results in a livelock. Because a remover can't cancel any indefinite
blocking within foreign files, this is a problem.
Resolve this by introducing support for more granular protection on a
per-file basis.
This is implemented by introducing an 'active_users' refcount_t to the
per-file struct debugfs_fsdata state. At file creation time, it is set to
one and a debugfs_remove() will drop that initial reference. The new
debugfs_file_get() and debugfs_file_put(), intended to be used in place of
former debugfs_use_file_start() and debugfs_use_file_finish(), increment
and decrement it respectively. Once the count drops to zero,
debugfs_file_put() will signal a completion which is possibly being waited
for from debugfs_remove().
Thus, as long as there is a debugfs_file_get() not yet matched by a
corresponding debugfs_file_put() around, debugfs_remove() will block.
Actual users of debugfs_use_file_start() and -finish() will get converted
to the new debugfs_file_get() and debugfs_file_put() by followup patches.
Fixes:
49d200deaa68 ("debugfs: prevent access to removed files' private data")
Reported-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Nicolai Stange <nicstange@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Nicolai Stange [Mon, 30 Oct 2017 23:15:47 +0000 (00:15 +0100)]
[RAMEN9610-9428] debugfs: add support for more elaborate ->d_fsdata
Currently, the user provided fops, "real_fops", are stored directly into
->d_fsdata.
In order to be able to store more per-file state and thus prepare for more
granular file removal protection, wrap the real_fops into a dynamically
allocated container struct, debugfs_fsdata.
A struct debugfs_fsdata gets allocated at file creation and freed from the
newly intoduced ->d_release().
Finally, move the implementation of debugfs_real_fops() out of the public
debugfs header such that struct debugfs_fsdata's declaration can be kept
private.
Signed-off-by: Nicolai Stange <nicstange@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Sangkyu Kim [Tue, 11 Dec 2018 10:53:46 +0000 (19:53 +0900)]
[RAMEN9610-9421][COMMON] arm64: erd9610_defconfig: Enable CPUSET.
Change-Id: I04fbc512631fae5d66858df6a5031ee4908a3fdd
Signed-off-by: lakkyung.jung <lakkyung.jung@samsung.com>
Sangkyu Kim [Tue, 11 Dec 2018 10:53:09 +0000 (19:53 +0900)]
[RAMEN9610-9421][COMMON] arm64: dts: add user-boost to cpufreq
Change-Id: Ib9afe433e0bb5360ae1b70911686d9e8b801c95d
Signed-off-by: Park Bumgyu <bumgyu.park@samsung.com>
Park Bumgyu [Thu, 23 Aug 2018 06:51:04 +0000 (15:51 +0900)]
[RAMEN9610-9421][COMMON] cpufreq: acme: request prefer perf at cpufreq_min_limit
Change-Id: I6b78a333c52708b59bf84c664795e370d9d8de7c
Signed-off-by: Park Bumgyu <bumgyu.park@samsung.com>
Sangkyu Kim [Tue, 11 Dec 2018 10:52:28 +0000 (19:52 +0900)]
[RAMEN9610-9421][COMMON] arm64: dts: add prefer perf service
Change-Id: Ic2921700f7c1e5574d4e71c9c4ef55eec2a22410
Signed-off-by: Park Bumgyu <bumgyu.park@samsung.com>
Park Bumgyu [Mon, 20 Aug 2018 01:29:10 +0000 (10:29 +0900)]
[RAMEN9610-9421][COMMON] sched: ems: support prefer perf service
Change-Id: Ida3e81c598a22e984839533e62604ffd20c94dc3
Signed-off-by: Park Bumgyu <bumgyu.park@samsung.com>
Park Bumgyu [Mon, 13 Aug 2018 06:21:27 +0000 (15:21 +0900)]
[RAMEN9610-9421][COMMON] sched: ems: introduce ems service
Change-Id: I6e0cc8b8db43035c5c933ed292f443c9a67e4520
Signed-off-by: Park Bumgyu <bumgyu.park@samsung.com>
Park Bumgyu [Thu, 23 Aug 2018 06:35:16 +0000 (15:35 +0900)]
[RAMEN9610-9421][COMMON] cpufreq: acme: remove code related with HMP boost.
Change-Id: I87bda7bd22f54fadefac19582340d9d3e1a09ca5
Signed-off-by: Park Bumgyu <bumgyu.park@samsung.com>
Soohyun Kim [Fri, 16 Nov 2018 05:43:17 +0000 (14:43 +0900)]
[RAMEN9610-9418][COMMON] soc: cpupm: Prevent bug fix.
This patch can prevent out-of-bound case when access pm->modes[i].
Change-Id: I6770fed95af0541426356b804b8c9f8ae43bbea2
Signed-off-by: Soohyun Kim <soohyuni.kim@samsung.com>
Youngtae Lee [Mon, 15 Oct 2018 06:21:01 +0000 (15:21 +0900)]
[RAMEN9610-9418][COMMON] sched: frt: Fix zero dividing bug
Change-Id: Id738b19174de909113aa8c5224a3e57f1762073d
Signed-off-by: Youngtae Lee <yt0729.lee@samsung.com>
Youngtae Lee [Mon, 8 Oct 2018 06:57:12 +0000 (15:57 +0900)]
[RAMEN9610-9418][COMMON] sched: frt: fix cpumask warnning bug
Change-Id: I27eb08fdcbe6ce7b35a09a38b2aa2fb4b90e76a7
Signed-off-by: Youngtae Lee <yt0729.lee@samsung.com>
Daeyeong Lee [Mon, 15 Oct 2018 06:38:00 +0000 (15:38 +0900)]
[RAMEN9610-9418][COMMON] sched: ems: Disable ontime when capacity of coregroup is zero
Change-Id: I22c3b9d97ca5b5f598436cfb06062b9cb24f2ff6
Signed-off-by: Daeyeong Lee <daeyeong.lee@samsung.com>
Daeyeong Lee [Mon, 15 Oct 2018 06:36:06 +0000 (15:36 +0900)]
[RAMEN9610-9418][COMMON] sched: ems: Modify get_cpu_max_capacity not to access NULL point
Change-Id: I2a88779e24ba4f30d0423224d3cdc78aea6e586a
Signed-off-by: Daeyeong Lee <daeyeong.lee@samsung.com>
Sangkyu Kim [Fri, 9 Nov 2018 09:49:41 +0000 (18:49 +0900)]
[RAMEN9610-9418][COMMON] ems: frt: fix initialize variable for check condition
Change-Id: I330d6250f3a8873ffd0bdbb1bca524db6ca56d7d
Signed-off-by: Sangkyu Kim <skwith.kim@samsung.com>
Munjoo Kim [Mon, 26 Nov 2018 08:46:36 +0000 (17:46 +0900)]
[RAMEN9610-9418][RAMEN9610-8873][COMMON] CFI Enabling
CFI enabling with Enforce mode.
Change-Id: I366bc2e9efb5a3646754b1919a3b7c9eb00cd490
Signed-off-by: Munjoo Kim <munjoo.kim@samsung.com>
Munjoo Kim [Tue, 20 Nov 2018 07:09:55 +0000 (16:09 +0900)]
[RAMEN9610-9418][RAMEN9610-8873][COMMON] LTO Enabling
Modify mk files for LLVM related configuration.
Change-Id: Ieeac412c933be2d209f18071bd7a049d5516d6ad
Signed-off-by: Munjoo Kim <munjoo.kim@samsung.com>
Taekki Kim [Tue, 27 Nov 2018 12:06:39 +0000 (21:06 +0900)]
[RAMEN9610-9418][COMMON] soc: samsung: add __nocfi for indirect functions
Change-Id: I521ba2eb7be97ffb5bd606471ff31362df0cb976
Signed-off-by: Taekki Kim <taekki.kim@samsung.com>
Youngmin Nam [Fri, 23 Nov 2018 07:33:53 +0000 (16:33 +0900)]
[RAMEN9610-9418][COMMON] debug: change function prototype related debug
Change-Id: Ie89372fef3256a2f623aaba99cf713bb3fbecfce
Signed-off-by: Youngmin Nam <youngmin.nam@samsung.com>
Signed-off-by: Youngwan Kim <y103.kim@samsung.com>
Munjoo Kim [Thu, 22 Nov 2018 07:39:00 +0000 (16:39 +0900)]
[RAMEN9610-9418][COMMON] soc: samsung: change function prototype related chipid
Change-Id: I4265419af8bf317ec18ee40845bc4cc75e85df83
Yoojin Park [Tue, 27 Nov 2018 08:55:01 +0000 (17:55 +0900)]
[RAMEN9610-9418][R10P0][KERNEL]gpu: To avoid CFI failure.
CFI is Control Flow Integrity.
CFI will make graph to verify the code-flows.
If code-flows don't match with graph generated, It will make the failure.
Change-Id: Ic04be25e9858e432968b08e0c9203b778a0e20b0
Signed-off-by: Yoojin Park <yoojin1.park@samsung.com>
Hyunju Kang [Mon, 26 Nov 2018 07:47:57 +0000 (16:47 +0900)]
[RAMEN9610-9418][9610] soc: samsung: epx: prevent CFI failed.
Change-Id: Icbfea62ec6502f815499ad5ef2cbe2ee98d8170b
Signed-off-by: Hyunju Kang <hjtop.kang@gmail.com>
Hyunju Kang [Mon, 26 Nov 2018 07:11:17 +0000 (16:11 +0900)]
[RAMEN9610-9418][COMMON] cpuidle: profiler: change function prototype
Change-Id: Id2bed6596e613a4388c547734bf611b1cdaa75d1
Signed-off-by: Hyunju Kang <hjtop.kang@gmail.com>
Jiyoung Jeong [Tue, 27 Nov 2018 00:35:09 +0000 (09:35 +0900)]
[RAMEN9610-9418][RAMEN9610-8873][COMMON] drivers: modem_if: change return type for CFI
static int vnet_xmit => static netdev_tx_t vnet_xmit
Change-Id: Ib03e51dfd6906ca7fea714fdbd4e4bd199101d4b
Signed-off-by: Jiyoung Jeong <ji_0.jeong@samsung.com>
dan huh [Wed, 28 Nov 2018 10:59:22 +0000 (19:59 +0900)]
[RAMEN9610-9418][COMMON] fimc-is2: Add a __nocfi at function to happen cfi error
PR JIRA ID: CPR-462
Change-Id: I4e666ac118a4f397e632cad4fe3cb28a4ca40ff7
Signed-off-by: dan huh <dan.huh@samsung.com>
dan huh [Wed, 28 Nov 2018 10:58:29 +0000 (19:58 +0900)]
[RAMEN9610-9418][9610] fimc-is2: Add __nocfi at happen to cfi error function
PR JIRA ID: CPR-462
Change-Id: I71dff7c73114f58a8120aca2de1848f757c09e9f
Signed-off-by: dan huh <dan.huh@samsung.com>
Kim Haksoo [Fri, 30 Nov 2018 09:09:25 +0000 (18:09 +0900)]
[RAMEN9610-9418][9610] fbdev: dpu: modify decon driver to avoid CFI failure
Change-Id: If59ee07e7604e06612f72e2593bf659fa39d40bf
Signed-off-by: Kim Haksoo <herb@samsung.com>
Pilsun Jang [Thu, 10 May 2018 09:52:39 +0000 (18:52 +0900)]
[RAMEN9610-9416][Common] ASoC: abox: added holding dram path while fw is booting
Change-Id: I4063053d31eabadafab92a3290e206044b2ff5fd
Signed-off-by: Pilsun Jang <pilsun.jang@samsung.com>
sanghwan park [Tue, 11 Dec 2018 11:05:12 +0000 (20:05 +0900)]
[RAMEN9610-9358][9610] vipx: increase memory size for dram fw and mailbox latency
Change-Id: Ibdd068ec2f761269ff96c5d07d769ef8e3422db1
Signed-off-by: sanghwan park <senius.park@samsung.com>
Namhyung Kim [Tue, 30 Oct 2018 11:08:09 +0000 (20:08 +0900)]
[RAMEN9610-9352][MODAP-10498][COMMON] drivers: net: add pptp files
Change-Id: Ie2a406146d11403dd87233f2e972fca5ad991a6d
Signed-off-by: Namhyung Kim <nh78.kim@samsung.com>