Provide a securable way to open a file for filling
Test: incfs_test passes
Bug: 138149732
Signed-off-by: Paul Lawrence <paullawrence@google.com>
Change-Id: Ib4b6fd839ad30ce08e31121d19e2c0d7066d302f
Filling blocks is not equivalent to writing a file, since they are
constrained by the root hash. selinux policy may wish to treat them
differently, for instance.
Test: incfs_test passes
Bug: 138149732
Signed-off-by: Paul Lawrence <paullawrence@google.com>
Change-Id: Ic369b84b92547b1cfefe422bd881c4e466090aed
They provide no value and simply duplicate a check in copy_from/to_user
Test: incfs_test passes
Bug: 138149732
Signed-off-by: Paul Lawrence <paullawrence@google.com>
Change-Id: Icc6054a2d6a495c9a03cd1507dda1ab8ca0b0dc4
There were still a little bit of negative IBAT value via ADC
in not-charging case due to using the external Rsense.
This made IBAT less than next_cc_max and fv_uv kept
bumping up to the max one.
Fix this by setting a small negative current(ibat_zero_offset)
as the threshold (instead of 0uA) and entering MSC_DSG for
"not-charging" case.
Bug: 159100733
Signed-off-by: Jack Wu <wjack@google.com>
Change-Id: Ia56c972ba62a41ab4dae437973b4968df4e8ea51
Align the sample rate with other projects (from 64 to 8).
It could enlarge the energy data range to delay data overflow.
Bug: 159324671
Change-Id: I2990568116d898e07025f5647db3694aa9f6b2d4
Signed-off-by: Xavier_Cheng <xacheng@google.com>
Define the mdp transfer time of sofef00 and sofef01 display.
Bug: 159300095
Test: suspend/resume, scrolling, app launching/closing, playing video
Signed-off-by: tangrobin <tangrobin@google.com>
Change-Id: I8bfed2b11746e6a5a4208b7c63354803f628d40e
This is required for devices launching with R.
Bug: 158733532
Test: builds
Signed-off-by: Martijn Coenen <maco@google.com>
Change-Id: Ib493c22f4ff7a1f3464607c031d75509122717b6
Use an entrophy provided by bootloader from /chosen/rng-seed
in DT to increase kernels's initial device randomness.
Bug: 150154308
Test: boot bramble
Change-Id: Ie57601d3ac5f33bbcdc3e40d703df821c87640c3
Signed-off-by: Chenglu Lin <chenglulin@google.com>
module_param doesn't work. Get device serial from saved_command_line
Bug: 159078410
Signed-off-by: Jenny Ho <hsiufangho@google.com>
Change-Id: I2c2d14c88fb86e35aa585c67932d1adde3a54674
Remove ineffective defconfig.
This patch remove configs in update_abi_config() which did not change configs.
Bug: 157102855
Signed-off-by: lucaswei <lucaswei@google.com>
Change-Id: I565b350d88e8ba3da32bf8d25eb0a229153f8501
We currently deal with ARM64_ERRATUM_1188873 by always trapping EL0
accesses for both instruction sets. Although nothing wrong comes out
of that, people trying to squeeze the last drop of performance from
buggy HW find this over the top. Oh well.
Let's change the mitigation by flipping the counter enable bit
on return to userspace. Non-broken HW gets an extra branch on
the fast path, which is hopefully not the end of the world.
The arch timer workaround is also removed.
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
(cherry picked from commit 0f80cad3124f986d0e46c14d46b8da06d87a2bf4)
Signed-off-by: Chiawei Wang <chiaweiwang@google.com>
Bug: 154668398
Test: bionic-benchmarks64 --bionic_xml vdso.xml
Change-Id: Ia357f14a7ec740fe283c60c357163dfd22745e84
From 6.4.4.2 Structured VDM:
"Structured VDMs Shall only be used when an Explicit Contract is in place
with the following exception: o Prior to establishing an Explicit Contract
a Source May issue Discover Identity Messages, to a Cable Plug
using SOP’ Packets, as an Initiator (see Section 8.3.3.22.3).
• Either Port May be an Initiator of Structured VDMs except for the Enter
Mode and Exit Mode Commands which Shall only be initiated by the DFP."
The above implies that when PD3.0 link is established PD3.0 sinks
can send out discover identity command/AMS once PD negotiation is done.
However, since sink has to wait for SinkTxOk before starting the AMS and
the source might be trying to attempt Discover identity to the sink,
reaching SNK_READY on the sink side does not imply TCPM's first attempt
to read discover identity goes through. To solve this, the patch adds
a retry logic to initiate Discover Identity AMS.
Bug: 138061779
Signed-off-by: Badhri Jagan Sridharan <badhri@google.com>
Change-Id: I1be7530295f9056bd4cde012891b3a1da783d1d9
When partner identity is set by the TCPM, the sysfs nodes seem
to get updated, however userspace misses the update as the
uevent does not get sent. Call kobject_uevent to fix this.
Bug: 138061779
Signed-off-by: Badhri Jagan Sridharan <badhri@google.com>
Change-Id: I6d1d67de82222b90218118b388356311991d81e3
Bug: 124259099
Test: virtual thermal zone can be read
Change-Id: I160e994b5acc58fcd32f542638fa2b39ac842f14
Signed-off-by: TeYuan Wang <kamewang@google.com>
This reverts commit 53b665d139.
Reason for revert: The original issue (b/158635600) has been fixed by pa/1699349. Revert this commit to enable pa/1695246.
Bug: 158635600
Bug: 154668398
Bug: 157925983
Test: run bionic-benchmarks --bionic_xml vdso.xml
Test: check the time returned from time()
e.g. adb bugreport and check the date suffix
in the file name
Signed-off-by: Chiawei Wang <chiaweiwang@google.com>
Change-Id: Iedc676521d489da98aab87b9aa126b0905badb17
All of these modules have the PROBE_PREFER_ASYNCHRONOUS flag set for
their respective drivers, meaning they probed asynchronously in non-GKI
configs. However, when a module is loaded, this flag will not apply
correctly unless the module itself is given the .async_probe=1 flag at
load time. Fix this. Local tests show ~500ms boot time savings.
This change affects the following:
* snd_soc_cs35l41_i2c
* i2c_qcom_geni
* st21nfc
* spmi_pmic_arb
* ufs_qcom
Bug: 158789858
Signed-off-by: J. Avila <elavila@google.com>
Change-Id: I8abc47ec9efa1efab69879b7ce8ca401398127ab
The normal range is 0x0~0x3FF. The HBM range is 0x0~0x194. When
HBM enabled, the level will change from 0x3FF(normal range) to
0x0 (HBM range). It caused a little flicker. Fixed the HBM
range (0x194~0x194) to avoid the symptom.
Bug: 158957266
Test: Test adaptive brightness from normal to HBM
Change-Id: I83f0520abcfc85e873764bab34148f6c19317fe7
Signed-off-by: linpeter <linpeter@google.com>
In __cvdso_clock_getres_time32(), when clock_getres_fallback()
fails, it's not required to update the struct timespec res.
Bug: 159086668
Bug: 154668398
Test: run cts -m CtsBionicTestCases -t time#clock_getres_unknown
Signed-off-by: Chiawei Wang <chiaweiwang@google.com>
Change-Id: Ibb7279d4520658d5c8be76e721249d3c62267d9f
Enable binderfs. This allows binder stats to be available when debugfs
is disabled by moving binder, hwbinder and vndbinder devices to their
own filesystem in /dev/binderfs.
Test: build, boot, binder devices present in /dev/binderfs
Bug: 136497735
Change-Id: Ibee4712b4e718d234a8b64d778cb2001d14cf4fd
Signed-off-by: Hridya Valsaraju <hridya@google.com>
Added one on top of merge commit.
"f2fs: avoid checkpatch error"
* aosp/upstream-f2fs-stable-linux-4.19.y:
f2fs: attach IO flags to the missing cases
f2fs: add node_io_flag for bio flags likewise data_io_flag
f2fs: remove unused parameter of f2fs_put_rpages_mapping()
f2fs: handle readonly filesystem in f2fs_ioc_shutdown()
f2fs: avoid utf8_strncasecmp() with unstable name
f2fs: don't return vmalloc() memory from f2fs_kmalloc()
f2fs: fix retry logic in f2fs_write_cache_pages()
f2fs: fix wrong discard space
f2fs: compress: don't compress any datas after cp stop
f2fs: remove unneeded return value of __insert_discard_tree()
f2fs: fix wrong value of tracepoint parameter
f2fs: protect new segment allocation in expand_inode_data
f2fs: code cleanup by removing ifdef macro surrounding
writeback: Avoid skipping inode writeback
f2fs: avoid inifinite loop to wait for flushing node pages at cp_error
f2fs: compress: fix zstd data corruption
f2fs: add compressed/gc data read IO stat
f2fs: fix potential use-after-free issue
f2fs: compress: don't handle non-compressed data in workqueue
f2fs: remove redundant assignment to variable err
f2fs: refactor resize_fs to avoid meta updates in progress
f2fs: use round_up to enhance calculation
f2fs: introduce F2FS_IOC_RESERVE_COMPRESS_BLOCKS
f2fs: Avoid double lock for cp_rwsem during checkpoint
f2fs: report delalloc reserve as non-free in statfs for project quota
f2fs: Fix wrong stub helper update_sit_info
f2fs: compress: let lz4 compressor handle output buffer budget properly
f2fs: remove blk_plugging in block_operations
f2fs: introduce F2FS_IOC_RELEASE_COMPRESS_BLOCKS
f2fs: shrink spinlock coverage
f2fs: correctly fix the parent inode number during fsync()
f2fs: introduce mempool for {,de}compress intermediate page allocation
f2fs: introduce f2fs_bmap_compress()
f2fs: support fiemap on compressed inode
f2fs: support partial truncation on compressed inode
f2fs: remove redundant compress inode check
f2fs: use strcmp() in parse_options()
f2fs: Use the correct style for SPDX License Identifier
Conflicts:
fs/f2fs/data.c
fs/f2fs/dir.c
Bug: 154167995
Change-Id: I04ec97a9cafef2d7b8736f36a2a8d244965cae9a
Signed-off-by: Jaegeuk Kim <jaegeuk@google.com>
Currently when calculating boosted util for a cpu, it uses a fixed
value of 1024 for calculation. So when top-app tasks moved to LC,
which has much lower capacity than BC, the freq calculated will be
high even the cpu util is low. This results in higher power
consumption, especially on arch which has more little cores than
big cores. By replacing the fixed value of 1024 with actual cpu
capacity will reduce the freq calculated on LC.
Bug: 152925197
Test: boosted util reduced on little cores
Signed-off-by: Rick Yiu <rickyiu@google.com>
Change-Id: I80cdd08a2c7fa5e674c43bfc132584d85c14622b
With the introduction of placement hint patch, boosted tasks will not
scheduled from big cores. We tune capacity margin to let important
boosted tasks get scheduled on big cores. However, the capacity margin
affects all group of tasks, so that non-boosted tasks get more chances
to be scheduled on big cores, too. This could be solved by separating
capacity margin for boosted tasks.
Bug: 152925197
Test: margin set correctly
Signed-off-by: Rick Yiu <rickyiu@google.com>
Change-Id: I0e059c56efa9bc8513f0ef4b0f6ab8f5d04a592a
Remove ION_POOL_AUTO_REFILL before we get actual
improvement data.
Bug: 154228928
Test: boot
Signed-off-by: Martin Liu <liumartin@google.com>
Change-Id: I9fac856b1f170a00d6ead61d8fcf36ca148dd0db
According to redfin evt thermal result, the highest skin-therm is only
39degC for idle+18W charging case. So, adjust more aggressive wired
chargig behavior after 39degC which can benefit thermal UX, and will
not impact charging KPI.
Bug: 153850974
Test: verified charging throttling with emul_temp
Change-Id: I1473efa343cf55e989d4e9493c747787f3b3e078
Signed-off-by: TeYuan Wang <kamewang@google.com>
Useful for testing, debugging, when combined with network namespaces
in particular.
Generated via:
echo 'CONFIG_VETH=y' >> arch/arm64/configs/redbull_defconfig
make ARCH=arm64 redbull_defconfig
make ARCH=arm64 savedefconfig
mv defconfig arch/arm64/configs/redbull_defconfig
(and then undid unrelated changes)
Test: built
Bug: 158835517
Signed-off-by: Maciej Żenczykowski <maze@google.com>
Change-Id: I5f03a721c78b127f2290b873705a730ffdc71c06
Use GKI_RAMDISK_PREBUILT_BINARY to pack ramdisk into boot.img and
initramfs.cpio into vendor_boot.img.
Bug: 157624527
Signed-off-by: Robin Peng <robinpeng@google.com>
Change-Id: Ie56ad6f6926221238598d56108074aca45c41557
Leaf changes summary: 1 artifact changed
Changed leaf types summary: 1 leaf type changed
Removed/Changed/Added functions summary: 0 Removed, 0 Changed, 0 Added function
Removed/Changed/Added variables summary: 0 Removed, 0 Changed, 0 Added variable
'struct sk_buff at skbuff.h:666:1' changed:
type size changed from 1920 to 1984 (in bits)
1 data member insertion:
'nf_bridge_info* sk_buff::nf_bridge', at offset 960 (in bits) at skbuff.h:715:1
there are data member changes:
'unsigned int sk_buff::len' offset changed from 960 to 1024 (in bits) (by +64 bits)
'unsigned int sk_buff::data_len' offset changed from 992 to 1056 (in bits) (by +64 bits)
'__u16 sk_buff::mac_len' offset changed from 1024 to 1088 (in bits) (by +64 bits)
'__u16 sk_buff::hdr_len' offset changed from 1040 to 1104 (in bits) (by +64 bits)
'__u16 sk_buff::queue_mapping' offset changed from 1056 to 1120 (in bits) (by +64 bits)
'__u8 sk_buff::__cloned_offset[]' offset changed from 1072 to 1136 (in bits) (by +64 bits)
'__u8 sk_buff::pfmemalloc' offset changed from 1072 to 1136 (in bits) (by +64 bits)
'__u8 sk_buff::__pkt_type_offset[]' offset changed from 1088 to 1152 (in bits) (by +64 bits)
'__u32 sk_buff::headers_start[]' offset changed from 1088 to 1152 (in bits) (by +64 bits)
'__u8 sk_buff::ooo_okay' offset changed from 1088 to 1152 (in bits) (by +64 bits)
'__u8 sk_buff::csum_valid' offset changed from 1096 to 1160 (in bits) (by +64 bits)
'__u8 sk_buff::ipvs_property' offset changed from 1104 to 1168 (in bits) (by +64 bits)
'__u16 sk_buff::tc_index' offset changed from 1120 to 1184 (in bits) (by +64 bits)
while looking at anonymous data member 'union {__wsum csum; struct {__u16 csum_start; __u16 csum_offset;};}':
the internal name of that anonymous data memberchanged from:
__anonymous_union__6
to:
__anonymous_union__8
This is usually due to an anonymous member type being added or removed from the containing type
and offset changed from 1152 to 1216 (in bits) (by +64 bits)
'__u32 sk_buff::priority' offset changed from 1184 to 1248 (in bits) (by +64 bits)
'int sk_buff::skb_iif' offset changed from 1216 to 1280 (in bits) (by +64 bits)
'__u32 sk_buff::hash' offset changed from 1248 to 1312 (in bits) (by +64 bits)
'__be16 sk_buff::vlan_proto' offset changed from 1280 to 1344 (in bits) (by +64 bits)
'__u16 sk_buff::vlan_tci' offset changed from 1296 to 1360 (in bits) (by +64 bits)
while looking at anonymous data member 'union {unsigned int napi_id; unsigned int sender_cpu;}':
the internal name of that anonymous data memberchanged from:
__anonymous_union__7
to:
__anonymous_union__9
This is usually due to an anonymous member type being added or removed from the containing type
and offset changed from 1312 to 1376 (in bits) (by +64 bits)
'__u32 sk_buff::secmark' offset changed from 1344 to 1408 (in bits) (by +64 bits)
while looking at anonymous data member 'union {__u32 mark; __u32 reserved_tailroom;}':
the internal name of that anonymous data memberchanged from:
__anonymous_union__8
to:
__anonymous_union__10
This is usually due to an anonymous member type being added or removed from the containing type
and offset changed from 1376 to 1440 (in bits) (by +64 bits)
while looking at anonymous data member 'union {__be16 inner_protocol; __u8 inner_ipproto;}':
the internal name of that anonymous data memberchanged from:
__anonymous_union__9
to:
__anonymous_union__11
This is usually due to an anonymous member type being added or removed from the containing type
and offset changed from 1408 to 1472 (in bits) (by +64 bits)
'__u16 sk_buff::inner_transport_header' offset changed from 1424 to 1488 (in bits) (by +64 bits)
'__u16 sk_buff::inner_network_header' offset changed from 1440 to 1504 (in bits) (by +64 bits)
'__u16 sk_buff::inner_mac_header' offset changed from 1456 to 1520 (in bits) (by +64 bits)
'__be16 sk_buff::protocol' offset changed from 1472 to 1536 (in bits) (by +64 bits)
'__u16 sk_buff::transport_header' offset changed from 1488 to 1552 (in bits) (by +64 bits)
'__u16 sk_buff::network_header' offset changed from 1504 to 1568 (in bits) (by +64 bits)
'__u16 sk_buff::mac_header' offset changed from 1520 to 1584 (in bits) (by +64 bits)
'u64 sk_buff::android_kabi_reserved1' offset changed from 1536 to 1600 (in bits) (by +64 bits)
'__u32 sk_buff::headers_end[]' offset changed from 1536 to 1600 (in bits) (by +64 bits)
'u64 sk_buff::android_kabi_reserved2' offset changed from 1600 to 1664 (in bits) (by +64 bits)
'sk_buff_data_t sk_buff::tail' offset changed from 1664 to 1728 (in bits) (by +64 bits)
'sk_buff_data_t sk_buff::end' offset changed from 1696 to 1760 (in bits) (by +64 bits)
'unsigned char* sk_buff::head' offset changed from 1728 to 1792 (in bits) (by +64 bits)
'unsigned char* sk_buff::data' offset changed from 1792 to 1856 (in bits) (by +64 bits)
'unsigned int sk_buff::truesize' offset changed from 1856 to 1920 (in bits) (by +64 bits)
'refcount_t sk_buff::users' offset changed from 1888 to 1952 (in bits) (by +64 bits)
293 impacted interfaces
Signed-off-by: Mark Salyzyn <salyzyn@google.com>
Bug: 157994070
Change-Id: I9bcff125376362fd0c4673b8786473b186f97bde
Remove condition for including struct sk_buff members based on
CONFIG_BRIDGE_NETFILTER config.
Bug: 151840548
Bug: 157994070
Test: build
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Change-Id: Iee626843e107e8d64c3c6c4a1cc9c08f4e38f5af
Merged-In: Iee626843e107e8d64c3c6c4a1cc9c08f4e38f5af
Allow other build targets to compile kernel and the boot.img with
prebuilt ramdisk. This also gives the ability to reuse the mixed
flashing mechanism described in b/151989050.
Bug: 157624527
Signed-off-by: Robin Peng <robinpeng@google.com>
Change-Id: I1fae17cf4b9e280b23edc52f7453946546d8c63c
Qualcomm au011 change condition in Kconfig, this will enable
CONFIG_DYNAMIC_DEBUG with CONFIG_PROC_FS=y. Aligned to android
common kernel, we disable CONFIG_DYANAMIC_DEBUG in defconfig.
Bug: 157858241
Signed-off-by: lucaswei <lucaswei@google.com>
Change-Id: I70100023afacaba5195826fb74beedb49d05d07a
