am: 784935b9c2 -s ours
am skip reason: change_id Ia6366834613d1e12498fa90377e79f62a2149776 with SHA1 a8e4b89768 is in history
Change-Id: I67fc6a91f63086417c79b13c5403761bf6cf86e5
There appears to be a kernel bug that causes SIGHUP and SIGCONT to be
sent to the parent process group we spawn from if the process group
contains stopped jobs (e.g. the parent itself, because of wait_for_gdb).
Call setsid in all of our children to prevent this from happening.
Bug: http://b/31124563
Test: adb shell 'setprop debug.debuggerd.wait_for_gdb 1; killall -ABRT surfaceflinger'
Change-Id: I1a48d70886880a5bfbe2deb80d48deece55faf09
(cherry picked from commit 18cb681247)
Some USB controllers only support 16k writes, so drop down to that.
Bug: http://b/130622010
Test: treehugger
Change-Id: I836d5b2d0d6fcae05f290af11a6a19c6e8d7c1ba
(cherry picked from commit 770a6a4d46)
A fork() was historically added in case of fs_mgr crashing or leaking
memory, but this should not be the case with fs_mgr, and a fork() only
hides any such problem, instead of allowing us to address it
directly.
Test: boot
Change-Id: If7ee4807757048258a6ea9a79a24cebbacc530cc
procrank's usage in its output shows 20x increase from
last year. This is because it is keeping the process
maps around until termination. Fix that by getting rid
of ProcMemInfo objects when done parsing /proc/<pid>/maps,pagemap.
Note that the total allocations do not change and have not necessarily
regressed from Pie.
Bug: 130672819
Test: adb shell procrank | grep 'procrank\|cmdline'
Change-Id: Ib7bf960ed1d053347fcfc0c8aee9019607a1eb01
Signed-off-by: Sandeep Patil <sspatil@google.com>
am: 6569f35ae0 -s ours
am skip reason: change_id I9a140c7d9d93e1dd43c34c8cf066f4a62e2bf604 with SHA1 e8678cf883 is in history
Change-Id: I3ff5682fe9b4cd89065f4236b4f664a7aa0e6428
Some app may have different cgroup settings in cpuset and schedtune for its
threads, so let blkio follow cpuset only, which represents the app's current
state more accurately. Otherwise, if that thread is doing IO, then its
performance will be affected because its blkio group is in lower priority
group as schedtune.
ex: an app is now in top-app, but some thread of it set schedtune group to
background, and blkio follows schedtune because it is called later.
Main thread:
6:schedtune:/top-app
5:memory:/
4:cpuset:/top-app
3:cpuacct:/uid_1000/pid_8766
2:cpu:/
1:blkio:/
0::/
Some thread:
6:schedtune:/background
5:memory:/
4:cpuset:/top-app
3:cpuacct:/uid_1000/pid_8766
2:cpu:/
1:blkio:/background
0::/
Bug: 124727032
Test: blkio has same settings with cpuset
Change-Id: I9a140c7d9d93e1dd43c34c8cf066f4a62e2bf604
Merged-In: I9a140c7d9d93e1dd43c34c8cf066f4a62e2bf604
am: b2032ff720 -s ours
am skip reason: change_id Id3ab65b3860b6ea6cfec310ab13652009c81f415 with SHA1 81f1385589 is in history
Change-Id: I92f678cc6d11dd526f8c5c17f65cf0033b487ced
/sbin was traditionally used for static binaries on the ramdisk for
Android, but now everything is a shared binary, so this directory is
empty and we do not want to encourage creation of new libraries in
this directory.
Bug: 73660730
Test: build
Change-Id: I0d0aa052e1eaf529d18921c45169473df0ee51ff
Some app may have different cgroup settings in cpuset and schedtune for its
threads, so let blkio follow cpuset only, which represents the app's current
state more accurately. Otherwise, if that thread is doing IO, then its
performance will be affected because its blkio group is in lower priority
group as schedtune.
ex: an app is now in top-app, but some thread of it set schedtune group to
background, and blkio follows schedtune because it is called later.
Main thread:
6:schedtune:/top-app
5:memory:/
4:cpuset:/top-app
3:cpuacct:/uid_1000/pid_8766
2:cpu:/
1:blkio:/
0::/
Some thread:
6:schedtune:/background
5:memory:/
4:cpuset:/top-app
3:cpuacct:/uid_1000/pid_8766
2:cpu:/
1:blkio:/background
0::/
Bug: 124727032
Test: blkio has same settings with cpuset
Change-Id: I9a140c7d9d93e1dd43c34c8cf066f4a62e2bf604
Some USB controllers only support 16k writes, so drop down to that.
Bug: http://b/130622010
Test: treehugger
Change-Id: I836d5b2d0d6fcae05f290af11a6a19c6e8d7c1ba
There appears to be a kernel bug that causes SIGHUP and SIGCONT to be
sent to the parent process group we spawn from if the process group
contains stopped jobs (e.g. the parent itself, because of wait_for_gdb).
Call setsid in all of our children to prevent this from happening.
Bug: http://b/31124563
Test: adb shell 'setprop debug.debuggerd.wait_for_gdb 1; killall -ABRT surfaceflinger'
Change-Id: I1a48d70886880a5bfbe2deb80d48deece55faf09
