Age | Commit message (Collapse) | Author |
|
am: 5079d0aabc
Change-Id: Ica4813dcf76197b8866307c7c0ba5fc4a56fd6e3
|
|
Use the uapi ones instead, so I don't have to think about whether ours
are correct.
Bug: http://b/31044648
Change-Id: Ibe4a078ad3223419301a5f585e8ed24dd7366ce2
|
|
am: a2632be956
* commit 'a2632be956cf05558cc7825830ffd5f9f093206d':
Add oom score tracking to procrank
C++-ify procrank
Change-Id: Ic2baa7e0144480a2400299e2b981c158e4709caa
|
|
procrank -o will sort by the oom score for each process, and show
cuumulative totals of memory stats for processes whose score is below
each lowmemorykiller oom score threshold. This is particularly useful
with procrank -o -c to see the amount of page cache used lowmemorykiller
threshold.
Change-Id: I26e290c0835472630f70ed933de43bfbfd7bdfb0
|
|
Change-Id: I1b006d2af04f21c0c02ac2d7629340d91b876886
|
|
Starting with Linux 4.1, reading from /dev/block/zram*/mem_used_total
logs this warning:
"Attribute mem_used_total (and others) will be removed. See zram
documentation."
mem_used_total and several related counters are now consolidated under a
new node mm_stat. Prefer that when it's available.
Bug: 25951511
Change-Id: Ida732ee56a1e39e01c390f61bb40b3f52f643004
Signed-off-by: Greg Hackmann <ghackmann@google.com>
|
|
am: 6ca5d69923
* commit '6ca5d6992307ac6d063e4a7a1f0457afb88e033c':
procrank: add proportional swap accounting
|
|
Shared pages are reported in each of the sharing process swapped pages.
Compute a proportional swap usage to get a meaningful value of what each
process has in swap. Report also process unique pages swapped out.
In case ZRAM is used compute the compression ration and report the
actual RAM use of the swapped pages.
Bug: 25392275
Change-Id: I3a28c7812a09a02e1a604593615f5c6ad0340f9f
|
|
am: 7ad5b7d135
* commit '7ad5b7d13581832472d9f8da576c50f72c203bf2':
procrank: Fix some printf alignements
|
|
Bug: 25392275
Change-Id: I8ea4d01310190c34b68d3f3140ea336ad82df665
|
|
Bug: 25392275
Change-Id: I6d2a0b03abba555b3667f6ccf12896caeecf27aa
|
|
Change-Id: If75e007111f82178e372c88ab4dce62fd8ff7873
|
|
|
|
Use uint64_t for sizes.
Change-Id: Ic686b4cbf4406aec1e9b01810924488bff63722e
|
|
Change-Id: Ic339a05a3091ec6b33b68bc5e76082de3513e9c7
Signed-off-by: Ashok Bhat <ashok.bhat@arm.com>
|
|
Add the -c, -C, and -k arguments to only show pages that are
cached (backed by storage), non-cached (backed by ram or swap),
or have been collapsed by KSM respectively.
Change-Id: I707570ad406106b0aca230d18f9945901d193aca
|
|
Add a column showing swapped pages per process, but only show it
if swapped pages exist in at least one process.
Change-Id: I51a12fb8999c2a73418277fd575e7109478667d5
|
|
adb bugreport > b & sleep 5; kill $?
will show that a bunch of tools will SIGPIPE when trying to output data
which is normal.
We don't have a way on linux (bsd yes), to set NOSIGPIPE on the file desc
passed down via fork/exec.
Bug: 6447319
Change-Id: I1634a00a155fd8a3fd827aa4a77e94518534336c
|
|
Change-Id: I0dac539c220fda867ce166d78dc898a8ec7aa18a
|
|
Now print a sum of all pss and uss at then end of the output, and
then print information from /proc/meminfo to help put those in
context.
Change-Id: I323d219820830e9435bb4dcfd78cbb9a3d6c129c
|
|
pm_process_usage can return an error, and leave procs[i]->usage
unitialized. Detect the error case and print a warning. Also
make the initialization of procs[i]->usage to 0 unconditional,
so nothing will be printed in the final procrank stats when an
error is detected.
Change-Id: I03f90ae6a5ebb201b6e9e43593cec225e1a8ded0
|
|
Change-Id: I172667643e13e2e1c278fd4c0156d5a2445d79f7
|
|
We have phones that have more than 256 PIDs running at one time right
now, so don't segmentation fault when that condition happens.
Free all proc info after it ahs been used.
Also if a process goes away before we read its process name, just print
"<unknown>" instead.
Change-Id: Ia3899be61b047852f62bf3cc6f30eb2fcd797a61
|
|
|
|
|
|
|