OSDN Git Service

android-x86/system-extras.git
12 years agoPass file descriptor instead of file name
Anatol Pomazau [Wed, 11 Jan 2012 23:12:27 +0000 (15:12 -0800)]
Pass file descriptor instead of file name

Passing a file descriptor to make_ext4fs_internal() is more flexible.
We can use tmpfile() to create a temporary file.
tmpfile() is better than other solutions because it unlinks the file
right after creating it, so closing fd effectively removes temp file.
Thus we don't have to worry about large temp files accidently left
on the filesystem in case of the program crash.

Change-Id: I44146704572c314e1d6cfca7ce918efa7fb92a7a

12 years agoRevert "Pass file descriptor instead of file name"
Anatol Pomazau [Fri, 3 Feb 2012 18:57:40 +0000 (10:57 -0800)]
Revert "Pass file descriptor instead of file name"

This reverts commit 93eb1dc9e68d2e9dea94f56d8bce478c1a52b354

12 years agoPass file descriptor instead of file name
Anatol Pomazau [Wed, 11 Jan 2012 23:12:27 +0000 (15:12 -0800)]
Pass file descriptor instead of file name

Passing a file descriptor to make_ext4fs_internal() is more flexible.
We can use tmpfile() to create a temporary file.
tmpfile() is better than other solutions because it unlinks the file
right after creating it, so closing fd effectively removes temp file.
Thus we don't have to worry about large temp files accidently left
on the filesystem in case of the program crash.

Change-Id: I44109e98854c314e1d6cfca7ce918efa7fb92a7a

12 years agoMerge "Fix various issues in procrank."
Selim Gurun [Wed, 25 Jan 2012 17:30:27 +0000 (09:30 -0800)]
Merge "Fix various issues in procrank."

12 years agoam 178509ef: am fa351ab2: Only do bugreport on user build with ADB enabled
Kenny Root [Wed, 25 Jan 2012 16:57:41 +0000 (08:57 -0800)]
am 178509ef: am fa351ab2: Only do bugreport on user build with ADB enabled

* commit '178509efc415f365ad09c11df2d68b51ac32aca4':
  Only do bugreport on user build with ADB enabled

12 years agoam fa351ab2: Only do bugreport on user build with ADB enabled
Kenny Root [Tue, 24 Jan 2012 23:37:09 +0000 (15:37 -0800)]
am fa351ab2: Only do bugreport on user build with ADB enabled

* commit 'fa351ab265957fa8815df3c4ca1f3c105f253e8b':
  Only do bugreport on user build with ADB enabled

12 years agoOnly do bugreport on user build with ADB enabled
Kenny Root [Tue, 24 Jan 2012 17:20:19 +0000 (09:20 -0800)]
Only do bugreport on user build with ADB enabled

This makes it less likely that users that don't know what a bugreport is
will trigger it, but leaves the possibility for developers to trigger a
bugreport with a key combo.

Bug: 5866877
Change-Id: Ifc3c5126ec9342880a3635845a89c5a4a40156d2

12 years agoOnly do bugreport on user build with ADB enabled
Kenny Root [Tue, 24 Jan 2012 17:20:19 +0000 (09:20 -0800)]
Only do bugreport on user build with ADB enabled

This makes it less likely that users that don't know what a bugreport is
will trigger it, but leaves the possibility for developers to trigger a
bugreport with a key combo.

Bug: 5866877
Change-Id: Ifc3c5126ec9342880a3635845a89c5a4a40156d2

12 years agoFix various issues in procrank.
Selim Gurun [Sat, 21 Jan 2012 03:44:54 +0000 (19:44 -0800)]
Fix various issues in procrank.

Fixed these problems:
1. Page swapped bit was not extracted correctly.
2. Pages were ignored when page present bit is not set.
3. Bit operations were not correct.
4. There was a compiler warning about unsigned/signed comparision.
5. Line limit was too short for the map file. This was causing procrank
   to generate a warning and remove the related process from results.

Change-Id: Ifed3913a49b15f59010cfa842090a13228074df9

12 years agoam 9883083a: am 412d55c7: Merge "fix fd leaking and deallocation freed resource"
Jean-Baptiste Queru [Fri, 20 Jan 2012 23:52:02 +0000 (15:52 -0800)]
am 9883083a: am 412d55c7: Merge "fix fd leaking and deallocation freed resource"

* commit '9883083afdb09845b0364f04c18d4fff24a89036':
  fix fd leaking and deallocation freed resource

12 years agoam 412d55c7: Merge "fix fd leaking and deallocation freed resource"
Jean-Baptiste Queru [Thu, 19 Jan 2012 01:58:19 +0000 (17:58 -0800)]
am 412d55c7: Merge "fix fd leaking and deallocation freed resource"

* commit '412d55c7ee5ce42dc52e0c03c3e0d1c5c2d9d870':
  fix fd leaking and deallocation freed resource

12 years agoMerge "fix fd leaking and deallocation freed resource"
Jean-Baptiste Queru [Wed, 18 Jan 2012 23:43:21 +0000 (15:43 -0800)]
Merge "fix fd leaking and deallocation freed resource"

12 years agoam b30136b7: (-s ours) DO NOT MERGE Add iteration option to micro_bench
Eric Rowe [Thu, 12 Jan 2012 19:34:20 +0000 (11:34 -0800)]
am b30136b7: (-s ours) DO NOT MERGE Add iteration option to micro_bench

* commit 'b30136b7f51858db1a3272bf91801fe7faf16309':
  DO NOT MERGE Add iteration option to micro_bench

12 years agoMerge "libc: Add pthread and stdio benchmark test programs."
David 'Digit' Turner [Thu, 15 Dec 2011 11:00:39 +0000 (03:00 -0800)]
Merge "libc: Add pthread and stdio benchmark test programs."

12 years agoDO NOT MERGE Add iteration option to micro_bench
Eric Rowe [Fri, 9 Dec 2011 01:11:00 +0000 (17:11 -0800)]
DO NOT MERGE Add iteration option to micro_bench

Change-Id: Ie7ab13a58dd4585e1c489cb407e59a76bed8955b
Bug-Id: http://b/issue?id=5734427

12 years agoAdd iteration option to micro_bench
Eric Rowe [Fri, 9 Dec 2011 01:11:00 +0000 (17:11 -0800)]
Add iteration option to micro_bench

Change-Id: Ie7ab13a58dd4585e1c489cb407e59a76bed8955b
Bug-Id: http://b/issue?id=5734427

12 years agolibc: Add pthread and stdio benchmark test programs.
David 'Digit' Turner [Wed, 7 Dec 2011 11:45:44 +0000 (12:45 +0100)]
libc: Add pthread and stdio benchmark test programs.

This patch adds two test programs to system/extras/test/bionic/libc

The first one can be used to benchmark simple stdio operation
(for now fread() / fwrite()).

The second one can be used to benchmark misc. pthread operations.
It can also easily be built with other toolchains for comparison
with other C library / pthread implementations.

Change-Id: I3e611244f3ff95e7485d8b8d6f91e5b8bab2de4b

12 years agoMerge "making sure it is built into the nativetest path so it gets picked up by conti...
Tsu Chiang Chuang [Tue, 29 Nov 2011 16:59:47 +0000 (08:59 -0800)]
Merge "making sure it is built into the nativetest path so it gets picked up by continous tests."

12 years agomaking sure it is built into the nativetest path so it gets picked up by
Tsu Chiang Chuang [Tue, 29 Nov 2011 01:41:30 +0000 (17:41 -0800)]
making sure it is built into the nativetest path so it gets picked up by
continous tests.

Change-Id: I6c6841b629691dd6825025886cf8169afbc3df98

12 years agoDon't assume that the stack is always at 0x1000000.
Nick Kralevich [Fri, 18 Nov 2011 18:37:48 +0000 (10:37 -0800)]
Don't assume that the stack is always at 0x1000000.

With our ASLR changes, the stack is now located at a random
location in memory. Remove the hardcoded assumption that the
stack will always be between 0x10000000 and 0x40000000.

Change-Id: Ic0b6d9e3ae844f057ace81a982cec7e2b344e179

12 years agoMerge "Fix divide-by-zero in test-fb-refresh."
Iain Merrick [Tue, 8 Nov 2011 11:22:20 +0000 (03:22 -0800)]
Merge "Fix divide-by-zero in test-fb-refresh."

12 years agoRemove libraries from static library module.
Ying Wang [Tue, 8 Nov 2011 02:20:37 +0000 (18:20 -0800)]
Remove libraries from static library module.

Static library module does not link any library.
There is no host library named libz.

Change-Id: I952b2bbda9206fa8c97f3a3c397dcc59a1a42f20

12 years agoam 0dbb4f81: am 8be30e05: Merge "Include the shared memory and slab memory in procran...
Dianne Hackborn [Mon, 7 Nov 2011 15:51:36 +0000 (15:51 +0000)]
am 0dbb4f81: am 8be30e05: Merge "Include the shared memory and slab memory in procrank output." into ics-mr0

* commit '0dbb4f8155df7852810e257de767b1f701f2fe3c':
  Include the shared memory and slab memory in procrank output.

12 years agoFix divide-by-zero in test-fb-refresh.
Iain Merrick [Mon, 7 Nov 2011 15:46:48 +0000 (15:46 +0000)]
Fix divide-by-zero in test-fb-refresh.

On some platforms, pixclock is 0, so this tool was crashing with SIGFPE.

Change-Id: I51bcec2cf46b19c161aff6ec892ad2bdb0bd844f

12 years agoam 230b99fb: (-s ours) am 2c9d3f92: (-s ours) DO NOT MERGE. Integrate: showmap now...
Dianne Hackborn [Sat, 5 Nov 2011 19:33:34 +0000 (19:33 +0000)]
am 230b99fb: (-s ours) am 2c9d3f92: (-s ours) DO NOT MERGE.  Integrate: showmap now includes the number of maps in a row.

* commit '230b99fb850bbbe17b8ee41d1968062d72895fdd':
  DO NOT MERGE.  Integrate: showmap now includes the number of maps in a row.

12 years agoam 42ec7355: showmap now includes the number of maps in a row.
Dianne Hackborn [Sat, 5 Nov 2011 16:40:08 +0000 (16:40 +0000)]
am 42ec7355: showmap now includes the number of maps in a row.

* commit '42ec73551a7ed4798de53cb8ed3b34fa964dde98':
  showmap now includes the number of maps in a row.

12 years agoam 8be30e05: Merge "Include the shared memory and slab memory in procrank output...
Dianne Hackborn [Sat, 5 Nov 2011 07:32:29 +0000 (07:32 +0000)]
am 8be30e05: Merge "Include the shared memory and slab memory in procrank output." into ics-mr0

* commit '8be30e05a19e861a4605149e266e9b3e73542d1b':
  Include the shared memory and slab memory in procrank output.

12 years agoMerge "Include the shared memory and slab memory in procrank output." into ics-mr0
Dianne Hackborn [Sat, 5 Nov 2011 07:31:02 +0000 (00:31 -0700)]
Merge "Include the shared memory and slab memory in procrank output." into ics-mr0

12 years agoInclude the shared memory and slab memory in procrank output.
Dianne Hackborn [Sat, 5 Nov 2011 03:08:19 +0000 (20:08 -0700)]
Include the shared memory and slab memory in procrank output.

Change-Id: I0dac539c220fda867ce166d78dc898a8ec7aa18a

12 years agoam 2c9d3f92: (-s ours) DO NOT MERGE. Integrate: showmap now includes the number...
Dianne Hackborn [Fri, 4 Nov 2011 21:01:39 +0000 (21:01 +0000)]
am 2c9d3f92: (-s ours) DO NOT MERGE.  Integrate: showmap now includes the number of maps in a row.

* commit '2c9d3f9293fc4c511e8f031cdade3f0cb9d93127':
  DO NOT MERGE.  Integrate: showmap now includes the number of maps in a row.

12 years agoDO NOT MERGE. Integrate: showmap now includes the number of maps in a row.
Dianne Hackborn [Fri, 4 Nov 2011 02:17:31 +0000 (19:17 -0700)]
DO NOT MERGE.  Integrate: showmap now includes the number of maps in a row.

Change-Id: I02c52cbb08b429881e1191971ecf097b2187e793

12 years agoshowmap now includes the number of maps in a row.
Dianne Hackborn [Fri, 4 Nov 2011 02:17:31 +0000 (19:17 -0700)]
showmap now includes the number of maps in a row.

Change-Id: Ie8c5d01322ec7a2995a35ea050b38cc299568075

12 years agomake ext4 utils callable from C++
Doug Zongker [Fri, 28 Oct 2011 17:21:08 +0000 (10:21 -0700)]
make ext4 utils callable from C++

Change-Id: Ibbcff4cd2b07e3bd044459e498e4228a3608c27f

12 years agoMerge "qtaguid: Converting test to GTest so that it will be run continously by TF."
Tsu Chiang Chuang [Fri, 28 Oct 2011 16:42:04 +0000 (09:42 -0700)]
Merge "qtaguid: Converting test to GTest so that it will be run continously by TF."

12 years agoRename LOG() to ALOG() DO NOT MERGE
Steve Block [Wed, 12 Oct 2011 16:28:18 +0000 (17:28 +0100)]
Rename LOG() to ALOG()  DO NOT MERGE

See https://android-git.corp.google.com/g/#/c/141576

Bug: 5449033
Change-Id: I00b3eb4a1ebf0a52389064b3138b15c410a7cd97

12 years agoBring screenshot back to bugreport
Guang Zhu [Thu, 20 Oct 2011 21:48:17 +0000 (14:48 -0700)]
Bring screenshot back to bugreport

Since the key cord for screen shot is now strictly enforced, we no
longer get a free screen shot with bugreport. This change to bugreport
script will capture a screen shot first, then proceed to dumpstate.
SendBug is also updated accordingly to take 2 parameters: one for
bugreport path, the other for screen shot path.

Change-Id: I20fcea2f3a207b0dcd5341c8646c694dac12ec54

12 years agoqtaguid: Converting test to GTest so that it will be run continously by TF.
Tsu Chiang Chuang [Tue, 20 Sep 2011 00:08:25 +0000 (17:08 -0700)]
qtaguid: Converting test to GTest so that it will be run continously by TF.

Change-Id: I067ed277f168e466d09572fb6c21e51e0bdaa9f2

12 years agotests: add socket tagging test for qtaguid kernel netfilter module
JP Abgrall [Fri, 9 Sep 2011 21:35:17 +0000 (14:35 -0700)]
tests: add socket tagging test for qtaguid kernel netfilter module

Add a small test to verify that socket tagging is not hopelessly
broken.

Test for
 - invalid: arg count, command, tags.
 - double tagging
 - acct_tag changing
 - optional uid
 - 64bit unsigned vs signed
 - invalid sock_fd
 - multiple entries
 - untagging
 - uid/gid checking + max tags.

It tries its best to allow concurrent invocations, but it just
can not be sure that it always detects the in-kernel addresses of
the sockets it has tagged.

Change-Id: I4e372fcddbd0b1cd6ba5d8be4692e0f31ec71a99

12 years agoAdd screenshot to bugreport email if available
Alon Albert [Tue, 13 Sep 2011 18:11:47 +0000 (11:11 -0700)]
Add screenshot to bugreport email if available

Only works for email bugreport for now. Might add BugReported support later - requiers either
changes to BugReporter tool or zipping the file with is not great.

Change-Id: I1941a8436f6b5ba31b04334a89cad075f3bc836e

12 years agoBugReporter integration
Guang Zhu [Fri, 19 Aug 2011 00:48:43 +0000 (17:48 -0700)]
BugReporter integration

Queries PackageManager for a predefined intent, if it's resolvable,
hand the bugreport over to BugReporter. Otherwise same as before:
compose an intent to send out in mail with attachment

Change-Id: Ib7a5b49551ae9285ed46464277168487520d798b

12 years agoUpdate to follow framework change.
Dianne Hackborn [Tue, 16 Aug 2011 01:25:31 +0000 (18:25 -0700)]
Update to follow framework change.

Change-Id: I1247ba30e6e2e424a4cce12063a3c6af5db035e9

13 years agoadditional scripts for emailing bugreports after capture
Guang Zhu [Wed, 20 Jul 2011 01:44:41 +0000 (18:44 -0700)]
additional scripts for emailing bugreports after capture

First script calls dumpsys to generate the bugreport, second
script launches a cmd line tool that queries account manager
for possible accounts to set as "to:" field and launches an
intent to mail the bugreport out

This is done in 2 steps because it's more convenient to look
up account for emailing with a framework context, but to make
sure bugreport is captured, using dumpsys from shell script
is more reliable in case Dalvik VM crashes.

Change-Id: Ibb54ac793bbd59f016db9fe98118155168c9371e

13 years agoAdd information about totals to procrank.
Dianne Hackborn [Mon, 18 Jul 2011 19:41:50 +0000 (12:41 -0700)]
Add information about totals to procrank.

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

13 years agoMerge "showmap enhancements"
Jeff Brown [Thu, 14 Jul 2011 09:29:07 +0000 (02:29 -0700)]
Merge "showmap enhancements"

13 years agoshowmap enhancements
Jeff Brown [Mon, 11 Jul 2011 22:29:54 +0000 (15:29 -0700)]
showmap enhancements

Fixed a bug where maps were being coalesced even when they were
not necessarily coalesced.

Enhanced the BSS segment detection heuristic so that it takes
the map start / end addresses into account and displays the
inferred map type directly.

Automatically sort all maps by name or address as appropriate.

Fixed a bug parsing maps with paths that contain whitespace.

Enhanced the -a argument to show the same columns as the normal
mode but just prefixed with the virtual memory map information.

Change-Id: Ice78afb0c5b597683cb1a1ba65e3c10f10258f7c

13 years agoMerge "Round down the filesystem size to a multiple of the block size"
Ken Sumrall [Thu, 14 Jul 2011 00:34:22 +0000 (17:34 -0700)]
Merge "Round down the filesystem size to a multiple of the block size"

13 years agoRound down the filesystem size to a multiple of the block size
Ken Sumrall [Thu, 14 Jul 2011 00:27:07 +0000 (17:27 -0700)]
Round down the filesystem size to a multiple of the block size

If it's not a multiple of the block size, then errors are generated
when making sparse images.  This doesn't change the size of the generated
filesystem, because partial blocks are not included in the filesystem.

Change-Id: Ifc00ef7961f651ce22c6400759eadce2775d7811

13 years agoMerge "Handle EOF when reading /proc/<pid>/pagemap"
Colin Cross [Wed, 13 Jul 2011 20:49:01 +0000 (13:49 -0700)]
Merge "Handle EOF when reading /proc/<pid>/pagemap"

13 years agoHandle EOF when reading /proc/<pid>/pagemap
Colin Cross [Wed, 13 Jul 2011 20:35:20 +0000 (13:35 -0700)]
Handle EOF when reading /proc/<pid>/pagemap

Instead of handling maps with the name "[vectors]" specially,
silently ignore EOF when reading from /proc/<pid>/pagemap, which
occurs any time a a mapping is outside of the userspace range.

Change-Id: I674ade1eab6fd7732c6d9e120d0750cca5415b25

13 years agoMerge changes I5f0758bb,I03f90ae6
Colin Cross [Wed, 13 Jul 2011 06:11:44 +0000 (23:11 -0700)]
Merge changes I5f0758bb,I03f90ae6

* changes:
  Ignore the vector page in /proc/<pid>/maps
  Detect an error case and prevent printing an uninitialized variable

13 years agoIgnore the vector page in /proc/<pid>/maps
Colin Cross [Wed, 13 Jul 2011 05:32:43 +0000 (22:32 -0700)]
Ignore the vector page in /proc/<pid>/maps

In kernel 2.6.37, the vector page was added to /proc/<pid>/maps,
but because it is located above TASK_SIZE (usually 0xbf000000),
it is considered to be in the kernel's address space, not the
process', so it doesn't show up in /proc/<pid>/pagemap.

When the vector page is detected, using the name "[vectors]",
remove it from the map.

Change-Id: I5f0758bbe5d2b927056fa9fee684fea63dd0fa4b

13 years agoDetect an error case and prevent printing an uninitialized variable
Colin Cross [Wed, 13 Jul 2011 05:30:14 +0000 (22:30 -0700)]
Detect an error case and prevent printing an uninitialized variable

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

13 years agoMerge "Remove the simulator target from all makefiles. Bug: 5010576"
Jeff Brown [Tue, 12 Jul 2011 19:16:46 +0000 (12:16 -0700)]
Merge "Remove the simulator target from all makefiles. Bug: 5010576"

13 years agoRemove the simulator target from all makefiles.
Jeff Brown [Tue, 12 Jul 2011 05:12:33 +0000 (22:12 -0700)]
Remove the simulator target from all makefiles.
Bug: 5010576

Change-Id: I5a7df94d99dcb6f1de98bd9b8f659fd51a491c82

13 years agoMerge "script for capturing and mailing bugreport"
Guang Zhu [Mon, 11 Jul 2011 23:45:41 +0000 (16:45 -0700)]
Merge "script for capturing and mailing bugreport"

13 years agoscript for capturing and mailing bugreport
Guang Zhu [Fri, 8 Jul 2011 18:21:58 +0000 (11:21 -0700)]
script for capturing and mailing bugreport

Change-Id: Ie6b25393bfd61be54dc6b58b81bffbc2d26b40f5

13 years agoMore robust parsing for smaps
Kenny Root [Wed, 6 Jul 2011 21:59:57 +0000 (14:59 -0700)]
More robust parsing for smaps

Change-Id: Ica273bf302bbc5025e48aa62eff5bf64fbdcac77

13 years agoUpdate simg2img to handle CHUNK_TYPE_FILL.
Ken Sumrall [Thu, 30 Jun 2011 19:15:00 +0000 (12:15 -0700)]
Update simg2img to handle CHUNK_TYPE_FILL.

Change-Id: Ia4d178b535e1c154cf2b8e826ea1fba342b05658

13 years agoBug fixes to make_ext4fs
Ken Sumrall [Thu, 30 Jun 2011 03:28:30 +0000 (20:28 -0700)]
Bug fixes to make_ext4fs

Fix definition of uuid struct.
Properly set the block_group number in each backup copy of the superblock.
Adjust the info.total_inodes field after rounding up the inodes per block group.
Add the option -t to initialize all the inode tables.  If also specified with
the -s option, use the CHUNK_TYPE_FILL sparse records to initialize them.

Change-Id: Idd8bcee1b9bde3e82ad8da89ef974fbc12d7a6c6

13 years agoAdd test/debugging code to ext4fixup
Ken Sumrall [Thu, 23 Jun 2011 01:34:28 +0000 (18:34 -0700)]
Add test/debugging code to ext4fixup

Add debugging test code to specify where to bail partway through
Add a script to drive the test code for automated testing.  This
found 4 bugs!

Change-Id: I14dc8b1e2c9d2d2f332346958d488feaf768d396

13 years agoFixes for ext4fixup.
Ken Sumrall [Wed, 22 Jun 2011 23:50:24 +0000 (16:50 -0700)]
Fixes for ext4fixup.

Add check for filesystem that needs the journal run
Fix check for a cleanly unmounted filesystem.
Fix computation of new inode number
Fix check in two places for high bit being set on inode num
Only run sanity check pass if filesystem fixup state is unset
  (i.e. not partway through the conversion)

Change-Id: Ib5521e7f3c04d32c02d9890644a99378e3a3659e

13 years agoext4fixup tool to fix wonky filesystems made by old version of make_ext4fs
Ken Sumrall [Fri, 10 Jun 2011 04:24:42 +0000 (21:24 -0700)]
ext4fixup tool to fix wonky filesystems made by old version of make_ext4fs

The filesystems made by make_ext4fs for the Gingerbread and initial Honeycomb
releases did not round up the number of inodes/block_group to use all the
space available.  This worked, but newer kernels have additional code that
refuses to write to these filesystems.  There is a kernel patch available
to fix the problem in the kernel, but this tool was also written to fixup
the filesystems to be compliant with the new kernels.

Change-Id: I317b02ee2132ce6a0daeb4d1b40ad5ef57a0fb32

13 years agoA few small cleanups to setup_fs.c
Ken Sumrall [Mon, 6 Jun 2011 20:24:13 +0000 (13:24 -0700)]
A few small cleanups to setup_fs.c

Based on code review comments when the code was moved from
crespo specific code to ext4 generic code.

Change-Id: Ia09261563863f6db1d5eadb5a3677b0b38382b93

13 years agofix fd leaking and deallocation freed resource
Robert CH Chou [Sat, 4 Jun 2011 14:04:53 +0000 (22:04 +0800)]
fix fd leaking and deallocation freed resource

Change-Id: Ibee4d7000bce5bc8c2e45339d4d53a242569ae4e

13 years agoMove the setup_fs program from device/samsung/crespo to ext4_utils.
Ken Sumrall [Fri, 3 Jun 2011 21:38:27 +0000 (14:38 -0700)]
Move the setup_fs program from device/samsung/crespo to ext4_utils.

Prime also uses this, so it's no longer crespo specfic.

Change-Id: Idd35ee85599717a06902f696a73e5a4bd6522ac6

13 years agoReconcile with honeycomb-release
The Android Open Source Project [Fri, 20 May 2011 21:19:44 +0000 (14:19 -0700)]
Reconcile with honeycomb-release

Change-Id: Ic2b2fe51f052e8b2a2d04f058354490e9cea07ae

13 years agoMerge "Make inodes per group a multiple of inodes per block"
Colin Cross [Sun, 1 May 2011 02:08:09 +0000 (19:08 -0700)]
Merge "Make inodes per group a multiple of inodes per block"

13 years agoMake inodes per group a multiple of inodes per block
Colin Cross [Fri, 29 Apr 2011 23:45:48 +0000 (16:45 -0700)]
Make inodes per group a multiple of inodes per block

Kernel version 2.6.37 and later refuse to write to filesystems
for which inodes per group is not divisible by inodes per block.
Round the inodes per group up.

Change-Id: I9eb4fb5dd616a8abfd2f1fdc4b7d05538a2fc8fe

13 years agotests: fb-simple: add 32bit support
Dima Zavin [Tue, 11 Jan 2011 18:46:25 +0000 (10:46 -0800)]
tests: fb-simple: add 32bit support

Add a commandline argument to provide bpp.

test-fb-simple <bpp> <w> <h> <buffer#>

Valid values for bpp are 16 and 32.

Change-Id: I60485700c379746948abda6f03d617d9f605fb7b
Signed-off-by: Dima Zavin <dima@android.com>
13 years agoProperly compute the size of a directory
Ken Sumrall [Thu, 31 Mar 2011 23:11:52 +0000 (16:11 -0700)]
Properly compute the size of a directory

When computing the size of a directory that is being created,
make sure to include the size of the null entry at the end of
the list of entries in that directory.

Change-Id: I7deb05a170911cd0cc8517600c9747f64915606f

13 years agoam 2ae7663e: Implement poor man\'s exception handling to prevent errors from killling...
Ken Sumrall [Fri, 25 Mar 2011 05:37:01 +0000 (22:37 -0700)]
am 2ae7663e: Implement poor man\'s exception handling to prevent errors from killling recovery

* commit '2ae7663e1e064000356ee9e49ebd08bdddc545f9':
  Implement poor man's exception handling to prevent errors from killling recovery

13 years agoImplement poor man's exception handling to prevent errors from killling recovery
Ken Sumrall [Thu, 24 Mar 2011 05:08:53 +0000 (22:08 -0700)]
Implement poor man's exception handling to prevent errors from killling recovery

make_ext4fs was originally designed to be a standalone program, and the
error() macro calls exit() after printing a diagnostic.  However, when
it was turned into a library that recovery links against, calling exit()
kills the entire recovery process, which the kernel starts back up,
which dies, and the cycle continues.

This changes make_ext4fs to use setjmp() early, and changes error to
use longjmp to return to that point.  When longjmp is called, the
error handler just returns an error.  If it was invoked as a standalone
program, this causes the program to exit with an error.  If it was
invoked as a library function, the error return is gracefully handled
by recovery.

Change-Id: Iad50f46cac0ac399b5cdfd083d5712f88a10e48e

13 years agoMerge commit '427c3a2177eb4d34c88406bb86107548737168cc' into honeycomb-release
The Android Automerger [Wed, 23 Mar 2011 19:19:41 +0000 (12:19 -0700)]
Merge commit '427c3a2177eb4d34c88406bb86107548737168cc' into honeycomb-release

13 years agoam 8c00dae6: am 427c3a21: Make a failure of the BLKDISCARD ioctl non-fatal.
Ken Sumrall [Wed, 23 Mar 2011 03:42:37 +0000 (20:42 -0700)]
am 8c00dae6: am 427c3a21: Make a failure of the BLKDISCARD ioctl non-fatal.

* commit '8c00dae657ce7941121ec218d07adeab9ee3d8a1':
  Make a failure of the BLKDISCARD ioctl non-fatal.

13 years agoam e2e962fd: am 3e620597: Fix to properly wipe filesystems larger than 4 gigabytes.
Ken Sumrall [Wed, 23 Mar 2011 03:40:59 +0000 (20:40 -0700)]
am e2e962fd: am 3e620597: Fix to properly wipe filesystems larger than 4 gigabytes.

* commit 'e2e962fdf6fa3d29986b92438e2f9c2b8e9d2284':
  Fix to properly wipe filesystems larger than 4 gigabytes.

13 years agoam 427c3a21: Make a failure of the BLKDISCARD ioctl non-fatal.
Ken Sumrall [Wed, 23 Mar 2011 03:37:38 +0000 (20:37 -0700)]
am 427c3a21: Make a failure of the BLKDISCARD ioctl non-fatal.

* commit '427c3a2177eb4d34c88406bb86107548737168cc':
  Make a failure of the BLKDISCARD ioctl non-fatal.

13 years agoam 3e620597: Fix to properly wipe filesystems larger than 4 gigabytes.
Ken Sumrall [Wed, 23 Mar 2011 03:35:56 +0000 (20:35 -0700)]
am 3e620597: Fix to properly wipe filesystems larger than 4 gigabytes.

* commit '3e620597671c7efa678c1cd804202cff513bc751':
  Fix to properly wipe filesystems larger than 4 gigabytes.

13 years agoMake a failure of the BLKDISCARD ioctl non-fatal.
Ken Sumrall [Wed, 23 Mar 2011 03:18:02 +0000 (20:18 -0700)]
Make a failure of the BLKDISCARD ioctl non-fatal.

Failure to erase the partition before making a new filesystem
should not be fatal, especially since this is called from recovery,
and if make_ext4fs() dies, recovery dies.

Change-Id: I1b50b8c486d403a5d9cbd5e9de73cf752d8ef8f7

13 years agoFix to properly wipe filesystems larger than 4 gigabytes.
Ken Sumrall [Sat, 12 Feb 2011 02:36:42 +0000 (18:36 -0800)]
Fix to properly wipe filesystems larger than 4 gigabytes.

Pass the wipe size to the wipe function in a 64 bit int.

Change-Id: Ia6a0c70c8ddacaae35d17732d61b126219e5c158

13 years agosystem/extras: remove LOCAL_PRELINK_MODULE
Iliyan Malchev [Mon, 14 Mar 2011 21:02:18 +0000 (14:02 -0700)]
system/extras: remove LOCAL_PRELINK_MODULE

Change-Id: Iafd936bcb30e3fcc18c522a701d567dd744f2fa3
Signed-off-by: Iliyan Malchev <malchev@google.com>
13 years agoFix to properly wipe filesystems larger than 4 gigabytes.
Ken Sumrall [Sat, 12 Feb 2011 02:36:42 +0000 (18:36 -0800)]
Fix to properly wipe filesystems larger than 4 gigabytes.

Pass the wipe size to the wipe function in a 64 bit int.

Change-Id: Iac95693efa2d9cadbe77aa83b5ff01d7fc8d8b70

13 years agoFix for X86 target
Kenny Root [Tue, 8 Feb 2011 01:18:53 +0000 (17:18 -0800)]
Fix for X86 target

The X86 target couldn't compile some tests that use assembly, but it can
compile other things that now depend on some projects in the test subdir
(namely libtestUtil for OpenGL tests)

Also fatblock was including <sys/endian.h> directly instead of
<endian.h> which broke some stuff on X86

Change-Id: Ib71a6e548189ca2eacfd0caa7dbd17d200efd0ea

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Sat, 5 Feb 2011 06:25:55 +0000 (22:25 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Sat, 5 Feb 2011 05:55:36 +0000 (21:55 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Sat, 5 Feb 2011 01:35:20 +0000 (17:35 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Fri, 4 Feb 2011 19:49:03 +0000 (11:49 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Fri, 4 Feb 2011 18:25:58 +0000 (10:25 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Fri, 4 Feb 2011 15:18:44 +0000 (07:18 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Thu, 3 Feb 2011 23:26:14 +0000 (15:26 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Thu, 3 Feb 2011 19:22:57 +0000 (11:22 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Thu, 3 Feb 2011 15:17:16 +0000 (07:17 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Thu, 3 Feb 2011 00:12:36 +0000 (16:12 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Wed, 2 Feb 2011 15:21:27 +0000 (07:21 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Tue, 1 Feb 2011 18:11:56 +0000 (10:11 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Tue, 1 Feb 2011 01:24:22 +0000 (17:24 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agotests:bionic: small improvement to dlclose test.
David 'Digit' Turner [Sun, 2 Jan 2011 09:05:21 +0000 (10:05 +0100)]
tests:bionic: small improvement to dlclose test.

Change-Id: I2fa0e06fab770a42b95006751baca07bb0e19d1e

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Sun, 30 Jan 2011 21:37:28 +0000 (13:37 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agomerge in honeycomb-release history after reset to honeycomb
The Android Automerger [Sun, 30 Jan 2011 17:58:42 +0000 (09:58 -0800)]
merge in honeycomb-release history after reset to honeycomb

13 years agoam bb079ff7: ext4_utils: Fix non-linux builds
Colin Cross [Sat, 29 Jan 2011 04:00:41 +0000 (20:00 -0800)]
am bb079ff7: ext4_utils: Fix non-linux builds

* commit 'bb079ff767d5dd76be2d8753a3aef192b75b5664':
  ext4_utils: Fix non-linux builds

13 years agoext4_utils: Fix non-linux builds
Colin Cross [Sat, 29 Jan 2011 03:54:36 +0000 (19:54 -0800)]
ext4_utils: Fix non-linux builds

Change-Id: I5a22083645055abc36a3addc4eae45b40a7a6b2b