| d2c8f52189f8c2a13b88a107c6495f9d83196d2d |
|
28-Feb-2012 |
Andrew Hsieh <andrewhsieh@google.com> |
Fixed two 64-bit porting issues; Make pid/tid type consistent
1. In printf, use "%zu" for variable of type size_t
2. Print tid in %5d
3. Make type of pid/tid in AndroidLogEntry and logger_entry consistent
Change-Id: I3e3d9536ee58823f349a4734ae093d30eabe1bfe
/system/core/liblog/logprint.c
|
| a820a0e587b1f6be7283588b744cd33ba5723bdf |
|
27-Oct-2011 |
Jeff Sharkey <jsharkey@android.com> |
Gracefully handle truncated log messages.
Bug: 5522726
Change-Id: I8637c7da854ec1ecb321632c45ee9bc2edc94a67
/system/core/liblog/logprint.c
|
| e1ede1530ff21d3b8920c0cbbebb42ccff4fa22d |
|
19-Oct-2011 |
Nick Kralevich <nnk@google.com> |
liblog: do better checks of log messages.
Testing:
The following test cases all passed and generated log entries:
# echo -n '\03foo\0bar\0' > /dev/log/main
# echo -n '\03\0bar\0' > /dev/log/main
# echo -n '\03\0a\0' > /dev/log/main
The following entries were successfully processed by
logcat but produced no log entries:
# echo -n '\03\0\0' > /dev/log/main
# echo -n '\03a\0\0' > /dev/log/main
# echo -n '\03b\0\0' > /dev/log/main
Also tested the pathological error condition:
cat /dev/urandom > /dev/log/main
which produced many "+++ LOG: malformed log entry" errors.
Bug: 5478600
Change-Id: I53bc79507242dcfc14445746c29edf47be0a90b4
/system/core/liblog/logprint.c
|
| 63f4a84104c228c74cdf132f5cef4f5184ae04bb |
|
17-Oct-2011 |
Nick Kralevich <nnk@google.com> |
liblog: ensure that the message length is consistent
When parsing log entries which may have embedded \0s, it's
possible for entry->messageLen to not be the actual
length of the string in entry->message. Detect this condition.
Bug: 5417417
Change-Id: I712cac7696af7831e24765b5a1b345d6ff5fb407
/system/core/liblog/logprint.c
|
| 4bf3c02e026077d14a4512c7c2f71937da3c2d50 |
|
01-Oct-2011 |
Kenny Root <kroot@google.com> |
Add checking for log entry format
The log tag may be zero length if corrupted, so check for this
condition.
Change-Id: I7616226dabe78a85859b0ab53aca08f734dbdd84
/system/core/liblog/logprint.c
|
| 7df6349a63b89d279131ec7dd7e6ef1b065854f5 |
|
09-Mar-2010 |
The Android Open Source Project <initial-contribution@android.com> |
merge from open-source master
Change-Id: I698b3c16bdef8751cc857b00f815cb2f95465431
|
| e2bf2ea4d2846031edfc52b942ad53e5467243f6 |
|
01-Mar-2010 |
Joe Onorato <joeo@android.com> |
Make logcat print both the main and system buffers by default. Make SLOGx macros work.
/system/core/liblog/logprint.c
|
| b45b5c9f227473050ef785d11e518e947c8754fb |
|
11-Feb-2010 |
Keith Preston <keithpre@gmail.com> |
Fix Heap Corruption from too long of a TAG
snprintf has a weird return value. It returns what would have been written given a large enough buffer.
In the case that the prefix is longer then our buffer(128), it messes up the calculations below possibly causing heap corruption.
To avoid this we double check and set the length at the maximum (size minus null byte
/system/core/liblog/logprint.c
|
| abb9638e3696e393d6dc45439feda91073e52103 |
|
04-Nov-2008 |
Alexey Tarasov <tarasov@dodologics.com> |
1. added macro definitions to AndroidConfig.h
2. removed unnecessary include of <alloca.h>
/system/core/liblog/logprint.c
|
| 4f6e8d7a00cbeda1e70cc15be9c4af1018bdad53 |
|
21-Oct-2008 |
The Android Open Source Project <initial-contribution@android.com> |
Initial Contribution
/system/core/liblog/logprint.c
|