| %|
--
1 100.000 no-vmlinux
2011/3/4 某因幡 tewil...@gmail.com
The log looks like:
qemu: fatal: Unimplemented cp15 register write (c9, c12, {0, 0})
R00= R01=1791a2cf R02=c0028000 R03=000e
R04=c0028000 R05= R06=002eda00 R07=
R08=c01ed584 R09=
your local changes and
make sure that all sample files are pulled from the device to the host.
-Ben
On Sat, Mar 5, 2011 at 9:56 PM, 某因幡 tewil...@gmail.com wrote:
Hi, I managed to run oprofile on a SE x10i.
But I only see the log below. Nothing else than no-vmlinux. Why?
My steps are:
1
and debug from there.
-Ben
On Sat, Mar 5, 2011 at 11:06 PM, 某因幡 tewil...@gmail.com wrote:
Hi, the fixed one is only replace python24 to python on the first line.
2011/3/6 Ben Cheng bcch...@android.com
opcontrol --status reports that there are 47 samples, but opimport_pull
only
=6013 -ZC- A svc32
Aborted
2011/3/3 某因幡 tewil...@gmail.com:
Hi, I recompiled goldfish kernel with oprofile support.
And then I did the following:
1. push oprofiled and opcontrol to /data/local and chmod, export PATH
2. pull /proc/kallsyms and find out _text and _etext
3. echo /data
Hi, I recompiled goldfish kernel with oprofile support.
And then I did the following:
1. push oprofiled and opcontrol to /data/local and chmod, export PATH
2. pull /proc/kallsyms and find out _text and _etext
3. echo /data/vmlinux
4. opcontrol --vmlinux=/vmlinux --kernel-range=${_text},${_etext}
my code is like this, holder is a SurfaceHolder, and colors is a int[]:
canvas = (*env)-CallObjectMethod(env, holder,
mid_SurfaceHolder_lockCanvas);
// JNI_TRUE for ARGB and JNI_FALSE for RGB565
(*env)-CallVoidMethod(env, canvas, mid_Canvas_drawBitmap,