[Xenomai-core] Re: [PATCH] ipipe-tracer-v5

2005-12-28 Thread Philippe Gerum

Jan Kiszka wrote:

Hi Philippe,

this is a revised version of the tracer patch, addressing your concerns
hopefully in the right way. Changes since the previous release:

 * move __ipipe_init_trace_proc() declaration to linux/ipipe.h
 * define __BUILTIN_RETURN_ADDRESSx in linux/ipipe.h unless some arch
   defines BROKEN_BUILTIN_RETURN_ADDRESS and provides its own variants
 * improve ipipe_tsc2ns() for i386 so that it makes use of do_div (and
   also accepts 64 bit input)
 * move ipipe_tsc2us() to asm-i386/ipipe.h
 * fix a display bug in the output of frozen
 * reformatting
 * remove inclusion of linux/ipipe_trace.h from linux/sched.h

NOTE: The last change makes the local inclusion of ipipe_trace.h
necessary when instrumenting some code. Previous instrumentation patches
are likely broken now.

I think it should be ready for apply now, shouldn't it?



Yes; you guys did a nice job. v5 is on the merge queue, and will be 
applied tomorrow. We'll work on that basis for enhancements/fixes. Thanks.


--

Philippe.

___
Xenomai-core mailing list
Xenomai-core@gna.org
https://mail.gna.org/listinfo/xenomai-core


[Xenomai-core] Re: [PATCH] ipipe-tracer-v5

2005-12-28 Thread Philippe Gerum

Jan Kiszka wrote:

Hi Philippe,

this is a revised version of the tracer patch, addressing your concerns
hopefully in the right way. Changes since the previous release:

 * move __ipipe_init_trace_proc() declaration to linux/ipipe.h
 * define __BUILTIN_RETURN_ADDRESSx in linux/ipipe.h unless some arch
   defines BROKEN_BUILTIN_RETURN_ADDRESS and provides its own variants
 * improve ipipe_tsc2ns() for i386 so that it makes use of do_div (and
   also accepts 64 bit input)
 * move ipipe_tsc2us() to asm-i386/ipipe.h
 * fix a display bug in the output of frozen
 * reformatting
 * remove inclusion of linux/ipipe_trace.h from linux/sched.h

NOTE: The last change makes the local inclusion of ipipe_trace.h
necessary when instrumenting some code. Previous instrumentation patches
are likely broken now.

I think it should be ready for apply now, shouldn't it?



Yes; you guys did a nice job. v5 is on the merge queue, and will be 
applied tomorrow. We'll work on that basis for enhancements/fixes. Thanks.


--

Philippe.