Re: Re: [RFC PATCH 4/4] ftrace/x86: Add support for -mfentry to x86_64

2012-08-10 Thread Masami Hiramatsu
(2012/08/09 22:46), Steven Rostedt wrote: > Peter and Masami > > During my final tests, I found that this change breaks the > !DYNAMIC_FTRACE config. That is, when we don't do the run-time updates > of mcount calls to nops, the compiler will use fentry but the code still > uses mcount. Ah,

Re: Re: [RFC PATCH 4/4] ftrace/x86: Add support for -mfentry to x86_64

2012-08-10 Thread Masami Hiramatsu
(2012/08/09 22:46), Steven Rostedt wrote: Peter and Masami During my final tests, I found that this change breaks the !DYNAMIC_FTRACE config. That is, when we don't do the run-time updates of mcount calls to nops, the compiler will use fentry but the code still uses mcount. Ah, right. we

Re: [RFC PATCH 4/4] ftrace/x86: Add support for -mfentry to x86_64

2012-08-09 Thread Steven Rostedt
On Thu, 2012-08-09 at 09:46 -0400, Steven Rostedt wrote: > Peter and Masami > > During my final tests, I found that this change breaks the > !DYNAMIC_FTRACE config. That is, when we don't do the run-time updates > of mcount calls to nops, the compiler will use fentry but the code still > uses

Re: [RFC PATCH 4/4] ftrace/x86: Add support for -mfentry to x86_64

2012-08-09 Thread Steven Rostedt
Peter and Masami During my final tests, I found that this change breaks the !DYNAMIC_FTRACE config. That is, when we don't do the run-time updates of mcount calls to nops, the compiler will use fentry but the code still uses mcount. I fixed this in the patch below. But as you two have acked and

Re: [RFC PATCH 4/4] ftrace/x86: Add support for -mfentry to x86_64

2012-08-09 Thread Masami Hiramatsu
(2012/08/08 4:38), Steven Rostedt wrote: > From: Steven Rostedt > > If the kernel is compiled with gcc 4.6.0 which supports -mfentry, > then use that instead of mcount. > > With mcount, frame pointers are forced with the -pg option and we > get something like: > > : >55

Re: [RFC PATCH 4/4] ftrace/x86: Add support for -mfentry to x86_64

2012-08-09 Thread Masami Hiramatsu
(2012/08/08 4:38), Steven Rostedt wrote: From: Steven Rostedt srost...@redhat.com If the kernel is compiled with gcc 4.6.0 which supports -mfentry, then use that instead of mcount. With mcount, frame pointers are forced with the -pg option and we get something like:

Re: [RFC PATCH 4/4] ftrace/x86: Add support for -mfentry to x86_64

2012-08-09 Thread Steven Rostedt
Peter and Masami During my final tests, I found that this change breaks the !DYNAMIC_FTRACE config. That is, when we don't do the run-time updates of mcount calls to nops, the compiler will use fentry but the code still uses mcount. I fixed this in the patch below. But as you two have acked and

Re: [RFC PATCH 4/4] ftrace/x86: Add support for -mfentry to x86_64

2012-08-09 Thread Steven Rostedt
On Thu, 2012-08-09 at 09:46 -0400, Steven Rostedt wrote: Peter and Masami During my final tests, I found that this change breaks the !DYNAMIC_FTRACE config. That is, when we don't do the run-time updates of mcount calls to nops, the compiler will use fentry but the code still uses mcount.