This is the mail archive of the systemtap@sourceware.org mailing list for the systemtap project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Any rescue other than recompiling the kernel when the kernel function is inlined?


Shiyao Ma <i@introo.me> writes:

> [...]
> Yes. fche says it's related to `-fvar-tracking-assignments'.

(It -might- be related.  That may just give you access to more data
values here and there.  .inline functions don't have noted .return
points, so it won't help with that.  Try statement probes addressed
by source file:line.)

> Since it's only related to dwarf, I'd guess it's feasible to recompile
> a kernel and use staprun on the product machine.

So compiling a new copy of the kernel, with a little different flags,
hoping that the executable parts match exactly, but more debuginfo is
generated on the side?  That -might- work, though you may have to
suppress other systemtap safety features (related to build-id) that are
meant to prevent such mismatches.  "at your own risk".

- FChE


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]