On Mon, Mar 19, 2012 at 5:23 PM, Petr Machata <[email protected]> wrote: > Randy MacLeod <[email protected]> writes: >> OE-core/YoctoProject.org. > > Hmm, don't you target MIPS among other arches? I'd be very surprised if > that stuff still worked. The changes for multi-threading were manifold, > and that MIPS stuff is all ifdef'd away and full of duplicate code. I'm > pretty sure I missed some edge case somewhere. It doesn't help that I > don't know MIPS at all.
Ok, I'll look at how the situation compares to: meta-oe/recipes-devtools/ltrace/ltrace_0.5.3.bb For a tool such as ltrace we may just accept it's limitations on certain targets at least to start. Once it's in the system, other developers might be convinced to fix the problems on their respective hosts. > I'm afraid the situation is similar, though not > that bleak, in case of ARM. Yes, a friend confirms that on arm the syscall tracing is okay but full tracing is not. Thanks for the tips. -- ../Randy/.. _______________________________________________ Ltrace-devel mailing list [email protected] http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/ltrace-devel
