On Thu, 2019-07-11 at 11:11 +0800, Mao Han wrote: > On Wed, Jul 10, 2019 at 11:33:23PM +0200, Mark Wielaard wrote: > > On Mon, 2019-06-03 at 16:16 +0800, Mao Han wrote: > > > test binary: > > > https://github.com/c-sky/test-result/blob/master/elfutils/divzero > > > https://github.com/c-sky/test-result/blob/master/elfutils/core > > > > I was looking at these, but they seem to have a wrong e_machine > > value > > of EM_RCE, not EM_CSKY? Is that deliberate? > > > > I used an no upstream toolchain to verify coredump backtrace, which > use > the EM_CSKY_OLD equals to 39. We have changed the EM number to 252 in > our newly released toolchain with compatibility to EM_CSKY_OLD. > See binutils/include/elf/common.h: > /* C-SKY historically used 39, the same value as MCORE, from which > the > architecture was derived. */ > #define EM_CSKY_OLD EM_MCORE
Aha. But that is unfortunate on 2 levels. First EM_MCORE isn't an official e_machine name. The value 39 is associated with EM_RCE for the Morotola RCE. I assume those are the same thing? Second, it makes it hard to check the elfutils backend since you'll have to comment out all EM_RCE usage, and change the EM_CSKY to the old value. Do you have any binaries/core files generated by an upstream toolchain that uses the assigned ELF values? That would be really helpful running some tests. Thanks, Mark
