Re: tpm-test.patch does not cleanly apply

Jonathan McCune wrote on 2012-04-11:
> Hello list,
> 
> In both tboot-1.7.0 and the latest revision in the mercurial repo
> (299:950fec11ef90, dated 1/15/2012):
> 
> tboot-1.7.0 $ patch --dry-run -p1 < test-patches/tpm-test.patch
> patching file tboot/common/tpm.c Hunk #1 succeeded at 2121 with fuzz 1 (offset 418 lines).
> patching file tboot/include/tpm.h
> Hunk #1 succeeded at 272 with fuzz 2 (offset 15 lines).
> patching file tboot/txt/txt.c
> Hunk #1 FAILED at 475.
> Hunk #2 succeeded at 740 with fuzz 2 (offset 162 lines).
> 1 out of 2 hunks FAILED -- saving rejects to file tboot/txt/txt.c.rej
> 
> Changing the fuzz factor with -F3 causes the patch to "apply", but it's faulty.
> The call to tpm_unit_test_before_senter() ends up in the wrong place.
> 
> $ patch --dry-run -p1 -F3 < test-patches/tpm-test.patch
> 
> I found putting the call to tpm_unit_test_before_senter() at the end
> of the function txt_prepare_cpu() in tboot/txt/txt.c seems reasonable.
> 
> If desired I could prepare and post a proper patch, but it seems a
> very trivial edit.

Jon,

Thanks for raising this issue. The better place to put the call to tpm_unit_test_before_senter() should
be in tboot/common/tboot.c:begin_lauch(), and before "if ( !is_tpm_ready(0) )".