[Bug gold/28098] New: Linking with with libcudart_static.a produces an empty eh_frame_hdr section
https://sourceware.org/bugzilla/show_bug.cgi?id=28098 Bug ID: 28098 Summary: Linking with with libcudart_static.a produces an empty eh_frame_hdr section Product: binutils Version: 2.38 (HEAD) Status: UNCONFIRMED Severity: normal Priority: P2 Component: gold Assignee: ccoutant at gmail dot com Reporter: michael.krasnyk at gmail dot com CC: ian at airs dot com Target Milestone: --- Hi, linking with with libcudart_static.a produces an empty eh_frame_hdr section. Steps to reproduce: curl -s https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/cuda-cudart-dev-11-2_11.2.152-1_amd64.deb | dpkg --fsys-tarfile - | tar xOf - ./usr/local/cuda-11.2/targets/x86_64-linux/lib/libcudart_static.a > libcudart_static.a sha256sum libcudart_static.a 3ca014141258ae33abb175aa88a8282507a04d8a16d8f8dc9340e7831c5831e6 libcudart_static.a cat > main.c
[Bug gold/28098] Linking with with libcudart_static.a produces an empty eh_frame_hdr section
https://sourceware.org/bugzilla/show_bug.cgi?id=28098 --- Comment #1 from Michael Krasnyk --- Created attachment 13559 --> https://sourceware.org/bugzilla/attachment.cgi?id=13559&action=edit add filtering of R_X86_64_NONE entries Relocation section '.rela.eh_frame' has an R_X86_64_NONE entry 00013200 000a0002 R_X86_64_PC32 .text + 5ae80 00013238 R_X86_64_NONE 0 00013258 001e0002 R_X86_64_PC32 .text._ZN15CUOSdlsymLoaderIPFimPKcEED2Ev + 0 Please consider the patch that adds filtering R_X86_64_NONE entries in Track_relocs::advance -- You are receiving this mail because: You are on the CC list for the bug.
[Bug gold/28098] Linking with with libcudart_static.a produces an empty eh_frame_hdr section
https://sourceware.org/bugzilla/show_bug.cgi?id=28098 Alan Modra changed: What|Removed |Added Ever confirmed|0 |1 Status|UNCONFIRMED |ASSIGNED Assignee|ccoutant at gmail dot com |amodra at gmail dot com Last reconfirmed||2021-07-18 --- Comment #2 from Alan Modra --- Thanks for the analysis and patch. I'm going to ask you to do just a little more work. First, change the patch to skip r_info==0 relocs, which is R_*_NONE (usually, there are some targets that stupidly define r_type of 0 to be something other than a no-operation relocations) *and* r_sym of zero. That will match ld.bfd. Also, put a suitable comment in reloc.h, something like "..would be skipped, excluding r_info==0 relocs." and perhaps expand on this a little in the function comment in reloc.cc, mentioning R_*_NONE and r_sym of zero. Finally, check over all uses of advance() in gold to see that it isn't used to count relocs for an output array or something like that, which might be broken by your change. If you get to that point, post the patch to binut...@sourceware.org and I'll commit it. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug gold/28098] Linking with with libcudart_static.a produces an empty eh_frame_hdr section
https://sourceware.org/bugzilla/show_bug.cgi?id=28098 --- Comment #3 from Michael Krasnyk --- Thanks for your comments, Alan! I have sent the updated patch to the mailing list. To duplicate here from the email: Result values of Reloc::advance are used at five places https://sourceware.org/git/?p=binutils-gdb.git;a=blob;f=gold/ehframe.cc;h=8a19b21055924ae17c40f9f7c2ac4e421bf58392;hb=HEAD#l665 https://sourceware.org/git/?p=binutils-gdb.git;a=blob;f=gold/ehframe.cc;h=8a19b21055924ae17c40f9f7c2ac4e421bf58392;hb=HEAD#l688 https://sourceware.org/git/?p=binutils-gdb.git;a=blob;f=gold/ehframe.cc;h=8a19b21055924ae17c40f9f7c2ac4e421bf58392;hb=HEAD#l889 https://sourceware.org/git/?p=binutils-gdb.git;a=blob;f=gold/ehframe.cc;h=8a19b21055924ae17c40f9f7c2ac4e421bf58392;hb=HEAD#l922 https://sourceware.org/git/?p=binutils-gdb.git;a=blob;f=gold/ehframe.cc;h=8a19b21055924ae17c40f9f7c2ac4e421bf58392;hb=HEAD#l1018 for early termination of do_add_ehframe_input_section and at one place https://sourceware.org/git/?p=binutils-gdb.git;a=blob;f=gold/ehframe.cc;h=8a19b21055924ae17c40f9f7c2ac4e421bf58392;hb=HEAD#l919 in an assertion, so this change should not change behavior but prevent earlier termination if the relocation section has an entry with r_info = 0. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug gold/28098] Linking with with libcudart_static.a produces an empty eh_frame_hdr section
https://sourceware.org/bugzilla/show_bug.cgi?id=28098 --- Comment #4 from Michael Krasnyk --- Created attachment 13561 --> https://sourceware.org/bugzilla/attachment.cgi?id=13561&action=edit updated patch -- You are receiving this mail because: You are on the CC list for the bug.
