[lld] size comparison script

[Sean Silva via llvm-commits]

Is there a reason it uses readelf instead of an LLVM equivalent?

On Tue, Oct 20, 2015 at 1:05 PM, Rafael Espíndola <
llvm-commits at lists.llvm.org> wrote:

> This a small script I wrote for comparing the size of two ELF binaries.
>
> For example, this is what it prints when comparing a gold and a lld
> linked clang:
>
> .rodata                13979004 24179137
> .eh_frame               1967596  2475516
> .comment                    100    83161
> .rela.dyn                   480    20400
> .data                   3728960  3729104
> .bss                     144124   144152
> .hash                      2184     2192
> .interp                      28       28
> .init                        26       26
> .got                      18864    18864
> .tm_clone_table               0        0
> .note.ABI-tag                32       32
> .fini                         9        9
> .note.GNU-stack               0        0
> .fini_array                   8        8
> .init_array                1880     1880
> .jcr                          8        8
> .note.gnu.gold-version       28        0
> .got.plt                   2048     2016
> .plt                       4064     4000
> .shstrtab                   298      231
> .symtab                 5667000  5666928
> .rela.plt                  6072     5976
> .dynamic                    592      448
> .dynsym                    6744     6552
> .gnu.version_r              336        0
> .dynstr                    5315     4962
> .gnu.version                562        0
> .text                  41531081 41527673
> .strtab                 7841007  6802265
>
> So yes, SHF_STRING handling is next for me :-)
>
> Cheers,
> Rafael
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20151020/e2f82bf9/attachment.html>