[llvm] r287786 - llvm-readobj: Use hash tables to print dynamic symbols.

Wed Nov 23 11:48:22 PST 2016

>Is this actually compatible with what the GNU tools emit? If so, fine (but could've >been mentioned in the commit message, I think). If not, I don't see why you >need to implement this in the GNU mode and not in readobj "default" mode. >Does this make sense?

Good point. Yes it is compatible to what GNU readelf does. This only changes the order in which the symbols are printed  as per buckets and chains in hash.

Do you want me to have this implemented in LLVM style?
--
Hemant Kulkarni
khemant at codeaurora.org
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by the Linux Foundation

-----Original Message-----
From: davide.italiano at gmail.com [mailto:davide.italiano at gmail.com] On Behalf Of Davide Italiano
Sent: Wednesday, November 23, 2016 1:38 PM
To: Hemant Kulkarni <khemant at codeaurora.org>
Cc: llvm-commits <llvm-commits at lists.llvm.org>
Subject: Re: [llvm] r287786 - llvm-readobj: Use hash tables to print dynamic symbols.

On Wed, Nov 23, 2016 at 11:33 AM, Hemant Kulkarni <khemant at codeaurora.org> wrote:
> This is not objdump change. This is ELFDumper change in llvm-readobj.
>

This is a change to llvm-readobj GNU style output, so I suppose it's something compatible with either objdump or readelf, no?

> The GNU style used symbols by using DT_* tags based sections alone. With this change, GNU style will go over the hash sections and print using the DT_* tags based sections. AFAIK there is no tool that can help validate if there is any issue with these linker created hash sections without running the binary.
>

Is this actually compatible with what the GNU tools emit? If so, fine (but could've been mentioned in the commit message, I think). If not, I don't see why you need to implement this in the GNU mode and not in readobj "default" mode. Does this make sense?

--
Davide

"There are no solved problems; there are only problems that are more or less solved" -- Henri Poincare