This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: rtld: resolve incoming STT_GNU_IFUNC only after DT_INIT

From: Florian Weimer <fw at deneb dot enyo dot de>
To: John Reiser <jreiser at bitwagon dot com>
Cc: libc-alpha at sourceware dot org, Florian Weimer <fweimer at redhat dot com>
Date: Wed, 01 Nov 2017 18:24:31 +0100
Subject: Re: rtld: resolve incoming STT_GNU_IFUNC only after DT_INIT
Authentication-results: sourceware.org; auth=none
References: <9b03e519-3ec8-b603-dc19-b25379cbc9a5@bitwagon.com> <87h8ueufc1.fsf@mid.deneb.enyo.de> <86b5833b-b75f-e42c-523c-25dcde71b1e5@bitwagon.com>

* John Reiser:

> However, a relocation from some other module to a STT_GNU_IFUNC symbol
> in the current shared library cannot be intercepted like that.
> That's why DT_INIT should be called before such a relocation is resolved,
> so that the IFUNC code can be de-compressed into the right place
> before it gets called.

I don't think it's safe at all to call DT_INIT during relocation
processing.  You need to explain why you think it is.

I really don't see how we can make this happen.  As I said, DT_INIT
may need the relocations which result from IFUNC processing.

Follow-Ups:
- Re: rtld: resolve incoming STT_GNU_IFUNC only after DT_INIT
  - From: John Reiser

References:
- rtld: resolve incoming STT_GNU_IFUNC only after DT_INIT
  - From: John Reiser
- Re: rtld: resolve incoming STT_GNU_IFUNC only after DT_INIT
  - From: Florian Weimer
- Re: rtld: resolve incoming STT_GNU_IFUNC only after DT_INIT
  - From: John Reiser

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]