This is the mail archive of the
mailing list for the binutils project.
Re: [ARM] PR ld/21402, only override the symbol dynamic decision on undefined weak symbol
- From: Jiong Wang <wong dot kwongyuan dot tools at gmail dot com>
- To: Renlin Li <renlin dot li at foss dot arm dot com>
- Cc: "binutils at sourceware dot org" <binutils at sourceware dot org>, Nicholas Clifton <nickc at redhat dot com>, Ramana Radhakrishnan <Ramana dot Radhakrishnan at arm dot com>
- Date: Wed, 27 Sep 2017 10:07:20 +0100
- Subject: Re: [ARM] PR ld/21402, only override the symbol dynamic decision on undefined weak symbol
- Authentication-results: sourceware.org; auth=none
- References: <59C26C78.email@example.com> <CAAfDdZ2zoce2=C1tKuN0f62T3xHvZUTaHzxnqcDa6ogjS_cxjQ@mail.gmail.com> <59C40506.firstname.lastname@example.org>
> * testsuite/ld-arm/tls-app.d: Update address.
> * testsuite/ld-arm/tls-app.r: Remove relocations.
I have concerns on these testcases changes.
They look not like something straightforward to understand.
Could you explain why some dynamic TLS relocations are removed? This
patch doesn't touch TLS relaxation, so how could some dynamic TLS
relocations been resolved at static linking time?
My concern is this patch has break some internal logic that caused
runtime TLS relocations missing, this will cause wrong tls address
> * testsuite/ld-arm/unresolved-1-dyn.d: Update.
This looks to me is redundant dynamic relocation that the linker
should not generate it at all.