From: Szabolcs Nagy <nsz@port70.net>
To: Adhemerval Zanella Netto <adhemerval.zanella@linaro.org>
Cc: Szabolcs Nagy <szabolcs.nagy@arm.com>, libc-alpha@sourceware.org
Subject: Re: [PATCH] aarch64: Remove ld.so __tls_get_addr plt usage
Date: Sat, 6 Apr 2024 19:40:16 +0200 [thread overview]
Message-ID: <20240406174016.GD3766212@port70.net> (raw)
In-Reply-To: <0af468ae-bf65-4e20-8bd7-af85eba2e5d9@linaro.org>
* Adhemerval Zanella Netto <adhemerval.zanella@linaro.org> [2024-04-05 13:29:48 -0300]:
>
>
> On 05/04/24 11:58, Szabolcs Nagy wrote:
> > The 04/05/2024 09:35, Adhemerval Zanella wrote:
> >> Use the hidden alias instead.
> >>
> >> Checked on aarch64-linux-gnu.
> >
> > does this change behaviour in case __tls_get_addr is interposed?
> >
> >
>
> Do we support symbol interposition for symbols in implementation namespace?
> If so, we will need to revert x86 to use the same semantic and document it
> properly.
i suspect this will partially break the sanitizers
and possibly other similar tools that hook internals
so i hoped at least for better commit msg pointing
out that this is deliberate behaviour change.
>
> >> ---
> >> sysdeps/aarch64/dl-tlsdesc.S | 3 ++-
> >> sysdeps/unix/sysv/linux/aarch64/localplt.data | 2 --
> >> 2 files changed, 2 insertions(+), 3 deletions(-)
> >>
> >> diff --git a/sysdeps/aarch64/dl-tlsdesc.S b/sysdeps/aarch64/dl-tlsdesc.S
> >> index 9b253b39dd..4febf2ad21 100644
> >> --- a/sysdeps/aarch64/dl-tlsdesc.S
> >> +++ b/sysdeps/aarch64/dl-tlsdesc.S
> >> @@ -220,7 +220,7 @@ _dl_tlsdesc_dynamic:
> >> SAVE_Q_REGISTERS
> >>
> >> mov x0, x1
> >> - bl __tls_get_addr
> >> + bl HIDDEN_JUMPTARGET(__tls_get_addr)
> >>
> >> mrs x1, tpidr_el0
> >> sub PTR_REG (0), PTR_REG (0), PTR_REG (1)
> >> @@ -246,5 +246,6 @@ _dl_tlsdesc_dynamic:
> >> b 1b
> >> cfi_endproc
> >> .size _dl_tlsdesc_dynamic, .-_dl_tlsdesc_dynamic
> >> + .hidden HIDDEN_JUMPTARGET(__tls_get_addr)
> >> # undef NSAVEXREGPAIRS
> >> #endif
> >> diff --git a/sysdeps/unix/sysv/linux/aarch64/localplt.data b/sysdeps/unix/sysv/linux/aarch64/localplt.data
> >> index 5d217cc50d..5dd07472df 100644
> >> --- a/sysdeps/unix/sysv/linux/aarch64/localplt.data
> >> +++ b/sysdeps/unix/sysv/linux/aarch64/localplt.data
> >> @@ -9,5 +9,3 @@ libm.so: matherr
> >> # If outline atomics are used, libgcc (built outside of glibc) may
> >> # call __getauxval using the PLT.
> >> libc.so: __getauxval ?
> >> -# The dynamic loader needs __tls_get_addr for TLS.
> >> -ld.so: __tls_get_addr
> >> --
> >> 2.34.1
> >>
next prev parent reply other threads:[~2024-04-06 17:40 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-04-05 12:35 Adhemerval Zanella
2024-04-05 14:58 ` Szabolcs Nagy
2024-04-05 16:29 ` Adhemerval Zanella Netto
2024-04-06 17:40 ` Szabolcs Nagy [this message]
2024-04-08 8:04 ` Florian Weimer
2024-04-07 20:29 ` Cristian Rodríguez
2024-04-08 7:26 ` Szabolcs Nagy
2024-04-08 16:57 ` Adhemerval Zanella Netto
2024-04-09 8:30 ` Szabolcs Nagy
2024-04-09 14:03 ` Adhemerval Zanella Netto
2024-04-09 14:05 ` H.J. Lu
2024-04-09 14:11 ` Palmer Dabbelt
2024-04-09 14:46 ` H.J. Lu
2024-04-09 17:50 ` Fangrui Song
2024-04-10 7:29 ` Szabolcs Nagy
2024-04-10 8:23 ` Florian Weimer
2024-04-10 15:46 ` enh
2024-04-15 11:41 ` Florian Weimer
2024-04-15 20:22 ` Adhemerval Zanella Netto
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20240406174016.GD3766212@port70.net \
--to=nsz@port70.net \
--cc=adhemerval.zanella@linaro.org \
--cc=libc-alpha@sourceware.org \
--cc=szabolcs.nagy@arm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).