From: Tamar Christina <Tamar.Christina@arm.com>
To: "binutils@sourceware.org" <binutils@sourceware.org>
Cc: nd <nd@arm.com>, Richard Earnshaw <Richard.Earnshaw@arm.com>,
Marcus Shawcroft <Marcus.Shawcroft@arm.com>
Subject: Re: [PATCH][Binutils][AArch64] When DF_BIND_NOW don't use TLSDESC GOT value.
Date: Mon, 01 Apr 2019 13:00:00 -0000 [thread overview]
Message-ID: <20190401130020.GA7304@arm.com> (raw)
In-Reply-To: <DB6PR0802MB230904F995F0F469BB346D39FF5A0@DB6PR0802MB2309.eurprd08.prod.outlook.com>
[-- Attachment #1: Type: text/plain, Size: 4495 bytes --]
Hi All,
When using DF_BIND_NOW on AArch64 we don't reserve the GOT slot for a TLSDESC,
but we still emitted DT_TLSDESC_GOT and DT_TLSDESC_PLT. This caused random
memory corruption as the "special" value of (bfd_vma)-1 would be set for
dt_tlsdesc_got.
Since we don't have a value of dt_tlsdesc_got I also don't emit DT_TLSDESC_PLT
now becuase it would point to an incomplete PLT. To be able to write the PLT
entry DT_TLSDESC_GOT is needed and since we don't have one we can't write the
PLT entry either.
It is my understanding that GLIBC doesn't need these two entries when not lazy
loading. Conversely AArch32 does not reserve neither the GOT not the PLT slot
when doing DF_BIND_NOW.
AArch32 does not need these checks because these values are initialized to 0
and so the if (...) checks don't pass, but on AArch64 these are initialized
to (bfd_vma)-1 and thus we need some extra checks.
build on native hardware and regtested on
aarch64-none-elf, aarch64-none-elf (32 bit host),
aarch64-none-linux-gnu, aarch64-none-linux-gnu (32 bit host)
Cross-compiled and regtested on
aarch64-none-linux-gnu, aarch64_be-none-linux-gnu
and no issues, also fixes 3 other linker testcases.
Ok for master? and for backport to binutils-2.32?
Thanks,
Tamar
bfd/ChangeLog:
2019-04-01 Tamar Christina <tamar.christina@arm.com>
PR ld/24302
* elfnn-aarch64.c (elfNN_aarch64_size_dynamic_sections): Don't emit
DT_TLSDESC_GOT and DT_TLSDESC_PLT when DF_BIND_NOW.
(elfNN_aarch64_finish_dynamic_sections): Don't write PLT if DF_BIND_NOW.
ld/ChangeLog:
2019-04-01 Tamar Christina <tamar.christina@arm.com>
PR ld/24302
* testsuite/ld-aarch64/aarch64-elf.exp: Add new test.
* testsuite/ld-aarch64/tls-relax-gdesc-le-now.d: New test.
The 03/29/2019 19:16, Tamar Christina wrote:
> Hi All,
>
> I will post a slight modification on Monday after all checks finish for an small thing I noticed after sending this out.
>
> Please hold on review until then.
>
> Thanks,
> Tamar
>
> > -----Original Message-----
> > From: binutils-owner@sourceware.org <binutils-owner@sourceware.org>
> > On Behalf Of Tamar Christina
> > Sent: Friday, March 29, 2019 18:58
> > To: binutils@sourceware.org
> > Cc: nd <nd@arm.com>; Richard Earnshaw <Richard.Earnshaw@arm.com>;
> > Marcus Shawcroft <Marcus.Shawcroft@arm.com>
> > Subject: [PATCH][Binutils][AArch64] When DF_BIND_NOW don't use
> > TLSDESC GOT value.
> >
> > Hi All,
> >
> > When using DF_BIND_NOW on AArch64 we don't reserve the GOT slot for a
> > TLSDESC, but we still emitted DT_TLSDESC_GOT and DT_TLSDESC_PLT. This
> > caused random memory corruption as the "special" value of (bfd_vma)-1
> > would be set for dt_tlsdesc_got.
> >
> > Since we don't have a value of dt_tlsdesc_got I also don't emit
> > DT_TLSDESC_PLT now becuase it would point to an incomplete PLT. To be
> > able to write the PLT entry DT_TLSDESC_GOT is needed and since we don't
> > have one we can't write the PLT entry either.
> >
> > It is my understanding that GLIBC doesn't need these two entries when not
> > lazy loading. Conversely AArch32 does not reserve neither the GOT not the
> > PLT slot when doing DF_BIND_NOW.
> >
> > AArch32 does not need these checks because these values are initialized to 0
> > and so the if (...) checks don't pass, but on AArch64 these are initialized to
> > (bfd_vma)-1 and thus we need some extra checks.
> >
> > build on native hardware and regtested on
> > aarch64-none-elf, aarch64-none-elf (32 bit host),
> > aarch64-none-linux-gnu, aarch64-none-linux-gnu (32 bit host)
> >
> > Cross-compiled and regtested on
> > aarch64-none-linux-gnu, aarch64_be-none-linux-gnu
> >
> > and no issues, also fixes 3 other linker testcases.
> >
> > Ok for master? and for backport to binutils-2.32?
> >
> > Thanks,
> > Tamar
> >
> > bfd/ChangeLog:
> >
> > 2019-03-29 Tamar Christina <tamar.christina@arm.com>
> >
> > PR ld/24302
> > * elfnn-aarch64.c (elfNN_aarch64_size_dynamic_sections): Don't
> > emit
> > DT_TLSDESC_GOT and DT_TLSDESC_PLT when DF_BIND_NOW.
> > (elfNN_aarch64_finish_dynamic_sections): Don't write PLT if
> > DF_BIND_NOW.
> >
> > ld/ChangeLog:
> >
> > 2019-03-29 Tamar Christina <tamar.christina@arm.com>
> >
> > PR ld/24302
> > * testsuite/ld-aarch64/aarch64-elf.exp: Add new test.
> > * testsuite/ld-aarch64/tls-relax-gdesc-le-now.d: New test.
> >
> > --
--
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: rb10866.patch --]
[-- Type: text/x-diff; name="rb10866.patch", Size: 3481 bytes --]
diff --git a/bfd/elfnn-aarch64.c b/bfd/elfnn-aarch64.c
index 57a723d547734b4a4b5074fb7c90811e9bd56184..9d4df11f9d42c39fad5742f92a5edb561e809259 100644
--- a/bfd/elfnn-aarch64.c
+++ b/bfd/elfnn-aarch64.c
@@ -9064,13 +9064,13 @@ elfNN_aarch64_size_dynamic_sections (bfd *output_bfd ATTRIBUTE_UNUSED,
if (htab->root.splt->size == 0)
htab->root.splt->size += htab->plt_header_size;
- htab->tlsdesc_plt = htab->root.splt->size;
- htab->root.splt->size += htab->tlsdesc_plt_entry_size;
-
/* If we're not using lazy TLS relocations, don't generate the
- GOT entry required. */
+ GOT and PLT entry required. */
if (!(info->flags & DF_BIND_NOW))
{
+ htab->tlsdesc_plt = htab->root.splt->size;
+ htab->root.splt->size += htab->tlsdesc_plt_entry_size;
+
htab->dt_tlsdesc_got = htab->root.sgot->size;
htab->root.sgot->size += GOT_ENTRY_SIZE;
}
@@ -9174,6 +9174,7 @@ elfNN_aarch64_size_dynamic_sections (bfd *output_bfd ATTRIBUTE_UNUSED,
return FALSE;
if (htab->tlsdesc_plt
+ && !(info->flags & DF_BIND_NOW)
&& (!add_dynamic_entry (DT_TLSDESC_PLT, 0)
|| !add_dynamic_entry (DT_TLSDESC_GOT, 0)))
return FALSE;
@@ -9686,6 +9687,7 @@ elfNN_aarch64_finish_dynamic_sections (bfd *output_bfd,
case DT_TLSDESC_GOT:
s = htab->root.sgot;
+ BFD_ASSERT (htab->dt_tlsdesc_got != (bfd_vma)-1);
dyn.d_un.d_ptr = s->output_section->vma + s->output_offset
+ htab->dt_tlsdesc_got;
break;
@@ -9705,8 +9707,9 @@ elfNN_aarch64_finish_dynamic_sections (bfd *output_bfd,
this_hdr.sh_entsize = htab->plt_entry_size;
- if (htab->tlsdesc_plt)
+ if (htab->tlsdesc_plt && !(info->flags & DF_BIND_NOW))
{
+ BFD_ASSERT (htab->dt_tlsdesc_got != (bfd_vma)-1);
bfd_put_NN (output_bfd, (bfd_vma) 0,
htab->root.sgot->contents + htab->dt_tlsdesc_got);
diff --git a/ld/testsuite/ld-aarch64/aarch64-elf.exp b/ld/testsuite/ld-aarch64/aarch64-elf.exp
index ba3ce36191f49b2ce25880b7a61408b6e8875884..e69aedd0d4d1b434b659a9d426cf86c528610608 100644
--- a/ld/testsuite/ld-aarch64/aarch64-elf.exp
+++ b/ld/testsuite/ld-aarch64/aarch64-elf.exp
@@ -256,6 +256,7 @@ run_dump_test "tls-relax-all-ilp32"
run_dump_test "tls-relax-gd-le"
run_dump_test "tls-relax-gd-le-ilp32"
run_dump_test "tls-relax-gdesc-le"
+run_dump_test "tls-relax-gdesc-le-now"
run_dump_test "tls-relax-gdesc-le-ilp32"
run_dump_test "tls-relax-gd-ie"
run_dump_test "tls-relax-gd-ie-ilp32"
diff --git a/ld/testsuite/ld-aarch64/tls-relax-gdesc-le-now.d b/ld/testsuite/ld-aarch64/tls-relax-gdesc-le-now.d
new file mode 100644
index 0000000000000000000000000000000000000000..f1565e9c11f0136a5183694ecc22191617bf5cc2
--- /dev/null
+++ b/ld/testsuite/ld-aarch64/tls-relax-gdesc-le-now.d
@@ -0,0 +1,19 @@
+#source: tls-relax-gdesc-le.s
+#ld: -shared -z now
+#readelf: -dr
+#...
+ 0x.+ \(STRTAB\) \s+0x.+
+ 0x.+ \(SYMTAB\) \s+0x.+
+ 0x.+ \(STRSZ\) \s+.+ \(bytes\)
+ 0x.+ \(SYMENT\) \s+.+ \(bytes\)
+ 0x.+ \(PLTGOT\) \s+0x.+
+ 0x.+ \(PLTRELSZ\) \s+.+ \(bytes\)
+ 0x.+ \(PLTREL\) \s+RELA
+ 0x.+ \(JMPREL\) \s+0x.+
+ 0x.+ \(BIND_NOW\) \s+
+ 0x.+ \(FLAGS_1\) \s+ Flags: NOW
+ 0x.+ \(NULL\) \s+ 0x0
+
+Relocation section '\.rela\.plt' at offset .+ contains 1 entry:
+ Offset Info Type Sym\. Value Sym\. Name \+ Addend
+.+ .+ R_AARCH64_TLSDESC 0
next prev parent reply other threads:[~2019-04-01 13:00 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-03-29 18:58 Tamar Christina
2019-03-29 19:16 ` Tamar Christina
2019-04-01 13:00 ` Tamar Christina [this message]
2019-04-10 16:19 ` Nick Clifton
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=20190401130020.GA7304@arm.com \
--to=tamar.christina@arm.com \
--cc=Marcus.Shawcroft@arm.com \
--cc=Richard.Earnshaw@arm.com \
--cc=binutils@sourceware.org \
--cc=nd@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).