Use static TLS for libdyninstAPI_RT.so
This commit is contained in:
parent
2ee1cda802
commit
2dcffbf04a
51
dyninst-static-tls.patch
Normal file
51
dyninst-static-tls.patch
Normal file
@ -0,0 +1,51 @@
|
||||
From c4dca1b9bca83b180138dfa4d7e5a65736cccb31 Mon Sep 17 00:00:00 2001
|
||||
From: Josh Stone <jistone@redhat.com>
|
||||
Date: Tue, 21 Jun 2016 16:06:09 -0700
|
||||
Subject: [PATCH] RTlib: use static TLS for the tramp guard (#99)
|
||||
|
||||
With dynamic TLS, the call to __tls_get_addr() could reach additional
|
||||
instrumented code, infinitely recursing back to check the tramp guard.
|
||||
Static TLS is a limited resource, but this case in RTlib is special
|
||||
enough to warrant it for safety alone, nevermind performance.
|
||||
|
||||
(cherry picked from commit 73cd0019856eca0636e652e402f9eaed6ba9dc61)
|
||||
|
||||
Make the tramp guard bigger to avoid a glibc bug.
|
||||
|
||||
Having just one byte of TLS induces glibc bug 14898 with the definition
|
||||
of FORCED_DYNAMIC_TLS_OFFSET. Bumping to two bytes avoids this.
|
||||
|
||||
Fixes #101
|
||||
|
||||
(cherry picked from commit d1b4334e13e7cc4a3ee52a34c9d3f63ebff129b2)
|
||||
---
|
||||
dyninstAPI_RT/src/RTcommon.c | 11 +++++++++--
|
||||
1 file changed, 9 insertions(+), 2 deletions(-)
|
||||
|
||||
diff --git a/dyninstAPI_RT/src/RTcommon.c b/dyninstAPI_RT/src/RTcommon.c
|
||||
index ab125ff63f0e..f91ff12c943c 100644
|
||||
--- a/dyninstAPI_RT/src/RTcommon.c
|
||||
+++ b/dyninstAPI_RT/src/RTcommon.c
|
||||
@@ -117,10 +117,17 @@ int fakeTickCount;
|
||||
#if defined(_MSC_VER)
|
||||
#define TLS_VAR __declspec(thread)
|
||||
#else
|
||||
-#define TLS_VAR __thread
|
||||
+// Note, the initial-exec model gives us static TLS which can be accessed
|
||||
+// directly, unlike dynamic TLS that calls __tls_get_addr(). Such calls risk
|
||||
+// recursing back to us if they're also instrumented, ad infinitum. Static TLS
|
||||
+// must be used very sparingly though, because it is a limited resource.
|
||||
+// *** This case is very special -- do not use IE in general libraries! ***
|
||||
+#define TLS_VAR __thread __attribute__ ((tls_model("initial-exec")))
|
||||
#endif
|
||||
|
||||
-TLS_VAR int DYNINST_tls_tramp_guard = 1;
|
||||
+// It's tempting to make this a char, but glibc < 2.17 hits a bug:
|
||||
+// https://sourceware.org/bugzilla/show_bug.cgi?id=14898
|
||||
+static TLS_VAR short DYNINST_tls_tramp_guard = 1;
|
||||
|
||||
DLLEXPORT int DYNINST_lock_tramp_guard()
|
||||
{
|
||||
--
|
||||
2.7.4
|
||||
|
@ -2,7 +2,7 @@ Summary: An API for Run-time Code Generation
|
||||
License: LGPLv2+
|
||||
Name: dyninst
|
||||
Group: Development/Libraries
|
||||
Release: 4%{?dist}
|
||||
Release: 5%{?dist}
|
||||
URL: http://www.dyninst.org
|
||||
Version: 9.1.0
|
||||
Exclusiveos: linux
|
||||
@ -12,6 +12,7 @@ ExclusiveArch: %{ix86} x86_64 ppc ppc64
|
||||
Source0: http://www.paradyn.org/release%{version}/DyninstAPI-%{version}.tgz
|
||||
Source1: http://www.paradyn.org/release%{version}/Testsuite-%{version}.tgz
|
||||
Patch1: dyninst-export.patch
|
||||
Patch2: dyninst-static-tls.patch
|
||||
|
||||
%global dyninst_base DyninstAPI-%{version}
|
||||
%global testsuite_base Testsuite-%{version}
|
||||
@ -83,6 +84,7 @@ making sure that dyninst works properly.
|
||||
%setup -q -T -D -a 1
|
||||
|
||||
%patch1 -p0 -b .export
|
||||
%patch2 -d %{dyninst_base} -p1 -b .static-tls
|
||||
|
||||
%build
|
||||
|
||||
@ -171,6 +173,9 @@ find %{buildroot}%{_libdir}/dyninst/testsuite/ \
|
||||
%attr(644,root,root) %{_libdir}/dyninst/testsuite/*.a
|
||||
|
||||
%changelog
|
||||
* Tue Jun 21 2016 Josh Stone <jistone@redhat.com> - 9.1.0-5
|
||||
- Use static TLS for libdyninstAPI_RT.so
|
||||
|
||||
* Thu Mar 10 2016 William Cohen <wcohen@redhat.com> - 9.1.0-4
|
||||
- Export libdyninstAPI_RT_init_maxthreads (ref rhbz1315841)
|
||||
|
||||
|
Loading…
Reference in New Issue
Block a user