gdb/gdb-rhbz1971096-glibc2.34-2.patch
Kevin Buettner 71264766c5 Backport upstream patches which fix multi-threaded debugging for
glibc-2.34 (RHBZ 1971096, Simon Marchi, Kevin Buettner).
2021-06-14 23:24:14 -07:00

150 lines
5.9 KiB
Diff

From FEDORA_PATCHES Mon Sep 17 00:00:00 2001
From: Kevin Buettner <kevinb@redhat.com>
Date: Mon, 14 Jun 2021 19:20:07 -0700
Subject: gdb-rhbz1971096-glibc2.34-2.patch
;; Backport patch fixing gdb on glibc-2.34 machines w/ regard to
;; libthread_db initialization.
libthread_db initialization changes related to upcoming glibc-2.34
This commit makes some adjustments to accomodate the upcoming
glibc-2.34 release. Beginning with glibc-2.34, functionality formerly
contained in libpthread has been moved to libc. For the time being,
libpthread.so still exists in the file system, but it won't show up in
ldd output and therefore won't be able to trigger initialization of
libthread_db related code. E.g...
Fedora 34 / glibc-2.33.9000:
[kev@f34-2 gdb]$ ldd testsuite/outputs/gdb.threads/tls/tls
linux-vdso.so.1 (0x00007ffcf94fa000)
libstdc++.so.6 => /lib64/libstdc++.so.6 (0x00007ff0ba9af000)
libm.so.6 => /lib64/libm.so.6 (0x00007ff0ba8d4000)
libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00007ff0ba8b9000)
libc.so.6 => /lib64/libc.so.6 (0x00007ff0ba6c6000)
/lib64/ld-linux-x86-64.so.2 (0x00007ff0babf0000)
Fedora 34 / glibc-2.33:
[kev@f34-1 gdb]$ ldd testsuite/outputs/gdb.threads/tls/tls
linux-vdso.so.1 (0x00007fff32dc0000)
libpthread.so.0 => /lib64/libpthread.so.0 (0x00007f815f6de000)
libstdc++.so.6 => /lib64/libstdc++.so.6 (0x00007f815f4bf000)
libm.so.6 => /lib64/libm.so.6 (0x00007f815f37b000)
libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00007f815f360000)
libc.so.6 => /lib64/libc.so.6 (0x00007f815f191000)
/lib64/ld-linux-x86-64.so.2 (0x00007f815f721000)
Note that libpthread is missing from the ldd output for the
glibc-2.33.9000 machine.
This means that (unless we happen to think of some entirely different
mechanism), we'll now need to potentially match "libc" in addition to
"libpthread" as libraries which might be thread libraries. This
accounts for the change made in solib.c. Note that the new code
attempts to match "/libc." via strstr(). That trailing dot (".")
avoids inadvertently matching libraries such as libcrypt (and
all the other many libraries which begin with "libc").
To avoid attempts to load libthread_db when encountering older
versions of libc, we now attempt to find "pthread_create" (which is a
symbol that we'd expect to be in any pthread library) in the
associated objfile. This accounts for the changes in
linux-thread-db.c.
I think that other small adjustments will need to be made elsewhere
too. I've been working through regressions on my glibc-2.33.9000
machine; I've fixed some fairly "obvious" changes in the testsuite
(which are in other commits). For the rest, it's not yet clear to me
whether the handful of remaining failures represent a problem in glibc
or gdb. I'm still investigating, however, I'll note that these are
problems that I only see on my glibc-2.33.9000 machine.
gdb/ChangeLog:
* solib.c (libpthread_name_p): Match "libc" in addition
to "libpthread".
* linux-thread-db.c (libpthread_objfile_p): New function.
(libpthread_name_p): Adjust preexisting callers to use
libpthread_objfile_p().
diff --git a/gdb/linux-thread-db.c b/gdb/linux-thread-db.c
--- a/gdb/linux-thread-db.c
+++ b/gdb/linux-thread-db.c
@@ -800,6 +800,24 @@ check_thread_db (struct thread_db_info *info, bool log_progress)
return test_passed;
}
+/* Predicate which tests whether objfile OBJ refers to the library
+ containing pthread related symbols. Historically, this library has
+ been named in such a way that looking for "libpthread" in the name
+ was sufficient to identify it. As of glibc-2.34, the C library
+ (libc) contains the thread library symbols. Therefore we check
+ that the name matches a possible thread library, but we also check
+ that it contains at least one of the symbols (pthread_create) that
+ we'd expect to find in the thread library. */
+
+static bool
+libpthread_objfile_p (objfile *obj)
+{
+ return (libpthread_name_p (objfile_name (obj))
+ && lookup_minimal_symbol ("pthread_create",
+ NULL,
+ obj).minsym != NULL);
+}
+
/* Attempt to initialize dlopen()ed libthread_db, described by INFO.
Return true on success.
Failure could happen if libthread_db does not have symbols we expect,
@@ -1072,7 +1090,7 @@ try_thread_db_load_from_pdir (const char *subdir)
return false;
for (objfile *obj : current_program_space->objfiles ())
- if (libpthread_name_p (objfile_name (obj)))
+ if (libpthread_objfile_p (obj))
{
if (try_thread_db_load_from_pdir_1 (obj, subdir))
return true;
@@ -1181,7 +1199,7 @@ static bool
has_libpthread (void)
{
for (objfile *obj : current_program_space->objfiles ())
- if (libpthread_name_p (objfile_name (obj)))
+ if (libpthread_objfile_p (obj))
return true;
return false;
@@ -1295,7 +1313,7 @@ thread_db_new_objfile (struct objfile *objfile)
of the list of shared libraries to load, and in an app of several
thousand shared libraries, this can otherwise be painful. */
&& ((objfile->flags & OBJF_MAINLINE) != 0
- || libpthread_name_p (objfile_name (objfile))))
+ || libpthread_objfile_p (objfile)))
check_for_thread_db ();
}
diff --git a/gdb/solib.c b/gdb/solib.c
--- a/gdb/solib.c
+++ b/gdb/solib.c
@@ -906,12 +906,17 @@ Do you need \"set solib-search-path\" or \"set sysroot\"?"),
Uses a fairly simplistic heuristic approach where we check
the file name against "/libpthread". This can lead to false
- positives, but this should be good enough in practice. */
+ positives, but this should be good enough in practice.
+
+ As of glibc-2.34, functions formerly residing in libpthread have
+ been moved to libc, so "/libc." needs to be checked too. (Matching
+ the "." will avoid matching libraries such as libcrypt.) */
bool
libpthread_name_p (const char *name)
{
- return (strstr (name, "/libpthread") != NULL);
+ return (strstr (name, "/libpthread") != NULL
+ || strstr (name, "/libc.") != NULL );
}
/* Return non-zero if SO is the libpthread shared library. */