on the Variant II code, however arm64 uses Variant I. The former placed the
thread pointer after the data, pointing at the thread control block, while
the latter places these before said data.
Because of this we need to use the size of the previous entry to calculate
where to place the current entry. We also need to reserve 16 bytes at the
start for the thread control block.
This also fixes the value of TLS_TCB_SIZE to be correct. This is the size
of two unsigned longs, i.e. 2 * 8 bytes.
While here remove the bogus adjustment of the pointer in the
R_AARCH64_TLS_TPREL64 case. It should be the offset of the data relative
to the thread pointer, including the thread control block.
Sponsored by: ABT Systems Ltd
have TLS program header. This is needed on architectures with Variant I
tls, that is arm, arm64, mips, and powerpc. These place the thread control
block at the start of the buffer and, without this, this data may be
trashed.
This appears to not be an issue on mips or powerpc as they include a second
adjustment to move the thread local data, however this is on arm64 (with a
future change to fix placing this data), and should be on arm. I am unable
to trigger this on arm, even after changing the code to move the data
around to make it more likely to be hit. This is most likely because my
tests didn't use the variable in offset 0.
Reviewed by: kib
MFC after: 1 week
Sponsored by: ABT Systems Ltd
Clang emits SSE instructions on amd64 in the common path of
pthread_mutex_unlock. If the thread does not otherwise use SSE,
this usage incurs a context-switch of the FPU/SSE state, which
reduces the performance of multiple real-world applications by a
non-trivial amount (3-5% in one application).
Instead of this change, I experimented with eagerly switching the
FPU state at context-switch time. This did not help. Most of the
cost seems to be in the read/write of memory--as kib@ stated--and
not in the #NM handling. I tested on machines with and without
XSAVEOPT.
One counter-argument to this change is that most applications already
use SIMD, and the number of applications and amount of SIMD usage
are only increasing. This is absolutely true. I agree that--in
general and in principle--this change is in the wrong direction.
However, there are applications that do not use enough SSE to offset
the extra context-switch cost. SSE does not provide a clear benefit
in the current libthr code with the current compiler, but it does
provide a clear loss in some cases. Therefore, disabling SSE in
libthr is a non-loss for most, and a gain for some.
I refrained from disabling SSE in libc--as was suggested--because
I can't make the above argument for libc. It provides a wide variety
of code; each case should be analyzed separately.
https://lists.freebsd.org/pipermail/freebsd-current/2015-March/055193.html
Suggestions from: dim, jmg, rpaulo
Approved by: kib (mentor)
MFC after: 2 weeks
Sponsored by: Dell Inc.
_rtld_bind. The compiler may generate code using these registers and not
save them. Unfortunately, as we make use of libc, we are unable to disallow
rtld from using floating-point register without also doing the same for the
parts of libc we use, or by limiting what _rtld_bind is able to call.
Obtained from: ABT Systems Ltd
Sponsored by: The FReeBSD Foundation
location pointer when the return value doesn't fit in a register, e.g. when
returning a struct.
Obtained from: ABT Systems Ltd
Sponsored by: The FreeBSD Foundation
When enough time has passed for users to update their userland the kernel
fix will be applied. This will change the ABI to have x0 point to the args
and sp be correctly aligned.
It is expected this compatibility code can be removed when the kernel and
qemu usermode emulation have both been updated for the new ABI.
This fixes clang failures, and most likely other crashes.
Obtained from: ABT Systems Ltd
Sponsored by: The FreeBSD Foundation
Off by default, build behaves normally.
WITH_META_MODE we get auto objdir creation, the ability to
start build from anywhere in the tree.
Still need to add real targets under targets/ to build packages.
Differential Revision: D2796
Reviewed by: brooks imp
The requirement is for a GCC-compatible compiler and not necessarily
GCC itself. However, we currently expect any compiler used for building
the whole of FreeBSD to be GCC-compatible and many things will break if
not; there's no longer a need to have an explicit test for this in rtld.
Reviewed by: imp, kib
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D2422
According to standard, the presence of the flags only means that the
object path must be resolved at the time object loading, instead of my
reading that the flag is required to enable token substitution at all.
The consequence is that -z origin linker flag is no longer required
for the token substitution in the run/rpath or the needed library
soname. It is only recommended if token substition is needed at
dlopen(3) time, since namecache might drop the required entries at the
time of resolution.
Found, reviewed and tested by: emaste
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
* Add VCREAT flag to indicate when a new file is being created
* Add VVERIFY to indicate verification is required
* Both VCREAT and VVERIFY are only passed on the MAC method vnode_check_open
and are removed from the accmode after
* Add O_VERIFY flag to rtld open of objects
* Add 'v' flag to __sflags to set O_VERIFY flag.
Submitted by: Steve Kiernan <stevek@juniper.net>
Obtained from: Juniper Networks, Inc.
GitHub Pull Request: https://github.com/freebsd/freebsd/pull/27
Relnotes: yes
as always participating in the global symbols namespace, regardless of
the way the object was brought into the process address space.
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
This is only an interim fix; MIPS should be using the MI code instead,
which does not have this issue.
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D661
ABI specifies that, for R_AARCH64_TLSDESC relocations, we use the symbol
value, addend, and object tls offset to calculate the offset from the tls
base. We then cache this value for future reference.
Differential Revision: https://reviews.freebsd.org/D2183
Reviewed by: kib
Sponsored by: The FreeBSD Foundation
still need libc_pic for a few things, but this is expected to be ready
soon.
Differential Revision: https://reviews.freebsd.org/D2136
Reviewed by: kib
Sponsored by: The FreeBSD Foundation
rtld on x86 to be hidden. This is a micro-optimization, which allows
intrinsic references inside rtld to be handled without indirection
through PLT. The visibility of rtld symbols for other objects in the
symbol namespace is controlled by a version script.
Reviewed by: kan, jilles
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
without any translation. If the file is a symbolic link, $ORIGIN may not be
expanded to the actual origin. Use realpath(3) to properly expand $ORIGIN
to its absolute path.
Reviewed by: kib
MFC after: 1 week
Avoid use of register variables, which some compilers (e.g. clang)
don't like. It makes the code a little clearer as well.
This allows a clang 3.5 built powerpc world to run (tested in a jail).
MFC after: 1 week
The symbol leaked after r276630 since lib/libc/sys/openat.c defines
versions for openat using .symver (version script cannot assign two
versions to one symbol), and rtld uses openat. Instead, directly use
__sys_openat().
Reported and tested by: antoine
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
rtld-elf for powerpc 32 bit:
libexec/rtld-elf/powerpc/reloc.c:486:6: error: taking the absolute value of unsigned type 'Elf_Addr' (aka 'unsigned int') has no effect [-Werror,-Wabsolute-value]
if (abs(offset) < 32*1024*1024) { /* inside 32MB? */
^
libexec/rtld-elf/powerpc/reloc.c:486:6: note: remove the call to 'abs' since unsigned values cannot be negative
if (abs(offset) < 32*1024*1024) { /* inside 32MB? */
^~~
1 error generated.
Cast 'offset' to int, since that was intended, and should be safe to do
on architectures with 32-bit two's complement ints.
Reviewed by: kib
Differential Revision: https://reviews.freebsd.org/D1387
it exports to the debugger. It currently has two choices: it can use
a compiled-in path (/libexec/ld-elf.so.1) or it can use the path stored
in the interpreter path in the binary being executed. The runtime linker
currently prefers the second. However, this is usually wrong for compat32
binaries since the binary specifies the path of rtld on a 32-bit system
(/libexec/ld-elf.so.1) instead of the actual path (/libexec/ld-elf32.so.1).
For now, always assume the compiled in path (/libexec/ld-elf32.so.1) as
the rtld path and ignore the path in the binary for the 32-bit runtime
linker.
Differential Revision: https://reviews.freebsd.org/D1236
Reviewed by: kib
It is automatically set when -fPIC is passed to the compiler.
Reviewed by: dim, kib
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D1179
