ChangeLog:
2013-12-16 Will Newton <will.newton@linaro.org>
* manual/memory.texi (Malloc Examples): Mention aligned_alloc.
(Aligned Memory Blocks): Add documentation for aligned_alloc
and suggest it as an alternative to posix_memalign.
(Hooks for Malloc): Document __memalign_hook is also called
for aligned_alloc. (Summary of Malloc): Add summary for
aligned alloc. Document __memalign_hook is also called
for aligned_alloc.
The current documentation suggests using memalign and valloc which
are now considered obsolete, so suggest using posix_memalign instead.
Also document the possible error return and errno values for memalign
and posix_memalign and improve documentation of __memalign_hook.
ChangeLog:
2013-12-16 Will Newton <will.newton@linaro.org>
* manual/memory.texi (Malloc Examples): Clarify default
alignment documentation. Suggest posix_memalign rather
than memalign or valloc.
(Aligned Memory Blocks): Remove suggestion to use memalign
or valloc. Remove obsolete comment about BSD.
Document memalign errno values and mark the function obsolete.
Document posix_memalign returned error codes. Mark valloc
as obsolete. (Hooks for Malloc): __memalign_hook is also
called for posix_memalign and valloc.
(Summary of Malloc): Add posix_memalign to function summary.
__memalign_hook is also called for posix_memalign and valloc.
- Remove redundant mynumber union definitions
- Clean up a clumsy ternary operator
- Rename TAYLOR_SINCOS to TAYLOR_SIN since we're only expanding the
sin Taylor series in it.
Autoconf is tested for and run if needed only when --enable-maintainer-mode
is used on configure. This results in the autom4te.cache directory only
being written in the source directory during configure if automatic
autoconf usage is requested.
Fixes BZ #14120.
A sse42 version of strstr used pcmpistr instruction which is quite
ineffective. A faster way is look for pairs of characters which is uses
sse2, is faster than pcmpistr and for real strings a pairs we look for
are relatively rare.
For linear time complexity we use buy or rent technique which switches
to two-way algorithm when superlinear behaviour is detected.
For PPC64, all the wrappers at sysdeps are superfluous: they are
basically the same implementation from math/w_sqrt.c with the
'#ifdef _IEEE_LIBM'. And the power4 version just force the 'fsqrt'
instruction utilization with an inline assembly, which is already
handled by math_private.h __ieee754_sqrt implementation.
This patch adds GLIBC_2.3 and GLIBC_2.19 mark on Version.def so they are always
define in abi-versions.h. This fixes a build issue with fe_nomask.c in PPC64
LE where neither GLIBC_2_3 or GLIBC_2_19 are defined for SHLIB_COMPAT, resulting
in a wrong evaluation in the macro.
This patch add optimized __mpn_addmul, __mpn_addsub, __mpn_lshift, and
__mpn_mul_1 implementations for PowerPC64. They are originally from GMP
with adjustments for GLIBC.
`volatile int` means the same as 'int volatile', but that's not the
case for 'volatile char *' and 'char * volatile'. We won't need a
'char volatile *' or other complicated semantics for now.
This patch add static probes for setjmp/longjmp in the way gdb expects,fixing
the gdb.base/longjmp.exp gdb testcases.
It changes the symbol_name and use macros to to avoid change the probe names
and ending up adding more logic on GDB (since with the expected name
GDB work seamlessly).
This patch adds the ability to accept output arguments to functions
being benchmarked, by nesting the argument type in <> in the args
directive. It includes the sincos implementation as an example, where
the function would have the following args directive:
## args: double:<double *>:<double *>
This simply adds a definition for a static variable whose pointer gets
passed into the function, so it's not yet possible to pass something
more complicated like a pre-allocated string or array. That would be
a good feature to add if a function needs it.
The values in the input file will map only to the input arguments. So
if I had a directive like this for a function foo:
## args: int:<int *>:int:<int *>
and I have a value list like this:
1, 2
3, 4
5, 6
then the function calls generated would be:
foo (1, &out1, 2, &out2);
foo (3, &out1, 4, &out2);
foo (5, &out1, 6, &out2);
To avoid having a ELFv2 binary accidentally picking up an old ABI ld.so,
this patch bumps the soname to ld64.so.2.
In theory (or for testing purposes) this will also allow co-installing
ld.so versions for both ABIs on the same system. Note that the kernel
will already be able to load executables of both ABIs. However, there
is currently no plan to use that theoretical possibility in a any
supported distribution environment ...
Note that in order to check which ABI to use, we need to invoke the
compiler to check the _CALL_ELF macro; this is done in a new configure
check in sysdeps/unix/sysv/linux/powerpc/powerpc64/configure.ac,
replacing the hard-coded value of default-abi in the Makefile.
The ELFv2 ABI changes the calling convention by passing and returning
structures in registers in more cases than the old ABI:
http://gcc.gnu.org/ml/gcc-patches/2013-11/msg01145.htmlhttp://gcc.gnu.org/ml/gcc-patches/2013-11/msg01147.html
For the most part, this does not affect glibc, since glibc assembler
files do not use structure parameters / return values. However, one
place is affected: the LD_AUDIT interface provides a structure to
the audit routine that contains all registers holding function
argument and return values for the intercepted PLT call.
Since the new ABI now sometimes uses registers to return values
that were never used for this purpose in the old ABI, this structure
has to be extended. To force audit routines to be modified for the
new ABI if necessary, the patch defines v2 variants of the la_ppc64
types and routines.
In addition, the patch contains two unrelated changes to the
PLT trampoline routines: it fixes a bug where FPR return values
were stored in the wrong place, and it removes the unnecessary
save/restore of CR.
This updates glibc for the changes in the ELFv2 relating to the
stack frame layout. These are described in more detail here:
http://gcc.gnu.org/ml/gcc-patches/2013-11/msg01149.htmlhttp://gcc.gnu.org/ml/gcc-patches/2013-11/msg01146.html
Specifically, the "compiler and linker doublewords" were removed,
which has the effect that the save slot for the TOC register is
now at offset 24 rather than 40 to the stack pointer.
In addition, a function may now no longer necessarily assume that
its caller has set up a 64-byte register save area its use.
To address the first change, the patch goes through all assembler
files and replaces immediate offsets in instructions accessing the
ABI-defined stack slots by symbolic offsets. Those already were
defined in ucontext_i.sym and used in some of the context routines,
but that doesn't really seem like the right place for those defines.
The patch instead defines those symbolic offsets in sysdeps.h,
in two variants for the old and new ABI, and uses them systematically
in all assembler files, not just the context routines.
The second change only affected a few assembler files that used
the save area to temporarily store some registers. In those
cases where this happens within a leaf function, this patch
changes the code to store those registers to the "red zone"
below the stack pointer. Otherwise, the functions already allocate
a stack frame, and the patch changes them to add extra space in
these frames as temporary space for the ELFv2 ABI.
This is a follow-on to the previous patch to support the ELFv2 ABI in the
dynamic loader, split off into its own patch since it is just an optional
optimization.
In the ELFv2 ABI, most functions define both a global and a local entry
point; the local entry requires r2 to be already set up by the caller
to point to the callee's TOC; while the global entry does not require
the caller to know about the callee's TOC, but it needs to set up r12
to the callee's entry point address.
Now, when setting up a PLT slot, the dynamic linker will usually need
to enter the target function's global entry point. However, if the
linker can prove that the target function is in the same DSO as the
PLT slot itself, and the whole DSO only uses a single TOC (which the
linker will let ld.so know via a DT_PPC64_OPT entry), then it is
possible to actually enter the local entry point address into the
PLT slot, for a slight improvement in performance.
Note that this uncovered a problem on the first call via _dl_runtime_resolve,
because that routine neglected to restore the caller's TOC before calling
the target function for the first time, since it assumed that function
would always reload its own TOC anyway ...
This patch adds support for the ELFv2 ABI feature to remove function
descriptors. See this GCC patch for in-depth discussion:
http://gcc.gnu.org/ml/gcc-patches/2013-11/msg01141.html
This mostly involves two types of changes: updating assembler source
files to the new logic, and updating the dynamic loader.
After the refactoring in the previous patch, most of the assembler source
changes can be handled simply by providing ELFv2 versions of the
macros in sysdep.h. One somewhat non-obvious change is in __GI__setjmp:
this used to "fall through" to the immediately following __setjmp ENTRY
point. This is no longer safe in the ELFv2 since ENTRY defines both
a global and a local entry point, and you cannot simply fall through
to a global entry point as it requires r12 to be set up.
Also, makecontext needs to be updated to set up registers according to
the new ABI for calling into the context's start routine.
The dynamic linker changes mostly consist of removing special code
to handle function descriptors. We also need to support the new PLT
and glink format used by the the ELFv2 linker, see:
https://sourceware.org/ml/binutils/2013-10/msg00376.html
In addition, the dynamic linker now verifies that the dynamic libraries
it loads match its own ABI.
The hack in VDSO_IFUNC_RET to "synthesize" a function descriptor
for vDSO routines is also no longer necessary for ELFv2.
This is the first patch to support the new ELFv2 ABI in glibc.
As preparation, this patch simply refactors some of the powerpc64 assembler
code to move all code related to creating function descriptors (.opd section)
or using function descriptors (function pointer call) into a central place
in sysdep.h.
Note that most locations creating .opd entries were already using macros
in sysdep.h, this patch simply extends this to the remaining places.
No relevant change in generated code expected.
The TCB header on Intel contains a field __private_ss that is used
to efficiently implement the -fsplit-stack GCC feature.
In order to prepare for a possible future implementation of that
feature on powerpc64, we'd like to reserve a similar field in
the TCB header as well. (It would be good if this went in with
or before the ELFv2 patches to ensure that this field will be
available always in the ELFv2 environment.)
The field needs to be added at the front of tcbhead_t structure
to avoid changing the ABI; see the recent discussion when adding
the EBB fields.
This patch updates glibc in accordance with the binutils patch checked in here:
https://sourceware.org/ml/binutils/2013-10/msg00372.html
This changes the various R_PPC64_..._HI and _HA relocations to report
32-bit overflows. The motivation is that existing uses of @h / @ha
are to build up 32-bit offsets (for the "medium model" TOC access
that GCC now defaults to), and we'd really like to see failures at
link / load time rather than silent truncations.
For those rare cases where a modifier is needed to build up a 64-bit
constant, new relocations _HIGH / _HIGHA are supported.
The patch also fixes a bug in overflow checking for the R_PPC64_ADDR30
and R_PPC64_ADDR32 relocations.
The context established by "makecontext" has a link register pointing
back to an error path within the makecontext routine. This is currently
covered by the CFI FDE for makecontext itself, which is simply wrong
for the stack frame *inside* the context. When trying to unwind (e.g.
doing a backtrace) in a routine inside a context created by makecontext,
this can lead to uninitialized stack slots being accessed, causing the
unwinder to crash in the worst case.
Similarly, during parts of the "setcontext" routine, when the stack
pointer has already been switched to point to the new context, the
address range is still covered by the CFI FDE for setcontext. When
trying to unwind in that situation (e.g. backtrace from an async
signal handler for profiling), it is again possible that the unwinder
crashes.
Theses are all problems in existing code, but the changes in stack
frame layout appear to make the "worst case" much more likely in
the ELFv2 ABI context. This causes regressions e.g. in the libgo
testsuite on ELFv2.
This patch fixes this by ending the makecontext/setcontext FDEs
before those problematic parts of the assembler, similar to what
is already done on other platforms. This fixes the libgo
regression on ELFv2.
In case of power failure followed by filesystem issues locale-archive
can end-up containing all zeros. In that case all calls to setlocale()
generate a SIGFPE. This renders a system with a default non-C locale
unbootable.
Avoid this by ignoring the locale instead of generating a SIGFPE.
Joseph pointed out in the bug report (and in an earlier thread) that
systemtap probes cause build time warnings like the following:
../sysdeps/ieee754/dbl-64/e_atan2.c:602:4: warning: the address of
'p' will always evaluate as 'true' [-Waddress]
due to the fact that we're now passing non-weak variables to
LIBC_PROBE in the libm probes. This happens only on configurations
that do not enable systemtap. The macro definition of LIBC_PROBE in
this case only acts as a sanity checker to ensure that the number
parameters passed to LIBC_PROBE is equal to the argument count
parameter passed before it. This can be done in a much simpler manner
by just adding a macro definition for each number of arguments. I am
assuming here that we don't really want to bother with supporting
LIBC_PROBE with an indeterminate number of arguments and if there is a
need for a probe to have more data than what is currently supported (4
arguments), one could simply add an additional macro here.
Only gaih_inet() and gaih_inet_serv() use a special bit flag denoted
by the GAIH_OKIFUNSPEC macro. Only the return value of
gaih_inet_serv() is actively checked for the bit flag which is
redundant because it just copies the nonzero property of the value
otherwise returned. The return value of gaih_inet() is only checked
for being zero and then the bit flag is filtered out. As the bit flag
is set only for otherwise nonzero return values, it doesn't affect the
zero comparison. GAIH_EAI just an alias to ~GAIH_OKIFUNSPEC.
AF_INET lookup in hosts file uses _nss_files_gethostbyname2_r, which
is not capable of returning a canonical name if it has found one.
This change adds _nss_files_gethostbyname3_r, which wraps around
_nss_files_gethostbyname2_r and then returns result.h_name as the
canonical name.
The event code is PTRACE_EVENT_SECCOMP, not PTRAVE_EVENT_SECCOMP.
This patch fixes the V->C typo. There are no ABI issues since the
number remains the same for the code. Code using the old wrong
name will need to be updated.
This patch helps some math functions performance by adding the libc_fexxx
variant of inline functions to handle both FPU round and exception set/restore
and by using them on the libc_fexxx_ctx functions. It is based on already coded
fexxx family functions for PPC with fpu.
Here is the summary of performance improvements due this patch (measured on a
POWER7 machine):
Before:
cos(): ITERS:9.5895e+07: TOTAL:5116.03Mcy, MAX:77.6cy, MIN:49.792cy, 18744 calls/Mcy
exp(): ITERS:2.827e+07: TOTAL:5187.15Mcy, MAX:494.018cy, MIN:38.422cy, 5450.01 calls/Mcy
pow(): ITERS:6.1705e+07: TOTAL:5144.26Mcy, MAX:171.95cy, MIN:29.935cy, 11994.9 calls/Mcy
sin(): ITERS:8.6898e+07: TOTAL:5117.06Mcy, MAX:83.841cy, MIN:46.582cy, 16982 calls/Mcy
tan(): ITERS:2.9473e+07: TOTAL:5115.39Mcy, MAX:191.017cy, MIN:172.352cy, 5761.63 calls/Mcy
After:
cos(): ITERS:2.05265e+08: TOTAL:5111.37Mcy, MAX:78.754cy, MIN:24.196cy, 40158.5 calls/Mcy
exp(): ITERS:3.341e+07: TOTAL:5170.84Mcy, MAX:476.317cy, MIN:15.574cy, 6461.23 calls/Mcy
pow(): ITERS:7.6153e+07: TOTAL:5129.1Mcy, MAX:147.5cy, MIN:30.916cy, 14847.2 calls/Mcy
sin(): ITERS:1.58816e+08: TOTAL:5115.11Mcy, MAX:1490.39cy, MIN:22.341cy, 31048.4 calls/Mcy
tan(): ITERS:3.4964e+07: TOTAL:5114.18Mcy, MAX:177.422cy, MIN:146.115cy, 6836.68 calls/Mcy
As detailed in PR11157, the use of '__block' is known to interfere
with keywords in some environments, such as the Clang -fblocks extension.
Recently a similar issue was raised concerning the use of '__unused'
and a '__glibc' prefix was proposed to create a glibc implementation
namespace for these sorts of issues [1]. This patches takes that
approach.
[1] https://sourceware.org/ml/libc-alpha/2012-02/msg00047.html
[2] http://lists.debian.org/debian-glibc/2013/11/msg00020.html
On hppa and ia64, the macro DL_AUTO_FUNCTION_ADDRESS() uses the
variable fptr[2] in it's own scope.
The content of fptr[] is thus undefined right after the macro exits.
Newer gcc's (>= 4.7) reuse the stack space of this variable triggering
a segmentation fault in dl-init.c:69.
To fix this we rewrite the macros to make the call directly to init
and fini without needing to pass back a constructed function pointer.
Not all architectures define this value, and if they don't,
just let the test run the same as test-fpucw, with __fpu_control
set to _FPU_DEFAULT explicitly.
