As pointer out on the mailing list, the inline assembly code in
sysdeps/powerpc/ifunc-sel.h doesn't have a list of clobbered registers
and used wrong constraints.
This patch fixes that. I verified it doesn't introduce any change in the
generated code.
Changelog:
* sysdeps/powerpc/ifunc-sel.h (ifunc_sel): Add "11", "12", "cr0" to the
clobber list. Use "i" constraint instead of "X".
(ifunc_one): Add "12" to the clobber list. Use "i" constraint instead
of "X".
On 32-bit PowerPC GCC 6 always saves the PIC register on the stack in
the prologue and adjust the stack in the epilogue. It is therefore not
possible anymore to just exit the function in the inline asm code,
otherwise it corrupts the stack pointer. This causes the following tests
to fail when using GCC 6:
FAIL: elf/ifuncmain1
FAIL: elf/ifuncmain1pic
FAIL: elf/ifuncmain1picstatic
FAIL: elf/ifuncmain1pie
FAIL: elf/ifuncmain1staticpic
FAIL: elf/ifuncmain1staticpie
FAIL: elf/ifuncmain1vis
FAIL: elf/ifuncmain1vispic
FAIL: elf/ifuncmain1vispie
FAIL: elf/ifuncmain2pic
FAIL: elf/ifuncmain2picstatic
FAIL: elf/ifuncmain3
FAIL: elf/ifuncmain4picstatic
FAIL: elf/ifuncmain5
FAIL: elf/ifuncmain5picstatic
FAIL: elf/ifuncmain5staticpic
The solution is to replace the beqlr instructions by a beq to the end
of the inline asm code. This fixes all the above failures.
ChangeLog:
* sysdeps/powerpc/ifunc-sel.h (ifunc_sel): Replace beqlr instructions
by beq instructions jumping to the end of the function.
The alpha specific version of trunc and truncf always add and subtract
0x1.0p23 or 0x1.0p52 even for big values. This causes this kind of
errors in the testsuite:
Failure: Test: trunc_towardzero (0x1p107)
Result:
is: 1.6225927682921334e+32 0x1.fffffffffffffp+106
should be: 1.6225927682921336e+32 0x1.0000000000000p+107
difference: 1.8014398509481984e+16 0x1.0000000000000p+54
ulp : 0.5000
max.ulp : 0.0000
Change this by returning the input value when its absolute value is
greater than 0x1.0p23 or 0x1.0p52. NaN have to go through the add and
subtract operations to get possibly silenced.
Finally remove the code to handle inexact exception, trunc should never
generate such an exception.
Changelog:
* sysdeps/alpha/fpu/s_trunc.c (__trunc): Return the input value
when its absolute value is greater than 0x1.0p52.
[_IEEE_FP_INEXACT] Remove.
* sysdeps/alpha/fpu/s_truncf.c (__truncf): Return the input value
when its absolute value is greater than 0x1.0p23.
[_IEEE_FP_INEXACT] Remove.
The alpha version of rint wrongly return sNaN for sNaN input. Fix that
by checking for NaN and by returning the input value added with itself
in that case.
Changelog:
* sysdeps/alpha/fpu/s_rint.c (__rint): Add argument with itself
when it is a NaN.
* sysdeps/alpha/fpu/s_rintf.c (__rintf): Likewise.
The alpha version of floor wrongly return sNaN for sNaN input. Fix that
by checking for NaN and by returning the input value added with itself
in that case.
Finally remove the code to handle inexact exception, floor should never
generate such an exception.
Changelog:
* sysdeps/alpha/fpu/s_floor.c (__floor): Add argument with itself
when it is a NaN.
[_IEEE_FP_INEXACT] Remove.
* sysdeps/alpha/fpu/s_floorf.c (__floorf): Likewise.
The alpha version of ceil wrongly return sNaN for sNaN input. Fix that
by checking for NaN and by returning the input value added with itself
in that case.
Finally remove the code to handle inexact exception, ceil should never
generate such an exception.
Changelog:
* sysdeps/alpha/fpu/s_ceil.c (__ceil): Add argument with itself
when it is a NaN.
[_IEEE_FP_INEXACT] Remove.
* sysdeps/alpha/fpu/s_ceilf.c (__ceilf): Likewise.
The ceil, floor and trunc functions on sparc do not fully follow the
standard and trigger an inexact exception when presented a value which
is not an integer. Since glibc 2.24 this causes a few tests to fail,
for instance:
testing double (without inline functions)
Failure: ceil (lit_pi): Exception "Inexact" set
Failure: ceil (-lit_pi): Exception "Inexact" set
Failure: ceil (min_subnorm_value): Exception "Inexact" set
Failure: ceil (min_value): Exception "Inexact" set
Failure: ceil (0.1): Exception "Inexact" set
Failure: ceil (0.25): Exception "Inexact" set
Failure: ceil (0.625): Exception "Inexact" set
Failure: ceil (-min_subnorm_value): Exception "Inexact" set
Failure: ceil (-min_value): Exception "Inexact" set
Failure: ceil (-0.1): Exception "Inexact" set
Failure: ceil (-0.25): Exception "Inexact" set
Failure: ceil (-0.625): Exception "Inexact" set
I tried to fix that by using the same strategy than used on other
architectures, that is by saving the FSR register at the beginning
and restoring it at the end of the function. When doing so I noticed
a comment that this operation might be very costly, so I decided to
do some benchmarks.
The benchmarks below represent the time required to run each of the
function 60 millions of times with different input value. I have done
that in the basic V9 code, the VIS2 code, and using the default C
implementation of the libc, for both sparc32 and sparc64, on a Niagara
T1 based machine and an UltraSparc IIIi. Given I don't have access to a
more recent machine), I haven't been able to test the VIS3 version. Also
it should be noted that it doesn't make sense to do this benchmark for
V8 or earlier as in that case we use the default C implementation. The
results are available in the table below, the "+ fix" version correspond
to the one saving and restoring the FSR.
Niagara T1 / sparc32
--------------------
ceilf ceil floorf floor truncf trunc
V9 19.10 22.48 19.10 22.48 16.59 19.27
V9 + fix 19.77 23.34 19.77 23.33 17.27 20.12
VIS2 16.87 19.62 16.87 19.62
VIS2 + fix 17.55 20.47 17.55 20.47
C impl 11.39 13.80 11.40 13.80 10.88 10.84
Niagara T1 / sparc64
--------------------
ceilf ceil floorf floor truncf trunc
V9 18.14 22.23 18.14 22.23 15.64 19.02
V9 + fix 18.82 23.08 18.82 23.08 16.32 19.87
VIS2 15.92 19.37 15.92 19.37
VIS2 + fix 16.59 20.22 16.59 20.22
C impl 11.39 13.60 11.39 15.36 10.88 12.65
UltraSparc IIIi / sparc32
-------------------------
ceilf ceil floorf floor truncf trunc
V9 4.81 7.09 6.61 11.64 4.91 7.05
V9 + fix 7.20 10.42 7.14 10.54 6.76 9.47
VIS2 4.81 7.03 4.76 7.13
VIS2 + fix 6.76 9.51 6.71 9.63
C impl 3.88 8.62 3.90 9.45 3.57 6.62
UltraSparc IIIi / sparc64
-------------------------
ceilf ceil floorf floor truncf trunc
V9 3.48 4.39 3.48 4.41 3.01 3.85
V9 + fix 4.76 5.90 4.76 5.90 4.86 6.26
VIS2 2.95 3.61 2.95 3.61
VIS2 + fix 4.24 5.37 4.30 7.97
C impl 3.63 4.89 3.62 6.38 3.33 4.03
The first thing that should be noted is that the C implementation is
always faster on the Niagara T1 based machine. On the UltraSparc IIIi
the float version on sparc32 is also faster.
Coming back about the fix saving and restoring the FSR, it appears
it has a big impact as expected. In that case the C implementation is
always faster than the fixed implementations.
This patch therefore removes the sparc specific implementations in
favor of the generic ones.
Changelog:
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/Makefile
[$(subdir) = math] (libm-sysdep_routines): Remove.
[$(subdir) = math && $(have-as-vis3) = yes] (libm-sysdep_routines):
Remove s_ceilf-vis3, s_ceil-vis3, s_floorf-vis3, s_floor-vis3,
s_truncf-vis3, s_trunc-vis3.
* sysdeps/sparc/sparc64/fpu/multiarch/Makefile: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_ceil-vis2.S: Delete
file.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_ceil-vis3.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_ceil.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_ceilf-vis2.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_ceilf-vis3.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_ceilf.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_floor-vis2.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_floor-vis3.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_floor.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_floorf-vis2.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_floorf-vis3.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_floorf.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_trunc-vis3.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_trunc.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_truncf-vis3.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_truncf.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/s_ceil.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/s_ceilf.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/s_floor.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/s_floorf.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/s_trunc.S: Likewise.
* sysdeps/sparc/sparc32/sparcv9/fpu/s_truncf.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_ceil-vis2.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_ceil-vis3.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_ceil.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_ceilf-vis2.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_ceilf-vis3.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_ceilf.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_floor-vis2.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_floor-vis3.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_floor.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_floorf-vis2.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_floorf-vis3.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_floorf.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_trunc-vis3.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_trunc.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_truncf-vis3.S: Likewise.
* sysdeps/sparc/sparc64/fpu/multiarch/s_truncf.S: Likewise.
* sysdeps/sparc/sparc64/fpu/s_ceil.S: Likewise.
* sysdeps/sparc/sparc64/fpu/s_ceilf.S: Likewise.
* sysdeps/sparc/sparc64/fpu/s_floor.S: Likewise.
* sysdeps/sparc/sparc64/fpu/s_floorf.S: Likewise.
* sysdeps/sparc/sparc64/fpu/s_trunc.S: Likewise.
* sysdeps/sparc/sparc64/fpu/s_truncf.S: Likewise.
Don't compile do_test with -mavx, -mavx nor -mavx512 since they won't run
on non-AVX machines.
[BZ #20384]
* sysdeps/x86_64/fpu/Makefile (extra-test-objs): Add
test-double-libmvec-sincos-avx-main.o,
test-double-libmvec-sincos-avx2-main.o,
test-double-libmvec-sincos-main.o,
test-float-libmvec-sincosf-avx-main.o,
test-float-libmvec-sincosf-avx2-main.o and
test-float-libmvec-sincosf-main.o.
test-float-libmvec-sincosf-avx512-main.o.
($(objpfx)test-double-libmvec-sincos): Also link with
$(objpfx)test-double-libmvec-sincos-main.o.
($(objpfx)test-double-libmvec-sincos-avx): Also link with
$(objpfx)test-double-libmvec-sincos-avx-main.o.
($(objpfx)test-double-libmvec-sincos-avx2): Also link with
$(objpfx)test-double-libmvec-sincos-avx2-main.o.
($(objpfx)test-float-libmvec-sincosf): Also link with
$(objpfx)test-float-libmvec-sincosf-main.o.
($(objpfx)test-float-libmvec-sincosf-avx): Also link with
$(objpfx)test-float-libmvec-sincosf-avx2-main.o.
[$(config-cflags-avx512) == yes] (extra-test-objs): Add
test-double-libmvec-sincos-avx512-main.o and
($(objpfx)test-double-libmvec-sincos-avx512): Also link with
$(objpfx)test-double-libmvec-sincos-avx512-main.o.
($(objpfx)test-float-libmvec-sincosf-avx512): Also link with
$(objpfx)test-float-libmvec-sincosf-avx512-main.o.
(CFLAGS-test-double-libmvec-sincos.c): Removed.
(CFLAGS-test-float-libmvec-sincosf.c): Likewise.
(CFLAGS-test-double-libmvec-sincos-main.c): New.
(CFLAGS-test-double-libmvec-sincos-avx-main.c): Likewise.
(CFLAGS-test-double-libmvec-sincos-avx2-main.c): Likewise.
(CFLAGS-test-float-libmvec-sincosf-main.c): Likewise.
(CFLAGS-test-float-libmvec-sincosf-avx-main.c): Likewise.
(CFLAGS-test-float-libmvec-sincosf-avx2-main.c): Likewise.
(CFLAGS-test-float-libmvec-sincosf-avx512-main.c): Likewise.
(CFLAGS-test-double-libmvec-sincos-avx.c): Set to -DREQUIRE_AVX.
(CFLAGS-test-float-libmvec-sincosf-avx.c ): Likewise.
(CFLAGS-test-double-libmvec-sincos-avx2.c): Set to
-DREQUIRE_AVX2.
(CFLAGS-test-float-libmvec-sincosf-avx2.c ): Likewise.
(CFLAGS-test-double-libmvec-sincos-avx512.c): Set to
-DREQUIRE_AVX512F.
(CFLAGS-test-float-libmvec-sincosf-avx512.c): Likewise.
* sysdeps/x86_64/fpu/test-double-libmvec-sincos.c: Rewritten.
* sysdeps/x86_64/fpu/test-float-libmvec-sincosf.c: Likewise.
* sysdeps/x86_64/fpu/test-double-libmvec-sincos-avx-main.c: New
file.
* sysdeps/x86_64/fpu/test-double-libmvec-sincos-avx2-main.c:
Likewise.
* sysdeps/x86_64/fpu/test-double-libmvec-sincos-avx512-main.c:
Likewise.
* sysdeps/x86_64/fpu/test-double-libmvec-sincos-main.c:
Likewise.
* sysdeps/x86_64/fpu/test-float-libmvec-sincosf-avx-main.c:
Likewise.
* sysdeps/x86_64/fpu/test-float-libmvec-sincosf-avx2-main.c:
Likewise.
* sysdeps/x86_64/fpu/test-float-libmvec-sincosf-avx512-main.c:
Likewise.
* sysdeps/x86_64/fpu/test-float-libmvec-sincosf-main.c:
Likewise.
This partly reverts commit f8238ae3c7
that regenerated the ulps, to make the max ulps good for gcc-5,
gcc-6 and gcc-trunk as well.
* sysdeps/aarch64/libm-test-ulps: Updated.
If the default memcpy variant is called with a length of >1MB on 31bit,
r13 is clobbered as the algorithm is switching to mvcle. The mvcle code
returns without restoring r13. All other cases are restoring r13.
If memcpy is called from outside libc the ifunc resolver will only select
this variant if running on machines older than z10.
Otherwise or if memcpy is called from inside libc, this default memcpy
variant is called.
The testcase timezone/tst-tzset is triggering this issue in some combinations
of gcc versions and optimization levels.
This bug was introduced in commit 04bb21ac93
and thus is a regression compared to former glibc 2.23 release.
This patch removes the usage of r13 at all. Thus it is not saved and restored.
The base address for execute-instruction is now stored in r5 which is obtained
after r5 is not needed anymore as 256byte block counter.
ChangeLog:
* sysdeps/s390/s390-32/memcpy.S (memcpy): Eliminate the usage
of r13 as it is not restored in mvcle case.
If a function passes in a variable named "ret", the code will miscompile
when it declares a local ret variable. In some cases, it's even a build
failure like so:
../sysdeps/unix/sysv/linux/spawni.c: In function '__spawni_child':
../sysdeps/unix/sysv/linux/spawni.c:289:5: error: address of register variable 'ret' requested
while (write_not_cancel (p, &ret, sizeof ret) < 0)
Compile i386 rtld-*.os with -mno-sse -mno-mmx -mfpmath=387 so that no
code in ld.so uses mm/xmm/ymm/zmm registers on i386 since the first 3
mm/xmm/ymm/zmm registers are used to pass vector parameters which must
be preserved.
* sysdeps/i386/Makefile (rtld-CFLAGS): New.
[subdir == elf] (CFLAGS-.os): Replace -mno-sse -mno-mmx
-mfpmath=387 with $(rtld-CFLAGS).
[subdir != elf] (CFLAGS-.os): Compile rtld-*.os with
$(rtld-CFLAGS).
During the sincos consolidation I made two mistakes, one was a logical
error due to which cos(0x1.8475e5afd4481p+0) returned
sin(0x1.8475e5afd4481p+0) instead.
The second issue was an error in negating inputs for the correct
quadrants for sine. I could not find a suitable test case for this
despite running a program to search for such an input for a couple of
hours.
Following patch fixes both issues. Tested on x86_64. Thanks to Matt
Clay for identifying the issue.
[BZ #20357]
* sysdeps/ieee754/dbl-64/s_sin.c (sloww): Fix up condition
to call __mpsin/__mpcos and to negate values.
* math/auto-libm-test-in: Add test.
* math/auto-libm-test-out: Regenerate.
This patch changes both the nptl and libc Linux raise implementation
to avoid the issues described in BZ#15368. The strategy used is
summarized in bug report first comment:
1. Block all signals (including internal NPTL ones);
2. Get pid and tid directly from syscall (not relying on cached
values);
3. Call tgkill;
4. Restore old signal mask.
Tested on x86_64 and i686.
[BZ #15368]
* sysdeps/unix/sysv/linux/nptl-signals.h
(__nptl_clear_internal_signals): New function.
(__libc_signal_block_all): Likewise.
(__libc_signal_block_app): Likewise.
(__libc_signal_restore_set): Likewise.
* sysdeps/unix/sysv/linux/pt-raise.c (raise): Use Linux raise.c
implementation.
* sysdeps/unix/sysv/linux/raise.c (raise): Reimplement to not use
the cached pid/tid value in pthread structure.
64-bit off_t in pread64, preadv, pwrite64 and pwritev syscalls is passed
in one 64-bit register for both x32 and x86-64. Since the inline
asm statement only passes long, which is 32-bit for x32, in registers,
64-bit off_t is truncated to 32-bit on x32. Since __ASSUME_PREADV and
__ASSUME_PWRITEV are defined unconditionally, these syscalls can be
implemented in syscalls.list to pass 64-bit off_t in one 64-bit register.
Tested on x86-64 and x32 with off_t > 4GB on pread64/pwrite64 and
preadv64/pwritev64.
[BZ #20348]
* sysdeps/unix/sysv/linux/x86_64/syscalls.list: Add pread64,
preadv64, pwrite64 and pwritev64.
Since _dl_tlsdesc_dynamic is called via PLT, we need to add 8 bytes for
push in the PLT entry to align the stack.
[BZ #20309]
* configure.ac (have-mtls-dialect-gnu2): Set to yes if
-mtls-dialect=gnu2 works.
* configure: Regenerated.
* elf/Makefile [have-mtls-dialect-gnu2 = yes]
(tests): Add tst-gnu2-tls1.
(modules-names): Add tst-gnu2-tls1mod.
($(objpfx)tst-gnu2-tls1): New.
(tst-gnu2-tls1mod.so-no-z-defs): Likewise.
(CFLAGS-tst-gnu2-tls1mod.c): Likewise.
* elf/tst-gnu2-tls1.c: New file.
* elf/tst-gnu2-tls1mod.c: Likewise.
* sysdeps/x86_64/dl-tlsdesc.S (_dl_tlsdesc_dynamic): Add 8
bytes for push in the PLT entry to align the stack.
Define LO_HI_LONG to skip pos_h since it is ignored by kernel:
static inline loff_t pos_from_hilo(unsigned long high, unsigned long low)
{
#define HALF_LONG_BITS (BITS_PER_LONG / 2)
return (((loff_t)high << HALF_LONG_BITS) << HALF_LONG_BITS) | low;
}
where size of loff_t == size of long.
[BZ #20349]
* sysdeps/unix/sysv/linux/x86_64/sysdep.h (LO_HI_LONG): New.
AArch64 uses HWCAP bits but they are not defined in sys/auxv.h.
This patch adds a copy of the linux v4.6 arm64 uapi asm/hwcap.h
definitions.
* sysdeps/unix/sysv/linux/aarch64/bits/hwcap.h: New.
When glibc is built with --enable-profile, the ENTRY of
asm functions includes CALL_MCOUNT for profiling.
(matters for binaries static linked against libc_p.a.)
CALL_MCOUNT did not save/restore argument registers
around the _mcount call so it clobbered them.
(it is enough to only save/restore the arguments passed
to a given asm function, but that would be too many asm
changes so it is simpler to always save all argument
registers in this macro.)
float args are not saved: mcount does not clobber the
float regs and currently no asm function takes float
arguments anyway.
[BZ #18707]
* sysdeps/aarch64/Makefile (CFLAGS-mcount.c): Add -mgeneral-regs-only.
* sysdeps/aarch64/sysdep.h (CALL_MCOUNT): Save argument registers.
The p{read,write}v{64} consolidation patch [1] added a wrong guard
for LO_HI_LONG definition. It currently uses both
'__WORDSIZE == 64' and 'defined __ASSUME_WORDSIZE64_ILP32' to set
the value to be passed in one argument, otherwise it will be split
in two.
However it fails on MIPS64n32 where syscalls n32 uses the compat
implementation in the kernel meaning the off_t arguments are passed
in two separate registers.
GLIBC already defines a macro for such cases (__OFF_T_MATCHES_OFF64_T),
so this patch uses it instead.
Checked on x86_64, i686, x32, aarch64, armhf, and s390.
* sysdeps/unix/sysv/linux/sysdep.h
[__WORDSIZE == 64 || __ASSUME_WORDSIZE64_ILP32] (LO_HI_LONG): Remove
guards.
* misc/tst-preadvwritev-common.c: New file.
* misc/tst-preadvwritev.c: Use tst-preadvwritev-common.c.
* misc/tst-preadvwritev64.c: Use tst-preadwritev-common.c and add
a check for files larger than 2GB.
[1] 4751bbe2ad
This patch removes the __ASSUME_OFF_DIFF_OFF64 define introduced in
p{read,write} consolidation patch. This define was added based on
the idea 32 bits ports would continue to follow previous off{64}_t
definition where off_t size differs from off64_t one.
However, with recent AArch64/ILP32 patch submission and also with
discussion for RISCV kernel interface, 32 bits ports now may aim
to use off_t and off64_t with the same size as 64 bits.
So current assumption for both p{read,write} and p{read,write}v
are not compatible with new type definition. This patch now makes
the syscall wrappers to only depend on __OFF_T_MATCHES_OFF64_T to
define the default and 64-suffix variant, as follow:
<function>.c
#ifndef __OFF_T_MATCHES_OFF64_T
/* build <function> */
#endif
and
<function>64.c
/* build <function>64 */
#ifdef __OFF_T_MATCHES_OFF64_T
weak_alias (fallocate64, fallocate)
#endif
Tested on x86_64, i686, x32, and armhf.
* sysdeps/unix/sysv/linux/mips/kernel-features.h
(__ASSUME_OFF_DIFF_OFF64): Remove define.
* sysdeps/unix/sysv/linux/pread.c
[__WORDSIZE != 64 || __ASSUME_OFF_DIFF_OFF64] (pread): Replace by
__OFF_T_MATCHES_OFF64_T.
* sysdeps/unix/sysv/linux/pread64.c
[__WORDSIZE != 64 || __ASSUME_OFF_DIFF_OFF64] (pread64): Likewise.
* sysdeps/unix/sysv/linux/preadv.c
[__WORDSIZE != 64 || __ASSUME_OFF_DIFF_OFF64] (preadv): Likewise.
* sysdeps/unix/sysv/linux/preadv64.c
[__WORDSIZE != 64 || __ASSUME_OFF_DIFF_OFF64] (preadv64): Likewise.
* sysdeps/unix/sysv/linux/pwrite.c
[__WORDSIZE != 64 || __ASSUME_OFF_DIFF_OFF64] (pwrite): Likewise.
* sysdeps/unix/sysv/linux/pwrite64.c
[__WORDSIZE != 64 || __ASSUME_OFF_DIFF_OFF64] (pwrite64): Likewise.
* sysdeps/unix/sysv/linux/pwritev.c
[__WORDSIZE != 64 || __ASSUME_OFF_DIFF_OFF64] (pwritev): Likewise.
* sysdeps/unix/sysv/linux/pwritev64.c
[__WORDSIZE != 64 || __ASSUME_OFF_DIFF_OFF64] (pwritev64): Likewise.
The previous uses of this symbol were all in wordsize-32 code.
In commit eeddfa91cb ("Consolidate off_t/off64_t syscall
argument passing") it was expanded to be used in pread/pwrite.
Accordingly, we only define it in 32-bit compilation modes now.
Both tilepro and tilegx32 follow this convention for the
kernel ABI. tilegx64 follows it for passing 128-bit values,
but there are no such ABIs in the kernel.
Commit a6a4395d fixed modf implementation by compiling s_modf.c and
s_modff.c with -fsignaling-nans. However these files are also included
from the pre-POWER5+ implementation, and thus these files should also
be compiled with -fsignaling-nans.
Changelog:
[BZ #20240]
* sysdeps/powerpc/powerpc32/power4/fpu/multiarch/Makefile
(CFLAGS-s_modf-ppc32.c): New variable.
(CFLAGS-s_modff-ppc32.c): Likewise.
* sysdeps/powerpc/powerpc64/fpu/multiarch/Makefile
(CFLAGS-s_modf-ppc64.c): Likewise.
(CFLAGS-s_modff-ppc64.c): Likewise.
On s390, the current prelink undo code in elf_machine_lazy_rel()
has the requirement, that the plt stubs use the first got slots
after the 3 reserved ones.
In case of undoing prelink, the plt got slots are reset to the correct
addresses whithin the corresponding plt-stub. Therefore the address
is calculated by the address of the first plt-stub-address which
was written by prelink (see l->l_mach.plt) to got[1] and index of
current relocation multiplied with 32 (=size of one plt slot).
The index was calculated with ¤t-got-slot - &got[3].
This patch removes the requirement, that the plt-got-slots are
starting at got[3]. The index is now calculated with
¤t-reloc - &reloc[0]. The first struct Elf64_Rela is stored
at DT_JMPREL.
This patch is needed to prepare for partial relro support.
Ulrich Weigand suggested this approach to use DT_JMPREL - Thanks.
ChangeLog:
* sysdeps/s390/linkmap.h (struct link_map_machine):
Remove member gotplt and add member jmprel.
* sysdeps/s390/s390-32/dl-machine.h
(elf_machine_runtime_setup): Setup member jmprel with DT_JMPREL
instead of gotplt with &got[3].
(elf_machine_lazy_rel): Calculate address with reloc and jmprel.
* sysdeps/s390/s390-64/dl-machine.h: Likewise.
If the input values are unaligned and if there are null characters in the
memory before the starting address of the input values, strcasecmp
gives incorrect return code. Fixed it by adding mask the bits that
are not part of the string.
In a reference to PR ld/19908 make ld.so respect symbol export classes
aka visibility and treat STV_HIDDEN and STV_INTERNAL symbols as local,
preventing such symbols from preempting exported symbols.
According to the ELF gABI[1] neither STV_HIDDEN nor STV_INTERNAL symbols
are supposed to be present in linked binaries:
"A hidden symbol contained in a relocatable object must be either
removed or converted to STB_LOCAL binding by the link-editor when the
relocatable object is included in an executable file or shared object."
"An internal symbol contained in a relocatable object must be either
removed or converted to STB_LOCAL binding by the link-editor when the
relocatable object is included in an executable file or shared object."
however some GNU binutils versions produce such symbols in some cases.
PR ld/19908 is one and we also have this note in scripts/abilist.awk:
so clearly there is linked code out there which contains such symbols
which is prone to symbol table misinterpretation, and it'll be more
productive if we handle this gracefully, under the Robustness Principle:
"be liberal in what you accept, and conservative in what you produce",
especially as this is a simple (STV_HIDDEN|STV_INTERNAL) => STB_LOCAL
mapping.
References:
[1] "System V Application Binary Interface - DRAFT - 24 April 2001",
The Santa Cruz Operation, Inc., "Symbol Table",
<http://www.sco.com/developers/gabi/2001-04-24/ch4.symtab.html>
* sysdeps/generic/ldsodefs.h
(dl_symbol_visibility_binds_local_p): New inline function.
* elf/dl-addr.c (determine_info): Treat hidden and internal
symbols as local.
* elf/dl-lookup.c (do_lookup_x): Likewise.
* elf/dl-reloc.c (RESOLVE_MAP): Likewise.
nearbyint and nearbyintf should not trigger inexact exceptions, but
should still trigger an invalid exception for a sNaN input.
The SPARC specific implementations of these functions save the FSR at
the beginning of the function and restore it at the end to not trigger
an inexact exception. This however doesn't work for an sNaN input which
need to trigger an invalid exception. Fix that by adding a fcmp
instruction using the input value before saving FSR, so that an invalid
exception is triggered for a sNaN input.
This fixes the math/test-nearbyint-except test on SPARC.
Changelog:
* sparc/sparc32/sparcv9/fpu/s_nearbyint.S (__nearbyint): Trigger an
invalid exception for a sNaN input.
* sparc/sparc32/sparcv9/fpu/s_nearbyintf.S (__nearbyintf): Likewise.
* sparc/sparc32/sparcv9/fpu/multiarch/s_nearbyint-vis3.S
(__nearbyint_vis3): Likewise
* sparc/sparc32/sparcv9/fpu/multiarch/s_nearbyintf-vis3.S
(__nearbyintf_vis3): Likewise
* sparc/sparc64/fpu/s_nearbyint.S (__nearbyint): Likewise.
* sparc/sparc64/fpu/s_nearbyintf.S (__nearbyintf): Likewise.
* sparc/sparc64/fpu/multiarch/s_nearbyint-vis3.S (__nearbyint_vis3):
Likewise.
* sparc/sparc64/fpu/multiarch/s_nearbyintf-vis3.S (__nearbyintf_vis3):
Likewise.
If assembler doesn't support AVX512DQ, _dl_runtime_resolve_avx is used
to save the first 8 vector registers, which only saves the lower 256
bits of vector register, for lazy binding. When it is called on AVX512
platform, the upper 256 bits of ZMM registers are clobbered. Parameters
passed in ZMM registers will be wrong when the function is called the
first time. This patch requires binutils 2.24, whose assembler can store
and load ZMM registers, to build x86-64 glibc. Since mathvec library
needs assembler support for AVX512DQ, we disable mathvec if assembler
doesn't support AVX512DQ.
[BZ #20139]
* config.h.in (HAVE_AVX512_ASM_SUPPORT): Renamed to ...
(HAVE_AVX512DQ_ASM_SUPPORT): This.
* sysdeps/x86_64/configure.ac: Require assembler from binutils
2.24 or above.
(HAVE_AVX512_ASM_SUPPORT): Removed.
(HAVE_AVX512DQ_ASM_SUPPORT): New.
* sysdeps/x86_64/configure: Regenerated.
* sysdeps/x86_64/dl-trampoline.S: Make HAVE_AVX512_ASM_SUPPORT
check unconditional.
* sysdeps/x86_64/multiarch/ifunc-impl-list.c: Likewise.
* sysdeps/x86_64/multiarch/memcpy.S: Likewise.
* sysdeps/x86_64/multiarch/memcpy_chk.S: Likewise.
* sysdeps/x86_64/multiarch/memmove-avx512-no-vzeroupper.S:
Likewise.
* sysdeps/x86_64/multiarch/memmove-avx512-unaligned-erms.S:
Likewise.
* sysdeps/x86_64/multiarch/memmove.S: Likewise.
* sysdeps/x86_64/multiarch/memmove_chk.S: Likewise.
* sysdeps/x86_64/multiarch/mempcpy.S: Likewise.
* sysdeps/x86_64/multiarch/mempcpy_chk.S: Likewise.
* sysdeps/x86_64/multiarch/memset-avx512-no-vzeroupper.S:
Likewise.
* sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S:
Likewise.
* sysdeps/x86_64/multiarch/memset.S: Likewise.
* sysdeps/x86_64/multiarch/memset_chk.S: Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_d_cos8_core_avx512.S: Check
HAVE_AVX512DQ_ASM_SUPPORT instead of HAVE_AVX512_ASM_SUPPORT.
* sysdeps/x86_64/fpu/multiarch/svml_d_exp8_core_avx512.S:
Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_d_log8_core_avx512.S:
Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_d_pow8_core_avx512.S:
Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_d_sin8_core_avx512.S:
Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_d_sincos8_core_avx512.:
Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_s_cosf16_core_avx512.S:
Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_s_expf16_core_avx512.S:
Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S:
Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_s_powf16_core_avx512.S:
Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx51:
Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_s_sinf16_core_avx512.S:
Likewise.
current vector function declaration "#pragma omp declare simd notinbranch",
according to which vector sincos should have vector of pointers for second and
third parameters. It is fixed with implementation as wrapper to version
having second and third parameters as pointers.
[BZ #20024]
* sysdeps/x86/fpu/test-math-vector-sincos.h: New.
* sysdeps/x86_64/fpu/multiarch/svml_d_sincos2_core_sse4.S: Fixed ABI
of this implementation of vector function.
* sysdeps/x86_64/fpu/multiarch/svml_d_sincos4_core_avx2.S: Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_d_sincos8_core_avx512.S: Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx512.S:
Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_s_sincosf4_core_sse4.S: Likewise.
* sysdeps/x86_64/fpu/multiarch/svml_s_sincosf8_core_avx2.S: Likewise.
* sysdeps/x86_64/fpu/svml_d_sincos2_core.S: Likewise.
* sysdeps/x86_64/fpu/svml_d_sincos4_core.S: Likewise.
* sysdeps/x86_64/fpu/svml_d_sincos4_core_avx.S: Likewise.
* sysdeps/x86_64/fpu/svml_d_sincos8_core.S: Likewise.
* sysdeps/x86_64/fpu/svml_s_sincosf16_core.S: Likewise.
* sysdeps/x86_64/fpu/svml_s_sincosf4_core.S: Likewise.
* sysdeps/x86_64/fpu/svml_s_sincosf8_core.S: Likewise.
* sysdeps/x86_64/fpu/svml_s_sincosf8_core_avx.S: Likewise.
* sysdeps/x86_64/fpu/test-double-vlen2-wrappers.c: Use another wrapper
for testing vector sincos with fixed ABI.
* sysdeps/x86_64/fpu/test-double-vlen4-avx2-wrappers.c: Likewise.
* sysdeps/x86_64/fpu/test-double-vlen4-wrappers.c: Likewise.
* sysdeps/x86_64/fpu/test-double-vlen8-wrappers.c: Likewise.
* sysdeps/x86_64/fpu/test-float-vlen16-wrappers.c: Likewise.
* sysdeps/x86_64/fpu/test-float-vlen4-wrappers.c: Likewise.
* sysdeps/x86_64/fpu/test-float-vlen8-avx2-wrappers.c: Likewise.
* sysdeps/x86_64/fpu/test-float-vlen8-wrappers.c: Likewise.
* sysdeps/x86_64/fpu/test-double-libmvec-sincos-avx.c: New test.
* sysdeps/x86_64/fpu/test-double-libmvec-sincos-avx2.c: Likewise.
* sysdeps/x86_64/fpu/test-double-libmvec-sincos-avx512.c: Likewise.
* sysdeps/x86_64/fpu/test-double-libmvec-sincos.c: Likewise.
* sysdeps/x86_64/fpu/test-float-libmvec-sincosf-avx.c: Likewise.
* sysdeps/x86_64/fpu/test-float-libmvec-sincosf-avx2.c: Likewise.
* sysdeps/x86_64/fpu/test-float-libmvec-sincosf-avx512.c: Likewise.
* sysdeps/x86_64/fpu/test-float-libmvec-sincosf.c: Likewise.
* sysdeps/x86_64/fpu/Makefile: Added new tests.
Commits d81f90cc and 89faa0340 replaced called to __isnan and __isinf
by the corresponding GCC builtins. In turns GCC emits calls to _Qp_cmp.
We should therefore add _Qp_cmp to localplt.data as otherwise the
elf/check-localplt test fails with:
Extra PLT reference: libc.so: _Qp_cmp
A similar change has already been done for SPARC32 in commit 6ef1cb95.
Changelog:
* sysdeps/unix/sysv/linux/sparc/sparc64/localplt.data: Add _Qp_cmp.
This implementation is based on the one already used at
sysdeps/x86_64/fpu/e_expf.S.
This implementation improves the performance by ~14% on average in synthetic
benchmarks at the cost of decreasing accuracy to 1 ULP.
The patched change fixes a regression for executables compiled with the
-p option and linked with gcrt1.o. The executables crash on startup.
This regression was introduced in 2.22 and was noticed in the gcc testsuite.
Although the Enhanced REP MOVSB/STOSB (ERMS) implementations of memmove,
memcpy, mempcpy and memset aren't used by the current processors, this
patch adds Prefer_ERMS check in memmove, memcpy, mempcpy and memset so
that they can be used in the future.
* sysdeps/x86/cpu-features.h (bit_arch_Prefer_ERMS): New.
(index_arch_Prefer_ERMS): Likewise.
* sysdeps/x86_64/multiarch/memcpy.S (__new_memcpy): Return
__memcpy_erms for Prefer_ERMS.
* sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S
(__memmove_erms): Enabled for libc.a.
* ysdeps/x86_64/multiarch/memmove.S (__libc_memmove): Return
__memmove_erms or Prefer_ERMS.
* sysdeps/x86_64/multiarch/mempcpy.S (__mempcpy): Return
__mempcpy_erms for Prefer_ERMS.
* sysdeps/x86_64/multiarch/memset.S (memset): Return
__memset_erms for Prefer_ERMS.
For some reasons I have not investigated yet, tst-mode-switch-1 hangs on
a MIPS UTM-8 machine running an o32 userland and a 3.6.1 kernel.
This patch changes the test so that it runs under the test-skeleton
framework, causing the test to fail after a timeout instead of hanging
the whole testsuite. At the same time, also change the tst-mode-switch-2
and tst-mode-switch-3 tests.
Changelog:
* sysdeps/mips/tst-mode-switch-1.c (main): Converted to ...
(do_test): ... this.
(TEST_FUNCTION): New macro.
Include test-skeleton.c.
* sysdeps/mips/tst-mode-switch-2.c (main): Likewise.
* sysdeps/mips/tst-mode-switch-3.c (main): Likewise.
As discussed in
<https://sourceware.org/ml/libc-alpha/2016-05/msg00577.html>, TS
18661-1 disallows ceil, floor, round and trunc functions from raising
the "inexact" exception, in accordance with general IEEE 754 semantics
for when that exception is raised. Fixing this for x87 floating point
is more complicated than for the other versions of these functions,
because they use the frndint instruction that raises "inexact" and
this can only be avoided by saving and restoring the whole
floating-point environment.
As I noted in
<https://sourceware.org/ml/libc-alpha/2016-06/msg00128.html>, I have
now implemented a GCC option -fno-fp-int-builtin-inexact for GCC 7,
such that GCC will inline these functions on x86, without caring about
"inexact", when the default -ffp-int-builtin-inexact is in effect.
This allows users to get optimized code depending on the options they
pass to the compiler, while making the out-of-line functions follow TS
18661-1 semantics and avoid "inexact".
This patch duly fixes the out-of-line trunc function implementations
to avoid "inexact", in the same way as the nearbyint implementations.
I do not know how the performance of implementations such as these
based on saving the environment and changing the rounding mode
temporarily compares to that of the C versions or SSE 4.1 versions (of
course, for 32-bit x86 SSE implementations still need to get the
return value in an x87 register); it's entirely possible other
implementations could be faster in some cases.
Tested for x86_64 and x86.
[BZ #15479]
* sysdeps/i386/fpu/s_trunc.S (__trunc): Save and restore
floating-point environment rather than just control word.
* sysdeps/i386/fpu/s_truncf.S (__truncf): Likewise.
* sysdeps/i386/fpu/s_truncl.S (__truncl): Save and restore
floating-point environment, with "invalid" exceptions merged in,
rather than just control word.
* sysdeps/x86_64/fpu/s_truncl.S (__truncl): Likewise.
* math/libm-test.inc (trunc_test_data): Do not allow spurious
"inexact" exceptions.
As discussed in
<https://sourceware.org/ml/libc-alpha/2016-05/msg00577.html>, TS
18661-1 disallows ceil, floor, round and trunc functions from raising
the "inexact" exception, in accordance with general IEEE 754 semantics
for when that exception is raised. Fixing this for x87 floating point
is more complicated than for the other versions of these functions,
because they use the frndint instruction that raises "inexact" and
this can only be avoided by saving and restoring the whole
floating-point environment.
As I noted in
<https://sourceware.org/ml/libc-alpha/2016-06/msg00128.html>, I have
now implemented a GCC option -fno-fp-int-builtin-inexact for GCC 7,
such that GCC will inline these functions on x86, without caring about
"inexact", when the default -ffp-int-builtin-inexact is in effect.
This allows users to get optimized code depending on the options they
pass to the compiler, while making the out-of-line functions follow TS
18661-1 semantics and avoid "inexact".
This patch duly fixes the out-of-line floor function implementations
to avoid "inexact", in the same way as the nearbyint implementations.
I do not know how the performance of implementations such as these
based on saving the environment and changing the rounding mode
temporarily compares to that of the C versions or SSE 4.1 versions (of
course, for 32-bit x86 SSE implementations still need to get the
return value in an x87 register); it's entirely possible other
implementations could be faster in some cases.
Tested for x86_64 and x86.
[BZ #15479]
* sysdeps/i386/fpu/s_floor.S (__floor): Save and restore
floating-point environment rather than just control word.
* sysdeps/i386/fpu/s_floorf.S (__floorf): Likewise.
* sysdeps/i386/fpu/s_floorl.S (__floorl): Save and restore
floating-point environment, with "invalid" exceptions merged in,
rather than just control word.
* sysdeps/x86_64/fpu/s_floorl.S (__floorl): Likewise.
* math/libm-test.inc (floor_test_data): Do not allow spurious
"inexact" exceptions.
As discussed in
<https://sourceware.org/ml/libc-alpha/2016-05/msg00577.html>, TS
18661-1 disallows ceil, floor, round and trunc functions from raising
the "inexact" exception, in accordance with general IEEE 754 semantics
for when that exception is raised. Fixing this for x87 floating point
is more complicated than for the other versions of these functions,
because they use the frndint instruction that raises "inexact" and
this can only be avoided by saving and restoring the whole
floating-point environment.
As I noted in
<https://sourceware.org/ml/libc-alpha/2016-06/msg00128.html>, I have
now implemented a GCC option -fno-fp-int-builtin-inexact for GCC 7,
such that GCC will inline these functions on x86, without caring about
"inexact", when the default -ffp-int-builtin-inexact is in effect.
This allows users to get optimized code depending on the options they
pass to the compiler, while making the out-of-line functions follow TS
18661-1 semantics and avoid "inexact".
This patch duly fixes the out-of-line ceil function implementations to
avoid "inexact", in the same way as the nearbyint implementations.
I do not know how the performance of implementations such as these
based on saving the environment and changing the rounding mode
temporarily compares to that of the C versions or SSE 4.1 versions (of
course, for 32-bit x86 SSE implementations still need to get the
return value in an x87 register); it's entirely possible other
implementations could be faster in some cases.
Tested for x86_64 and x86.
[BZ #15479]
* sysdeps/i386/fpu/s_ceil.S (__ceil): Save and restore
floating-point environment rather than just control word.
* sysdeps/i386/fpu/s_ceilf.S (__ceilf): Likewise.
* sysdeps/i386/fpu/s_ceill.S (__ceill): Save and restore
floating-point environment, with "invalid" exceptions merged in,
rather than just control word.
* sysdeps/x86_64/fpu/s_ceill.S (__ceill): Likewise.
* math/libm-test.inc (ceil_test_data): Do not allow spurious
"inexact" exceptions.
Commit 43c29487 tried to fix the vfork aliases in libpthread.so on MIPS
and SPARC, but failed to do it correctly, introducing an ABI change.
This patch does the remaining changes needed to align the MIPS and SPARC
vfork implementations with the other architectures. That way the the
alpha version of pt-vfork.S works correctly for MIPS and SPARC. The
changes for alpha were done in 82aab97c.
Changelog:
* sysdeps/unix/sysv/linux/mips/vfork.S (__vfork): Rename into
__libc_vfork.
(__vfork) [IS_IN (libc)]: Remove alias.
(__libc_vfork) [IS_IN (libc)]: Define as an alias.
* sysdeps/unix/sysv/linux/sparc/sparc32/vfork.S: Likewise.
* sysdeps/unix/sysv/linux/sparc/sparc64/vfork.S: Likewise.
atomic_compare_and_exchange_bool_rel and
catomic_compare_and_exchange_bool_rel are removed and replaced with the
new C11-like atomic_compare_exchange_weak_release. The concurrent code
in nscd/cache.c has not been reviewed yet, so this patch does not add
detailed comments.
* nscd/cache.c (cache_add): Use new C11-like atomic operation instead
of atomic_compare_and_exchange_bool_rel.
* nptl/pthread_mutex_unlock.c (__pthread_mutex_unlock_full): Likewise.
* include/atomic.h (atomic_compare_and_exchange_bool_rel,
catomic_compare_and_exchange_bool_rel): Remove.
* sysdeps/aarch64/atomic-machine.h
(atomic_compare_and_exchange_bool_rel): Likewise.
* sysdeps/alpha/atomic-machine.h
(atomic_compare_and_exchange_bool_rel): Likewise.
* sysdeps/arm/atomic-machine.h
(atomic_compare_and_exchange_bool_rel): Likewise.
* sysdeps/mips/atomic-machine.h
(atomic_compare_and_exchange_bool_rel): Likewise.
* sysdeps/tile/atomic-machine.h
(atomic_compare_and_exchange_bool_rel): Likewise.
