SHA256
1
0
forked from pool/glibc
glibc/fix-locking-in-_IO_cleanup.patch

162 lines
4.8 KiB
Diff
Raw Normal View History

Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
Always do locking when accessing streams (bug 15142)
During exit, skip files that are currently locked to avoid deadlock.
[BZ #15142]
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
* include/bits/libio.h (_IO_ftrylockfile) [_IO_MTSAVE_IO]: Define.
* libio/genops.c (_IO_flush_all_lockp): Make static. Rename
argument to skip_locked, callers changed. Skip files that are
locked if skip_locked.
(_IO_unbuffer_all): Lock files before access, but skip locked
files.
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
* libio/libioP.h (_IO_flush_all_lockp): Don't declare.
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
Index: glibc-2.27/include/bits/libio.h
===================================================================
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
--- glibc-2.27.orig/include/bits/libio.h
+++ glibc-2.27/include/bits/libio.h
@@ -33,11 +33,15 @@ libc_hidden_proto (_IO_vfscanf)
if (((_fp)->_flags & _IO_USER_LOCK) == 0) _IO_lock_lock (*(_fp)->_lock)
# define _IO_funlockfile(_fp) \
if (((_fp)->_flags & _IO_USER_LOCK) == 0) _IO_lock_unlock (*(_fp)->_lock)
+# define _IO_ftrylockfile(_fp) \
+ (((_fp)->_flags & _IO_USER_LOCK) == 0 ? _IO_lock_trylock (*(_fp)->_lock) : 0)
# else
# define _IO_flockfile(_fp) \
if (((_fp)->_flags & _IO_USER_LOCK) == 0) _IO_flockfile (_fp)
# define _IO_funlockfile(_fp) \
if (((_fp)->_flags & _IO_USER_LOCK) == 0) _IO_funlockfile (_fp)
+# define _IO_ftrylockfile(_fp) \
+ (((_fp)->_flags & _IO_USER_LOCK) == 0 ? _IO_ftrylockfile (_fp) : 0)
# endif
#endif /* _IO_MTSAFE_IO */
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
Index: glibc-2.27/libio/genops.c
===================================================================
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
--- glibc-2.27.orig/libio/genops.c
+++ glibc-2.27/libio/genops.c
@@ -744,8 +744,8 @@ _IO_get_column (_IO_FILE *fp)
#endif
-int
-_IO_flush_all_lockp (int do_lock)
+static int
+_IO_flush_all_lockp (bool skip_locked)
{
int result = 0;
struct _IO_FILE *fp;
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
@@ -758,7 +758,16 @@ _IO_flush_all_lockp (int do_lock)
for (fp = (_IO_FILE *) _IO_list_all; fp != NULL; fp = fp->_chain)
{
run_fp = fp;
- if (do_lock)
+ if (skip_locked)
+ {
+ /* Skip files that are currently locked. */
+ if (_IO_ftrylockfile (fp))
+ {
+ run_fp = NULL;
+ continue;
+ }
+ }
+ else
_IO_flockfile (fp);
if (((fp->_mode <= 0 && fp->_IO_write_ptr > fp->_IO_write_base)
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
@@ -769,8 +778,7 @@ _IO_flush_all_lockp (int do_lock)
&& _IO_OVERFLOW (fp, EOF) == EOF)
result = EOF;
- if (do_lock)
- _IO_funlockfile (fp);
+ _IO_funlockfile (fp);
run_fp = NULL;
}
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
@@ -787,7 +795,7 @@ int
_IO_flush_all (void)
{
/* We want locking. */
- return _IO_flush_all_lockp (1);
+ return _IO_flush_all_lockp (false);
}
libc_hidden_def (_IO_flush_all)
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
@@ -852,22 +860,18 @@ _IO_unbuffer_all (void)
for (fp = (_IO_FILE *) _IO_list_all; fp; fp = fp->_chain)
{
+ run_fp = fp;
+ /* Skip files that are currently locked. */
+ if (_IO_ftrylockfile (fp))
+ {
+ run_fp = NULL;
+ continue;
+ }
+
if (! (fp->_flags & _IO_UNBUFFERED)
/* Iff stream is un-orientated, it wasn't used. */
&& fp->_mode != 0)
{
-#ifdef _IO_MTSAFE_IO
- int cnt;
-#define MAXTRIES 2
- for (cnt = 0; cnt < MAXTRIES; ++cnt)
- if (fp->_lock == NULL || _IO_lock_trylock (*fp->_lock) == 0)
- break;
- else
- /* Give the other thread time to finish up its use of the
- stream. */
- __sched_yield ();
-#endif
-
if (! dealloc_buffers && !(fp->_flags & _IO_USER_BUF))
{
fp->_flags |= _IO_USER_BUF;
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
@@ -881,16 +885,14 @@ _IO_unbuffer_all (void)
if (fp->_mode > 0)
_IO_wsetb (fp, NULL, NULL, 0);
-
-#ifdef _IO_MTSAFE_IO
- if (cnt < MAXTRIES && fp->_lock != NULL)
- _IO_lock_unlock (*fp->_lock);
-#endif
}
/* Make sure that never again the wide char functions can be
used. */
fp->_mode = -1;
+
+ _IO_funlockfile (fp);
+ run_fp = NULL;
}
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
#ifdef _IO_MTSAFE_IO
@@ -916,9 +918,9 @@ libc_freeres_fn (buffer_free)
int
_IO_cleanup (void)
{
- /* We do *not* want locking. Some threads might use streams but
- that is their problem, we flush them underneath them. */
- int result = _IO_flush_all_lockp (0);
+ /* We want to skip locked streams. Some threads might use streams but
+ that is their problem, we don't flush those. */
+ int result = _IO_flush_all_lockp (true);
/* We currently don't have a reliable mechanism for making sure that
C++ static destructors are executed in the correct order.
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
Index: glibc-2.27/libio/libioP.h
===================================================================
Accepting request 572614 from home:Andreas_Schwab:Factory - Update to glibc 2.27 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf, sinf, sincosf and tan with FMA * Optimized x86-64 trunc and truncf for processors with SSE4.1 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf * In order to support faster and safer process termination the malloc API family of functions will no longer print a failure address and stack backtrace after detecting heap corruption * The abort function terminates the process immediately, without flushing stdio streams * On platforms where long double has the IEEE binary128 format (aarch64, alpha, mips64, riscv, s390 and sparc), the math library now implements _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 These are the same interfaces added in version 2.26 for some platforms where this format is supported but is not the format of long double * On platforms with support for _Float64x (aarch64, alpha, i386, ia64, mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now implements interfaces for that type, as defined by ISO/IEC TS 18661-3:2015 * The math library now implements interfaces for the _Float32, _Float64 and _Float32x types, as defined by ISO/IEC TS 18661-3:2015 * glibc now implements the memfd_create and mlock2 functions on Linux * Support for memory protection keys was added * The copy_file_range function was added * The ldconfig utility now processes `include' directives using the C/POSIX collation ordering * Support for two grammatical forms of month names has been added * Support for the RISC-V ISA running on Linux has been added * Statically compiled applications attempting to load locales compiled for the OBS-URL: https://build.opensuse.org/request/show/572614 OBS-URL: https://build.opensuse.org/package/show/Base:System/glibc?expand=0&rev=491
2018-02-05 11:33:11 +01:00
--- glibc-2.27.orig/libio/libioP.h
+++ glibc-2.27/libio/libioP.h
@@ -486,7 +486,6 @@ extern int _IO_new_do_write (_IO_FILE *,
extern int _IO_old_do_write (_IO_FILE *, const char *, _IO_size_t);
extern int _IO_wdo_write (_IO_FILE *, const wchar_t *, _IO_size_t);
libc_hidden_proto (_IO_wdo_write)
-extern int _IO_flush_all_lockp (int);
extern int _IO_flush_all (void);
libc_hidden_proto (_IO_flush_all)
extern int _IO_cleanup (void);