1 2020-06-01 Uroš Bizjak <ubizjak@gmail.com>
4 * config/fpu-387.h (struct fenv): Add __attribute__ ((gcc_struct)).
6 2020-05-29 H.J. Lu <hjl.tools@gmail.com>
9 * configure: Regenerated.
11 2020-05-29 Jakub Jelinek <jakub@redhat.com>
14 * Makefile.am (i_findloc0_c): Add findloc0_i10.c.
15 (i_findloc1_c): Add findloc1_i10.c.
16 * gfortran.map (GFORTRAN_10.2): New symbol version, export
17 _gfortran_{,m,s}findloc{0,1}_c10 symbols.
18 * Makefile.in: Regenerated.
19 * generated/findloc0_c10.c: Generated.
20 * generated/findloc1_c10.c: Generated.
22 2020-05-28 Harald Anlauf <anlauf@gmx.de>
25 * io/unit.c (unlock_unit): Guard by check for NULL pointer.
27 2020-05-26 Harald Anlauf <anlauf@gmx.de>
28 Steven G. Kargl <kargl@gcc.gnu.org>
31 * io/transfer.c (st_wait_async): Do not dereference NULL pointer.
33 2020-05-26 Harald Anlauf <anlauf@gmx.de>
36 * io/transfer.c (finalize_transfer): Generate runtime error for
37 namelist input/output to unformatted file.
39 2020-05-23 Thomas Koenig <tkoenig@gcc.gnu.org>
42 * io/async.c (async_wait_id): Generate error if ID is higher
43 than the highest current ID.
44 * runtime/error.c (translate_error): Handle LIBERROR_BAD_WAIT_ID.
46 2020-05-21 H.J. Lu <hongjiu.lu@intel.com>
48 * m4/matmul.m4: Don't include <config/i386/cpuinfo.h>. Use
49 __builtin_cpu_is/__builtin_cpu_supports
50 * generated/matmul_c10.c: Regenerated.
51 * generated/matmul_c16.c: Likewise.
52 * generated/matmul_c4.c: Likewise.
53 * generated/matmul_c8.c: Likewise.
54 * generated/matmul_i1.c: Likewise.
55 * generated/matmul_i16.c: Likewise.
56 * generated/matmul_i2.c: Likewise.
57 * generated/matmul_i4.c: Likewise.
58 * generated/matmul_i8.c: Likewise.
59 * generated/matmul_r10.c: Likewise.
60 * generated/matmul_r16.c: Likewise.
61 * generated/matmul_r4.c: Likewise.
62 * generated/matmul_r8.c: Likewise.
64 2020-05-15 H.J. Lu <hongjiu.lu@intel.com>
67 * configure: Regenerated.
69 2020-05-14 Thomas Koenig <tkoenig@gcc.gnu.org>
72 * io/close.c (close_status): Add CLOSE_INVALID.
73 (st_close): Return early on invalid STATUS parameter.
75 2020-05-14 H.J. Lu <hongjiu.lu@intel.com>
77 * configure: Regenerated.
79 2020-05-06 Uroš Bizjak <ubizjak@gmail.com>
81 * config/fpu-387.h (__math_force_eval): Remove.
82 (__math_force_eval_div): New define.
83 (local_feraiseexcept): Use __math_force_eval_div to use
84 generic division to generate INVALID, DIVZERO and INEXACT
86 (struct fenv): Define named struct instead of typedef.
88 2020-05-01 Uroš Bizjak <ubizjak@gmail.com>
90 * config/fpu-387.h (__math_force_eval): New define.
91 (local_feraiseexcept): Use __math_force_eval to evaluate
92 generic division to generate INVALID and DIVZERO exceptions.
94 2020-04-22 Fritz Reese <foreese@gcc.gnu.org>
96 * intrinsics/trigd.c, intrinsics/trigd_lib.inc, intrinsics/trigd.inc:
97 Guard against unavailable math functions.
98 Use suffixes from kinds.h based on the REAL kind.
100 2020-04-22 Jakub Jelinek <jakub@redhat.com>
104 * configure.ac: Add math func checks for fmaf, fma and fmal. Add
105 HAVE_INLINE_BUILTIN_COPYSIGN check.
106 * c99_protos.h (copysign, fmaf, fma, fmal): Provide fallback
108 (HAVE_COPYSIGN, HAVE_FMAF, HAVE_FMA, HAVE_FMAL): Define if not
109 defined and fallback version is provided.
110 * intrinsics/c99_functions.c (copysign, fmaf, fma, fmal): Provide
111 fallback implementations if possible
112 * configure: Regenerated.
113 * config.h.in: Regenerated.
115 2020-04-19 Uroš Bizjak <ubizjak@gmail.com>
117 * config/fpu-387.h (local_feraiseexcept) [__SSE_MATH__]:
118 Remove unneeded assignments to volatile memory.
120 2020-04-01 Fritz Reese <foreese@gcc.gnu.org>
121 Steven G. Kargl <kargl@gcc.gnu.org>
124 * Makefile.am, Makefile.in: New make rule for intrinsics/trigd.c.
125 * gfortran.map: New routines for {sind, cosd, tand}X{r4, r8, r10, r16}.
126 * intrinsics/trigd.c, intrinsics/trigd_lib.inc, intrinsics/trigd.inc:
127 New files. Defines native degree-valued trig functions.
129 2020-02-18 Thomas Koenig <tkoenig@gcc.gnu.org>
132 * io/async.c (destroy_adv_cond): Do not destroy lock.
133 (async_io): Make sure au->lock is locked for finishing of thread.
134 Do not lock/unlock around signalling emptysignal. Unlock au->lock
136 (init_adv_cond): Do not initialize lock.
137 (enqueue_transfer): Unlock after signal.
138 (enqueue_done_id): Likewise.
139 (enqueue_done): Likewise.
140 (enqueue_close): Likewise.
141 (enqueue_data_transfer): Likewise.
142 (async_wait_id): Do not lock/unlock around signalling au->work.
143 (async_wait): Unlock after signal.
144 * io/async.h (SIGNAL): Add comment about needed au->lock.
145 Remove locking/unlocking of advcond->lock.
146 (WAIT_SIGNAL_MUTEX): Add comment. Remove locking/unlocking of
147 advcond->lock. Unlock mutex only at the end. Loop on
148 __ghread_cond_wait returning zero.
149 (REVOKE_SIGNAL): Add comment. Remove locking/unlocking of
151 (struct adv_cond): Remove mutex from struct.
153 2020-02-12 Sandra Loosemore <sandra@codesourcery.com>
158 * configure: Regenerated.
160 2020-01-24 Maciej W. Rozycki <macro@wdc.com>
162 * configure.ac: Handle `--with-toolexeclibdir='.
163 * Makefile.in: Regenerate.
164 * aclocal.m4: Regenerate.
165 * configure: Regenerate.
167 2020-01-17 Jerry DeLisle <jvdelisle@gcc.gnu.org>
170 * io/unit.c (set_internal_unit): Set round and sign flags
173 2020-01-17 Jerry DeLisle <jvdelisle@gcc.gnu.org>
176 * io/format.c (parse_format_list): Zero width not allowed with
178 * io/write_float.def (build_float_string): Include range of
179 higher exponent values that require wider width.
181 2020-01-01 Jerry DeLisle <jvdelisle@gcc.gnu.org>
184 * io/format.c (parse_format_list): Implement the E0 exponent
185 width to provide smallest possible width for exponent fields.
186 Refactor code for correct parsing and better readability of the
188 * io/io.h (write_real_w0): Change interface to pass in pointer
190 * io/transfer.c: Update all calls to write_real_w0 to use the
192 * io/write.c ((write_real_w0): Use the new interface with fnode
193 to access both the decimal precision and exponent widths used in
195 * io/write_float.def (build_float_string): Use the passed in
196 exponent width to calculate the used width in the case of E0.
198 2020-01-01 Jakub Jelinek <jakub@redhat.com>
200 Update copyright years.
202 Copyright (C) 2020 Free Software Foundation, Inc.
204 Copying and distribution of this file, with or without modification,
205 are permitted in any medium without royalty provided the copyright
206 notice and this notice are preserved.