This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
[RFC v5 19/21] RISC-V: Fix llrint and llround missing exceptions on RV32
- From: Alistair Francis <alistair dot francis at wdc dot com>
- To: libc-alpha at sourceware dot org
- Cc: arnd at arndb dot de, adhemerval dot zanella at linaro dot org, fweimer at redhat dot com, palmer at sifive dot com, macro at wdc dot com, zongbox at gmail dot com, alistair dot francis at wdc dot com, alistair23 at gmail dot com
- Date: Thu, 29 Aug 2019 09:50:51 -0700
- Subject: [RFC v5 19/21] RISC-V: Fix llrint and llround missing exceptions on RV32
- Ironport-sdr: xE8hdSU3UsERFrK/U1en/89/vMGI5wi1J+KkaKb3A4Sh3oxRW157+5QPFHo3ZopKCtGOh7XaAi BFaXgkZ6WUueRaW8H6vd04wh3L1dFtkCiH5HOkG37osQREpSynPi0aNUxtcvY17FJ035Di2nyJ M6jApffTjX9VmUTLQGp8xO/WzEQDw7ur4DBTVr0Tx9qgXlUj3Wd8T18Ykr/YatS+k2umGuMAuU hmddsBelU0YUQOAH88yT4xhcdoIJiUS6rgTMsMlWhZzPOgMnqw2NB1o/8IN/SHTB3YcLORCRQ9 6QM=
- Ironport-sdr: qlN59ohknsOBr4Ovd2CAv1u11QDHJO1NIU/B4a9WyMM5zALjT+8P//w0oajBUpXzHAVJQI3AKs NKhcUfz0RPeavmbQhkn84wlCDGf2VfClwc4nUI5JrgljvhOS2pAGaJr9nbk6w1oXm7n8DL2d+4 gUG6whilm82UJXRRzg9qVvN7ma2VnZkIscgLtH1bhYQWvQmuwyFJ9QU1NCLMTZX1EZok5NC7Cg S7N+2DZlBfTZ4H4C2UrO5dNMZH+ZVN7GQEm2XzbLSpdRTCq9KR6jcqaG/+Hg1eDrWblWRHKz5G abIm6iy+SX9EmY2+4/qw4ViQ
- Ironport-sdr: kFm1rSIICnylI0c7NL9bom5q3TncXa7lD1OfQRImp189wPgOBlvigGVQiIDITttc35S3/LV2P9 6zyNdwCR/RisfWDlEuZa12oPeK6i+T5BWJoV9sn6Mlp+0isaORKKajWnclEwTK5H/tWIQI8/1j 6j3oITrUEIRwPanZ7Dr87+E4z2WIGkOkjjHztw2pIA8LQuFCKZTJ8Xhqw3gTb1reikhVr39m0T jJppufSdiKZFwXBohhpzLwjGXoLyGhCWbLjEQUE5ftbK2Pek90bdp3bxmQWKKBoDgpgdmFHKps ELE=
- References: <cover.1567097252.git.alistair.francis@wdc.com>
- Wdcironportexception: Internal
From: Zong Li <zongbox@gmail.com>
Similar to the recent fix for MIPS, ARM and S/390, RV32 is missing
correct exception on overflow from llrint and llround functions because
cast from floating-point types to long long do not result in correct
exceptions on overflow.
2018-11-29 Zong Li <zong@andestech.com>
* sysdeps/riscv/rv32/fix-fp-int-convert-overflow.h: New file.
---
.../riscv/rv32/fix-fp-int-convert-overflow.h | 38 +++++++++++++++++++
1 file changed, 38 insertions(+)
create mode 100644 sysdeps/riscv/rv32/fix-fp-int-convert-overflow.h
diff --git a/sysdeps/riscv/rv32/fix-fp-int-convert-overflow.h b/sysdeps/riscv/rv32/fix-fp-int-convert-overflow.h
new file mode 100644
index 00000000000..42155048382
--- /dev/null
+++ b/sysdeps/riscv/rv32/fix-fp-int-convert-overflow.h
@@ -0,0 +1,38 @@
+/* Fix for conversion of floating point to integer overflow. ARM version.
+ Copyright (C) 2015-2018 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, see
+ <http://www.gnu.org/licenses/>. */
+
+#ifndef FIX_FP_INT_CONVERT_OVERFLOW_H
+#define FIX_FP_INT_CONVERT_OVERFLOW_H 1
+
+/* The generic libgcc2.c conversions from floating point
+ to long long may not raise the correct exceptions on overflow (and
+ may raise spurious "inexact" exceptions even in non-overflow cases,
+ see <https://gcc.gnu.org/bugzilla/show_bug.cgi?id=59412>). */
+#define FIX_FLT_LONG_CONVERT_OVERFLOW 0
+#define FIX_FLT_LLONG_CONVERT_OVERFLOW 1
+
+#define FIX_DBL_LONG_CONVERT_OVERFLOW 0
+#define FIX_DBL_LLONG_CONVERT_OVERFLOW 1
+
+#define FIX_LDBL_LONG_CONVERT_OVERFLOW 0
+#define FIX_LDBL_LLONG_CONVERT_OVERFLOW 0
+
+#define FIX_FLT128_LONG_CONVERT_OVERFLOW 0
+#define FIX_FLT128_LLONG_CONVERT_OVERFLOW 0
+
+#endif /* fix-fp-int-convert-overflow.h */
--
2.22.0