projects / gcc.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 28b2417) | patch
fp-bit.c (pack_d): When using paired doubles to implement a long double...
authorRichard Sandiford <rsandifo@redhat.com>
Wed, 28 Jan 2004 07:21:21 +0000 (07:21 +0000)
committerRichard Sandiford <rsandifo@gcc.gnu.org>
Wed, 28 Jan 2004 07:21:21 +0000 (07:21 +0000)
commitfaceece35d90a25ebfb9156d94cc6e8aa60bdb86
tree4c12b2cd7e8aa5cda660f0dbb77db1ea2040a0c0tree
parent28b241764ff7466a8eacff6ebd766676ae998e9bcommit | diff
fp-bit.c (pack_d): When using paired doubles to implement a long double...

* config/fp-bit.c (pack_d): When using paired doubles to implement
a long double, round the high part separately.
(unpack_d): Fix the case in which the high part is a power of two
and the low part is a nonzero value of the opposite sign.

From-SVN: r76780
gcc/ChangeLog diff | blob | history
gcc/config/fp-bit.c diff | blob | history