diff options
| author | amodra <amodra@138bc75d-0d04-0410-961f-82ee72b054a4> | 2004-03-04 10:23:23 +0000 |
|---|---|---|
| committer | amodra <amodra@138bc75d-0d04-0410-961f-82ee72b054a4> | 2004-03-04 10:23:23 +0000 |
| commit | 773d69917577e5c6708f7005e2055bf7347a2ec1 (patch) | |
| tree | 2a5b34507a773fc68de86a6f3a19dec1a8df7b3e /gcc | |
| parent | 7262f0b8f13f7befe0dedc0755d61fc0a9fd8516 (diff) | |
| download | gcc-773d69917577e5c6708f7005e2055bf7347a2ec1.tar.gz | |
* real.c (encode_ibm_extended): Don't bother rounding low double.
* c-cppbuiltin.c (builtin_define_float_constants): Tweak MAX
when fmt->pnan < fmt->p.
git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/trunk@78900 138bc75d-0d04-0410-961f-82ee72b054a4
Diffstat (limited to 'gcc')
| -rw-r--r-- | gcc/ChangeLog | 6 | ||||
| -rw-r--r-- | gcc/c-cppbuiltin.c | 10 | ||||
| -rw-r--r-- | gcc/real.c | 4 |
3 files changed, 19 insertions, 1 deletions
diff --git a/gcc/ChangeLog b/gcc/ChangeLog index 148e55da340..071f579e3bd 100644 --- a/gcc/ChangeLog +++ b/gcc/ChangeLog @@ -1,3 +1,9 @@ +2004-03-04 Alan Modra <amodra@bigpond.net.au> + + * real.c (encode_ibm_extended): Don't bother rounding low double. + * c-cppbuiltin.c (builtin_define_float_constants): Tweak MAX + when fmt->pnan < fmt->p. + 2004-03-04 Eric Christopher <echristo@redhat.com> * config/mips/mips.h (FUNCTION_ARG_REGNO_P): Fix to check diff --git a/gcc/c-cppbuiltin.c b/gcc/c-cppbuiltin.c index 4e671d6b055..8573b8c5a46 100644 --- a/gcc/c-cppbuiltin.c +++ b/gcc/c-cppbuiltin.c @@ -193,6 +193,16 @@ builtin_define_float_constants (const char *name_prefix, const char *fp_suffix, if (i < n) *p++ = "08ce"[n - i]; sprintf (p, "p%d", fmt->emax * fmt->log2_b); + if (fmt->pnan < fmt->p) + { + /* This is an IBM extended double format made up of two IEEE + doubles. The value of the long double is the sum of the + values of the two parts. The most significant part is + required to be the value of the long double rounded to the + nearest double. Rounding means we need a slightly smaller + value for LDBL_MAX. */ + buf[4 + fmt->pnan / 4] = "7bde"[fmt->pnan % 4]; + } } sprintf (name, "__%s_MAX__", name_prefix); builtin_define_with_hex_fp_value (name, type, decimal_dig, buf, fp_suffix); diff --git a/gcc/real.c b/gcc/real.c index ed671db0e4d..d578ace0883 100644 --- a/gcc/real.c +++ b/gcc/real.c @@ -3248,7 +3248,9 @@ encode_ibm_extended (const struct real_format *fmt, long *buf, if (u.class == rvc_normal) { do_add (&v, &normr, &u, 1); - round_for_format (base_fmt, &v); + /* The low double won't need rounding, since we round to a 106 bit + mantissa before calling this function, and we've just + subtracted off the top 54 bits. (53+1 because u is rounded.) */ encode_ieee_double (base_fmt, &buf[2], &v); } else |