https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118901
Bug ID: 118901
Summary: RISC-V: bfloat16-complex.c:(.text.startup+0x5f6):
undefined reference to `__divbc3' when zfh or zvfh
Product: gcc
Version: 15.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: majin at gcc dot gnu.org
Target Milestone: ---
$ riscv64-unknown-linux-gnu-gcc gcc/testsuite/gcc.dg/torture/bfloat16-complex.c
-march=rv64gc_zfh -mabi=lp64d -mcmodel=medany
riscv64-unknown-linux-gnu/bin/ld: /tmp/cc3e0Vp7.o: in function `main':
bfloat16-complex.c:(.text+0x79e): undefined reference to `__mulbc3'
riscv64-unknown-linux-gnu/bin/ld: bfloat16-complex.c:(.text+0x842): undefined
reference to `__divbc3'
collect2: error: ld returned 1 exit status
In the floating-point complex test cases, ZFH has led to failures due to
invalid libcalls.
I performed some basic debugging and it seems that the issue is caused by the
function below returning FLT_EVAL_METHOD_PROMOTE_TO_FLOAT16. Perhaps someone
here can provide some suggestions.
```
/* Set the value of FLT_EVAL_METHOD.
ISO/IEC TS 18661-3 defines two values that we'd like to make use of:
0: evaluate all operations and constants, whose semantic type has at
most the range and precision of type float, to the range and
precision of float; evaluate all other operations and constants to
the range and precision of the semantic type;
N, where _FloatN is a supported interchange floating type
evaluate all operations and constants, whose semantic type has at
most the range and precision of _FloatN type, to the range and
precision of the _FloatN type; evaluate all other operations and
constants to the range and precision of the semantic type;
If we have the zfh/zhinx/zvfh extensions then we support _Float16
in native precision, so we should set this to 16. */
static enum flt_eval_method
riscv_excess_precision (enum excess_precision_type type)
{
switch (type)
{
case EXCESS_PRECISION_TYPE_FAST:
case EXCESS_PRECISION_TYPE_STANDARD:
return ((TARGET_ZFH || TARGET_ZHINX || TARGET_ZVFH)
? FLT_EVAL_METHOD_PROMOTE_TO_FLOAT16
: FLT_EVAL_METHOD_PROMOTE_TO_FLOAT);
case EXCESS_PRECISION_TYPE_IMPLICIT:
case EXCESS_PRECISION_TYPE_FLOAT16:
return FLT_EVAL_METHOD_PROMOTE_TO_FLOAT16;
default:
gcc_unreachable ();
}
return FLT_EVAL_METHOD_UNPREDICTABLE;
}
```
