diff options
| author | Luc Van Oostenryck <luc.vanoostenryck@gmail.com> | 2017-08-15 11:24:01 +0200 |
|---|---|---|
| committer | Luc Van Oostenryck <luc.vanoostenryck@gmail.com> | 2018-06-23 07:46:40 +0200 |
| commit | 01951278a60490735677920c86f139d627b22dbf (patch) | |
| tree | 44e9168eacaf2053a4b96bc52a4fb38b288e07a2 /validation/expand | |
| parent | bf21036518acf0c3069c9c9e12b13fd09c8cdceb (diff) | |
| download | sparse-dev-01951278a60490735677920c86f139d627b22dbf.tar.gz | |
cast: specialize integer casts
Casts to integer used to be done with only 2 instructions:
OP_CAST & OP_SCAST.
Those are not very convenient as they don't reflect the real
operations that need to be done.
This patch specialize these instructions in:
- OP_TRUNC, for casts to a smaller type
- OP_ZEXT, for casts that need a zero extension
- OP_SEXT, for casts that need a sign extension
- Integer-to-integer casts of the same size are considered as
a NOPs and are, in fact, never emitted.
Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Diffstat (limited to 'validation/expand')
| -rw-r--r-- | validation/expand/builtin-expect.c | 8 |
1 files changed, 3 insertions, 5 deletions
diff --git a/validation/expand/builtin-expect.c b/validation/expand/builtin-expect.c index a80176a5..77f714ef 100644 --- a/validation/expand/builtin-expect.c +++ b/validation/expand/builtin-expect.c @@ -41,7 +41,7 @@ void *fptr(void *a) flia: .L0: <entry-point> - scast.32 %r2 <- (64) %arg1 + trunc.32 %r2 <- (64) %arg1 ret.32 %r2 @@ -54,7 +54,7 @@ flic: fila: .L4: <entry-point> - scast.64 %r6 <- (32) %arg1 + sext.64 %r6 <- (32) %arg1 ret.64 %r6 @@ -79,9 +79,7 @@ fils: fptr: .L12: <entry-point> - cast.64 %r12 <- (64) %arg1 - scast.64 %r13 <- (64) %r12 - ret.64 %r13 + ret.64 %arg1 * check-output-end |