suzanne.soy/racket
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

bc: fix compiler for set! as first subexpression of begin0

Browse Source 
An optimization relatively late in the BC bytecode compiler pipeline
was wrong for `begin0`. The transformation and bug must be a very old,
since it's intended to help the bytecode interpreter.

Thanks to Sage for reporting and Alexis for initial debugging.
This commit is contained in:
Matthew Flatt 2020-06-17 07:02:49 -06:00
parent 2de28c8cf3
commit 22069faebc
2 changed files with 14 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
pkgs/racket-test-core/tests/racket/syntax.rktl
View File

@ -2273,6 +2273,15 @@
(err/rt-test (write c (open-output-bytes))
exn:fail?))
;; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;; Regression test to make sure `(set! ...)` on a local variable
;; in the first position of ` begin0` is not miscompiled
(test (void) (let ([i 0])
(λ () (begin0
(set! i (add1 i))
(+ i 1)))))
;; ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
(report-errs)

8
racket/src/racket/src/resolve.c
View File

@ -506,13 +506,15 @@ static Scheme_Object *resolve_wcm(Scheme_Object *o, Resolve_Info *info)
static Scheme_Object *look_for_letv_change(Scheme_Sequence *s)
{
int i;
int i, start;
/* Change (begin e1 ... (set!-for-let [x 10] (void)) e2 ...)
to (begin e1 ... (set!-for-let [x 10] e2 ...)), which
avoids an unneeded recursive call in the evaluator */
for (i = 0; i < s->count - 1; i++) {
start = ((SCHEME_TYPE(s) == scheme_begin0_sequence_type) ? 1 : 0);
for (i = start; i < s->count - 1; i++) {
Scheme_Object *v;
v = s->array[i];
if (SAME_TYPE(SCHEME_TYPE(v), scheme_let_value_type)) {
@ -570,7 +572,7 @@ static Scheme_Object *resolve_sequence(Scheme_Object *o, Resolve_Info *info)
le = resolve_expr(s->array[i], info);
s->array[i] = le;
}
return look_for_letv_change(s);
}