eval: Avoid implementation-defined unsigned to signed conversions

Instead, use an additional int64_t member in the union. Since
exact-width integer types have no padding bits or trap representations,
and use two's-complement representation, we can portably access an
int64_t union member stored as uint64_t and vice-versa.

This allows us to reinterpret the value without invoking potentially
implementation-defined behavior of casting an unsigned integer to
a signed integer type which may not be able to represent its value.

2 files changed, 38 insertions, 37 deletions

diff --git a/cc.h b/cc.h
index 9597d6e..51b8f86 100644
--- a/cc.h
+++ b/cc.h
@@ -328,6 +328,7 @@ struct expr {
 		} ident;
 		union {
 			uint64_t u;
+			int64_t i;
 			double f;
 		} constant;
 		struct {
diff --git a/eval.c b/eval.c
index 2590687..958f517 100644
--- a/eval.c
+++ b/eval.c
@@ -17,13 +17,13 @@ cast(struct expr *expr)
 	else if (expr->type->prop & PROPINT && expr->type->basic.issigned)
 		size |= S;
 	switch (size) {
-	case 1:   expr->constant.u = (uint8_t)expr->constant.u;  break;
-	case 1|S: expr->constant.u = (int8_t)expr->constant.u;   break;
+	case 1:   expr->constant.u = (uint8_t)expr->constant.u; break;
+	case 1|S: expr->constant.u = (expr->constant.u & UINT8_MAX ^ INT8_MIN) - INT8_MIN; break;
 	case 2:   expr->constant.u = (uint16_t)expr->constant.u; break;
-	case 2|S: expr->constant.u = (int16_t)expr->constant.u;  break;
+	case 2|S: expr->constant.u = (expr->constant.u & UINT16_MAX ^ INT16_MIN) - INT16_MIN; break;
 	case 4:   expr->constant.u = (uint32_t)expr->constant.u; break;
-	case 4|S: expr->constant.u = (int32_t)expr->constant.u;  break;
-	case 4|F: expr->constant.f = (float)expr->constant.f;    break;
+	case 4|S: expr->constant.u = (expr->constant.u & UINT32_MAX ^ INT32_MIN) - INT32_MIN; break;
+	case 4|F: expr->constant.f = (float)expr->constant.f; break;
 	}
 }
 
@@ -37,47 +37,47 @@ binary(struct expr *expr, enum tokenkind op, struct expr *l, struct expr *r)
 		op |= S;
 	switch (op) {
 	case TMUL:
-	case TMUL|S:    expr->constant.u = l->constant.u * r->constant.u; break;
-	case TMUL|F:    expr->constant.f = l->constant.f * r->constant.f; break;
-	case TDIV:      expr->constant.u = l->constant.u / r->constant.u; break;
-	case TDIV|S:    expr->constant.u = (int64_t)l->constant.u / (int64_t)r->constant.u; break;
-	case TDIV|F:    expr->constant.f = l->constant.f / r->constant.f; break;
-	case TMOD:      expr->constant.u = l->constant.u % r->constant.u; break;
-	case TMOD|S:    expr->constant.u = (int64_t)l->constant.u % (int64_t)r->constant.u; break;
+	case TMUL|S:     expr->constant.u = l->constant.u * r->constant.u; break;
+	case TMUL|F:     expr->constant.f = l->constant.f * r->constant.f; break;
+	case TDIV:       expr->constant.u = l->constant.u / r->constant.u; break;
+	case TDIV|S:     expr->constant.i = l->constant.i / r->constant.i; break;
+	case TDIV|F:     expr->constant.f = l->constant.f / r->constant.f; break;
+	case TMOD:       expr->constant.u = l->constant.u % r->constant.u; break;
+	case TMOD|S:     expr->constant.i = l->constant.i % r->constant.i; break;
 	case TADD:
-	case TADD|S:    expr->constant.u = l->constant.u + r->constant.u; break;
-	case TADD|F:    expr->constant.f = l->constant.f + r->constant.f; break;
+	case TADD|S:     expr->constant.u = l->constant.u + r->constant.u; break;
+	case TADD|F:     expr->constant.f = l->constant.f + r->constant.f; break;
 	case TSUB:
-	case TSUB|S:    expr->constant.u = l->constant.u - r->constant.u; break;
-	case TSUB|F:    expr->constant.f = l->constant.f - r->constant.f; break;
+	case TSUB|S:     expr->constant.u = l->constant.u - r->constant.u; break;
+	case TSUB|F:     expr->constant.f = l->constant.f - r->constant.f; break;
 	case TSHL:
-	case TSHL|S:    expr->constant.u = l->constant.u << r->constant.u; break;
-	case TSHR:      expr->constant.u = l->constant.u >> r->constant.u; break;
-	case TSHR|S:    expr->constant.u = (int64_t)l->constant.u >> r->constant.u; break;
+	case TSHL|S:     expr->constant.u = l->constant.u << (r->constant.u & 63); break;
+	case TSHR:       expr->constant.u = l->constant.u >> (r->constant.u & 63); break;
+	case TSHR|S:     expr->constant.i = l->constant.i >> (r->constant.u & 63); break;
 	case TBAND:
-	case TBAND|S:   expr->constant.u = l->constant.u & r->constant.u; break;
+	case TBAND|S:    expr->constant.u = l->constant.u & r->constant.u; break;
 	case TBOR:
-	case TBOR|S:    expr->constant.u = l->constant.u | r->constant.u; break;
+	case TBOR|S:     expr->constant.u = l->constant.u | r->constant.u; break;
 	case TXOR:
-	case TXOR|S:    expr->constant.u = l->constant.u ^ r->constant.u; break;
-	case TLESS:     expr->constant.u = l->constant.u < r->constant.u; break;
-	case TLESS|S:   expr->constant.u = (int64_t)l->constant.u < (int64_t)r->constant.u; break;
-	case TLESS|F:   expr->constant.u = l->constant.f < r->constant.f; break;
-	case TGREATER:  expr->constant.u = l->constant.u > r->constant.u; break;
-	case TGREATER|S: expr->constant.u = (int64_t)l->constant.u > (int64_t)r->constant.u; break;
+	case TXOR|S:     expr->constant.u = l->constant.u ^ r->constant.u; break;
+	case TLESS:      expr->constant.u = l->constant.u < r->constant.u; break;
+	case TLESS|S:    expr->constant.u = l->constant.i < r->constant.i; break;
+	case TLESS|F:    expr->constant.u = l->constant.f < r->constant.f; break;
+	case TGREATER:   expr->constant.u = l->constant.u > r->constant.u; break;
+	case TGREATER|S: expr->constant.u = l->constant.i > r->constant.i; break;
 	case TGREATER|F: expr->constant.u = l->constant.f > r->constant.f; break;
-	case TLEQ:      expr->constant.u = l->constant.u <= r->constant.u; break;
-	case TLEQ|S:    expr->constant.u = (int64_t)l->constant.u <= (int64_t)r->constant.u; break;
-	case TLEQ|F:    expr->constant.u = l->constant.f <= r->constant.f; break;
-	case TGEQ:      expr->constant.u = l->constant.u >= r->constant.u; break;
-	case TGEQ|S:    expr->constant.u = (int64_t)l->constant.u >= (int64_t)r->constant.u; break;
-	case TGEQ|F:    expr->constant.u = l->constant.f >= r->constant.f; break;
+	case TLEQ:       expr->constant.u = l->constant.u <= r->constant.u; break;
+	case TLEQ|S:     expr->constant.u = l->constant.i <= r->constant.i; break;
+	case TLEQ|F:     expr->constant.u = l->constant.f <= r->constant.f; break;
+	case TGEQ:       expr->constant.u = l->constant.u >= r->constant.u; break;
+	case TGEQ|S:     expr->constant.u = l->constant.i >= r->constant.i; break;
+	case TGEQ|F:     expr->constant.u = l->constant.f >= r->constant.f; break;
 	case TEQL:
-	case TEQL|S:    expr->constant.u = l->constant.u == r->constant.u; break;
-	case TEQL|F:    expr->constant.u = l->constant.f == r->constant.f; break;
+	case TEQL|S:     expr->constant.u = l->constant.u == r->constant.u; break;
+	case TEQL|F:     expr->constant.u = l->constant.f == r->constant.f; break;
 	case TNEQ:
-	case TNEQ|S:    expr->constant.u = l->constant.u != r->constant.u; break;
-	case TNEQ|F:    expr->constant.u = l->constant.f != r->constant.f; break;
+	case TNEQ|S:     expr->constant.u = l->constant.u != r->constant.u; break;
+	case TNEQ|F:     expr->constant.u = l->constant.f != r->constant.f; break;
 	default:
 		fatal("internal error; unknown binary expression");
 	}