duskos

commit b8b4fd38490dd5a04b10e354f0c2840b1dfe25a8
parent c4ec74e921086d00e2e58b1e7eca1e8cc6ea4b25
Author: Virgil Dupras <hsoft@hardcoded.net>
Date:   Wed, 30 Nov 2022 11:10:32 -0500

ar/puff: remove false TODO

The problem came from ?: not guaranteeing eval shortcut, but that's by design.

Diffstat:
M
fs/ar/puff.c
 | 
1
-
M
fs/doc/cc/usage.txt
 | 
3
++-

2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/fs/ar/puff.c b/fs/ar/puff.c
@@ -700,7 +700,6 @@ int puff(unsigned char *dest,			/* pointer to destination pointer */
 	do {
 		last = bits(&s, 1);			/* one if last block */
 		type = bits(&s, 2);			/* block type 0..3 */
-		// TODO: the ?: ops fail on this
 		if (type == 0) err = stored(&s);
 		else if (type == 1) err = fixed(&s);
 		else if (type == 2) err = dynamic(&s);
diff --git a/fs/doc/cc/usage.txt b/fs/doc/cc/usage.txt
@@ -32,7 +32,8 @@ are a few differences:
   * "unsigned" always goes first
   * no "signed" (always default), no "auto"
 * No logical shortcut guarantee. In (a && b), b will be executed even if a
-  yields 0.
+  yields 0. This is also true of the "x ? y : z" operator. Both y and z are
+  evaluated regardless of x.
 * Number literals are the same as Dusk OS, so 12345, $1234 and 'X'. No 0x1234
   or 0o777.
 * string literals are not null-terminated, but "counted strings". The exact same