summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorLuke T. Shumaker <lukeshu@lukeshu.com>2025-04-17 12:42:05 -0600
committerLuke T. Shumaker <lukeshu@lukeshu.com>2025-04-18 10:03:30 -0600
commitb44e1813b9a39827ebe5b8ade6dd68713925c58b (patch)
treeb1d7e523513317c87c61c988e685a3640c281cfc
parentf5171dab8cfaf4864519779a6adb0c230b98b45b (diff)
lib9p_srv: Document more thoughts about improving stat()
-rw-r--r--lib9p/srv_include/lib9p/srv.h5
1 files changed, 4 insertions, 1 deletions
diff --git a/lib9p/srv_include/lib9p/srv.h b/lib9p/srv_include/lib9p/srv.h
index 026075e..82d4502 100644
--- a/lib9p/srv_include/lib9p/srv.h
+++ b/lib9p/srv_include/lib9p/srv.h
@@ -75,7 +75,10 @@ lo_interface lib9p_srv_dio;
/* FIXME: I don't like that the pointers returned by stat() and
* pread() have to remain live after they return. Perhaps a
- * `respond()`-callback? But that just reads as gross in C.
+ * `respond()`-callback? But that just reads as gross in C. Also,
+ * because stat is used by srv.c internally for things other than
+ * Tstat, a `respond()`-callback makes less sense for stat() than it
+ * does for read().
*
* FIXME: It would be nice if pread() could return more than 1 iovec.
*/