[PATCH] newlib: fclose: Use sfp lock while fp lock is active

Takashi Yano takashi.yano@nifty.ne.jp
Wed Jul 16 14:42:02 GMT 2025


With the commit 656df313e08a, if a thread acquires sfp lock after
another thread calls fclose() and fp lock is acquired, the first
thread falls into deadlock if it trys to acquire fp lock. This can
happen if the first thread calls __sft_lock_all() while the second
thread calls fclose().

This patch reverts the changes for newlib/libc/stdio/fclose.c in
the commit 656df313e08a.

Addresses: https://cygwin.com/pipermail/cygwin/2025-June/258323.html
Fixes: 656df313e08a ("* libc/stdio/fclose.c: Only use sfp lock to guard non-atomic changes of flags and fp lock.")
Reported-by: Takashi Yano <takashi.yano@nifty.ne.jp>
Reviewed-by:
Signed-off-by: Takashi Yano <takashi.yano@nifty.ne.jp>
---
 newlib/libc/stdio/fclose.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/newlib/libc/stdio/fclose.c b/newlib/libc/stdio/fclose.c
index 983ae2cc4..4c81cbf4b 100644
--- a/newlib/libc/stdio/fclose.c
+++ b/newlib/libc/stdio/fclose.c
@@ -72,6 +72,7 @@ _fclose_r (struct _reent *rptr,
   int __oldcancel;
   pthread_setcancelstate (PTHREAD_CANCEL_DISABLE, &__oldcancel);
 #endif
+  __sfp_lock_acquire ();
   if (!(fp->_flags2 & __SNLK))
     _flockfile (fp);
 
@@ -79,6 +80,7 @@ _fclose_r (struct _reent *rptr,
     {
       if (!(fp->_flags2 & __SNLK))
 	_funlockfile (fp);
+      __sfp_lock_release ();
 #ifdef _STDIO_WITH_THREAD_CANCELLATION_SUPPORT
       pthread_setcancelstate (__oldcancel, &__oldcancel);
 #endif
@@ -101,7 +103,6 @@ _fclose_r (struct _reent *rptr,
     FREEUB (rptr, fp);
   if (HASLB (fp))
     FREELB (rptr, fp);
-  __sfp_lock_acquire ();
   fp->_flags = 0;		/* release this FILE for reuse */
   if (!(fp->_flags2 & __SNLK))
     _funlockfile (fp);
-- 
2.45.1



More information about the Newlib mailing list