win32: special-case `ENOSPC` when writing to a pipe

Since c6d3cce6f3c4 (pipe_command(): handle ENOSPC when writing to a
pipe, 2022-08-17), one `write()` call that results in an `errno` value
`ENOSPC` (which typically indicates out of disk space, which makes
little sense in the context of a pipe) is treated the same as `EAGAIN`.

However, contrary to expectations, as diagnosed in
https://github.com/python/cpython/issues/101881#issuecomment-1428667015,
when writing to a non-blocking pipe on Windows, an `errno` value of
`ENOSPC` means something else: the write _fails_. Completely. Because
more data was provided than the internal pipe buffer can handle.
Somewhat surprising, considering that `write()` is allowed to write less
than the specified amount, e.g. by writing only as much as fits in that
buffer. But it doesn't, it writes no byte at all in that instance.

Let's handle this by manually detecting when an `ENOSPC` indicates that
a pipe's buffer is smaller than what needs to be written, and re-try
using the pipe's buffer size as `size` parameter.

It would be plausible to try writing the entire buffer in a loop,
feeding pipe buffer-sized chunks, but experiments show that trying to
write more than one buffer-sized chunk right after that will immediately
fail because the buffer is unlikely to be drained as fast as `write()`
could write again. And the whole point of a non-blocking pipe is to be
non-blocking.

Which means that the logic that determines the pipe's buffer size
unfortunately has to be run potentially many times when writing large
amounts of data to a non-blocking pipe, as there is no elegant way to
cache that information between `write()` calls. It's the best we can do,
though, so it has to be good enough.

This fix is required to let t3701.60 (handle very large filtered diff)
pass with the MSYS2 runtime provided by the MSYS2 project: Without this
patch, the failed write would result in an infinite loop. This patch is
not required with Git for Windows' variant of the MSYS2 runtime only
because Git for Windows added an ugly work-around specifically to avoid
a hang in that test case.

The diff is slightly chatty because it extends an already-existing
conditional that special-cases a _different_ `errno` value for pipes,
and because this patch needs to account for the fact that
`_get_osfhandle()` potentially overwrites `errno`.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

1 files changed, 15 insertions, 4 deletions

diff --git a/compat/mingw.c b/compat/mingw.c
index ec5280da16..bdb95e7ce0 100644
--- a/compat/mingw.c
+++ b/compat/mingw.c
@@ -705,13 +705,24 @@ ssize_t mingw_write(int fd, const void *buf, size_t len)
 {
 	ssize_t result = write(fd, buf, len);
 
-	if (result < 0 && errno == EINVAL && buf) {
+	if (result < 0 && (errno == EINVAL || errno == ENOSPC) && buf) {
+		int orig = errno;
+
 		/* check if fd is a pipe */
 		HANDLE h = (HANDLE) _get_osfhandle(fd);
-		if (GetFileType(h) == FILE_TYPE_PIPE)
+		if (GetFileType(h) != FILE_TYPE_PIPE)
+			errno = orig;
+		else if (orig == EINVAL)
 			errno = EPIPE;
-		else
-			errno = EINVAL;
+		else {
+			DWORD buf_size;
+
+			if (!GetNamedPipeInfo(h, NULL, NULL, &buf_size, NULL))
+				buf_size = 4096;
+			if (len > buf_size)
+				return write(fd, buf, buf_size);
+			errno = orig;
+		}
 	}
 
 	return result;