Re: [PATCH v2] mingw: emulate write(2) that fails with a EPIPE

2 messages, 2 authors, 2016-06-15 · open the first message on its own page

Re: [PATCH v2] mingw: emulate write(2) that fails with a EPIPE

From: Junio C Hamano <hidden>
Date: 2016-06-15 23:07:30

Johannes Schindelin [off-list ref] writes:
quoted hunk
On Windows, when writing to a pipe fails, errno is always
EINVAL. However, Git expects it to be EPIPE.

According to the documentation, there are two cases in which write()
triggers EINVAL: the buffer is NULL, or the length is odd but the mode
is 16-bit Unicode (the broken pipe is not mentioned as possible cause).
Git never sets the file mode to anything but binary, therefore we know
that errno should actually be EPIPE if it is EINVAL and the buffer is
not NULL.

See https://msdn.microsoft.com/en-us/library/1570wh78.aspx for more
details.

This works around t5571.11 failing with v2.6.4 on Windows.

Signed-off-by: Johannes Schindelin <redacted>
---
 compat/mingw.c | 17 +++++++++++++++++
 compat/mingw.h |  3 +++
 2 files changed, 20 insertions(+)
diff --git a/compat/mingw.c b/compat/mingw.c
index 90bdb1e..5edea29 100644
--- a/compat/mingw.c
+++ b/compat/mingw.c
@@ -394,6 +394,23 @@ int mingw_fflush(FILE *stream)
 	return ret;
 }
 
+#undef write
+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) {
+		/* check if fd is a pipe */
+		HANDLE h = (HANDLE) _get_osfhandle(fd);
+		if (GetFileType(h) == FILE_TYPE_PIPE)
+			errno = EPIPE;
+		else
+			errno = EINVAL;
+	}
+
+	return result;
+}
+
 int mingw_access(const char *filename, int mode)
 {
 	wchar_t wfilename[MAX_PATH];
diff --git a/compat/mingw.h b/compat/mingw.h
index 738865c..57ca477 100644
--- a/compat/mingw.h
+++ b/compat/mingw.h
@@ -210,6 +210,9 @@ FILE *mingw_freopen (const char *filename, const char *otype, FILE *stream);
 int mingw_fflush(FILE *stream);
 #define fflush mingw_fflush
 
+ssize_t mingw_write(int fd, const void *buf, size_t len);
+#define write mingw_write
+
 int mingw_access(const char *filename, int mode);
 #undef access
 #define access mingw_access

PLEASE DON'T DO THE BELOW TO THE SAME MESSAGE AS THE PATCH ITSELF.
"git apply" would not read and understand the next line as a natural
language sentence for obvious reasons.
quoted hunk
Interdiff vs v1:
diff --git a/compat/mingw.c b/compat/mingw.c
index 90bdb1e..5edea29 100644
--- a/compat/mingw.c
+++ b/compat/mingw.c
@@ -394,6 +394,23 @@ int mingw_fflush(FILE *stream)
 	return ret;
 }
 
+#undef write
+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) {
+		/* check if fd is a pipe */
+		HANDLE h = (HANDLE) _get_osfhandle(fd);
+		if (GetFileType(h) == FILE_TYPE_PIPE)
+			errno = EPIPE;
+		else
+			errno = EINVAL;
+	}
+
+	return result;
+}
+
 int mingw_access(const char *filename, int mode)
 {
 	wchar_t wfilename[MAX_PATH];
diff --git a/compat/mingw.h b/compat/mingw.h
index 2aca347..57ca477 100644
--- a/compat/mingw.h
+++ b/compat/mingw.h
@@ -210,22 +210,7 @@ FILE *mingw_freopen (const char *filename, const char *otype, FILE *stream);
 int mingw_fflush(FILE *stream);
 #define fflush mingw_fflush
 
-static inline 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) {
-		/* check if fd is a pipe */
-		HANDLE h = (HANDLE) _get_osfhandle(fd);
-		if (GetFileType(h) == FILE_TYPE_PIPE)
-			errno = EPIPE;
-		else
-			errno = EINVAL;
-	}
-
-	return result;
-}
-
+ssize_t mingw_write(int fd, const void *buf, size_t len);
 #define write mingw_write
 
 int mingw_access(const char *filename, int mode);

Re: [PATCH v2] mingw: emulate write(2) that fails with a EPIPE

From: Johannes Schindelin <hidden>
Date: 2016-06-15 23:07:31

Hi Junio,

On Thu, 17 Dec 2015, Junio C Hamano wrote:
PLEASE DON'T DO THE BELOW TO THE SAME MESSAGE AS THE PATCH ITSELF.
"git apply" would not read and understand the next line as a natural
language sentence for obvious reasons.

Johannes Schindelin [off-list ref] writes:
quoted
Interdiff vs v1:
Whoops. This is an obvious bug in my patch series submission script. Will
fix.

Sorry,
Dscho
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help