# [PATCH] mingw: do not set errno to 0 on success 3 messages from 2010-11-23 to 2010-11-24. Participants: Erik Faye-Lund, Junio C Hamano. Thread: https://gitlist.dev/t/25825 ## Erik Faye-Lund, 2010-11-23 19:53 Subject: [PATCH] mingw: do not set errno to 0 on success Message-ID: <1290541988-4608-1-git-send-email-kusmabite@gmail.com> URL: https://gitlist.dev/e/1290541988-4608-1-git-send-email-kusmabite%40gmail.com ``` Currently do_lstat always sets errno to 0 on success. This incorrectly overwrites previous errors. Fetch the error-code into a temporary variable instead, and assign that to errno on failure. Signed-off-by: Erik Faye-Lund --- A bug I found while hunting down another regression. maint-worthy, perhaps? compat/mingw.c | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/compat/mingw.c b/compat/mingw.c index f2d9e1f..b98e600 100644 --- a/compat/mingw.c +++ b/compat/mingw.c @@ -195,9 +195,10 @@ static inline time_t filetime_to_time_t(const FILETIME *ft) */ static int do_lstat(const char *file_name, struct stat *buf) { + int err; WIN32_FILE_ATTRIBUTE_DATA fdata; - if (!(errno = get_file_attr(file_name, &fdata))) { + if (!(err = get_file_attr(file_name, &fdata))) { buf->st_ino = 0; buf->st_gid = 0; buf->st_uid = 0; @@ -211,6 +212,7 @@ static int do_lstat(const char *file_name, struct stat *buf) buf->st_ctime = filetime_to_time_t(&(fdata.ftCreationTime)); return 0; } + errno = err; return -1; } -- 1.7.3.2 ``` ## Junio C Hamano, 2010-11-24 00:07 Subject: Re: [PATCH] mingw: do not set errno to 0 on success Message-ID: <7v4ob7vb1m.fsf@alter.siamese.dyndns.org> URL: https://gitlist.dev/e/7v4ob7vb1m.fsf%40alter.siamese.dyndns.org In-Reply-To: <1290541988-4608-1-git-send-email-kusmabite@gmail.com> ``` Erik Faye-Lund writes: > Currently do_lstat always sets errno to 0 on success. This incorrectly > overwrites previous errors. > > Fetch the error-code into a temporary variable instead, and assign that > to errno on failure. > > Signed-off-by: Erik Faye-Lund > --- > > A bug I found while hunting down another regression. maint-worthy, perhaps? I would say so, as long as the error return from get_file_attr() is trustworthy. > > compat/mingw.c | 4 +++- > 1 files changed, 3 insertions(+), 1 deletions(-) > > diff --git a/compat/mingw.c b/compat/mingw.c > index f2d9e1f..b98e600 100644 > --- a/compat/mingw.c > +++ b/compat/mingw.c > @@ -195,9 +195,10 @@ static inline time_t filetime_to_time_t(const FILETIME *ft) > */ > static int do_lstat(const char *file_name, struct stat *buf) > { > + int err; > WIN32_FILE_ATTRIBUTE_DATA fdata; > > - if (!(errno = get_file_attr(file_name, &fdata))) { > + if (!(err = get_file_attr(file_name, &fdata))) { > buf->st_ino = 0; > buf->st_gid = 0; > buf->st_uid = 0; > @@ -211,6 +212,7 @@ static int do_lstat(const char *file_name, struct stat *buf) > buf->st_ctime = filetime_to_time_t(&(fdata.ftCreationTime)); > return 0; > } > + errno = err; > return -1; > } > > -- > 1.7.3.2 ``` ## Erik Faye-Lund, 2010-11-24 09:54 Subject: Re: [PATCH] mingw: do not set errno to 0 on success Message-ID: URL: https://gitlist.dev/e/AANLkTi%3D8OnTJHgX3-g9uqLX2B3%2BeUQf0uykynrwW3L4E%40mail.gmail.com In-Reply-To: <7v4ob7vb1m.fsf@alter.siamese.dyndns.org> ``` On Wed, Nov 24, 2010 at 1:07 AM, Junio C Hamano wrote: > Erik Faye-Lund writes: > >> Currently do_lstat always sets errno to 0 on success. This incorrectly >> overwrites previous errors. >> >> Fetch the error-code into a temporary variable instead, and assign that >> to errno on failure. >> >> Signed-off-by: Erik Faye-Lund >> --- >> >> A bug I found while hunting down another regression. maint-worthy, perhaps? > > I would say so, as long as the error return from get_file_attr() is > trustworthy. It looks trustworthy to me; it returns 0 if GetFileAttributesExA succeeds, and either EACCES, ENAMETOOLONG, ENOMEM or ENOENT based on the result of GetLastError(). ```