unofficial mirror of bug-gnu-emacs@gnu.org 
 help / color / mirror / code / Atom feed
From: Bernardo <bernardo.bacic@pobox.com>
To: Robert Pluim <rpluim@gmail.com>, Eli Zaretskii <eliz@gnu.org>
Cc: 37884@debbugs.gnu.org
Subject: bug#37884: 27.0.50; Cannot write to a file in VirtualBox shared directory
Date: Mon, 28 Oct 2019 20:23:22 +1100	[thread overview]
Message-ID: <874kztdy0l.fsf@pobox.com> (raw)
In-Reply-To: <m2r22yp48o.fsf@gmail.com> (Robert Pluim's message of "Sun, 27 Oct 2019 17:01:11 +0100")

Robert Pluim <rpluim@gmail.com> writes:

>>>>>> On Sun, 27 Oct 2019 17:52:39 +0200, Eli Zaretskii <eliz@gnu.org> said:
>
>     >> From: Robert Pluim <rpluim@gmail.com>
>     >> Date: Sun, 27 Oct 2019 16:34:54 +0100
>     >> Cc: 37884@debbugs.gnu.org
>     >> 
>     >> Eli, if the results come back that using 0664 or similar on lockfiles
>     >> resolves this, would you be amenable to such a change?
>
>     Eli> I'm not sure I understand the proposal.  Is the suggestion to chmod
>     Eli> the file to 0644 before calling unlink?
>
> Yes, although of course this is assuming it works, one other option
> could be to just not mess with the file's permissions at all when
> initially creating the lockfile.
>

Hi Robert,

what you suggested (changing file mode) works fine; I haven't
investigated Eli's suggestion (w32.c)

in the end my test code looked like this (sans the headers)
int main()
{
	const char buff[] = "abc";
	int result = 0;
    
	int fd = open( "another_test", O_RDWR|O_CREAT|O_EXCL|O_CLOEXEC, 0600);
	result = write( fd , buff, 3 );
	printf( "write returned %d\n", result );
	result = fchmod( fd, 0444 );
	printf( "fchmod returned %d\n", result );
	close( fd );
	result = unlink( "another_test" );
	printf( "unlink returned %d\n", result );
	return 0;
}


and the output was:
write returned 3
fchmod returned 0
unlink returned -1

and when the file access mode was changed to 0644:
write returned 3
fchmod returned 0
unlink returned 0

and the lock file was removed successfully;

Afterwards, i modified a file in Emacs source tree like so:

$ git diff src/filelock.c
diff --git a/src/filelock.c b/src/filelock.c
index ff25d6475d..79eb8fa91e 100644
--- a/src/filelock.c
+++ b/src/filelock.c
@@ -403,7 +403,7 @@ create_lock_file (char *lfname, char *lock_info_str, bool force)
          lock_info_len = strlen (lock_info_str);
          err = 0;
          if (emacs_write (fd, lock_info_str, lock_info_len) != lock_info_len
-             || fchmod (fd, S_IRUSR | S_IRGRP | S_IROTH) != 0)
+             || fchmod (fd, S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH) != 0)
            err = errno;
          /* There is no need to call fsync here, as the contents of
             the lock file need not survive system crashes.  */

and Emacs is happy again, no problems with writing to files in
VirtualBox shared directory.

Please let me know if you want me to test anything else.

--
Cheers, Bernardo





  parent reply	other threads:[~2019-10-28  9:23 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-23 10:28 bug#37884: 27.0.50; Cannot write to a file in VirtualBox shared directory Bernardo
2019-10-24 16:10 ` Robert Pluim
     [not found]   ` <87blu2emi3.fsf@pobox.com>
2019-10-27 15:34     ` Robert Pluim
2019-10-27 15:52       ` Eli Zaretskii
2019-10-27 16:01         ` Robert Pluim
2019-10-27 16:34           ` Eli Zaretskii
2019-10-28  9:23           ` Bernardo [this message]
2019-10-28 12:32             ` Robert Pluim
2019-10-29  8:50               ` Bernardo
2019-10-29  9:04                 ` Andreas Schwab
2019-10-29 13:41                   ` Robert Pluim
2019-10-30  8:44                     ` Bernardo
2019-10-30 13:31                       ` Robert Pluim
2019-10-30 16:25                         ` Eli Zaretskii
2019-10-31  1:41                           ` Paul Eggert
2019-10-31  7:38                             ` Robert Pluim
2019-10-31 14:37                               ` Eli Zaretskii
2019-10-31 21:24                               ` Paul Eggert

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://www.gnu.org/software/emacs/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=874kztdy0l.fsf@pobox.com \
    --to=bernardo.bacic@pobox.com \
    --cc=37884@debbugs.gnu.org \
    --cc=eliz@gnu.org \
    --cc=rpluim@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://git.savannah.gnu.org/cgit/emacs.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).