From: Will Page <will.page@ni.com>
To: <yocto@yoctoproject.org>
Cc: Seebs <seebs@seebs.net>
Subject: [pseudo][PATCH] Fix to fcntl guts to ignore flags that can be ORed into cmd
Date: Fri, 15 Sep 2017 19:33:53 -0700 [thread overview]
Message-ID: <20170916023353.GA16344@canismajor> (raw)
[-- Attachment #1: Type: text/plain, Size: 93 bytes --]
I previously submitted this patch to the OE-core mailing list, and was
directed over here.
[-- Attachment #2: 0001-Fix-to-fcntl-guts-to-ignore-flags-that-can-be-ORed-i.patch --]
[-- Type: text/x-diff, Size: 2094 bytes --]
From c29b7aecfe671013897119b1dee2731900439285 Mon Sep 17 00:00:00 2001
From: Will Page <Will.Page@ni.com>
Date: Fri, 15 Sep 2017 13:07:44 -0700
Subject: [pseudo][PATCH] Fix to fcntl guts to ignore flags that can be ORed
into cmd
The fcntl guts switch on "cmd" parameter to identify the fcntl
command being issued, but isn't aware of the file creation flags that
can be ORed in.
This change masks out the flags from the command only for the switch
statement so that it can correctly determine whether it needs to pass
"lock" or "arg". When real_fcntl() is called, the original value is
still passed on. This corrects an issue observed using pseudo on modern
linux desktops resulting in "pseudo: unknown fcntl argument 1030,
assuming long argument." diagnostics in the output. Decimal 1030 is
0x0406, which translates to O_NOCTTY | F_SETLK, and should call
"rc = real_fcntl(fd, cmd, lock);", but instead falls through to the
default case instead calling "rc = real_fcntl(fd, cmd, arg);".
Tested the change using perftest - without this patch, the issue is
trivially reproducible on my Ubuntu Xenial system, and after patching it
emitted no "unknown fcntl argument" diagnostics.
Signed-off-by: Will Page <Will.Page@ni.com>
---
ports/linux/guts/fcntl.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/ports/linux/guts/fcntl.c b/ports/linux/guts/fcntl.c
index 639fd24..d278a8c 100644
--- a/ports/linux/guts/fcntl.c
+++ b/ports/linux/guts/fcntl.c
@@ -8,6 +8,10 @@
*/
long arg;
int save_errno;
+ /* some bits can be ORed into the cmd should be explicitly ignored
+ * see fcntl documentation on F_SETFL */
+ int o_mode_mask = (O_RDONLY | O_WRONLY | O_RDWR) |
+ (O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC);
/* we don't know whether we need lock or arg; grab both, which
* should be safe enough on Linuxy systems. */
@@ -15,7 +19,7 @@
arg = va_arg(ap, long);
va_end(ap);
- switch (cmd) {
+ switch (cmd & ~(o_mode_mask)) {
case F_DUPFD:
#ifdef F_DUPFD_CLOEXEC
case F_DUPFD_CLOEXEC:
--
2.7.4
next reply other threads:[~2017-09-16 3:23 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-09-16 2:33 Will Page [this message]
-- strict thread matches above, loose matches on Subject: below --
2017-09-15 22:27 [pseudo][PATCH] Fix to fcntl guts to ignore flags that can be ORed into cmd Will Page
2017-09-15 22:48 ` Burton, Ross
2017-09-16 2:22 ` Will Page
2017-09-16 17:59 ` Seebs
2017-09-15 23:10 ` Seebs
2017-09-18 13:24 ` Richard Purdie
2017-09-18 15:15 ` Will Page
2017-09-18 15:39 ` Will Page
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
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170916023353.GA16344@canismajor \
--to=will.page@ni.com \
--cc=seebs@seebs.net \
--cc=yocto@yoctoproject.org \
/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.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.