This is the mail archive of the cygwin mailing list for the Cygwin project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Intermittent perl crash (Attn: coreutils and bash maintainer)

On Tue, 18 Dec 2007, Michael Kairys wrote:

> "Larry Hall (Cygwin)" <reply-to-list-only-lh@XXXXXX.XXX> wrote in message

<>.  Thanks.

> > <>
> Thanks... I read with partial understanding but could not come up with any
> action items...
> My resolution was to verify that my temp is in fact writeable and then
> comment out the check in rebaseall.
> So far so good: no fork failures.

I'm guessing one can reconstruct your permissions as follows:

$ cd /tmp
$ mkdir nw
$ setfacl -m g:Users:rwx nw
$ chmod a-rwx nw
$ ls -ld nw
d---------+ 2 igor root 0 Dec 18 10:08 nw/
$ touch nw/foo
$ ls -l nw/foo
-rw------- 1 igor root 0 Dec 18 10:08 nw/foo
$ if [ -w nw ]; then echo "Yes"; else echo "No"; fi

One solution is, of course, "chmod u+rwx /tmp".  The other is to change
rebaseall to try creating a file anyway and checking the result.

There was some talk of changing coreutils to use "access" for testing
permissions, which would probably solve this as well.  However, even if I
use "/bin/test -w nw", I get a "No".  Has this change ever happened?  If
it ever does, bash will also need to be rebuilt so that its builtins use
the same interface.  It helps that both have the same maintainer.
      |\      _,,,---,,_ |
ZZZzz /,`.-'`'    -.  ;-;;,_		Igor Peshansky, Ph.D. (name changed!)
     |,4-  ) )-,_. ,\ (  `'-'		old name: Igor Pechtchanski
    '---''(_/--'  `-'\_) fL	a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

"That which is hateful to you, do not do to your neighbor.  That is the whole
Torah; the rest is commentary.  Go and study it." -- Rabbi Hillel

Unsubscribe info:
Problem reports:

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]