RedHat/Fedora patches (Was: Re: [mtools] Mtools 3.9.10 for VMS)

[Moritz Barsnick]

Hi,

Alain wrote:
> Moritz Barsnick wrote:
> > BTW, Fedora core 5 (and other Linux distributions?) has a bunch of
> > patches in the mtools code. Are they worth a look?
> > mtools-3.9.6-config.patch
> > mtools-3.9.6-paths.patch
> > mtools-3.9.7-bigdisk.patch
> > mtools-3.9.9-badc.patch
> > mtools-3.9.9-noargs.patch
>
> Judging from the version numbers, they look somewhat obsolete...

Well, they still apply them against the current version. They just
haven't re-built the patches.

a) mtools-3.9.6-config.patch
  - uncomments i.e. activates the floppy stuff in mtools.conf (a: and b:)
  - adds the flags "1.44m mformat_only" to a: and b:
-> Distribution specific, irrelevant.

b) mtools-3.9.6-paths.patch
  - changes all mentions of /usr/local/etc to /etc in mtools.5 and
    mtools.texi
-> This is basically distribution specific clarification, as those are
   originally explicitly examples according to the wording. 
   Nevertheless, one _could_ create mtools.texi from mtools.texi.in and
   have the paths fixed by configure...but you'd have to re-create the
   man pages from texinfo then as well. Not really relevant.

c) mtools-3.9.7-bigdisk.patch
  - changes the flags in mtools.conf back from
    "exclusive 1.44m mformat_only" to "exclusive mformat_only"
-> Distribution specific, actually reverts a distribution specific
   patch, d'uh! Irrelevant.

d) mtools-3.9.9-badc.patch
  - changes a "&" to "&&" in mainloop.c, IMO a correct fix (though
    _probably_ doesn't have any effect)
  - fixes a typo in usage message in mpartition.c
-> Can and should be used. Attached here.

e) mtools-3.9.9-noargs.patch
  - fixes case of missing arguments to scripts/mcomp, which is probably
    fair. IMO there could be more fixes in that script (like "" around
    vars).
-> Should be used. Attached here.

> > mtools-gcc4.patch
> > mtools-memset.patch
>
> Maybe these could still be relevant?

f) mtools-gcc4.patch
  - changes two syntaxes such as (stream)->Class->read()
                              to ((stream)->Class->read)()
    I think this fixes this bug report:
    https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=151135
    ("error: macro "read" passed 4 arguments, but takes just 3")
    Might only be relevant if read is a macro somewhere somehow?
-> I don't know gcc4, this patch doesn't change anything for me.
   Perhaps it's obsoleted with stable gcc4, but it doesn't hurt,
   does it? Attached here.

g) mtools-memset.patch
  - fixes two (very) wrong memset() calls in fat.c; look like valid
    fixes indeed!
-> Should be used. Attached here.


I recommend that you check these attached patches and consider including
them. From what I can see, these four look valid. (They all apply fine
against your most recent 20060531 patch.)

I will try to follow up with my personal patches when I get around to
it. I have to check whether there was anything relevant there...

Thanks for listening!
Kind regards,
Moritz

mtools-3.9.9-badc.patch
Description: Text document

mtools-3.9.9-noargs.patch
Description: Text document

mtools-gcc4.patch
Description: Text document

mtools-memset.patch
Description: Text document

_______________________________________________
mtools mailing list
address@hidden
http://www.tux.org/mailman/listinfo/mtools