Hi Szaka and Vincent,
Szabolcs Szakacsits wrote:
> Hello Vincent,
>
> On Thu, 23 Oct 2008, Vincent Untz wrote:
>
>
>> (I'm not subscribed to the list -- please cc me on reply)
>>
>
> Ok. Thank you for the patches.
>
>
>> I'm updating the ntfs-3g package in openSUSE, and I see that we have two
>> patches here. I have absolutely no idea if they have been sent upstream
>>
>
> Yes, Bernhard Kaindl sent them and they were/are discussed.
>
>
>> and/or if they're okay. I'd appreciate if someone could take a few
>> minutes to give an opinion about this.
>>
>> The first one is ntfs-3g-warnings.patch and should be relatively safe.
>>
There are two more warnings in mft.c about having no prototype
for two functions added lately : ntfs_mft_attr_extend() and
ntfs_mft_rec_alloc(). Apparently you do not address them,
in fact it is just a matter of declaring these functions as static.
I also have a patch for ntfs-3g in an error formatting message
("ntfs_attr_pwrite partial write" etc.) for variables to be formatted
as %lld though not cast to long long. This warning only shows on
64-bit compiles because in that situation the variables are plain
longs (I have not the exact patch at hand as my own version is
different).
Also there are a lot of warnings in fuse-lite, but I have never
looked at them.
>
> Afair, we weren't sure with some casts if it's a openSUSE pre-release gcc
> bug or not, so the issue was pending.
>
>
>> It just makes sure that ntfs-3g can be compiled "-Wformat
>> -Wformat-security -W -Wno-sign-compare -Werror". My guess is that we
>> compile it this way for extra security. Some changes, like the
>> fprintf() ones, are indeed good practice.
>>
>
> Yes, it is a good practice.
>
> However not every apple is the same. All the cases are safe (constant
> strings without any printf format conversion specifier) but gcc can't
> detect this, though it could. They basically point to a compiler bug.
> Gcc warns mindlessly when it could warn only when there is really a
> security risk (e.g. unvalidated user input).
>
> The relevant strings are format strings, and not part of a variable
> argument list. It's really very unfortunate to clutter and obfuscate the
> source code because of bogus and incorrect gcc warnings.
>
> But I'm afraid we don't have choice, we must life with it, so they will
> be applied, thanks.
>
> The last two hunks of the patch were already fixed in the CVS.
>
>
>> The second one is ntfs-3g-utf8-fallback.patch. There's a header
>> explaining the goal of the patch.
>>
>
> We would like to have built-in, default utf8 support, instead of using
> "locale=". This work was ongoing by Bernhard Kaindl (original patch) and
> recently by Jean-Pierre Andre. An improved patch is already included in
> the advanced (ownership, permission, acl support) ntfs-3g branch.
>
> Jean-Pierre, do you have feedback how well the new UTF8 converter works in
> general, and espcially with non-European languages?
>
No such feedback to date, and no known problem.
I can just tell I can now display non-latin chars.
The extension I made in Bernhard Kaindl implementation
was taking into account surrogate pairs (for rare
Unicode characters).
Regards
Jean-Pierre
-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
ntfs-3g-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ntfs-3g-devel
