[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: fatfs errors
From: |
M. Gerards |
Subject: |
Re: fatfs errors |
Date: |
Thu, 12 Dec 2002 12:28:49 +0100 |
User-agent: |
Internet Messaging Program (IMP) 3.1 |
> inode.c:60: conflicting types for
> 'diskfs_cached_lookup'
> :/include/hurd/diskfs.h:738:previous declaration of
> 'diskfs_cached_lookup'
>
> inode.c:in function 'read_node':
> inode.c:267:warning:'buflen' might be used
> uninitialized in this function.
> whats the problem?......
fatfs is now a part of the Hurd, could you please try the CVS version?
I think you aren't using a Hurd build with 64 bits interfaces. You should
upgrade the Hurd in that case or change all ino64_t to ino_t in the fatfs
sources.
There are still some problems with fatfs:
- Several deadlocks, two of them in diskfs_cached_lookup.
(When I was debugging fatfs yesterday it didn't deadlock... Are recursive
locks supported ATM?)
- A problem in diskfs_get_dirents. It crashed after returning. I attached GDB
to fatfs and wanted to get a backtrace (bt), but that wasn't possible (Couldn't
read data from address foo was the error IIRC). Can someone please help me with
this? This problem exists since the change to the 64 bits interfaces...
- Lack of write support, although I've implemented most stuff in my tree at
home. It will take some time because rename is difficult to implement.
Thanks you for testing fatfs! :)
-------------
Marco Gerards
- fatfs errors, Luis Araujo, 2002/12/11
- Re: fatfs errors,
M. Gerards <=