>-Original Message-
>From: Joe Orton [mailto:[EMAIL PROTECTED]
>Sent: Friday, April 30, 2004 3:08 AM
>
>Looks fine, though it doesn't handle errors in apr_atoi64
>itself so it would be good to surround the calls with a single
>errno = 0 / ... /
>if (errno) return -1; check too.
>
It s
On Thu, Apr 29, 2004 at 09:43:34AM -0700, Mathihalli, Madhusudan wrote:
> >-Original Message-
> >From: Joe Orton [mailto:[EMAIL PROTECTED]
> [SNIP]
> >
> >> What if the user really sent a
> >> large value for a small file ? Instead of erroring out -
> >thanks to the
> >> overflow mechanism
>-Original Message-
>From: Joe Orton [mailto:[EMAIL PROTECTED]
[SNIP]
>
>> What if the user really sent a
>> large value for a small file ? Instead of erroring out -
>thanks to the
>> overflow mechanism, we'll probably end up serving a sane result -
>> Should we leave it that way ?
>
>Oh,
On Thu, Apr 29, 2004 at 09:00:24AM -0700, Mathihalli, Madhusudan wrote:
>
> >-Original Message-
> >From: Joe Orton [mailto:[EMAIL PROTECTED]
> [SNIP]
>
> >apr_off_t is the right type to use since these are file offsets.
> >parse_byterange should probably check for integer overflow when
>
>-Original Message-
>From: Joe Orton [mailto:[EMAIL PROTECTED]
[SNIP]
>apr_off_t is the right type to use since these are file offsets.
>parse_byterange should probably check for integer overflow when
>sizeof(apr_off_t) != sizeof(apr_int64_t), but if you have a >2gb file
>and a 32-bit apr
On Wed, Apr 28, 2004 at 05:03:14PM -0700, Mathihalli, Madhusudan wrote:
> Hello,
> On my HP-UX 11i box (64-bit os), if I build a 32-bit app
> (default), the apr_off_t is a 4-byte entity and apr_int64_t is a
> 8-byte entity. I'm sure more than one person has experienced
> som
