Reporting VLC usage on RedHat-7.1

[Samuel Hocevar]

On Tue, Jul 24, 2001, David Madore wrote:

> I could send you a dozen crash reports, but I won't do so just now
> because I first want to know whether they're of any value - basically,
> they are so easy to reproduce that they can hardly be unkown.

   I don't think it would hurt to be reminded these bugs, even if
they are already known by most developers. There are old bugs which
keep getting fixed and reappearing and should probably be taken more
seriously.

> Similar to segfaults, I have observed many cases where the video would
> just stop playing (while the sound went on unperturbed).  I assume
> this is a case of one thread freezing up or dying mysteriously
> (another thing which could be either due to the libpthread bugs or to
> the program itself).  Basically, whenever I try to seek to a specific
> location on the DVD by moving the position cursor, I have problems
> like this.  Once the video freezes (the same image stays on display),
> it never seems to start again, whatever I try.

   Could you confirm this still happens with the current CVS version ?
Christophe's video decoder fixes and this night's libdvdcss fix improved
things a lot here.

> i.e. asm statements with no output operands and no "volatile" keyword.
> Well the compiler would be quite in its rights to simply optimize them
> away, wouldn't it?  After all, they are officially supposed to do
> nothing at all.  Surely there's something wrong, here.

   Adding volatile to the statement did not help here, but I'll have
a look at the generated assembly to see what we may be doing wrong. I
don't really know how inline asm works with gcc though, so if anyone
more knowledgeable has any clue, feel free to share ideas!

Regards,
-- 
Sam.