Re: [PATCH] Destroy resources when destroying input and output

2014-05-14 Thread Pekka Paalanen
On Wed, 14 May 2014 19:12:20 +0100 Neil Roberts wrote: > After looking at it a bit more I don't think we can really make > the automatic zombie idea work. The trouble is that > libwayland-server would need to know when the client has > destroyed the proxy in order to destroy the zombie resource.

Re: [PATCH] Destroy resources when destroying input and output

2014-05-14 Thread Neil Roberts
After looking at it a bit more I don't think we can really make the automatic zombie idea work. The trouble is that libwayland-server would need to know when the client has destroyed the proxy in order to destroy the zombie resource. However the destroy semantics are interface-dependent so only an

Re: [PATCH] Destroy resources when destroying input and output

2014-05-09 Thread Neil Roberts
Jason Ekstrand writes: > Most of the magic there is in allowing resources with no handler in > libwayland-server. The patch would be about 4 lines. Right now, > client-side wl_proxy objects are allowed to have a NULL implementation > and there's no problem; server-side, this is not currently allo

Re: [PATCH] Destroy resources when destroying input and output

2014-05-09 Thread Jason Ekstrand
On Fri, May 9, 2014 at 1:37 AM, Pekka Paalanen wrote: > Looking at it in general, there is one more fun complication. > > If the inert object has requests in its interface, that create new > objects, the server cannot just ignore those requests. I think the > server will need to actually create ne

Re: [PATCH] Destroy resources when destroying input and output

2014-05-09 Thread Pekka Paalanen
On Fri, 09 May 2014 14:33:58 +0100 Neil Roberts wrote: > Perhaps we should consider applying the patch anyway even though it's > not ideal. Currently if a client uses a dead output in a request such as > xdg_surface.set_output Weston will end up with a weston_output pointer > that points to freed

Re: [PATCH] Destroy resources when destroying input and output

2014-05-09 Thread Neil Roberts
Perhaps we should consider applying the patch anyway even though it's not ideal. Currently if a client uses a dead output in a request such as xdg_surface.set_output Weston will end up with a weston_output pointer that points to freed memory. This could cause the compositor to crash. That is worse

Re: [PATCH] Destroy resources when destroying input and output

2014-05-08 Thread Pekka Paalanen
On Thu, 8 May 2014 19:32:12 -0500 Jason Ekstrand wrote: > On Tue, Dec 31, 2013 at 3:37 AM, Pekka Paalanen wrote: > > > You cannot just go and destroy wl_resources, because there are > > clients using them - that is why the wl_resources exist in the > > first place. Clients use all protocol obje

Re: [PATCH] Destroy resources when destroying input and output

2014-05-08 Thread Jason Ekstrand
On Tue, Dec 31, 2013 at 3:37 AM, Pekka Paalanen wrote: > On Wed, 25 Dec 2013 17:02:09 +0100 > Mariusz Ceier wrote: > > > Some structures containing resources list are freed before > > resources on that list are destroyed, and that triggers invalid > > read/write in compositor.c:3103 indirectly c

Re: [PATCH] Destroy resources when destroying input and output

2013-12-31 Thread Pekka Paalanen
On Wed, 25 Dec 2013 17:02:09 +0100 Mariusz Ceier wrote: > Some structures containing resources list are freed before > resources on that list are destroyed, and that triggers invalid > read/write in compositor.c:3103 indirectly called from > text-backend.c:938 when running weston under valgrind.

[PATCH] Destroy resources when destroying input and output

2013-12-25 Thread Mariusz Ceier
Some structures containing resources list are freed before resources on that list are destroyed, and that triggers invalid read/write in compositor.c:3103 indirectly called from text-backend.c:938 when running weston under valgrind. This patch destroys resources on these lists before such structur