~alan-griffiths/mir/fix-1716436 : changes

~alan-griffiths/mir/fix-1716436 » Changes from revision 4239

From Revision 4219 to 4200

Rev	Summary	Authors	Tags	Date
4219	Update licences to (L)GPL3 or (L)GPL2. Update licences to (L)GPL3 or (L)GPL2. Approved by mir-ci-bot, Chris Halse Rogers.	Alan Griffiths		6 years ago
4218	This is an ugly sticking plaster over a CI problem: valgrind... This is an ugly sticking plaster over a CI problem: valgrind complains about uuid_generate() on artful amd64	Alan Griffiths		6 years ago
4217	[mirtest] Remove spurious link library. [mirtest] Remove spurious link library. Approved by mir-ci-bot.	Alan Griffiths		6 years ago
4216	[mesa-kms] Respect display orientation when painting cursor.... [mesa-kms] Respect display orientation when painting cursor. (LP: #1610078). Fixes: https://bugs.launchpad.net/bugs/1610078, https://bugs.launchpad.net/bugs/1706718. Approved by mir-ci-bot, Gerry Boland.	Alan Griffiths		6 years ago
4215	mcl::BufferVault: Fix lock inversion. mcl::BufferVault: Fix lock inversion. ~BufferVault takes BufferVault::mutex and then calls AtomicCallback::set_callback() on each buffer, which takes AtomicCallback::mutex. The RPC subsystem calls buffer->callback(), which takes AtomicCallback::mutex and then calls BufferVault::wire_transfer_inbound... which immediately takes BufferVault::mutex. This is clearly a disaster waiting to happen. Instead, move to a two-phase destruction approach: 1) Lock BufferVault::mutex, set a flag telling callbacks to bail, and collect all the existing buffers. 2) Drop the lock, and use buffer->set_callback() to wait on any thread already in a callback. 3) Don't bother telling the server to release any buffers; they're associated with the BufferStream that we're in the process of destroying, and will get cleaned up there. Approved by mir-ci-bot, Alan Griffiths.	Christopher James Ha...		6 years ago
4214	[mir_demo_client_display_config] add orientation changing . [mir_demo_client_display_config] add orientation changing . Approved by Chris Halse Rogers, mir-ci-bot.	Alan Griffiths		6 years ago
4213	Handle mir_event_type_close_window in examples (LP: #1706004... Handle mir_event_type_close_window in examples (LP: #1706004). Fixes: https://bugs.launchpad.net/bugs/1706004. Approved by Gerry Boland, mir-ci-bot.	Alan Griffiths		6 years ago
4212	Drop BufferStream::suitable_for_cursor Drop BufferStream::suitable_for_cursor This is a per-buffer, not per-bufferstream, property. Approved by mir-ci-bot, Alan Griffiths, Brandon Schaefer.	Christopher James Ha...		6 years ago
4211	[mir_demo_client_target] exit on receipt of mir_event_type_c... [mir_demo_client_target] exit on receipt of mir_event_type_close_window. (LP: #1705439). Fixes: https://bugs.launchpad.net/bugs/1705439. Approved by Gerry Boland, mir-ci-bot.	Alan Griffiths		6 years ago
4210	Tidy up some code touched while investigating lp:1705417. Tidy up some code touched while investigating lp:1705417. Approved by Brandon Schaefer, mir-ci-bot.	Alan Griffiths		6 years ago
4209	Respect cursor hotspot when hosted on Mir. (LP: #1705284). F... Respect cursor hotspot when hosted on Mir. (LP: #1705284). Fixes: https://bugs.launchpad.net/bugs/1705284. Approved by Brandon Schaefer, mir-ci-bot.	Alan Griffiths		6 years ago
4208	Move full responsibility for buffer IPC into the frontend. Move full responsibility for buffer IPC into the frontend. Currently buffer IPC is split between the frontend and various things which shared an mf::BufferMap. This branch consolidates everything so that, outside the SessionMediator, everything is dealing with a std::shared_ptr<mg::Buffer>. In addition to consolidating IPC into the frontend where it belongs, this makes it easy to submit client-allocated buffers. Approved by mir-ci-bot, Alan Griffiths.	Christopher James Ha...		6 years ago
4207	RPC: Don't require the server ACK client's buffer-release re... RPC: Don't require the server ACK client's buffer-release requests. There's no particularly good reason to require the client to wait for server acknowledgement before freeing the buffer that it wants to release, so don't. Approved by mir-ci-bot, Alan Griffiths.	Christopher James Ha...		6 years ago
4206	Only notify resize events for valid surfaces. Only notify resize events for valid surfaces. Approved by Chris Halse Rogers, mir-ci-bot.	Alan Griffiths		6 years ago
4205	Remove obsolete & broken example code. Fixes: https://bugs.l... Remove obsolete & broken example code. Fixes: https://bugs.launchpad.net/bugs/1663130. Approved by Chris Halse Rogers, mir-ci-bot.	Alan Griffiths		6 years ago
4204	Don't leak DRM fds in platform-eglstream probe. Don't leak DRM fds in platform-eglstream probe. Approved by mir-ci-bot, Gerry Boland.	Christopher James Ha...		6 years ago
4203	The protocol version needs to change we dropped message Rais... The protocol version needs to change we dropped message RaiseRequest (and subsumed it into RequestOperation). Approved by mir-ci-bot, Gerry Boland.	Alan Griffiths	v0.27.0	6 years ago
4202	Handle initial window size being imposed by shell in eglapps... Handle initial window size being imposed by shell in eglapps. Fixes: https://bugs.launchpad.net/bugs/1701308. Approved by Gerry Boland, mir-ci-bot.	Alan Griffiths		6 years ago
4201	Fix nested passthrough to respect the size of renderables. F... Fix nested passthrough to respect the size of renderables. Fixes: https://bugs.launchpad.net/bugs/1700981. Approved by Brandon Schaefer, mir-ci-bot.	Alan Griffiths		6 years ago
4200	Add missing GBM initialization for Mir on Mir on Mesa that g... Add missing GBM initialization for Mir on Mir on Mesa that got lost when we split out rendering platforms. Fixes: https://bugs.launchpad.net/bugs/1699484. Approved by Brandon Schaefer, Gerry Boland, mir-ci-bot.	Alan Griffiths		6 years ago