<html> <head> <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1"> <style type="text/css" style="display:none;"> P {margin-top:0;margin-bottom:0;} </style> </head> <body dir="ltr"> <div style="font-family: Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> Aha! That's the problem: I call notify_*() from my own thread which reads a socket fd from a remote server. <span style="color: rgb(0, 0, 0); font-family: Calibri, Helvetica, sans-serif; font-size: 12pt;">Sounds like the correct way to do this is use</span><span style="color: rgb(0, 0, 0); font-family: Calibri, Helvetica, sans-serif; font-size: 12pt;"> </span><span style="color: rgb(0, 0, 0); font-family: Calibri, Helvetica, sans-serif; font-size: 12pt;">wl_event_loop_add_fd() to register the socket, and read from the fd in the callback?</span></div> <div style="font-family: Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> <span style="color: rgb(0, 0, 0); font-family: Calibri, Helvetica, sans-serif; font-size: 12pt;"><br> </span></div> <div style="font-family: Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> <span style="color: rgb(0, 0, 0); font-family: Calibri, Helvetica, sans-serif; font-size: 12pt;">I have some non-input messages that need to remain async from weston's event loop (ex. some setup during compositor start), so I did a test by keeping the <span style="font-family: Calibri, Helvetica, sans-serif; background-color: rgb(255, 255, 255); display: inline !important">notify_*() event calls in my thread and then "dispatch" the event loop by sending a signal over a pipe which I registered with <span style="font-family: Calibri, Helvetica, sans-serif; background-color: rgb(255, 255, 255); display: inline !important">wl_event_loop_add_fd(). This appears to be working! I just want to make sure there's no obvious problems/race-conditions that come to mind with calling <span style="font-family: Calibri, Helvetica, sans-serif; background-color: rgb(255, 255, 255); display: inline !important">notify_*() asynchronously?</span></span></span></span></div> <div style="font-family: Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> <span style="color: rgb(0, 0, 0); font-family: Calibri, Helvetica, sans-serif; font-size: 12pt;"><span style="font-family: Calibri, Helvetica, sans-serif; background-color: rgb(255, 255, 255); display: inline !important"><span style="font-family: Calibri, Helvetica, sans-serif; background-color: rgb(255, 255, 255); display: inline !important"><span style="font-family: Calibri, Helvetica, sans-serif; background-color: rgb(255, 255, 255); display: inline !important"><br> </span></span></span></span></div> <div style="font-family: Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> <span style="color: rgb(0, 0, 0); font-family: Calibri, Helvetica, sans-serif; font-size: 12pt;"><span style="font-family: Calibri, Helvetica, sans-serif; background-color: rgb(255, 255, 255); display: inline !important"><span style="font-family: Calibri, Helvetica, sans-serif; background-color: rgb(255, 255, 255); display: inline !important">And thanks for all the help!! </span></span></span></div> <div style="font-family: Calibri, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);"> <br> </div> <div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0);"> <hr tabindex="-1" style="display:inline-block; width:98%;"> <b>From:</b> Pekka Paalanen<br> <b>Sent:</b> Thursday, February 13, 2020 4:33 AM<br> <b>To:</b> Josh Simonot<br> <b>Cc:</b> wayland-devel@lists.freedesktop.org<br> <b>Subject:</b> Re: repaint after notify_button() ? <div><br> </div> </div> <div class="BodyFragment"><font size="2"><span style="font-size:11pt;"> <div class="PlainText">On Wed, 12 Feb 2020 17:22:05 +0000<br> Josh Simonot <jsimonot@live.ca> wrote:<br> <br> > calling wl_event_loop_dispatch() is not a working solution: keeps<br> > tripping over a few different assertions causing crash, and doesn't<br> > seem to dispatch key events.<br> > <br> > I've seen:<br> > - page_flip_handler: Assertion `output->page_flip_pending' failed.<br> > - weston_output_finish_frame: Assertion `output->repaint_status ==<br> > REPAINT_AWAITING_COMPLETION' failed.<br> <br> Hi,<br> <br> you seem to be confusing some libwayland-server and libwayland-client<br> APIs (wl_display_dispatch_pending()). Both unfortunately have a 'struct<br> wl_display' type, but the two types are distinct and totally<br> incompatible. So make sure to use only server API with the server-side<br> library and objects.<br> <br> There should be no need to force-wake the event loop at all. It wakes<br> up automatically from epoll_wait() when any file descriptor (including<br> client sockets, timers and POSIX signals) registered with it is<br> readable (or sometimes writable, which ever is appropriate).<br> <br> Before sleeping in epoll, it also flushes out all pending messages out<br> to clients.<br> <br> Dispatching an event loop from random places at random times is a bad<br> idea. Especially if it happens during dispatch, which may well mess<br> things up: bad message ordering and maybe memory corruption. This means<br> there is really no place in Weston where you could actually<br> force-dispatch anything, everything already happens inside dispatch<br> calls except during compositor start and shutdown.<br> <br> You say the notify_*() functions are called in time. Next thing is to<br> check if the related events are sent out in time, and that you can<br> check with WAYLAND_DEBUG=server or using 'weston-debug proto' (needs<br> to start weston with --debug). If those are delayed, then we are sure<br> the issue is in the compositor. If those are in time, then the issue<br> cannot be in the compositor.<br> <br> >From what call chain to you call the notify_*() functions? An fd event<br> source of your own? Idle source? Timer source? Or simply from some<br> arbitrary function already in the DRM-backend?<br> <br> <br> Thanks,<br> pq<br> </div> </span></font></div> </body> </html>