Please take a look at the network charts in this profile: http://bit.ly/2Wfc5nb , or in this profile: http://bit.ly/2WaMOL0 These are profiles from bug 1522311. It seems to me that what's happening here is that the content process knows about network loads it wants to do many seconds before the parent process knows about them. And then the actual network loads start extremely late, long after they could have started. In fact, the network is idle during most of the time when the user is looking at a blank screen. What causes these network loads to start in the end? It looks like it's not happening until the *content process main thread* becomes available. Do these loads require a jump through the content process event loop before they get forwarded to the parent process?
Bug 1522316 Comment 0 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
Please take a look at the network charts in this profile: http://bit.ly/2Wd0Hs5 , or in this profile: http://bit.ly/2WaMOL0 These are profiles from bug 1522311. It seems to me that what's happening here is that the content process knows about network loads it wants to do many seconds before the parent process knows about them. And then the actual network loads start extremely late, long after they could have started. In fact, the network is idle during most of the time when the user is looking at a blank screen. What causes these network loads to start in the end? It looks like it's not happening until the *content process main thread* becomes available. Do these loads require a jump through the content process event loop before they get forwarded to the parent process?
Please take a look at the network charts in this profile: http://bit.ly/2Wd0Hs5, or in this profile: http://bit.ly/2WaMOL0 These are profiles from bug 1522311. It seems to me that what's happening here is that the content process knows about network loads it wants to do many seconds before the parent process knows about them. And then the actual network loads start extremely late, long after they could have started. In fact, the network is idle during most of the time when the user is looking at a blank screen. What causes these network loads to start in the end? It looks like it's not happening until the *content process main thread* becomes available. Do these loads require a jump through the content process event loop before they get forwarded to the parent process?
Please take a look at the network charts in this profile: http://bit.ly/2WgurnM, or in this profile: http://bit.ly/2WaMOL0 These are profiles from bug 1522311. It seems to me that what's happening here is that the content process knows about network loads it wants to do many seconds before the parent process knows about them. And then the actual network loads start extremely late, long after they could have started. In fact, the network is idle during most of the time when the user is looking at a blank screen. What causes these network loads to start in the end? It looks like it's not happening until the *content process main thread* becomes available. Do these loads require a jump through the content process event loop before they get forwarded to the parent process?