You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The memory track in the profiler shows memory allocated with malloc() and free(). However bookkeeping, fragmentation and a cache of "dirty memory" all mean that the actual amount of physical memory allocated to firefox is larger, and when freeing memory especially since Bug 1903758 this can "lag behind" the normal memory counter. mozjemalloc calls this "committed memory" this isn't an accurate name but there isn't a precise term.
The committed memory will always be larger than the memory that firefox has allocated. I want to enable it selectively because I think most engineers won't find it useful. What we expect to see is that when Firefox frees memory, first the existing counter will drop (as it does now) and then a second or so later when Firefox returns the memory to the OS so other programs can use it we should see the 2nd counter drop more sharply. I imagined rendering them both on the same plot with the existing graph "in front of" the new one - but if you have other ideas I'd be glad to hear them.
The memory track in the profiler shows memory allocated with malloc() and free(). However bookkeeping, fragmentation and a cache of "dirty memory" all mean that the actual amount of physical memory allocated to firefox is larger, and when freeing memory especially since Bug 1903758 this can "lag behind" the normal memory counter. mozjemalloc calls this "committed memory" this isn't an accurate name but there isn't a precise term.
The committed memory will always be larger than the memory that firefox has allocated. I want to enable it selectively because I think most engineers won't find it useful. What we expect to see is that when Firefox frees memory, first the existing counter will drop (as it does now) and then a second or so later when Firefox returns the memory to the OS so other programs can use it we should see the 2nd counter drop more sharply. I imagined rendering them both on the same plot with the existing graph "in front of" the new one - but if you have other ideas I'd be glad to hear them.
┆Issue is synchronized with this Jira Task
The text was updated successfully, but these errors were encountered: