668893 - Sort about:memory items of type COUNT alphabetically

Status: RESOLVED FIXED

(Toolkit :: about:memory, defect)

Description

[Justin Lebar (not reading bugmail)]

Split off from bug 657115 comment 38:

>     764,691 -- soft-page-faults
>       6,602 -- webgl-buffer-count
>          35 -- webgl-texture-count
>          27 -- hard-page-faults
>           4 -- webgl-context-count

This is unrelated to this bug... sorting the "counts" events by size doesn't
seem useful, alphabetical order would be better.  In which case,
hard-page-faults and soft-page-faults would be better renamed as
page-faults-hard and page-faults-soft.

Comment 1

[Nicholas Nethercote [inactive]]

I was hoping you'd file this for me :)  Thanks!

Comment 2

[Nicholas Nethercote [inactive]]

Attachment: patch

This patch changes the sort ordering and renames the page fault reporters.  Sample output:


Other Measurements

604,733,440 B -- vsize
 84,688,896 B -- resident
 42,991,616 B -- heap-committed
 39,926,938 B -- heap-used
  8,388,608 B -- js-gc-heap
  3,063,044 B -- heap-unused
    430,080 B -- heap-dirty
     48,320 B -- gfx-surface-image
           31 -- page-faults-hard
       22,109 -- page-faults-soft


Taras, will changing the name of a reporter cause problems for telemetry?

Comment 3

[Justin Lebar (not reading bugmail)]

Comment on attachment 544112 [details] [diff] [review]
patch

> // Compare two memory reporter nodes.  We want to group together measurements
> // with the same units, so sort first by the nodes' _units field, then sort by
>-// the amount if the units are equal.
>+// the amount if the units are equal.  Also, sort UNIT_COUNTS reporters in
>+// alphabetical order by path, because that gives the best looking output.

Nit: Rewrite this comment so it's consistent.  Perhaps:

 ... sort first by the nodes' _units field.  If the node has _units equal to
 UNITS_COUNT, sort second by the node's _path, ascending.  Otherwise, sort by
 the node's _amount, descending.

Comment 4

[Nicholas Nethercote [inactive]]

http://hg.mozilla.org/integration/mozilla-inbound/rev/7b51d251e2bd

Comment 5

[Marco Bonardo [:mak]]

http://hg.mozilla.org/mozilla-central/rev/7b51d251e2bd