Closed Bug 268798 Opened 20 years ago Closed 16 years ago

not flushing session cache during a memory-pressure notification

Categories

(Core :: Graphics: ImageLib, defect)

All
Windows XP
defect
Not set
normal

Tracking

()

RESOLVED FIXED

People

(Reporter: dougt, Assigned: dougt)

Details

Attachments

(1 file, 2 obsolete files)

For whatever reason, we only purge the chrome cache.
Attached patch patch v.1 (obsolete) — Splinter Review
calls ClearCache for the session cache as well.
Assignee: pavlov → dougt
Status: NEW → ASSIGNED
Attachment #165384 - Flags: review?(pavlov)
Comment on attachment 165384 [details] [diff] [review]
patch v.1

It seems like a reasonable thing to do when the memory-pressure one is called,
but not when the chrome ones are called.
Attachment #165384 - Flags: review?(pavlov) → review-
Attached patch patch v.2 (obsolete) — Splinter Review
Attachment #165384 - Attachment is obsolete: true
Attachment #165397 - Flags: review?(pavlov)
Comment on attachment 165397 [details] [diff] [review]
patch v.2

make that an 'else if'
Attachment #165397 - Flags: review?(pavlov) → review-
Attached patch v.3Splinter Review
Attachment #165397 - Attachment is obsolete: true
Attachment #165398 - Flags: review?(pavlov)
Attachment #165398 - Flags: review?(pavlov) → review+
Comment on attachment 165398 [details] [diff] [review]
v.3

david, can you sr?
Attachment #165398 - Flags: superreview?(dbaron)
Comment on attachment 165398 [details] [diff] [review]
v.3

sr=dbaron, if you're sure you don't want to clear both
Attachment #165398 - Flags: superreview?(dbaron) → superreview+
Comment on attachment 165398 [details] [diff] [review]
v.3

i am not sure we want to flush chrome since, in the case of Minimo, these files
are likely to be required very soon after we flush.
Attachment #165398 - Flags: approval1.8b4?
Attachment #165398 - Flags: approval1.8b4? → approval1.8b4+
Comment on attachment 165398 [details] [diff] [review]
v.3

Checking in imgCache.cpp;
/cvsroot/mozilla/modules/libpr0n/src/imgCache.cpp,v  <--  imgCache.cpp
new revision: 1.18; previous revision: 1.17
done
marking fixed.
Status: ASSIGNED → RESOLVED
Closed: 19 years ago
Resolution: --- → FIXED
Could this have caused an increase in Tp for all Win building machines ?

After this bug was checked in

Pacifica/Beast and Creature Tp rose just over 10%

For Beast, this patch was the only one that landed

before, consistent lower 180's
13535 2005:07:26:07:51:33 182
13536 2005:07:26:08:18:00 176 
13537 2005:07:26:08:44:42 201 <----
13538 2005:07:26:09:12:57 201
after, consistent lower 200's

the increase does not show on Mac/Linux machines

(if needed I can file a new bug)
is HeapMinimize really being called on windows?
(In reply to comment #12)
> is HeapMinimize really being called on windows?
I hope you aren't asking me, I wouldn't have a clue

I filed bug 305356 for the Tp regression.
this caused regression: 305356.
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
mass reassigning to nobody.
Assignee: dougt → nobody
Status: REOPENED → NEW
Doug, does more need to be done on this bug or can it be marked FIXED?
http://mxr.mozilla.org/seamonkey/source/modules/libpr0n/src/imgCache.cpp#349

Looks like it does the right thing.
Status: NEW → RESOLVED
Closed: 19 years ago16 years ago
Resolution: --- → FIXED
Assignee: nobody → doug.turner
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: