Closed Bug 1504495 Opened 7 years ago Closed 7 years ago

4 seconds of main thread work with SQLITE doing NSS stuff, on starting the profiler, and manually restoring previous session

Categories

(Core :: Security: PSM, defect)

defect
Not set
normal

Tracking

()

RESOLVED DUPLICATE of bug 1435141

People

(Reporter: mayankleoboy1, Unassigned)

Details

Attachments

(1 file)

Attached file aboutsupport.txt
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:65.0) Gecko/20100101 Firefox/65.0 Steps to reproduce: 1.Using my normal profile. "Restore previous session on startup" is disabled in preferences 2.Firefox profiler installed 3. Opened a tab to https://en.miui.com/forum-207-1.html 4. Close and restart firefox. Since previous session is not restored, a blank tab is opened. 5. Start the profiler 6. Go to the hamburger menu, and click on "Restore Previous Session" Actual results: main thread jank of 4 seconds. And i use a SSD Expected results: maybe not so? https://perfht.ml/2JCLpXF
Component: General → Places
Product: Core → Toolkit
tentative ni? from :mak, because of sqlite
Flags: needinfo?(mak77)
Summary: 4 seconds of main thread work with SQLITE on restoring previous session after startup → 4 seconds of main thread work with SQLITE on manually restoring previous session after startup
Please can you explain why you think this is places (bookmarks/history) sqlite? I could be reading it wrong, but from what I'm seeing, there's a lot of sqlite calls descending from CERT_FindUserCertsByUsage/CERT_GetCertNicknames which makes it look more like an NSS issue to me.
Flags: needinfo?(mayankleoboy1)
FWIW I just tried this on a fresh profile on the latest nightly and couldn't reproduce.
(In reply to Mark Banner (:standard8) from comment #2) > Please can you explain why you think this is places (bookmarks/history) > sqlite? > > I could be reading it wrong, but from what I'm seeing, there's a lot of > sqlite calls descending from CERT_FindUserCertsByUsage/CERT_GetCertNicknames > which makes it look more like an NSS issue to me. you are right, this does appear to be a NSS related. I didn't go deeper into the profile, and assumed sqlite=places. I will change the components accordingly.
Component: Places → Security: PSM
Flags: needinfo?(mayankleoboy1)
Flags: needinfo?(mak77)
Product: Toolkit → Core
:keeler, is this a dupe of bug 1370516 ?
Flags: needinfo?(dkeeler)
Summary: 4 seconds of main thread work with SQLITE on manually restoring previous session after startup → 4 seconds of main thread work with SQLITE doing NSS stuff, on manually restoring previous session after startup
The main-thread work might be exaggerated by enabling the profiler?
Summary: 4 seconds of main thread work with SQLITE doing NSS stuff, on manually restoring previous session after startup → 4 seconds of main thread work with SQLITE doing NSS stuff, on starting the profiler, and manually restoring previous session
This is not bug 1370516. This sounds like a combination of bug 1435141 and bug 1478148. The optimization suggested in bug 1503711 might also help.
Flags: needinfo?(dkeeler)
Status: NEW → RESOLVED
Closed: 7 years ago
Resolution: --- → DUPLICATE
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: