WebKit Bugzilla
New
Browse
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED FIXED
167324
Maintain ordering when doing speculative loads
https://bugs.webkit.org/show_bug.cgi?id=167324
Summary
Maintain ordering when doing speculative loads
Antti Koivisto
Reported
2017-01-23 14:26:30 PST
Currently we randomize the ordering of speculative loads because they are serialized as a hash map. It would be better to load in the same order as the requests were originally issued as that is likely to match the order the document needs them.
Attachments
patch
(7.73 KB, patch)
2017-01-23 14:31 PST
,
Antti Koivisto
cdumez
: review+
Details
Formatted Diff
Diff
View All
Add attachment
proposed patch, testcase, etc.
Antti Koivisto
Comment 1
2017-01-23 14:31:35 PST
Created
attachment 299539
[details]
patch
Chris Dumez
Comment 2
2017-01-23 14:53:33 PST
Comment on
attachment 299539
[details]
patch View in context:
https://bugs.webkit.org/attachment.cgi?id=299539&action=review
r=me
> Source/WebKit2/NetworkProcess/cache/NetworkCacheSubresourcesEntry.cpp:109 > + Vector<SubresourceInfo> result;
I think we should reserveInitialCapacity() and uncheckedAppend() since we know there will be at most subresourceLoads.size() items and the common case will be subresourceLoads.size() items.
> Source/WebKit2/NetworkProcess/cache/NetworkCacheSubresourcesEntry.h:49 > + , m_requestHeaders(request.httpHeaderFields())
It is a bit unfortunate that we are now copying those HashMaps for transient resources too now. I hope this does not show on PLT.
Antti Koivisto
Comment 3
2017-01-23 17:39:00 PST
https://trac.webkit.org/changeset/211068
Darin Adler
Comment 4
2017-01-23 18:02:28 PST
Comment on
attachment 299539
[details]
patch View in context:
https://bugs.webkit.org/attachment.cgi?id=299539&action=review
>> Source/WebKit2/NetworkProcess/cache/NetworkCacheSubresourcesEntry.cpp:109 >> + Vector<SubresourceInfo> result; > > I think we should reserveInitialCapacity() and uncheckedAppend() since we know there will be at most subresourceLoads.size() items and the common case will be subresourceLoads.size() items.
The patch landed with a call to reserveCapacity/append, but reserveInitialCapacity/uncheckedAppend are more efficient.
Antti Koivisto
Comment 5
2017-01-23 21:36:58 PST
Oops!
Antti Koivisto
Comment 6
2017-01-23 22:03:30 PST
Did that in
https://trac.webkit.org/r211076
Radar WebKit Bug Importer
Comment 7
2017-02-01 06:02:02 PST
<
rdar://problem/30306314
>
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug