Bug 150939 – Modern IDB: Make indexes actually index

• Component: WebCore Misc.
• Status: RESOLVED FIXED

[Brady Eidson]

Modern IDB: Make indexes actually index

[Brady Eidson]

Created attachment 265241 [details]
Patch v1

[Build Bot]

Comment on attachment 265241 [details]
Patch v1

Attachment 265241 [details] did not pass mac-wk2-ews (mac-wk2):
Output: http://webkit-queues.webkit.org/results/412198

New failing tests:
fast/hidpi/image-set-border-image-simple.html

[Build Bot]

Created attachment 265245 [details]
Archive of layout-test-results from ews107 for mac-mavericks-wk2

The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews.
Bot: ews107  Port: mac-mavericks-wk2  Platform: Mac OS X 10.9.5

[Alex Christensen]

Comment on attachment 265241 [details]
Patch v1

View in context: https://bugs.webkit.org/attachment.cgi?id=265241&action=review

r=me

> Source/WebCore/Modules/indexeddb/shared/IndexKey.cpp:45
> +    Vector<IDBKeyData> keys;
> +    keys.reserveCapacity(m_keys.size());

reserveInitialCapacity

[Brady Eidson]

(In reply to comment #4)
> Comment on attachment 265241 [details]
> Patch v1
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=265241&action=review
> 
> r=me
> 
> > Source/WebCore/Modules/indexeddb/shared/IndexKey.cpp:45
> > +    Vector<IDBKeyData> keys;
> > +    keys.reserveCapacity(m_keys.size());
> 
> reserveInitialCapacity

Good call. Thx

[Brady Eidson]

Created attachment 265263 [details]
Patch for landing

[WebKit Commit Bot]

Comment on attachment 265263 [details]
Patch for landing

Clearing flags on attachment: 265263

Committed r192294: <http://trac.webkit.org/changeset/192294>

[Darin Adler]

Comment on attachment 265241 [details]
Patch v1

View in context: https://bugs.webkit.org/attachment.cgi?id=265241&action=review

> Source/WebCore/Modules/indexeddb/server/IndexValueEntry.cpp:64
> +    if (m_unique && m_key && *m_key == key) {

This is wrong when m_unique is true but m_key is null or not equal. It will fall into the code below and deterrence the pointer as the wrong type!

[Darin Adler]

Comment on attachment 265263 [details]
Patch for landing

View in context: https://bugs.webkit.org/attachment.cgi?id=265263&action=review

> Source/WebCore/Modules/indexeddb/server/IndexValueEntry.cpp:64
> +    if (m_unique && m_key && *m_key == key) {

Still wrong here. Need a test case for this and need to fix this by using a nested if rather than just a single one.

[Darin Adler]

deterrence -> dereference

[Brady Eidson]

(In reply to comment #9)
> Comment on attachment 265263 [details]
> Patch for landing
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=265263&action=review
> 
> > Source/WebCore/Modules/indexeddb/server/IndexValueEntry.cpp:64
> > +    if (m_unique && m_key && *m_key == key) {
> 
> Still wrong here. Need a test case for this and need to fix this by using a
> nested if rather than just a single one.

This patch landed with this bug as you pointed out.

*BUT* it was caught in a new test and fixed in https://trac.webkit.org/changeset/192610