[Webkit-unassigned] [Bug 143872] [SOUP] Add initial implementation of NetworkProcess disk cache

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=143872

--- Comment #9 from Carlos Garcia Campos <cgarcia at igalia.com> ---
(In reply to comment #8)
> Comment on attachment 251014 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=251014&action=review
> 
> Awesome stuff Carlos!

Thanks.

> > Source/WebKit2/NetworkProcess/cache/NetworkCacheBlobStorage.cpp:34
> > +#include <fcntl.h>
> 
> Why this?

It's needed for open and its flags (O_CREAT, O_RDWR, etc.) in Linux, I guess it's not needed in mac, but doesn't hurt anyway.

> > Source/WebKit2/NetworkProcess/cache/NetworkCacheDataSoup.cpp:29
> > +#if ENABLE(NETWORK_CACHE)
> 
> Don't we need some USE(SOUP) or PLATFORM(LINUX) and the like?

I assumed NETWORK_CACHE is already enabled depending on the port.

> > Source/WebKit2/NetworkProcess/cache/NetworkCacheDataSoup.cpp:129
> > +    ASSERT(map && map != MAP_FAILED);
> 
> Perhaps better to split this in 2 ASSERTs in order to better spot the actual
> cause of failure.

Ok, I think I copied this one from mac impl :-)

> > Source/WebKit2/NetworkProcess/cache/NetworkCacheIOChannelSoup.cpp:58
> > +    ASSERT(m_outputStream || m_inputStream || m_ioStream);
> 
> Same here about splitting the ASSERT

Sure.

> > Source/WebKit2/NetworkProcess/cache/NetworkCacheIOChannelSoup.cpp:82
> > +    }
> 
> Can't we use Data directly instead of having to copy it from a SoupBuffer?

I'm not sure I understand the question. We create a Data with a SoupBuffer to avoid copying the data. The read soup buffer will be reused on next read, so we need to copy that data appending it to the Data.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20150420/750e3ba5/attachment.html>