Implement the filesystem proxy. This allows the FileRef tests to pass in the...

Implement the filesystem proxy. This allows the FileRef tests (the ones which
don't use FileIO which isn't don yet) to pass in the proxy.

Hook up the code from the URLLoader that downloads to a file. This allows all
URLLoader tests to pass in the proxy.

Change code in dispatcher that zeros out the array to take into account
padding. It was only zero-filling half of the array since sizeof(enum) * # elts
seems to be half as large as the actual array due to padding on 64-bit systems.

Make the aborted completion callbacks run asynchronously. This was caught by
one of the file ref tests. We really need a system for doing this better, but
I don't want to do that in this patch.

TEST=ppapi_tests run under proxy

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=75566

[vtl, 2011-02-21 19:06:00]

LGTM, though "invalid" may be clearer (as a type) than "none".

http://codereview.chromium.org/6543028/diff/6001/ppapi/c/dev/ppb_file_system_...
File ppapi/c/dev/ppb_file_system_dev.h (right):

http://codereview.chromium.org/6543028/diff/6001/ppapi/c/dev/ppb_file_system_...
ppapi/c/dev/ppb_file_system_dev.h:30: * TODO(brettw) clarify whether this must
have completed before a file can
I wonder what "completed" means -- the callback having been run or merely
scheduled to be run? (Probably, to discourage bad uses, it should be the
former.)

http://codereview.chromium.org/6543028/diff/8002/ppapi/proxy/ppb_file_system_...
File ppapi/proxy/ppb_file_system_proxy.cc (right):

http://codereview.chromium.org/6543028/diff/8002/ppapi/proxy/ppb_file_system_...
ppapi/proxy/ppb_file_system_proxy.cc:92: if (object->opened_)
I'm a little surprised that we allow multiple Open()s....

http://codereview.chromium.org/6543028/diff/8002/ppapi/proxy/ppb_file_system_...
ppapi/proxy/ppb_file_system_proxy.cc:97: return PP_ERROR_BADARGUMENT;
Is "bad argument" really right? (Probably a precedent has been set elsewhere. It
makes sense when |instance| is provided by the caller, but it seems wrong when
it comes from a valid resource object.)

http://codereview.chromium.org/6543028/diff/8002/webkit/plugins/ppapi/ppb_fil...
File webkit/plugins/ppapi/ppb_file_system_impl.cc (right):

http://codereview.chromium.org/6543028/diff/8002/webkit/plugins/ppapi/ppb_fil...
webkit/plugins/ppapi/ppb_file_system_impl.cc:85: return PP_FILESYSTEMTYPE_NONE;
Maybe PP_FILESYSTEMTYPE_INVALID would be more appropriate?

(I'm tempted to say that it should be
int32_t GetType(PP_Resource, PP_FileSystemType_Dev* type);
so that errors can be reported, but that's a pain to use.)

[brettw, 2011-02-21 19:19:10]

http://codereview.chromium.org/6543028/diff/6001/ppapi/c/dev/ppb_file_system_...
File ppapi/c/dev/ppb_file_system_dev.h (right):

http://codereview.chromium.org/6543028/diff/6001/ppapi/c/dev/ppb_file_system_...
ppapi/c/dev/ppb_file_system_dev.h:30: * TODO(brettw) clarify whether this must
have completed before a file can
I also wonder that. I added this to the TODO.

http://codereview.chromium.org/6543028/diff/8002/ppapi/proxy/ppb_file_system_...
File ppapi/proxy/ppb_file_system_proxy.cc (right):

http://codereview.chromium.org/6543028/diff/8002/ppapi/proxy/ppb_file_system_...
ppapi/proxy/ppb_file_system_proxy.cc:97: return PP_ERROR_BADARGUMENT;
I was setting this elsewhere. In many places I end up with a helper function
that retrieves both the object and the dispatcher, and want to return only one
value if that fails (since this is an internal failure, I wasn't worrying a lot
about passing error values around). So I did this here for consistency.

http://codereview.chromium.org/6543028/diff/8002/webkit/plugins/ppapi/ppb_fil...
File webkit/plugins/ppapi/ppb_file_system_impl.cc (right):

http://codereview.chromium.org/6543028/diff/8002/webkit/plugins/ppapi/ppb_fil...
webkit/plugins/ppapi/ppb_file_system_impl.cc:85: return PP_FILESYSTEMTYPE_NONE;
On 2011/02/21 19:06:00, vtl wrote:
> Maybe PP_FILESYSTEMTYPE_INVALID would be more appropriate?
> 
> (I'm tempted to say that it should be
> int32_t GetType(PP_Resource, PP_FileSystemType_Dev* type);
> so that errors can be reported, but that's a pain to use.)

Done.