Add support for chunked encoding in ChromeFrame for POST requests. This fixes...

Add support for chunked encoding in ChromeFrame for POST requests. This fixes the
URLRequestTestHTTP.TestPostChunkedDataBeforeStart net test failure in ChromeFrame.
To support chunked encoding we need to marshal the corresponding information in the
net::UploadData object to ensure that this object gets reconstructed correctly on the
other side (CF).

Disabled the URLRequestTestHTTP.TestPostChunkedDataAfterStart for ChromeFrame as this
test modifies the UploadData object after it has been marshaled over to CF which we don't
support in ChromeFrame.

BUG=none
TEST=Covered by existing net tests.


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=72723

[eroman, 2011-01-26 03:42:29]

LGTM on the src/net/* changes.

http://codereview.chromium.org/6357017/diff/10001/net/base/upload_data.h
File net/base/upload_data.h (right):

http://codereview.chromium.org/6357017/diff/10001/net/base/upload_data.h#newc...
net/base/upload_data.h:52: // marshaling.
nit: I believe the spelling is "marshalling" (two Ls).

http://codereview.chromium.org/6357017/diff/10001/net/base/upload_data.h#newc...
net/base/upload_data.h:102: // Sets whether this is the last chunk. Used during
IPC marshaling
ditto to above. Also please end in period.

[ananta, 2011-01-26 03:46:09]

http://codereview.chromium.org/6357017/diff/10001/net/base/upload_data.h
File net/base/upload_data.h (right):

http://codereview.chromium.org/6357017/diff/10001/net/base/upload_data.h#newc...
net/base/upload_data.h:52: // marshaling.
On 2011/01/26 03:42:29, eroman wrote:
> nit: I believe the spelling is "marshalling" (two Ls).

Done.

http://codereview.chromium.org/6357017/diff/10001/net/base/upload_data.h#newc...
net/base/upload_data.h:102: // Sets whether this is the last chunk. Used during
IPC marshaling
On 2011/01/26 03:42:29, eroman wrote:
> ditto to above. Also please end in period.

Done.

[amit, 2011-01-26 21:42:24]

lgtm

http://codereview.chromium.org/6357017/diff/1009/chrome/common/common_param_t...
File chrome/common/common_param_traits.cc (right):

http://codereview.chromium.org/6357017/diff/1009/chrome/common/common_param_t...
chrome/common/common_param_traits.cc:353: if (p.type() ==
net::UploadData::TYPE_BYTES ||
nit: It will be more readable to handle TYPE_CHUNK case in a separate 'if' or
even better in a switch/case since there are more of the if/else cases now.

http://codereview.chromium.org/6357017/diff/1009/chrome_frame/urlmon_upload_d...
File chrome_frame/urlmon_upload_data_stream.cc (right):

http://codereview.chromium.org/6357017/diff/1009/chrome_frame/urlmon_upload_d...
chrome_frame/urlmon_upload_data_stream.cc:31:
static_cast<uint64>(request_body_stream_->buf_len()));
if bug_len is ULONG you can remove both 'static_cast's

[ananta, 2011-01-26 22:19:58]

http://codereview.chromium.org/6357017/diff/1009/chrome/common/common_param_t...
File chrome/common/common_param_traits.cc (right):

http://codereview.chromium.org/6357017/diff/1009/chrome/common/common_param_t...
chrome/common/common_param_traits.cc:353: if (p.type() ==
net::UploadData::TYPE_BYTES ||
On 2011/01/26 21:42:24, amit wrote:
> nit: It will be more readable to handle TYPE_CHUNK case in a separate 'if' or
> even better in a switch/case since there are more of the if/else cases now.

Done.

http://codereview.chromium.org/6357017/diff/1009/chrome_frame/urlmon_upload_d...
File chrome_frame/urlmon_upload_data_stream.cc (right):

http://codereview.chromium.org/6357017/diff/1009/chrome_frame/urlmon_upload_d...
chrome_frame/urlmon_upload_data_stream.cc:31:
static_cast<uint64>(request_body_stream_->buf_len()));
On 2011/01/26 21:42:24, amit wrote:
> if bug_len is ULONG you can remove both 'static_cast's
Leaving this as is as per our discussion.