Change NaClIPCAdapter to save translated descriptors to the rewritten message.

Change NaClIPCAdapter to save translated descriptors to the rewritten message.
BUG=116317
TEST=manual
TBR=brettw@chromium.org
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=148431

[bbudge, 2012-07-25 01:15:56]

On 2012/07/25 00:05:47, bbudge1 wrote:

Fix Windows compile error.

[dmichael (off chromium), 2012-07-25 16:00:40]

http://codereview.chromium.org/10815087/diff/14001/chrome/nacl/nacl_ipc_adapt...
File chrome/nacl/nacl_ipc_adapter.cc (right):

http://codereview.chromium.org/10815087/diff/14001/chrome/nacl/nacl_ipc_adapt...
chrome/nacl/nacl_ipc_adapter.cc:148: int desc_count() const { return
static_cast<int>(descs_.size()); }
why not just use size_t for the return type?

http://codereview.chromium.org/10815087/diff/14001/chrome/nacl/nacl_ipc_adapt...
chrome/nacl/nacl_ipc_adapter.cc:183: int
NaClIPCAdapter::RewrittenMessage::TransferData(NaClImcTypedMsgHdr* msg) {
I think I liked |Read| better for the name. It's not really transferring
anything at this point.

http://codereview.chromium.org/10815087/diff/14001/chrome/nacl/nacl_ipc_adapt...
chrome/nacl/nacl_ipc_adapter.cc:362:
locked_data_.to_be_received_.push(rewritten_msg);
Should this part maybe be at the end of SaveMessage? If we end up with any kind
of error conditions below, it seems like it would be better to drop the message
than send it incomplete.

[bbudge, 2012-07-25 16:54:14]

http://codereview.chromium.org/10815087/diff/14001/chrome/nacl/nacl_ipc_adapt...
File chrome/nacl/nacl_ipc_adapter.cc (right):

http://codereview.chromium.org/10815087/diff/14001/chrome/nacl/nacl_ipc_adapt...
chrome/nacl/nacl_ipc_adapter.cc:148: int desc_count() const { return
static_cast<int>(descs_.size()); }
On 2012/07/25 16:00:40, dmichael wrote:
> why not just use size_t for the return type?

Done.

http://codereview.chromium.org/10815087/diff/14001/chrome/nacl/nacl_ipc_adapt...
chrome/nacl/nacl_ipc_adapter.cc:183: int
NaClIPCAdapter::RewrittenMessage::TransferData(NaClImcTypedMsgHdr* msg) {
On 2012/07/25 16:00:40, dmichael wrote:
> I think I liked |Read| better for the name. It's not really transferring
> anything at this point.

I was finding 'Read' to be confusing. But 2 against 1. Done.

http://codereview.chromium.org/10815087/diff/14001/chrome/nacl/nacl_ipc_adapt...
chrome/nacl/nacl_ipc_adapter.cc:362:
locked_data_.to_be_received_.push(rewritten_msg);
On 2012/07/25 16:00:40, dmichael wrote:
> Should this part maybe be at the end of SaveMessage? If we end up with any
kind
> of error conditions below, it seems like it would be better to drop the
message
> than send it incomplete.

Good idea. Done.

[dmichael (off chromium), 2012-07-25 17:14:01]

LGTM, though I'm hoping mseaborn can check the NaClDesc reference counting
stuff.

[Mark Seaborn, 2012-07-25 22:09:41]

LGTM