obsolete: SFI NaCl: Batch-open resource files

SFI NaCl: Batch-open resource files

This CL adds SFI NaCl support to https://codereview.chromium.org/649603004/.
SFI NaCl on Windows is still not supported.

* In the browser process, validate resource files in the same way
 as https://codereview.chromium.org/572973002/.
* In the NaCl plugin process, add a hook to the IPC adapter
 to intercept PpapiHostMsg_OpenResource messages from the untrusted
 side so that the trusted side can immediately return pre-opened
 descriptors without issuing a Chrome IPC to the renderer process.

BUG=nativeclient:3802
BUG=348232

[Yusuke Sato, 2014-11-18 20:12:19]

Patchset #1 (id:1) has been deleted

[Yusuke Sato, 2014-11-18 20:13:30]

Patchset #2 (id:40001) has been deleted

[Yusuke Sato, 2014-11-18 20:24:12]

yusukes@chromium.org changed reviewers:
+ mseaborn@chromium.org, teravest@chromium.org

[Yusuke Sato, 2014-11-18 20:24:13]

Please take a look.

[teravest, 2014-11-19 15:59:31]

https://codereview.chromium.org/728113002/diff/60001/components/nacl/browser/...
File components/nacl/browser/nacl_process_host.cc (right):

https://codereview.chromium.org/728113002/diff/60001/components/nacl/browser/...
components/nacl/browser/nacl_process_host.cc:933: file_paths[0] = file_path;
nit: This would be a bit easier to read now if file_path was renamed to
nexe_file_path, I think.

https://codereview.chromium.org/728113002/diff/60001/components/nacl/browser/...
components/nacl/browser/nacl_process_host.cc:980: close_delay);
Why is there a delay added here?

https://codereview.chromium.org/728113002/diff/60001/components/nacl/loader/n...
File components/nacl/loader/nacl_ipc_adapter.cc (right):

https://codereview.chromium.org/728113002/diff/60001/components/nacl/loader/n...
components/nacl/loader/nacl_ipc_adapter.cc:664: SaveMessage(*new_msg,
rewritten_msg.get());
I don't like that the critical section is made larger here, but this is probably
better than having an argument that says whether the lock is held coming in.

You should assert that the lock is acquired at the top of this function.

[Yusuke Sato, 2014-11-20 00:30:27]

https://codereview.chromium.org/728113002/diff/60001/components/nacl/browser/...
File components/nacl/browser/nacl_process_host.cc (right):

https://codereview.chromium.org/728113002/diff/60001/components/nacl/browser/...
components/nacl/browser/nacl_process_host.cc:933: file_paths[0] = file_path;
On 2014/11/19 15:59:30, teravest wrote:
> nit: This would be a bit easier to read now if file_path was renamed to
> nexe_file_path, I think.

Done.

https://codereview.chromium.org/728113002/diff/60001/components/nacl/browser/...
components/nacl/browser/nacl_process_host.cc:980: close_delay);
Removed.

I thought posting dozens of CloseFile tasks during plugin startup negatively
affects the plugin startup time, but I was wrong. Doing this PostTask seems just
fine (confirmed with chrome://tracing).

https://codereview.chromium.org/728113002/diff/60001/components/nacl/loader/n...
File components/nacl/loader/nacl_ipc_adapter.cc (right):

https://codereview.chromium.org/728113002/diff/60001/components/nacl/loader/n...
components/nacl/loader/nacl_ipc_adapter.cc:664: SaveMessage(*new_msg,
rewritten_msg.get());
On 2014/11/19 15:59:30, teravest wrote:
> I don't like that the critical section is made larger here, but this is
probably
> better than having an argument that says whether the lock is held coming in.
> 
> You should assert that the lock is acquired at the top of this function.

Done.

[Yusuke Sato, 2014-11-20 00:30:51]

Thanks. Please take another look.

[Yusuke Sato, 2015-03-17 18:16:40]

yusukes@chromium.org changed reviewers:
- mseaborn@chromium.org, teravest@chromium.org