Modify the PPAPI CDM interface to pass more info.

Modify the PPAPI CDM interface to pass more info.

Builds on interface work done by xhwang in issue 10857027.

Add init_data argument to AddKey.
Change the PPP Decrypt methods to take PP_EncryptedBlockInfo structs
instead of simple integer identifiers.
Change the PPB delivery methods to take PP_DecryptedBlockInfo structs
instead of simple integer identifiers.
Update the proxy code accordingly.

BUG=138139
TEST=none


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=152724

[Tom Finegan, 2012-08-17 22:43:06]

ddorwin, dmichael, xhwang,

Please give this CL a quick look-- it builds on the already lgtm'd interface
changes in https://chromiumcodereview.appspot.com/10857027/. 

Not ready for full review yet (i.e. I think there are probably a ton of
nitworthy things), but if you have a few minutes some early feedback is always
welcome. Thanks!

[dmichael (off chromium), 2012-08-17 22:56:54]

I'd rather wait until you think it's ready for review. At least, it would be
helpful if this patch was based off of xhwang's (e.g., using git, or just wait
until his lands)...  It's a bit hard to see what's changed as it is.

[xhwang, 2012-08-17 23:03:38]

On 2012/08/17 22:56:54, dmichael wrote:
> I'd rather wait until you think it's ready for review. At least, it would be
> helpful if this patch was based off of xhwang's (e.g., using git, or just wait
> until his lands)...  It's a bit hard to see what's changed as it is.

My interface CL can't land without this implmentation CL. But I agree we can
work on this code more and test it out before a full review.

[Tom Finegan, 2012-08-20 21:31:45]

dmichael, ddorwin, xhwang,

I was overly pessimistic about the state of this CL on Friday-- I think it's
ready for review. I've added a few comments about things that I'm not sure
about-- PTAL!

Thanks!

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/content_de...
File ppapi/proxy/content_decryptor_private_serializer-inl.h (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/content_de...
ppapi/proxy/content_decryptor_private_serializer-inl.h:4: 
Two questions about this file:

1) Is this file small enough that I shouldn't bother w/the -inl file name
suffix?

2) I think it belongs in ppapi/proxy, but I'm not sure. Does it, or should I
move the functions the the C++ wrapper?

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/ppb_instan...
File ppapi/proxy/ppb_instance_proxy.h (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/ppb_instan...
ppapi/proxy/ppb_instance_proxy.h:24: struct PP_DecryptedBlockInfo;
Should I include pp_content_decryptor.h up above instead of forward declaring
this?

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/ppp_conten...
File ppapi/proxy/ppp_content_decryptor_private_proxy.cc (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/ppp_conten...
ppapi/proxy/ppp_content_decryptor_private_proxy.cc:166: using
ppapi::proxy::SerializeBlockInfo;
Should I just do this at file scope?

[xhwang, 2012-08-20 22:19:31]

Nice patch, only a few comments.

http://codereview.chromium.org/10854209/diff/1/ppapi/proxy/content_decryptor_...
File ppapi/proxy/content_decryptor_private_serializer-inl.h (right):

http://codereview.chromium.org/10854209/diff/1/ppapi/proxy/content_decryptor_...
ppapi/proxy/content_decryptor_private_serializer-inl.h:18: // IPC. Return true
upon success, and false upon failure.
nit: Return+s+

http://codereview.chromium.org/10854209/diff/1/ppapi/proxy/content_decryptor_...
ppapi/proxy/content_decryptor_private_serializer-inl.h:26: bool
SerializeBlockInfo(const T* block_info,
Google style strongly recommends passing const ref as input parameter, so use
"const T& block_info"?

http://codereview.chromium.org/10854209/diff/1/ppapi/proxy/content_decryptor_...
ppapi/proxy/content_decryptor_private_serializer-inl.h:31: if
(serialized_block_info->size() != sizeof(T))
If we pass block_info as ref, we can use "sizeof(block_info)" here. (style guide
rule: Use sizeof(varname) instead of sizeof(type) whenever possible.)

Also, move this down after line 34?

http://codereview.chromium.org/10854209/diff/1/ppapi/proxy/ppb_instance_proxy.h
File ppapi/proxy/ppb_instance_proxy.h (right):

http://codereview.chromium.org/10854209/diff/1/ppapi/proxy/ppb_instance_proxy...
ppapi/proxy/ppb_instance_proxy.h:24: struct PP_DecryptedBlockInfo;
On 2012/08/20 21:31:45, tomf wrote:
> Should I include pp_content_decryptor.h up above instead of forward declaring
> this?

The rule is always forward declare if you can. So this is good.

http://codereview.chromium.org/10854209/diff/1/ppapi/proxy/ppp_content_decryp...
File ppapi/proxy/ppp_content_decryptor_private_proxy.cc (right):

http://codereview.chromium.org/10854209/diff/1/ppapi/proxy/ppp_content_decryp...
ppapi/proxy/ppp_content_decryptor_private_proxy.cc:324: const_cast<const
PP_EncryptedBlockInfo*>(&block_info));
Do you need this const cast?

[Tom Finegan, 2012-08-20 22:50:03]

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/content_de...
File ppapi/proxy/content_decryptor_private_serializer-inl.h (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/content_de...
ppapi/proxy/content_decryptor_private_serializer-inl.h:18: // IPC. Return true
upon success, and false upon failure.
On 2012/08/20 22:19:31, xhwang wrote:
> nit: Return+s+

Done, but did s/Return/Both functions return/ instead.

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/content_de...
ppapi/proxy/content_decryptor_private_serializer-inl.h:26: bool
SerializeBlockInfo(const T* block_info,
On 2012/08/20 22:19:31, xhwang wrote:
> Google style strongly recommends passing const ref as input parameter, so use
> "const T& block_info"?

Done.

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/content_de...
ppapi/proxy/content_decryptor_private_serializer-inl.h:31: if
(serialized_block_info->size() != sizeof(T))
On 2012/08/20 22:19:31, xhwang wrote:
> If we pass block_info as ref, we can use "sizeof(block_info)" here. (style
guide
> rule: Use sizeof(varname) instead of sizeof(type) whenever possible.)
> 
> Also, move this down after line 34?

Done/bug fixed.

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/ppp_conten...
File ppapi/proxy/ppp_content_decryptor_private_proxy.cc (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/ppp_conten...
ppapi/proxy/ppp_content_decryptor_private_proxy.cc:324: const_cast<const
PP_EncryptedBlockInfo*>(&block_info));
On 2012/08/20 22:19:31, xhwang wrote:
> Do you need this const cast?

Yeah, clang refuses to compile without it:

../../ppapi/proxy/ppp_content_decryptor_private_proxy.cc:321:5: error: no
matching function for call to 'CallWhileUnlocked'
    CallWhileUnlocked(ppp_decryptor_impl_->Decrypt,
    ^~~~~~~~~~~~~~~~~
../../ppapi/shared_impl/proxy_lock.h:103:12: note: candidate template ignored:
deduced conflicting types for parameter 'P3' ('const PP_EncryptedBlockInfo *'
vs. 'PP_EncryptedBlockInfo *')
ReturnType CallWhileUnlocked(ReturnType (*function)(P1, P2, P3),

[xhwang, 2012-08-21 00:39:27]

lgtm

[ddorwin, 2012-08-21 17:30:28]

webkit/media review.

https://chromiumcodereview.appspot.com/10854209/diff/5003/webkit/media/crypto...
File webkit/media/crypto/ppapi/cdm_wrapper.cc (right):

https://chromiumcodereview.appspot.com/10854209/diff/5003/webkit/media/crypto...
webkit/media/crypto/ppapi/cdm_wrapper.cc:154: DecryptedBlock decrypted_block;
Set encrypted_block_info on DecryptedBlock.

[dmichael (off chromium), 2012-08-21 17:37:07]

looks mostly good.

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/cpp/private/cont...
File ppapi/cpp/private/content_decryptor_private.h (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/cpp/private/cont...
ppapi/cpp/private/content_decryptor_private.h:38: const PP_EncryptedBlockInfo*
encrypted_block_info) = 0;
You should pass by const&, not const* here. The const* in the C interfaces is
only because C doesn't have references.

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/content_de...
File ppapi/proxy/content_decryptor_private_serializer-inl.h (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/content_de...
ppapi/proxy/content_decryptor_private_serializer-inl.h:4: 
On 2012/08/20 21:31:45, tomf wrote:
> Two questions about this file:
> 
> 1) Is this file small enough that I shouldn't bother w/the -inl file name
Yeah, we don't use the -inl suffix anywhere else in pepper.

> suffix?
> 
> 2) I think it belongs in ppapi/proxy, but I'm not sure. Does it, or should I
> move the functions the the C++ wrapper?
ppapi/proxy is the right place.

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/ppp_conten...
File ppapi/proxy/ppp_content_decryptor_private_proxy.cc (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/ppp_conten...
ppapi/proxy/ppp_content_decryptor_private_proxy.cc:166: using
ppapi::proxy::SerializeBlockInfo;
On 2012/08/20 21:31:45, tomf wrote:
> Should I just do this at file scope?
Why do you even need it? Aren't you in ppapi::proxy already?

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/ppp_conten...
ppapi/proxy/ppp_content_decryptor_private_proxy.cc:324: const_cast<const
PP_EncryptedBlockInfo*>(&block_info));
On 2012/08/20 22:50:03, tomf wrote:
> On 2012/08/20 22:19:31, xhwang wrote:
> > Do you need this const cast?
> 
> Yeah, clang refuses to compile without it:
> 
> ../../ppapi/proxy/ppp_content_decryptor_private_proxy.cc:321:5: error: no
> matching function for call to 'CallWhileUnlocked'
>     CallWhileUnlocked(ppp_decryptor_impl_->Decrypt,
>     ^~~~~~~~~~~~~~~~~
> ../../ppapi/shared_impl/proxy_lock.h:103:12: note: candidate template ignored:
> deduced conflicting types for parameter 'P3' ('const PP_EncryptedBlockInfo *'
> vs. 'PP_EncryptedBlockInfo *')
> ReturnType CallWhileUnlocked(ReturnType (*function)(P1, P2, P3),
Oops, yeah, we could also improve the CallWhileUnlocked templates (in
ppapi/shared_impl/proxy_lock.h") to handle this case. But this way is certainly
simpler. Adding const with const_cast is always safe.

https://chromiumcodereview.appspot.com/10854209/diff/1/webkit/media/crypto/pp...
File webkit/media/crypto/ppapi/cdm_wrapper.cc (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/webkit/media/crypto/pp...
webkit/media/crypto/ppapi/cdm_wrapper.cc:38: PP_DecryptedBlockInfo
decrypted_block_info;
Do you maybe still want a default constructor so you can zero out this struct?

https://chromiumcodereview.appspot.com/10854209/diff/1/webkit/media/crypto/pp...
webkit/media/crypto/ppapi/cdm_wrapper.cc:120: init_data.ByteLength());
surprised you don't get an unused variable warning/error from Windows compile.

I can't remember if it was this part where I complained before, but I still
think you won't want to use string when you talk to the real CDM, since you're
forced to copy if you do. Just some raw bytes would probably do. If this is just
a placeholder, though, it's fine.

https://chromiumcodereview.appspot.com/10854209/diff/1/webkit/media/crypto/pp...
webkit/media/crypto/ppapi/cdm_wrapper.cc:157: decrypted_block));
This copies potentially a lot of data... have you thought about ways to avoid
copying? FWIW, since you're only running out-of-process, you don't even need to
defer the callback. We enforce that the browser handles calls asynchronously in
some other places, like PostMessage. You can make the same assertion here, and
simplify some of this code, perhaps?

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/content...
File ppapi/proxy/content_decryptor_private_serializer-inl.h (right):

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/content...
ppapi/proxy/content_decryptor_private_serializer-inl.h:5: #ifndef
PPAPI_PROXY_CONTENT_DECRYPTOR_PRIVATE_SERIALIZER_INL_H_
I wouldn't use the -inl suffix

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/content...
ppapi/proxy/content_decryptor_private_serializer-inl.h:32: sizeof(T));
nit: You should prefer to use sizeof on a variable name instead of a type name,
i.e. sizeof(block_info)
http://google-styleguide.googlecode.com/svn/trunk/cppguide.xml#sizeof

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/content...
ppapi/proxy/content_decryptor_private_serializer-inl.h:46: if
(serialized_block_info.size() != sizeof(T))
sizeof(*block_info)

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/content...
ppapi/proxy/content_decryptor_private_serializer-inl.h:49:
std::memcpy(block_info, serialized_block_info.data(), sizeof(T));
sizeof(*block_info)

[xhwang, 2012-08-21 21:22:01]

http://codereview.chromium.org/10854209/diff/5003/ppapi/proxy/ppb_instance_pr...
File ppapi/proxy/ppb_instance_proxy.h (right):

http://codereview.chromium.org/10854209/diff/5003/ppapi/proxy/ppb_instance_pr...
ppapi/proxy/ppb_instance_proxy.h:7: 
nit: include what you use, add #include <string> here.

http://codereview.chromium.org/10854209/diff/5003/ppapi/proxy/ppp_content_dec...
File ppapi/proxy/ppp_content_decryptor_private_proxy.h (right):

http://codereview.chromium.org/10854209/diff/5003/ppapi/proxy/ppp_content_dec...
ppapi/proxy/ppp_content_decryptor_private_proxy.h:7: 
nit: include what you use, add #include <string> here.

[Tom Finegan, 2012-08-21 22:57:38]

ddorwin, dmichael, xhwang: PTAL, I think I've gotten to all the comments, but
with them sprinkled throughout the various patch sets I might have missed
something. 

Thanks!

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/cpp/private/cont...
File ppapi/cpp/private/content_decryptor_private.h (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/cpp/private/cont...
ppapi/cpp/private/content_decryptor_private.h:38: const PP_EncryptedBlockInfo*
encrypted_block_info) = 0;
On 2012/08/21 17:37:08, dmichael wrote:
> You should pass by const&, not const* here. The const* in the C interfaces is
> only because C doesn't have references.

Done.

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/content_de...
File ppapi/proxy/content_decryptor_private_serializer-inl.h (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/content_de...
ppapi/proxy/content_decryptor_private_serializer-inl.h:4: 
On 2012/08/21 17:37:08, dmichael wrote:
> On 2012/08/20 21:31:45, tomf wrote:
> > Two questions about this file:
> > 
> > 1) Is this file small enough that I shouldn't bother w/the -inl file name
> Yeah, we don't use the -inl suffix anywhere else in pepper.
> 
> > suffix?
> > 
> > 2) I think it belongs in ppapi/proxy, but I'm not sure. Does it, or should I
> > move the functions the the C++ wrapper?
> ppapi/proxy is the right place.

Done.

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/ppp_conten...
File ppapi/proxy/ppp_content_decryptor_private_proxy.cc (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/ppapi/proxy/ppp_conten...
ppapi/proxy/ppp_content_decryptor_private_proxy.cc:166: using
ppapi::proxy::SerializeBlockInfo;
On 2012/08/21 17:37:08, dmichael wrote:
> On 2012/08/20 21:31:45, tomf wrote:
> > Should I just do this at file scope?
> Why do you even need it? Aren't you in ppapi::proxy already?

Oops; question mooted.

https://chromiumcodereview.appspot.com/10854209/diff/1/webkit/media/crypto/pp...
File webkit/media/crypto/ppapi/cdm_wrapper.cc (right):

https://chromiumcodereview.appspot.com/10854209/diff/1/webkit/media/crypto/pp...
webkit/media/crypto/ppapi/cdm_wrapper.cc:38: PP_DecryptedBlockInfo
decrypted_block_info;
On 2012/08/21 17:37:08, dmichael wrote:
> Do you maybe still want a default constructor so you can zero out this struct?

Definitely, added. (memset abuse!)

https://chromiumcodereview.appspot.com/10854209/diff/1/webkit/media/crypto/pp...
webkit/media/crypto/ppapi/cdm_wrapper.cc:120: init_data.ByteLength());
On 2012/08/21 17:37:08, dmichael wrote:
> surprised you don't get an unused variable warning/error from Windows compile.
> 
> I can't remember if it was this part where I complained before, but I still
> think you won't want to use string when you talk to the real CDM, since you're
> forced to copy if you do. Just some raw bytes would probably do. If this is
just
> a placeholder, though, it's fine.

This is just a placeholder-- the CDM takes uint8* and a length.

https://chromiumcodereview.appspot.com/10854209/diff/1/webkit/media/crypto/pp...
webkit/media/crypto/ppapi/cdm_wrapper.cc:157: decrypted_block));
On 2012/08/21 17:37:08, dmichael wrote:
> This copies potentially a lot of data... have you thought about ways to avoid
> copying? FWIW, since you're only running out-of-process, you don't even need
to
> defer the callback. We enforce that the browser handles calls asynchronously
in
> some other places, like PostMessage. You can make the same assertion here, and
> simplify some of this code, perhaps?

Added TODO per offline discussion.

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/content...
File ppapi/proxy/content_decryptor_private_serializer-inl.h (right):

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/content...
ppapi/proxy/content_decryptor_private_serializer-inl.h:5: #ifndef
PPAPI_PROXY_CONTENT_DECRYPTOR_PRIVATE_SERIALIZER_INL_H_
On 2012/08/21 17:37:08, dmichael wrote:
> I wouldn't use the -inl suffix

Done.

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/content...
ppapi/proxy/content_decryptor_private_serializer-inl.h:32: sizeof(T));
On 2012/08/21 17:37:08, dmichael wrote:
> nit: You should prefer to use sizeof on a variable name instead of a type
name,
> i.e. sizeof(block_info)
> http://google-styleguide.googlecode.com/svn/trunk/cppguide.xml#sizeof

Doh! Missed this one when I fixed the other for xhwang. Done.

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/content...
ppapi/proxy/content_decryptor_private_serializer-inl.h:46: if
(serialized_block_info.size() != sizeof(T))
On 2012/08/21 17:37:08, dmichael wrote:
> sizeof(*block_info)

Done.

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/content...
ppapi/proxy/content_decryptor_private_serializer-inl.h:49:
std::memcpy(block_info, serialized_block_info.data(), sizeof(T));
On 2012/08/21 17:37:08, dmichael wrote:
> sizeof(*block_info)

Done.

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/ppb_ins...
File ppapi/proxy/ppb_instance_proxy.h (right):

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/ppb_ins...
ppapi/proxy/ppb_instance_proxy.h:7: 
On 2012/08/21 21:22:01, xhwang wrote:
> nit: include what you use, add #include <string> here.

Done.

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/ppp_con...
File ppapi/proxy/ppp_content_decryptor_private_proxy.h (right):

https://chromiumcodereview.appspot.com/10854209/diff/5003/ppapi/proxy/ppp_con...
ppapi/proxy/ppp_content_decryptor_private_proxy.h:7: 
On 2012/08/21 21:22:01, xhwang wrote:
> nit: include what you use, add #include <string> here.

Done.

https://chromiumcodereview.appspot.com/10854209/diff/5003/webkit/media/crypto...
File webkit/media/crypto/ppapi/cdm_wrapper.cc (right):

https://chromiumcodereview.appspot.com/10854209/diff/5003/webkit/media/crypto...
webkit/media/crypto/ppapi/cdm_wrapper.cc:154: DecryptedBlock decrypted_block;
On 2012/08/21 17:30:28, ddorwin wrote:
> Set encrypted_block_info on DecryptedBlock.

Done.

[dmichael (off chromium), 2012-08-21 23:05:14]

lgtm

https://chromiumcodereview.appspot.com/10854209/diff/11001/webkit/media/crypt...
File webkit/media/crypto/ppapi/cdm_wrapper.cc (right):

https://chromiumcodereview.appspot.com/10854209/diff/11001/webkit/media/crypt...
webkit/media/crypto/ppapi/cdm_wrapper.cc:54:
pp::Module::Get()->core()->CallOnMainThread(0, cb, PP_OK);
May be good to add a TODO comment here...  AIUI, this is only necessary because
PPAPI doesn't allow calls off the main thread yet. I'll be adding that soonish
(mid-late September, I think).

[ddorwin, 2012-08-21 23:18:25]

https://chromiumcodereview.appspot.com/10854209/diff/11001/webkit/media/crypt...
File webkit/media/crypto/ppapi/cdm_wrapper.cc (right):

https://chromiumcodereview.appspot.com/10854209/diff/11001/webkit/media/crypt...
webkit/media/crypto/ppapi/cdm_wrapper.cc:38:
std::memset(reinterpret_cast<void*>(&decrypted_block_info),
May be able to use a constructor. xhang is looking into this.
@dmichael, what is the normal Pepper practice?

https://chromiumcodereview.appspot.com/10854209/diff/11001/webkit/media/crypt...
webkit/media/crypto/ppapi/cdm_wrapper.cc:77: const PP_EncryptedBlockInfo&
encrypted_block_info) OVERRIDE;
This is no longer an override of PPP_ContentDecryptor_Private.

https://chromiumcodereview.appspot.com/10854209/diff/11001/webkit/media/crypt...
webkit/media/crypto/ppapi/cdm_wrapper.cc:176: // instead of passing a giant
std::string filled with encrypted data.
This will eventually be a PPB_Buffer_Dev ID and a raw pointer. So the issue may
go away.

[ddorwin, 2012-08-21 23:21:09]

lgtm

https://chromiumcodereview.appspot.com/10854209/diff/11001/webkit/media/crypt...
File webkit/media/crypto/ppapi/cdm_wrapper.cc (right):

https://chromiumcodereview.appspot.com/10854209/diff/11001/webkit/media/crypt...
webkit/media/crypto/ppapi/cdm_wrapper.cc:77: const PP_EncryptedBlockInfo&
encrypted_block_info) OVERRIDE;
On 2012/08/21 23:18:25, ddorwin wrote:
> This is no longer an override of PPP_ContentDecryptor_Private.

Nevermind. I was looking at the C header, not the C++ header.

[commit-bot: I haz the power, 2012-08-21 23:22:27]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tomfinegan@chromium.org/10854209/1025

[commit-bot: I haz the power, 2012-08-21 23:42:43]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tomfinegan@chromium.org/10854209/16003

[Tom Finegan (google), 2012-08-22 00:26:05]

On 2012/08/21 23:42:43, I haz the power (commit-bot) wrote:
> CQ is trying da patch. Follow status at
>
https://chromium-status.appspot.com/cq/tomfinegan%40chromium.org/10854209/16003

Rebased... CQ go!

[commit-bot: I haz the power, 2012-08-22 00:28:33]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tomfinegan@chromium.org/10854209/22001

[commit-bot: I haz the power, 2012-08-22 04:06:24]

Change committed as 152724