support loading pepper flash debugger dlls

Modifies the pepper flash component installer to also look in a path under the Windows system directory for a debugger version of the flash DLL. This DLL is not registered as a chrome component. It is configured to load after the non-debugger flash DLL.

BUG=345886
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=266761

[luken, 2014-04-23 22:30:46]

PTAL, thanks

[cpu_(ooo_6.6-7.5), 2014-04-24 20:42:04]

https://codereview.chromium.org/203993004/diff/80001/chrome/browser/component...
File chrome/browser/component_updater/pepper_flash_component_installer.cc
(right):

https://codereview.chromium.org/203993004/diff/80001/chrome/browser/component...
chrome/browser/component_updater/pepper_flash_component_installer.cc:101:
PathService::Get(chrome::DIR_PEPPER_FLASH_DEBUGGER_PLUGIN, &subdir);
any reason not to return the full path when
chrome::DIR_PEPPER_FLASH_DEBUGGER_PLUGIN is called?

https://codereview.chromium.org/203993004/diff/80001/chrome/browser/component...
chrome/browser/component_updater/pepper_flash_component_installer.cc:148:
L"???";
elif defined(ARCH_CPU_ARM_FAMILY) ?

https://codereview.chromium.org/203993004/diff/80001/chrome/browser/component...
chrome/browser/component_updater/pepper_flash_component_installer.cc:152:
Version* dll_version) {
DCHECK(BrowserThread::CurrentlyOn(BrowserThread::FILE));  ? I mean you cannot do
file IO in the UI Thread.

https://codereview.chromium.org/203993004/diff/80001/chrome/browser/component...
chrome/browser/component_updater/pepper_flash_component_installer.cc:155: if
(!base::PathExists(debugger_dir)) {
if the style of this file has single statement ifs applied consistently then
lets change this one. If on the other hand we are inconsistent then do as you
wish.

https://codereview.chromium.org/203993004/diff/80001/chrome/browser/component...
chrome/browser/component_updater/pepper_flash_component_installer.cc:168:
std::string file_name(path.BaseName().MaybeAsASCII());
I have the feeling that using base/strings/string_split.h will make this
prettier, splitting on '_'.  wdyt?

[luken, 2014-04-24 22:04:15]

https://codereview.chromium.org/203993004/diff/80001/chrome/browser/component...
File chrome/browser/component_updater/pepper_flash_component_installer.cc
(right):

https://codereview.chromium.org/203993004/diff/80001/chrome/browser/component...
chrome/browser/component_updater/pepper_flash_component_installer.cc:101:
PathService::Get(chrome::DIR_PEPPER_FLASH_DEBUGGER_PLUGIN, &subdir);
On 2014/04/24 20:42:04, cpu_(out until 23rd) wrote:
> any reason not to return the full path when
> chrome::DIR_PEPPER_FLASH_DEBUGGER_PLUGIN is called?

Done.

https://codereview.chromium.org/203993004/diff/80001/chrome/browser/component...
chrome/browser/component_updater/pepper_flash_component_installer.cc:148:
L"???";
On 2014/04/24 20:42:04, cpu_(out until 23rd) wrote:
> elif defined(ARCH_CPU_ARM_FAMILY) ?

Done.

https://codereview.chromium.org/203993004/diff/80001/chrome/browser/component...
chrome/browser/component_updater/pepper_flash_component_installer.cc:152:
Version* dll_version) {
On 2014/04/24 20:42:04, cpu_(out until 23rd) wrote:
> DCHECK(BrowserThread::CurrentlyOn(BrowserThread::FILE));  ? I mean you cannot
do
> file IO in the UI Thread.
> 

We are on the file IO thread, and the other function to find the regular flash
dll doesn't have the DCHECK, but I like the DCHECK so I added it to both. :)

https://codereview.chromium.org/203993004/diff/80001/chrome/browser/component...
chrome/browser/component_updater/pepper_flash_component_installer.cc:155: if
(!base::PathExists(debugger_dir)) {
On 2014/04/24 20:42:04, cpu_(out until 23rd) wrote:
> if the style of this file has single statement ifs applied consistently then
> lets change this one. If on the other hand we are inconsistent then do as you
> wish.

Done.

https://codereview.chromium.org/203993004/diff/80001/chrome/browser/component...
chrome/browser/component_updater/pepper_flash_component_installer.cc:168:
std::string file_name(path.BaseName().MaybeAsASCII());
On 2014/04/24 20:42:04, cpu_(out until 23rd) wrote:
> I have the feeling that using base/strings/string_split.h will make this
> prettier, splitting on '_'.  wdyt?
> 
I didn't know about this, yes!

[cpu_(ooo_6.6-7.5), 2014-04-25 18:13:23]

lgtm

[jam, 2014-04-28 17:47:38]

lgtm with comment

https://codereview.chromium.org/203993004/diff/160001/chrome/common/chrome_pa...
File chrome/common/chrome_paths.h (right):

https://codereview.chromium.org/203993004/diff/160001/chrome/common/chrome_pa...
chrome/common/chrome_paths.h:21: DIR_APP = PATH_START,        // Directory where
dlls and data reside.
nit: Please undo all the spacing changes to this file. i see you made it so that
each section has two spaces from the longest constant to the comment, but this
isn't maintainable because if someone adds a longer string, to do what you did
would involved changing all the lines (it also makes blaming a line harder)

[luken, 2014-04-28 17:57:16]

https://codereview.chromium.org/203993004/diff/160001/chrome/common/chrome_pa...
File chrome/common/chrome_paths.h (right):

https://codereview.chromium.org/203993004/diff/160001/chrome/common/chrome_pa...
chrome/common/chrome_paths.h:21: DIR_APP = PATH_START,        // Directory where
dlls and data reside.
On 2014/04/28 17:47:38, jam wrote:
> nit: Please undo all the spacing changes to this file. i see you made it so
that
> each section has two spaces from the longest constant to the comment, but this
> isn't maintainable because if someone adds a longer string, to do what you did
> would involved changing all the lines (it also makes blaming a line harder)
Will do. git cl format did this, I'll clean it up.

[cpu_(ooo_6.6-7.5), 2014-04-28 18:06:14]

The CQ bit was checked by cpu@chromium.org

[commit-bot: I haz the power, 2014-04-28 18:07:41]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/luken@chromium.org/203993004/180001

[commit-bot: I haz the power, 2014-04-28 18:32:57]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-04-28 18:32:58]

Try jobs failed on following builders:
  tryserver.chromium on linux_chromium_gn_rel

[cpu_(ooo_6.6-7.5), 2014-04-28 19:16:16]

The CQ bit was checked by cpu@chromium.org

[commit-bot: I haz the power, 2014-04-28 19:17:00]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/luken@chromium.org/203993004/180001

[commit-bot: I haz the power, 2014-04-28 19:45:16]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-04-28 19:45:17]

Try jobs failed on following builders:
  tryserver.chromium on linux_chromium_chromeos_rel

[cpu_(ooo_6.6-7.5), 2014-04-28 20:03:51]

The CQ bit was checked by cpu@chromium.org

[commit-bot: I haz the power, 2014-04-28 20:04:18]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/luken@chromium.org/203993004/180001

[commit-bot: I haz the power, 2014-04-29 07:22:11]

Change committed as 266761