summaryrefslogtreecommitdiffstats
path: root/content/browser/child_process_launcher.cc
Commit message (Collapse)AuthorAgeFilesLines
* Change IPC::ChannelMojo to use associated interfaces.sammc2016-03-071-2/+4
| | | | | | | | BUG=579813 Review URL: https://codereview.chromium.org/1669493005 Cr-Commit-Position: refs/heads/master@{#379669}
* Reland: Bootstrap Mojo IPC independent of Chrome IPCrockot2016-03-021-5/+34
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The original CL: https://crrev.com/eb199fbaecd8e5d294010dfbe8897b7d83aeab07 broke NaCL integration tests on some bots. After some extensive investigation this is somehow tied to an obscure dependency on the value of kIPCDescriptorMax. Moving kMojoIPCChannel to content_descriptors.h allows us to avoid changing that value and thus avoid the mysterious breakage. :| Original CL description follows: This initializes Mojo IPC very early in child process startup so we can start using cross-process pipes in content without a dependency on Chrome IPC. On all supported platforms, a new OS pipe handle is passed through ChildProcessLauncher and hooked up to the EDK on ChildThreadImpl startup to establish a Mojo IPC channel to the parent. A command-line argument is also to used to share a unique pipe token with the child, and this token is exchanged over the Mojo IPC channel to bootstrap a primordial message pipe. While this CL establishes these facilities in a way that can be used for all child processes, only render processes are brought up in this way for now. BUG=583557 TEST=Chrome still runs in mash with Mus-in-renderer support. TBR=ben@chromium.org,jschuh@chromium.org Review URL: https://codereview.chromium.org/1747053003 Cr-Commit-Position: refs/heads/master@{#378704}
* Revert of Bootstrap Mojo IPC independent of Chrome IPC (patchset #18 ↵rockot2016-03-011-34/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | id:340001 of https://codereview.chromium.org/1685183004/ ) Reason for revert: Breaking NaCl in a way that somehow was not caught in the CQ... :{ Will reland once that's sorted out. Original issue's description: > Bootstrap Mojo IPC independent of Chrome IPC > > This initializes Mojo IPC very early in child process > startup so we can start using cross-process pipes in > content without a dependency on Chrome IPC. > > On all supported platforms, a new OS pipe handle is > passed through ChildProcessLauncher and hooked up to > the EDK on ChildThreadImpl startup to establish a > Mojo IPC channel to the parent. > > A command-line argument is also to used to share a > unique pipe token with the child, and this token is > exchanged over the Mojo IPC channel to bootstrap a > primordial message pipe. > > While this CL establishes these facilities in a way > that can be used for all child processes, only > render processes are brought up in this way for now. > > BUG=583557 > TEST=Chrome still runs in mash with Mus-in-renderer support. > > Committed: https://crrev.com/eb199fbaecd8e5d294010dfbe8897b7d83aeab07 > Cr-Commit-Position: refs/heads/master@{#378272} TBR=ben@chromium.org,jam@chromium.org,jschuh@chromium.org,pennymac@chromium.org,tsepez@chromium.org # Skipping CQ checks because original CL landed less than 1 days ago. NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=583557 Review URL: https://codereview.chromium.org/1748973003 Cr-Commit-Position: refs/heads/master@{#378373}
* Bootstrap Mojo IPC independent of Chrome IPCrockot2016-02-291-5/+34
| | | | | | | | | | | | | | | | | | | | | | | | | | | This initializes Mojo IPC very early in child process startup so we can start using cross-process pipes in content without a dependency on Chrome IPC. On all supported platforms, a new OS pipe handle is passed through ChildProcessLauncher and hooked up to the EDK on ChildThreadImpl startup to establish a Mojo IPC channel to the parent. A command-line argument is also to used to share a unique pipe token with the child, and this token is exchanged over the Mojo IPC channel to bootstrap a primordial message pipe. While this CL establishes these facilities in a way that can be used for all child processes, only render processes are brought up in this way for now. BUG=583557 TEST=Chrome still runs in mash with Mus-in-renderer support. Review URL: https://codereview.chromium.org/1685183004 Cr-Commit-Position: refs/heads/master@{#378272}
* Windows: Expose handle inheritance to the sandboxed launcherrockot2016-02-261-1/+3
| | | | | | | | | | | | Handles to be inherited by a child process can now be set in a delegate passed to StartSandboxedProcess(), and/or through the TargetPolicy API. TEST=sbox_integration_tests, content_unittests BUG=588190 Review URL: https://codereview.chromium.org/1703953002 Cr-Commit-Position: refs/heads/master@{#378007}
* Pass both 32 and 64 bit snapshot and natives fds to child processes.tobiasjs2016-02-101-1/+28
| | | | | | | | | | | | | | | Child processes are in the best position to determine which files to use, therefore it is simplest just to provide both 32 and 64 bit versions from the parent. BUG=581409,455699 Committed: https://crrev.com/c560d75783aca05249092dd11503b53f7b631be1 Cr-Commit-Position: refs/heads/master@{#374371} Review URL: https://codereview.chromium.org/1665513002 Cr-Commit-Position: refs/heads/master@{#374643}
* Revert of Pass both 32 and 64 bit snapshot and natives fds to child ↵thakis2016-02-091-28/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | processes. (patchset #3 id:40001 of https://codereview.chromium.org/1665513002/ ) Reason for revert: Added a static initializer: https://build.chromium.org/p/chromium/builders/Linux/builds/71576 # v8_initializer.cc _GLOBAL__sub_I_v8_initializer.cc+0xf # v8_initializer.cc __cxa_atexit@plt [registers a dtor to run at exit] Original issue's description: > Pass both 32 and 64 bit snapshot and natives fds to child processes. > > Child processes are in the best position to determine which files > to use, therefore it is simplest just to provide both 32 and 64 > bit versions from the parent. > > BUG=581380,455699 > > Committed: https://crrev.com/c560d75783aca05249092dd11503b53f7b631be1 > Cr-Commit-Position: refs/heads/master@{#374371} TBR=jochen@chromium.org,torne@chromium.org,tobiasjs@chromium.org # Skipping CQ checks because original CL landed less than 1 days ago. NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=581380,455699 Review URL: https://codereview.chromium.org/1681003003 Cr-Commit-Position: refs/heads/master@{#374380}
* Pass both 32 and 64 bit snapshot and natives fds to child processes.tobiasjs2016-02-091-1/+28
| | | | | | | | | | | | Child processes are in the best position to determine which files to use, therefore it is simplest just to provide both 32 and 64 bit versions from the parent. BUG=581380,455699 Review URL: https://codereview.chromium.org/1665513002 Cr-Commit-Position: refs/heads/master@{#374371}
* Have each SandboxedProcessLauncherDelegate maintain a zygote.kerrnel2016-01-221-20/+27
| | | | | | | | | | | | | | | | | | | To improve component updates of PPAPI plugins, Chrome needs multiple zygotes. This will allow the PPAPI zygote to be recreated when a plugin is updated. This CL allows Chrome to maintain a zygote for each process type by having each SandboxedProcessLauncherDelegate maintain a class which can communicate with its respective zygote. This CL will be followed up with work to allow customization of zygotes, which will give Chrome the improve component update experience. BUG=569191 Committed: https://crrev.com/3c1e16b490255119b6f70f94d1716645e897b185 Cr-Commit-Position: refs/heads/master@{#370488} Review URL: https://codereview.chromium.org/1532423003 Cr-Commit-Position: refs/heads/master@{#371042}
* Revert of Have each SandboxedProcessLauncherDelegate maintain a zygote. ↵kerrnel2016-01-211-27/+20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (patchset #15 id:440001 of https://codereview.chromium.org/1532423003/ ) Reason for revert: Revert "Have each SandboxedProcessLauncherDelegate maintain a zygote." This reverts commit 3c1e16b490255119b6f70f94d1716645e897b185. This broke the Site Isolation FYI bot failure: https://build.chromium.org/p/chromium.fyi/builders/Site%20Isolation%20Linux/builds/7700 To re-land the CL, that failure needs to be diagnosed and corrected. Original issue's description: > Have each SandboxedProcessLauncherDelegate maintain a zygote. > > To improve component updates of PPAPI plugins, Chrome needs multiple zygotes. > This will allow the PPAPI zygote to be recreated when a plugin is updated. > This CL allows Chrome to maintain a zygote for each process type by having > each SandboxedProcessLauncherDelegate maintain a class which can communicate > with its respective zygote. This CL will be followed up with work to allow > customization of zygotes, which will give Chrome the improve component update > experience. > > BUG=569191 > > Committed: https://crrev.com/3c1e16b490255119b6f70f94d1716645e897b185 > Cr-Commit-Position: refs/heads/master@{#370488} TBR=mdempsky@chromium.org,thestig@chromium.org,avi@chromium.org,mseaborn@chromium.org # Skipping CQ checks because original CL landed less than 1 days ago. NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=569191 Review URL: https://codereview.chromium.org/1617213002 Cr-Commit-Position: refs/heads/master@{#370771}
* Have each SandboxedProcessLauncherDelegate maintain a zygote.kerrnel2016-01-201-20/+27
| | | | | | | | | | | | | | | | To improve component updates of PPAPI plugins, Chrome needs multiple zygotes. This will allow the PPAPI zygote to be recreated when a plugin is updated. This CL allows Chrome to maintain a zygote for each process type by having each SandboxedProcessLauncherDelegate maintain a class which can communicate with its respective zygote. This CL will be followed up with work to allow customization of zygotes, which will give Chrome the improve component update experience. BUG=569191 Review URL: https://codereview.chromium.org/1532423003 Cr-Commit-Position: refs/heads/master@{#370488}
* Convert Pass()→std::move() for Mac build.dcheng2016-01-131-2/+1
| | | | | | | | | | BUG=557422 R=avi@chromium.org TBR=jam@chromium.org Review URL: https://codereview.chromium.org/1579863003 Cr-Commit-Position: refs/heads/master@{#369135}
* Replace IOSurfaceManager by directly passing IOSurface Mach ports over ↵rsesek2016-01-051-4/+0
| | | | | | | | | | | | | | | | | | Chrome IPC. IOSurfaceManager was created as a mechanism to exchange IOSurfaces using Mach ports, at a time when Chrome IPC was unable to broker Mach port attachments. As part of fixing https://crbug.com/466437, Chrome IPC grew the capability to also also send Mach ports as part of a message. With this new capability, the custom Mach IPC channel used by IOSurfaceManager is no longer necessary, and its complexity can be removed. BUG=569226,323304 Review URL: https://codereview.chromium.org/1532813002 Cr-Commit-Position: refs/heads/master@{#367474}
* Switch to standard integer types in content/browser/.avi2015-12-251-0/+1
| | | | | | | | | BUG=138542 TBR=jam@chromium.org Review URL: https://codereview.chromium.org/1549113002 Cr-Commit-Position: refs/heads/master@{#366868}
* Change callers of ScopedGeneric::Pass() to std::move()danakj2015-12-031-7/+9
| | | | | | | | | R=Nico BUG=557422 Review URL: https://codereview.chromium.org/1467413007 Cr-Commit-Position: refs/heads/master@{#362853}
* Remove jank profiling from Context::Notify.erikchen2015-11-201-22/+0
| | | | | | | | | | Profiling is not being followed up on, so should be removed. BUG=465841 Review URL: https://codereview.chromium.org/1459263003 Cr-Commit-Position: refs/heads/master@{#360899}
* Bind Application in renderer.ben2015-11-191-59/+0
| | | | | | | | | | | | | | | | | | | This involved changing how we get the client handle to the renderer. In the first iteration I was passing this on the command line but that turns out not to work with the sandbox. So instead I an approach used by the Mojo-in-Chrome MojoApplication class and pass the primordial handle via Chrome IPC. I had to twiddle a bunch of BUILD.gn files in content to get this to work without crashing due to inconsistencies in how MOJO_SHELL_CLIENT was defined. R=jam@chromium.org,tsepez@chromium.org http://crbug.com/551253 Committed: https://crrev.com/3edb97198bc5fbc22c5cf13286e8af80449ddfb9 Cr-Commit-Position: refs/heads/master@{#360293} Committed: https://crrev.com/2c716f9d5ebff610641f4506c17ec81d4b89b7bb Cr-Commit-Position: refs/heads/master@{#360396} Review URL: https://codereview.chromium.org/1452823003 Cr-Commit-Position: refs/heads/master@{#360550}
* Revert of Bind Application in renderer (patchset #15 id:280001 of ↵alancutter2015-11-181-0/+59
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | https://codereview.chromium.org/1452823003/ ) Reason for revert: This change is making hundreds of webkit_tests crash on linux_chromium_rel_ng with: [1:1:1118/141621:108326049405:FATAL:lock.cc(32)] Check failed: owning_thread_ref_.is_null(). #0 0x00000067e2de base::debug::StackTrace::StackTrace() #1 0x00000057a71f logging::LogMessage::~LogMessage() #2 0x0000005eb69b base::Lock::CheckUnheldAndMark() #3 0x000000518e1c base::Lock::Acquire() #4 0x000000518b43 base::AutoLock::AutoLock() #5 0x0000005b4014 base::SequenceCheckerImpl::CalledOnValidSequencedThread() #6 0x000006f074ae IDMap<>::Lookup() #7 0x000006f0598c content::RendererBlinkPlatformImpl::SetPlatformEventObserverForTesting() #8 0x000008459264 content::SetMockGamepadProvider() Original issue's description: > Bind Application in renderer. > > This involved changing how we get the client handle to the renderer. In the first iteration I was passing this on the command line but that turns out not to work with the sandbox. So instead I an approach used by the Mojo-in-Chrome MojoApplication class and pass the primordial handle via Chrome IPC. > > I had to twiddle a bunch of BUILD.gn files in content to get this to work without crashing due to inconsistencies in how MOJO_SHELL_CLIENT was defined. > > R=jam@chromium.org,tsepez@chromium.org > http://crbug.com/551253 > > Committed: https://crrev.com/3edb97198bc5fbc22c5cf13286e8af80449ddfb9 > Cr-Commit-Position: refs/heads/master@{#360293} > > Committed: https://crrev.com/2c716f9d5ebff610641f4506c17ec81d4b89b7bb > Cr-Commit-Position: refs/heads/master@{#360396} TBR=jam@chromium.org,tsepez@chromium.org,ben@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true Review URL: https://codereview.chromium.org/1457173002 Cr-Commit-Position: refs/heads/master@{#360446}
* Bind Application in renderer.ben2015-11-181-59/+0
| | | | | | | | | | | | | | | | This involved changing how we get the client handle to the renderer. In the first iteration I was passing this on the command line but that turns out not to work with the sandbox. So instead I an approach used by the Mojo-in-Chrome MojoApplication class and pass the primordial handle via Chrome IPC. I had to twiddle a bunch of BUILD.gn files in content to get this to work without crashing due to inconsistencies in how MOJO_SHELL_CLIENT was defined. R=jam@chromium.org,tsepez@chromium.org http://crbug.com/551253 Committed: https://crrev.com/3edb97198bc5fbc22c5cf13286e8af80449ddfb9 Cr-Commit-Position: refs/heads/master@{#360293} Review URL: https://codereview.chromium.org/1452823003 Cr-Commit-Position: refs/heads/master@{#360396}
* Revert of Bind Application in renderer (patchset #14 id:260001 of ↵msramek2015-11-181-0/+59
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | https://codereview.chromium.org/1452823003/ ) Reason for revert: Speculatively reverting this, as this CL landed exactly at the time when renderer started crashing on linux_chromium_rel_ng. Original issue's description: > Bind Application in renderer. > > This involved changing how we get the client handle to the renderer. In the first iteration I was passing this on the command line but that turns out not to work with the sandbox. So instead I an approach used by the Mojo-in-Chrome MojoApplication class and pass the primordial handle via Chrome IPC. > > I had to twiddle a bunch of BUILD.gn files in content to get this to work without crashing due to inconsistencies in how MOJO_SHELL_CLIENT was defined. > > R=jam@chromium.org,tsepez@chromium.org > http://crbug.com/551253 > > Committed: https://crrev.com/3edb97198bc5fbc22c5cf13286e8af80449ddfb9 > Cr-Commit-Position: refs/heads/master@{#360293} TBR=jam@chromium.org,tsepez@chromium.org,ben@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=557732 Review URL: https://codereview.chromium.org/1460713002 Cr-Commit-Position: refs/heads/master@{#360331}
* Bind Application in renderer.ben2015-11-181-59/+0
| | | | | | | | | | | | | This involved changing how we get the client handle to the renderer. In the first iteration I was passing this on the command line but that turns out not to work with the sandbox. So instead I an approach used by the Mojo-in-Chrome MojoApplication class and pass the primordial handle via Chrome IPC. I had to twiddle a bunch of BUILD.gn files in content to get this to work without crashing due to inconsistencies in how MOJO_SHELL_CLIENT was defined. R=jam@chromium.org,tsepez@chromium.org http://crbug.com/551253 Review URL: https://codereview.chromium.org/1452823003 Cr-Commit-Position: refs/heads/master@{#360293}
* Move Shell connection to content.ben2015-11-151-1/+61
| | | | | | | | | | | | | | | | | | I want to be able to use Mojo in content, so I think it makes sense that the shell bind step lives in content too. 1. Moves MojoRunnerState to content/common and renames it to MojoShellConnection. This class will be available in any process that creates an impl prior to running the main message loop. 2. Modifies ChildProcessLauncher to create a platform channel pair and put the client handle on the command line of the child process, and registers the server handle with the external shell. This will allow the child process to bind an Application request if it wants. TODO: What if the child process doesn't bind it? What happens to the instance created in the external shell? TODO: ChildProcessLauncher is a convenient chokepoint. It means the shell handle is passed to all child processes created by content. Do we want to do this? I think it's OK for now since this code is only triggered if Chrome itself is run from within the external shell. Before we can move this to production I think CreateInstanceForHandle is going to have to take a CapabilityFilter. R=jam@chromium.org http://crbug.com/551253 Review URL: https://codereview.chromium.org/1442893002 Cr-Commit-Position: refs/heads/master@{#359770}
* Stop warnings resulting from unimplemented SetProcessBackgrounded().shrike2015-10-021-1/+3
| | | | | | | | | | | | | | | | | Process::SetProcessBackgrounded() is not implemented on certain platforms (e.g. OS X). Calling Process::SetProcessBackgrounded() on these systems results in a warning to the console. Callers should first check result of Process::CanBackgroundProcesses() before calling SetProcessBackgrounded(). This cl also fixes a small error in the documentation for SetProcessBackgrounded(). BUG=537301 Review URL: https://codereview.chromium.org/1378753002 Cr-Commit-Position: refs/heads/master@{#351943}
* Revert hidden renderer backgrounding on the Mac.shrike2015-09-241-8/+0
| | | | | | | | | | | Backgrouning hidden renderers on the Mac has had unexpected performance regressions. This cl removes the code that enables backgrounding. BUG=534908 Review URL: https://codereview.chromium.org/1358173002 Cr-Commit-Position: refs/heads/master@{#350611}
* Define a bootstrap sandbox policy for renderer processes.rsesek2015-09-181-7/+10
| | | | | | | | | | | | This also refactors the //content-level integration. Note that the sandbox is not yet enabled. BUG=367863,501128 R=avi@chromium.org Review URL: https://codereview.chromium.org/1351893003 Cr-Commit-Position: refs/heads/master@{#349735}
* Refactor the bootstrap sandbox process launching integration.rsesek2015-09-181-10/+12
| | | | | | | | | | | | | | | | | | | | There are three changes: - A LaunchOptions::PreExecDelegate is now used to perform the bootstrap port replacement in the new child. This removes sandbox-specific knowledge from //base. - The replacement bootstrap port is no longer registered with launchd. Instead, a new sandbox manager port is registered. Clients communicate with this server to get the replacement bootstrap port. - Using the above port, clients now perform a post-fork-pre-exec handshake to check in with the sandbox server. This removes the complicated PrepareToFork/FinishedFork interface. BUG=367863,388214 R=mark@chromium.org Review URL: https://codereview.chromium.org/1346923006 Cr-Commit-Position: refs/heads/master@{#349571}
* Cleanup: Remove unneeded base/memory/shared_memory.h #includes.thestig2015-09-151-2/+1
| | | | | | | | | | CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel TBR=tsepez@chromium.org Review URL: https://codereview.chromium.org/1341153003 Cr-Commit-Position: refs/heads/master@{#348961}
* Add new termination status for failed child process launch.wfh2015-09-021-0/+1
| | | | | | | | | | | | | | Treat renderer startup failure as a renderer crash for metrics. This makes situations where sandbox failures are happening en-masse without crash reports being produced easier to spot on the stability dashboard. BUG=526198,467920 Review URL: https://codereview.chromium.org/1320153002 Cr-Commit-Position: refs/heads/master@{#347060}
* Always use ENABLE_PLUGINS to indicate plugins support.hshi2015-08-291-0/+3
| | | | | | | | | | | | | | | | | We should rely on GYP_DEFINE "enable_plugins" in gyp files and C++ macro "ENABLE_PLUGINS" in source code to indicate plugins support. This removes hard-coded assumptions that certain OS does (or does not) support plugins. Also add GYP_DEFINE "enable_pdf" and macro "ENABLE_PDF" to indicate built-in PDF viewer plugin support. BUG=490433 TEST=trybot TBR=sievers@chromium.org Review URL: https://codereview.chromium.org/1318143002 Cr-Commit-Position: refs/heads/master@{#346284}
* Load non-locale .pak files directly from the .apk on Android (rather than ↵agrieve2015-06-241-1/+5
| | | | | | | | | | extracting on start-up). BUG=394502 Review URL: https://codereview.chromium.org/1181953002 Cr-Commit-Position: refs/heads/master@{#335964}
* Take 2: Moved logic for mapping child process FDs for ICU and V8 into ↵agrieve2015-06-191-10/+37
| | | | | | | | | | | | | | | | | | child_process_launcher.cc Used to be defined in each app's ContentBrowserClient, but since content/ is the one that receives the FDs, it makes sense that it should be the one to send them. This also removes ChildProcessLauncher::AppendMappedFileCommandLineSwitches as it is no longer needed. Changes MemoryMappedFile::Region to be a POD so that it doesn't create require static initializers. BUG=394502 Review URL: https://codereview.chromium.org/1185973003 Cr-Commit-Position: refs/heads/master@{#335207}
* Revert of Moved logic for mapping child process FDs for ICU and V8 into ↵mek2015-06-161-37/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | child_process_launcher.cc (patchset #20 id:380001 of https://codereview.chromium.org/1182443003/) Reason for revert: Adds new static initializers in http://build.chromium.org/p/chromium/buildstatus?builder=Linux%20x64&number=4979 # icu_util.cc cc::VertexShaderQuadAA::VertexShaderQuadAA() # icu_util.cc base::MemoryMappedFile::MemoryMappedFile() # icu_util.cc base::i18n::(anonymous namespace)::g_icudtl_region # icu_util.cc base::i18n::(anonymous namespace)::g_icudtl_mapped_file # icu_util.cc operator new(unsigned long) Original issue's description: > Moved logic for mapping child process FDs for ICU and V8 into child_process_launcher.cc > > Used to be defined in each app's ContentBrowserClient, but since > content/ is the one that receives the FDs, it makes sense that it should > be the one to send them. > > This also removes ChildProcessLauncher::AppendMappedFileCommandLineSwitches > as it is no longer needed. > > BUG=394502 > > Committed: https://crrev.com/228414fc8870f88f11ada7512e88ea6999890f56 > Cr-Commit-Position: refs/heads/master@{#334702} TBR=jochen@chromium.org,jam@chromium.org,jungshik@google.com,michaelbai@chromium.org,rmcilroy@chromium.org,thestig@chromium.org,agrieve@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=394502 Review URL: https://codereview.chromium.org/1187213002 Cr-Commit-Position: refs/heads/master@{#334719}
* Moved logic for mapping child process FDs for ICU and V8 into ↵agrieve2015-06-161-10/+37
| | | | | | | | | | | | | | | | | child_process_launcher.cc Used to be defined in each app's ContentBrowserClient, but since content/ is the one that receives the FDs, it makes sense that it should be the one to send them. This also removes ChildProcessLauncher::AppendMappedFileCommandLineSwitches as it is no longer needed. BUG=394502 Review URL: https://codereview.chromium.org/1182443003 Cr-Commit-Position: refs/heads/master@{#334702}
* Allow startup with missing V8 snapshot file.erikcorry2015-06-081-0/+5
| | | | | | | | | | | | | | | | | | We want to stop shipping the snapshot file, and instead we want to generate it on the client. This will reduce the download size. But since snapshot generation will be asynchronous in a utility process, it might not be present on the first few runs of the browser. This means we have to be able to start up without the snapshot file (just with the natives source file). This CL fixes Blink to cope with a missing snapshot file (V8 could already cope). R=rmcilroy@chromium.org, sky@chromium.org BUG= Review URL: https://codereview.chromium.org/1164483003 Cr-Commit-Position: refs/heads/master@{#333258}
* content: Pass IOSurface references using Mach IPC.reveman2015-06-041-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This removes the use of global IOSurfaces and instead passes ownership between processes using Mach IPC. The IOSurface GpuMemoryBuffer factory instance in the GPU process sends a synchronous Mach message to the browser process to register each IOSurface it creates. IOSurface registration messages are handled by the BrowserIOSurfaceManager class and child processes can use a Mach message to acquire a reference to an IOSurface that has been registered with the manager. The BrowserIOSurfaceManager class keeps track of the ownership of each IOSurface and prevents a child process from acquiring a reference to an IOSurface that it doesn't own. A unique unguessable token is generated for each child process that is allowed to use IOSurfaces. The token restricts what IOSurfaces a child process has access to and prevents a malicious process from gaining access to IOSurfaces it doesn't own. Security Considerations ----------------------- In general, this is a major improvement to security as it provides proper sand-boxing of IOSurfaces. Prior to this change, IOSurfaces were global and any process on the system (including all renderer processes) had access to all IOSurfaces. The renderer who owns the IOSurface is the only process (except for the browser and GPU) that has access to the IOSurface as a result of this change. Passing of IOSurface references to child processes require a Mach port to be open in the child process sandbox for sending messages to the browser. As a result, Mach message handling in the browser process (BrowserIOSurfaceManager::Handle*Request) requires validation and proper error handling to prevent a malicious renderer from exploiting this channel. BUG=323304 TEST=content_unittests --gtest_filter=GpuMemoryBuffer*/1, content_unittests --gtest_filter=BrowserIOSurfaceManagerTest.*, content_shell --enable-native-gpu-memory-buffers Review URL: https://codereview.chromium.org/1137453002 Cr-Commit-Position: refs/heads/master@{#332757}
* Pass file Regions along with FDs to child processes on Androidagrieve2015-06-031-2/+2
| | | | | | | | | | | | This is just sending kWholeFile for every FD atm, but will be used in the future to load v8 snapshots, icu data, and .pak files from the APK without needing to extract them first. BUG=394502 Review URL: https://codereview.chromium.org/1156183003 Cr-Commit-Position: refs/heads/master@{#332608}
* Add support for backgrounding processes on the Macshrike2015-06-021-1/+9
| | | | | | | | | | | | | | | | | Added process_mac.cc with implementations of IsProcessBackgrounded() and SetProcessBackgrounded(). BUG=460102 Originally Committed: https://crrev.com/e3bb10f7860a1d553c85293bd7d7615c0e7f0fd9 Reverted: https://crrev.com/ce6226a7ffe2c1cb7ac5f6cf34b56b8d217686b9 Second commit: https://crrev.com/0160d130f8a4462fa7bfb8a9924e476d31ba9a48 Second revert: https://crrev.com/93ef7cd278d450b06f4a95fad6577d05b67624aa Review URL: https://codereview.chromium.org/989703002 Cr-Commit-Position: refs/heads/master@{#332454}
* Revert of Add support for backgrounding processes on the Mac (patchset #22 ↵tnagel2015-05-191-9/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | id:420001 of https://codereview.chromium.org/989703002/) Reason for revert: This CL seems to be the cause of ExtensionApiNewTabTest.Tabs failures on Mac. Original issue's description: > Add support for backgrounding processes on the Mac > > Added process_mac.cc with implementations of IsProcessBackgrounded() and > SetProcessBackgrounded(). > > BUG=460102 > > Originally Committed: https://crrev.com/e3bb10f7860a1d553c85293bd7d7615c0e7f0fd9 > Reverted: https://crrev.com/ce6226a7ffe2c1cb7ac5f6cf34b56b8d217686b9 > > Committed: https://crrev.com/0160d130f8a4462fa7bfb8a9924e476d31ba9a48 > Cr-Commit-Position: refs/heads/master@{#330464} TBR=rsesek@chromium.org,gab@chromium.org,thakis@chromium.org,asvitkine@chromium.org,avi@chromium.org,kbr@chromium.org,shrike@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=460102 Review URL: https://codereview.chromium.org/1142913004 Cr-Commit-Position: refs/heads/master@{#330519}
* Add support for backgrounding processes on the Macshrike2015-05-191-1/+9
| | | | | | | | | | | | | | Added process_mac.cc with implementations of IsProcessBackgrounded() and SetProcessBackgrounded(). BUG=460102 Originally Committed: https://crrev.com/e3bb10f7860a1d553c85293bd7d7615c0e7f0fd9 Reverted: https://crrev.com/ce6226a7ffe2c1cb7ac5f6cf34b56b8d217686b9 Review URL: https://codereview.chromium.org/989703002 Cr-Commit-Position: refs/heads/master@{#330464}
* Revert of Add support for backgrounding processes on the Mac (patchset #21 ↵thakis2015-05-171-9/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | id:400001 of https://codereview.chromium.org/989703002/) Reason for revert: Speculative; several browser_tests started failing on the Mac 10.9 bots, and this looks like the most likely culprit: http://build.chromium.org/p/chromium.mac/builders/Mac10.9%20Tests/builds/1929 http://build.chromium.org/p/chromium.mac/builders/Mac10.9%20Tests%20%28dbg%29/builds/6014 Original issue's description: > Add support for backgrounding processes on the Mac > > Added process_mac.cc with implementations of IsProcessBackgrounded() and > SetProcessBackgrounded(). > > BUG=460102 TBR=rsesek@chromium.org,gab@chromium.org,asvitkine@chromium.org,avi@chromium.org,kbr@chromium.org,shrike@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=460102 Review URL: https://codereview.chromium.org/1128183011 Cr-Commit-Position: refs/heads/master@{#330294}
* Add support for backgrounding processes on the Macshrike2015-05-161-1/+9
| | | | | | | | | | | Added process_mac.cc with implementations of IsProcessBackgrounded() and SetProcessBackgrounded(). BUG=460102 Review URL: https://codereview.chromium.org/989703002 Cr-Commit-Position: refs/heads/master@{#330275}
* Simplify ChildProcessLaunchersievers2015-03-281-385/+250
| | | | | | | | | | | | | Remove the refcounted internal state object (nested class 'Context'). This refactor makes it more obvious what happens on what thread, and avoids the need to pass refptrs around. TBR=bradnelson@chromium.org BUG=469248 Review URL: https://codereview.chromium.org/1022703007 Cr-Commit-Position: refs/heads/master@{#322695}
* Drop queued IPC messages if the child process dies while spawning.lfg2015-03-231-0/+11
| | | | | | | | BUG=464633 Review URL: https://codereview.chromium.org/1018043004 Cr-Commit-Position: refs/heads/master@{#321878}
* Revert of Add DCHECKS() for current thread in ChildProcessLauncher (patchset ↵tkent2015-03-231-66/+57
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | #4 id:60001 of https://codereview.chromium.org/1002853003/) Reason for revert: Broke linux_blink_rel try bots. http://build.chromium.org/p/tryserver.blink/builders/linux_blink_rel/builds/53723/steps/webkit_tests%20%28with%20patch%29/logs/stdio > [20222:20222:0322/165944:1212144529:FATAL:child_process_launcher.cc(508)] Check failed: ::content::BrowserThread::CurrentlyOn(client_thread_id_). Must be called on Chrome_UIThread; actually called on CrBrowserMain. Original issue's description: > Add DCHECKS() for current thread in ChildProcessLauncher > > And rename some functions to indicate when they don't run > on the default (client) thread. > > This avoids misuse and makes it more obvious what threads we are on > given that this involves multiple threads (launcher thread, and > different client threads for renderers and other children - former > calls from UI thread, the latter from IO thread - > plus there are some Android quirks). > > BUG=469248 > > Committed: https://crrev.com/8efc3b5a5e7b24ffeed0e5576a725b77868eb9ab > Cr-Commit-Position: refs/heads/master@{#321648} TBR=rvargas@chromium.org,sievers@chromium.org NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=469248 Review URL: https://codereview.chromium.org/1025923002 Cr-Commit-Position: refs/heads/master@{#321723}
* Add DCHECKS() for current thread in ChildProcessLaunchersievers2015-03-201-57/+66
| | | | | | | | | | | | | | | | | And rename some functions to indicate when they don't run on the default (client) thread. This avoids misuse and makes it more obvious what threads we are on given that this involves multiple threads (launcher thread, and different client threads for renderers and other children - former calls from UI thread, the latter from IO thread - plus there are some Android quirks). BUG=469248 Review URL: https://codereview.chromium.org/1002853003 Cr-Commit-Position: refs/heads/master@{#321648}
* Add profiling to Context::Notify.erikchen2015-03-111-0/+22
| | | | | | | | BUG=465841 Review URL: https://codereview.chromium.org/995113002 Cr-Commit-Position: refs/heads/master@{#320034}
* Redefine base::Process:Terminate so that it can replace base::KillProcessrvargas2015-03-091-1/+1
| | | | | | | | BUG=417532 Review URL: https://codereview.chromium.org/983963002 Cr-Commit-Position: refs/heads/master@{#319776}
* removing all the code related to chrome://statscpu2015-02-121-8/+0
| | | | | | | | | | | | | this is all broken and has been for a while. Since then, 2012 to 2013 other, sytem-specific systems have been used. Note that there is a v8 test that is using the stats table. Hoping to sort this now. BUG=149742 Review URL: https://codereview.chromium.org/899753002 Cr-Commit-Position: refs/heads/master@{#315890}
* Enable clean child process exit in sanitizer builds on ChromeOS.earthdok2015-02-091-4/+3
| | | | | | | | | | | | This was blocked on a WebRTC issue for a long time. However, latest trybot runs appear green. BUG=360622, 324216, 307043 TBR=jam@chromium.org Review URL: https://codereview.chromium.org/227533008 Cr-Commit-Position: refs/heads/master@{#315350}
* Rename base::Process::pid() to Pid()rvargas2015-01-241-1/+1
| | | | | | | | | | | | | | | | | The method is not a simple accessor. BUG=417532 TBR=satorux@chromium.org (chromeos) TBR=siggi@chromium.org (browser_watcher) TBR=mseaborn@chromium.org (nacl) TBR=davidben@chromium.org (content) TBR=rsesek@chromium.org (sandbox) TBR=sadrul@chromium.org (ui) Review URL: https://codereview.chromium.org/864163003 Cr-Commit-Position: refs/heads/master@{#312978}
generated by cgit v1.1 at 2024-06-09 10:18:18 +0000