I am building chrome m99 milestone and cannot seem to find where it builds the webrtc library which it uses. when you goto webrtc.org its all chrome javascript examples of its usage case, but in reality native servers must stream this to make real use of it for low latency apps. certainly we arent packaging chrome apps on the server for this purpose, nor would it be possible to actually access video buffers efficiently using that.
i've attempted to build with targets webrtc, //gen/third_party/webrtc, i've attempted to gen directly in the webrtc dir (in third_party) which has a BUILD.gn, but you cannot. it errors and says there are missing build files:
C:\src\chromium\src\third_party\webrtc>gn gen out/Debug
ERROR at //.gn:9:1: Unable to load "C:/src/chromium/src/third_party/webrtc/build/dotfile_settings.gni".
import("//build/dotfile_settings.gni")
I see in its BUILD.gn it has a 'build_with_chromium' boolean but does not seem any way to do that. the last doc i saw on this was from 2017 in a blog post, and the libwebrtc project he used is no longer maintained. the guy went commercial but essentially it gloms together a bunch of .o's into a lib but none of the third party structure is the same anymore in m99.
inside chromium checkout:
# fetch branch head info so that we can checkout m99
# https://chromiumdash.appspot.com/branches (4844 branch ref = m99)
gclient sync --with_branch_heads --with_tags
git fetch
git checkout -b branch_4844 branch-heads/4844
gclient sync --with_branch_heads --with_tags
# use depot_tools of our own (in PATH)
set DEPOT_TOOLS_WIN_TOOLCHAIN=0
# is_component_build=true disables webrtc as i'm noticing in the main build in chrome, so i set it to false to be sure (i believe its false by default)
gn gen out/Default --args="is_component_build=false symbol_level=0 is_debug=true"
autoninja -C out/Default chrome
since chrome uses webrtc it should be built, but who knows where the objs are?
in libwebrtc (no longer maintained) circa m62 or so they selected these binaries for producing a lib:
file(GLOB_RECURSE _OBJ_EXCLUDED
${WEBRTC_OUTPUT_DIR}/obj/third_party/yasm/gen*/*${_OBJ_EXT}
${WEBRTC_OUTPUT_DIR}/obj/third_party/yasm/re2c/*${_OBJ_EXT}
${WEBRTC_OUTPUT_DIR}/obj/third_party/yasm/yasm/*${_OBJ_EXT}
${WEBRTC_OUTPUT_DIR}/obj/third_party/protobuf/protoc/*${_OBJ_EXT}
${WEBRTC_OUTPUT_DIR}/obj/third_party/protobuf/protobuf_full/*${_OBJ_EXT}
${WEBRTC_OUTPUT_DIR}/obj/webrtc/examples/*${_OBJ_EXT}
${WEBRTC_OUTPUT_DIR}/obj/webrtc/tools/*${_OBJ_EXT}
${WEBRTC_OUTPUT_DIR}/obj/webrtc/modules/video_capture/video_capture/video_capture_external${_OBJ_EXT}
${WEBRTC_OUTPUT_DIR}/obj/webrtc/modules/video_capture/video_capture/device_info_external${_OBJ_EXT})
it was made at a time where webrtc could be built on its own, but it just does not seem to be the case anymore unless there is a magical '1-2-3' that could build the hard work of these great
people who seem to not want to expose a library?
I've attempted to give it webrtc, /third_party/webrtc, //gen/third_party/webrtc. no avail it is just completely inaccessible as-is.
openwebrtc is not maintained, nor is libwebrtc which glommed this all together.
Related
I couldn't figure out a nice way to determine a suitable revision number to use with Puppeteer Sharp.
What I did was use the version lookup feature on the "OmahaProxy - Google Chrome" site. I looked-up the version of Chrome I'm running on my computer. [That seems like a reasonable starting point.] I assumed (guessed) that the "Branch Base Position" shown in the version info was a revision number.
I then opened the Chromium continuous builds archive and looked for a build for the revision around the revision number I found from the "OmahaProxy" site.
Is there a better way to find or pick a suitable revision number?
Puppeteer is always bundled with a specific revision of a specific version. I usually check the release information on Github where the expected Chromium version and revision is specified. For example:
v1.17.0
Big Changes
Chromium 76.0.3803.0 (r662092)
Then to download the right one,
Go to Chromium browser snapshots
Choose the directory of your platform (e.g., Linux_x64)
Copy the revision number into the "Filter:" field without the "r" (e.g., 662092)
Download the .zip file you need.
Additional info
The URL template below can be just plugged in with the right information:
https://commondatastorage.googleapis.com/chromium-browser-snapshots/index.html?prefix=<platform>/<revision>/
For example:
https://commondatastorage.googleapis.com/chromium-browser-snapshots/index.html?prefix=Linux_x64/662092/
The most common issue on Linux is missing dependencies, and the Puppeteer "Troubleshooting" document specifies all the Linux dependencies. (See apt instructions for Ubuntu.)
curl needs -L when downloading on the console. See Stackoverflow thread.
When one is using puppeteer-core,
you will then need to call puppeteer.connect(\[options\]) or
puppeteer.launch(\[options\]) with an explicit executablePath
option.
(from puppeteer vs puppeteer-core)
Puppeteer requires number which correspond Chromium browser build snapshot number. You can get the latest snapshot number here:
https://www.googleapis.com/download/storage/v1/b/chromium-browser-snapshots/o/Win_x64%2FLAST_CHANGE?generation=1528272852044880&alt=media
or you can view all available snapshots here:
https://commondatastorage.googleapis.com/chromium-browser-snapshots/index.html?prefix=Win_x64/
I can see my page get crash(see aw, snap page) with 20% proprobility after 10 mins(otherwise it runs well like forever)
so I tried:
1) CPU and memory check with task manager, and see no increasing(so no leakage).
2) enable crush log in the chrome://settings/
result:
2.1) see still nothing in the chrome://crashes page, not even a crush ID (0 crashes).
2.2) see nothing in the folder under path
C:/%User%/AppData/Local/Google/CrashReports (nothing in) nor
C:/%User%/AppData/Local/Google/Chrome/User Data/Crash Reports (folder not exist)
2.3) but indeed see DMP in the:
C:/%User%/AppData/Local/Google/Chrome/User Data/CrashPads/reports
but seems they are not readable, and it also seems not the correct address for crash logs
3) can get chrome log either by command line arguments, or using sawbuck, but found nothing but only 2 errors, one for sawbuck itself, and another saying can't send the report to google.
So the questions are:
1) are those DMP the crash logs(the default Dir for dump file has been changed for chrome v50)
2) how can I abstract information out of the DMP file, if chrome://crashes page shows nothing (for chrome on windows)
p.s. 2 usage pages are found at https://www.chromium.org/developers/decoding-crash-dumps
https://www.chromium.org/developers/crash-reports
but seems it's not for windows without a recompile of chrome's component, is there any 3rd party tools to interpret the DMP file?
env informations:
chrome version: 50.0.2661.02 m
; Host OS: windows 10
The crash dumps (.dmp files) in C:\Users\<user>\AppData\Local\Google\Chrome\User Data\Crashpad\reports can be read by standard Windows debuggers. WinDbg is one tool (provided by Microsoft) for analysing these dumps; it's not going to win any beauty contents, but it's powerful and gets the job done. The recommended way to obtain it is, somewhat bizarrely, the Windows Driver Kit.
You'll need debugging symbols to make sense of the results, and these aren't included in standard builds of Chrome. To get symbols for both Chrome and the Windows runtime, set the following as your Symbols path:
SRV*c:\symbols*https://msdl.microsoft.com/download/symbols;SRV*c:\symbols*https://chromium-browser-symsrv.commondatastorage.googleapis.com
There are numerous resources on using WinDbg on the web; this cheat sheet contains some useful commands to get you started.
Is there a tool / process by which I can decompile two .exes (one of which runs on a handheld device, the other which doesn't) so that I can get a glimpse into what differs/what the problem may be?
Of course, seeing that one has "00xA" where the other has "00xB" won't help me. I mean a way to see code that differs, or more likely, a compiled resource or config file difference, or some build option or so?
I'm almost positive the problem has nothing to do with code per se (if/while/switch statements, etc.) but with the project setup.
I cannot compare the working source, as it does not exist - I only have the .exe from which it came. The source I have available is a "reasonable facsimile" of it, but not exactly the same, at least not in project settings and configurations, etc.
UPDATE
Re: this page, I'm not familiar with some of the terminology; specifically, where it talks about a "FLATRELEASEDIR" - WTH?
Will this work to test my CE-bound* .exe on my PC:
Run depends.exe*
Open my CE-bound** .exe
Stare in wonder at the amazing plethora of errata and data
?
*Supposedly not to be confused with "adult diapers"
** "bound" as in, "that's where it's headed"
UPDATE 2
Here's what I see when running Dependency Walker on my app:
...but note my dilemma here.
UPDATE 3
Here's what I did to try to replicate my handheld environment on my PC as best as possible:
I wanted to just copy over the whole dad-burned thing (the entire handheld device representation from Windows Explorer), but it won't let me do that - it's not copyable, so I copied over just the "Program Files" folder (which contains the subfolder of my app and a couple of other folders which I don't know are standard or custom/peculiar to us).
This is from the handheld device where the .exe in question does run properly, so obviously (theoretically?) there should be no missing modules - Dependency Walker should run it and say all is well.
However, it doesn't; I get the same results as the one p[a,o]sted above.
Note: This is true even after selecting "Options > Configure Module Search Order... > The Application Directory > Move Up in Dependency Walker.
So since even an .exe that runs fine tests out as being "not ready for CE-time", how can I use this to accurately test which, if any, necessary files are missing?
Do I need to copy over other folders, too, or what?
UPDATE 4
So, the (supposedly?) missing modules are:
API-MS-WIN-APPMODEL-RUNTIME-L1-1-0.DLL
API-MS-WIN-CORE-WINRT-ERROR-L1-1-0.DLL
API-MS-WIN-CORE-WINRT-L1-1-0.DLL
API-MS-WIN-CORE-WINRT-ROBUFFER-L1-1-0.DLL
API-MS-WIN-CORE-WINRT-STRING-L1-1-0.DLL
API-MS-WIN-SHCORE-SCALING-L1-1-1.DLL
DCOMP.DLL
GPSVC.DLL
IESHIMS.DLL
If true, then why does the .exe run on the handheld?
Note: Other folders on the handheld are:
Application Data => only contains "desktop.ini"
BTExplorer
My Documents => contains a "desktop.ini" and several SQLCE databases (*.SDF)
profiles => contains an empty "default" folder
Recycled => contains a single ".lnk" file
Temp => contains a handful of ".tmp" files
Windows => contains a slew of folders and files, including several .DLLs, but none of those listed above
The "Program Files" folder has a ".NET Compact Framework" subfolder, but it is empty.
In the \Windows\AppMgr folder, there are some interesting files, such as:
Microsoft .NET Compact Framework.DAT
Microsoft .NET Compact Framework.DLL
...several other DAT/DLL pairs, including:
Symbol Managed Class Libraries.DAT (no corresponding DLL)
...and even a pair:
SSCS HHS.[DAT,DLL]
-- which is our company acronym (SSCS) and name of the project (HHS)
What generates these? From where are they referenced??
Another possible piece of the puzzle is that the handheld device on which the .exe does not work does not have a \Windows\AppMgr folder, and thus does not have those .DAT/.DLL files listed above...
UPDATE 5
Another oddity:
After reading this: "Files that install in the GAC go in the\Windows directory" in the book "Microsoft .NET Compact Framework" by Andy Wiggly, et al, I tried copying the Windows folder from the handheld device to the spot on my PC below the folder named "3910ProgramFilesFolderCopy" which looks like this in Explorer:
C:
3910ProgramFilesFolderCopy
Program Files
Windows <= tried to copy the "Windows" folder from the handeld on top of "3910ProgramFilesFolderCopy" so it would end up here
...but it only copies a few files before borborygmas occurs (electronic borborygmas, that is) and it tells me, "the system cannot find the file specified"
Heavens to Murgratroid!?! Why does it find some but not all? Looking at the Windows folder on the handheld in Windows Explorer, I see that many of the precious DLLs are greyed out:
Arggh, double-arg, and (&c)
For PC you could try .NET Reflector.
http://www.red-gate.com/products/dotnet-development/reflector/
You may try to use dependency walker to undestand wich DLL/APIs they need and check if those are supported on your device.
http://geekswithblogs.net/WindowsEmbeddedCookbook/archive/2011/02/02/using-dependency-walker.aspx
I'll explain what i'm trying to do (and why) first and then get more into the details.
I'm trying to get Registration Free COM activation working in the context of a language that is compiled JIT. I use Reg-Fre COM because i would like to avoid having to register my COM component and would like this application to be installable without administrative rights.
The JIT compiler is the main executable is signed and contains an embeded manifest. Latest versions of Windows prefer the embeded manifest over an external one, by default (it was the opposite on Windows XP).
Since the JIT is signed, i cannot extract/modify/update it's manifest without breaking the signing. Also, this would introduce complexity in the maintenance of our application (always making sure we embed a new manifest). And there's are also the legal implications of modifying a program for which i don't own the rights.
So, my idea was to try to get the WinSxS activation going thru a Win32 wrapper library for which i would provide a manifest that lists the dependencies. As a proof of concept, i decided to make a simple VB6 app, a C++ Wrapper Library and two COM library (1 in .NET and 1 in VB6).
The wrapper contains 3 methods, one that calls the VB6 COM, one that calls the .NET COM and one that returns a simple string.
As long as everything as my COM are registered, it works, of course.
If i unregister them, provide manifests for the COMs and add them as dependencies to the main executable's manifest, it works.
If i unregister them, provide manifests for the COMs and the C++ Wrapper and add the com as dependencies for the C++ Wrapper and then add the C++ Wrapper as the only dependency in the main executable's manifest, it works.
If i delete my main executable's manifest, the COM calls stop working but the simple string method still works.
Please note that for all my tests, i run a Powershell script that changes the Modified Date of all my files, making sure to trip the activation context cache (manifest).
From there, i started debugging using SXSTrace.exe and setting the COMPlus_LoadLogDir variable. I found out that if my manifest is external, it never gets taken into consideration. I also found out that if i embed it in the .DLL, sxstrace shows it is found, parsed and that the dependencies are found, their manifest found and i get the usual successful context activation lines in the log.
INFO: Parsing Manifest File C:\RegFreeComWrapper\bin\vbCom.MANIFEST.
INFO: Manifest Definition Identity is vbCom,type="win32",version="1.0.0.0".
INFO: Parsing Manifest File C:\RegFreeComWrapper\bin\netCom.MANIFEST.
INFO: Manifest Definition Identity is netCom,processorArchitecture="msil",version="1.0.0.0".
INFO: Activation Context generation succeeded.
End Activation Context Generation.
However, even tho the log says that, my COM calls still dont go thru.
When the call goes thru for the .NET COM call, i get some output in the COMPlus_CLRLoadLogDir configured path.
In this case, i don't get any output.
As soon as i add the manifest that only lists my C++ Wrapper as a dependency to my main executable, everything starts working again. I get a similar output from SXSTrace, but now the COM call works. As mentioned, in all cases, the C++ Wrapper's simple string method works. Meaning that even tho i have no manifest from my main executable, it can still find my .dll and call it.
Seems like the COMs don't work if i dont have the full manifest chain:
Application.Manifest > Wrapper.Manifest > COMs manifests
So, two things here...
1. Is the thing i'm trying to do even possible? (Having dependencies listed only in the manifest of a DLL that belongs to me, without touching the main executable's manifest)
2. If it is possible, i'm probably only missing a small thing to get everything going... what is it ? :)
I can provide all my code/executable as examples, should someone be interested in trying stuff with em.
Got it working!
I modified my C++ Wrapper to use the Activation Context API and force a context activation using 'ActivateActCtx'
http://msdn.microsoft.com/en-us/library/aa374151(v=vs.85).aspx
I wrapped my code with their example's code and made sure to reference my manifests and now all is working fine. No manifest on my main executable!
For some weeks now we have been fighting with an issue where at a small number of customers our Outlook addin gets unloaded and disabled for yet undetermined reasons. By "disabled" I mean that Outlook changes the following registry value from 3 to 2 which in effect means that the addin will not be loaded on next startup:
HKEY_LOCAL_MACHINE\Software\Microsoft\Office\Outlook\Addins\[OurAddin.sProgID]\LoadBehavior
There is no error message and neither do any exceptions show up in the log files that our addin produces itself.
I have already found the following page which specifically deals with the LoadBehavior change issue: http://blogs.msdn.com/vsod/archive/2008/04/22/Troubleshooting-com-add-in-load-failures.aspx
However, none of the possible reasons proposed there appear to be applicable:
The addin is not merely listed in the Disabled Items list.
There are no unhandled exceptions neither in the IDTExtensibility2 methods nor anywhere else in the code. All code is wrapped in try/catch equivalents and all exception output is emitted only via OutputDebugString or into a log file.
The error appears to be independent of anti-virus software, i.e. it also occurs with it disabled.
Disabling all other addins also has no effect on the error.
So, what else can cause Outlook to disable an addin?
Some more details / observations:
We haven't been able to reproduce the issue in our test environments so far so we haven't yet been able to attach a debugger while the issue occurs.
The issue never occurs while we try to watch what happens via remote support (TeamViewer). I suspect this is because TeamViewer uses a hook DLL that injects itself into all running processes (including Outlook) and thus affects the memory layout, timing, thread order, whatever.
Whenever we compile a new version of the addin to try out something new the addin will typically work fine for a couple of hours or even days only to eventually get disabled again. Once this has happened all subsequent attempts to get the addin to load on that machine (by manually changing back the LoadBehavior value) will fail (i.e. LoadBehaviour will simply change back to 2) until we compile and deploy another build (or try to watch using TeamViewer - see above).
Typically the addin will get unloaded right on Outlook startup though occasionally it also does happen after Outlook has already been running for some time. The log file in those cases looks completely inconspicious - the addin simply goes through the regular shutdown steps just as if Outlook had been closed normally.
As far as I can tell from our log files and by observing the issue via SysInternals ProcessMonitor, when the addin get disabled on Outlook startup (rather than during the session) the DLL gets unloaded even before the COM object (i.e. the addin) gets instantiated (log messages in the constructor never show up).
We have put OutputDebugString messages in initialization sections (this a Delphi DLL). None of them show up when the addin fails to load.
Only a very small fraction of our customers is affected by this issue. We have several tens of thousands of installations from whom we haven't received any reports about this.
UPDATE: It seems that often (but not always) one of the last things that gets logged before the addin gets unloaded is an exception with text "OLE error 800A01A8". That exception gets caught by a global exception handler built into the framework I'm using (Add-in-Express) and does not appear originate from anywhere it my own code every single method of which is by now entirely wrapped in try..catch. This typically occurs right after I set the visibility of my CommandBarButtons from an Inspector's Activate event handler.
Common properties of all affected machines:
Windows XP Professional, up-to-date patch level
Outlook 2003 Professional, up-to-date patch level
varying versions of McAfee Virus Scan (though disabling it has no effect - see above)
Users are members of the local Administrators group
One more thing to note which very probably is significant as well (though maybe not as much as I first thought):
We are using a licensing / copy protection module from a third-party vendor which wraps the compiled DLL in a "shell" and only unpacks it on-the-fly. Ever since I found out that the addin gets unloaded even before any of our own code gets executed this has been my prime suspect. However, while the vendor confirmed that there may be unhandled exceptions in their code a log file produced by a special debug version of the protection shell showed that the unpacking process completed successfully and control was already handed back to the protected DLL before Outlook unloaded the addin. So it appears that whatever causes Outlook to unload our addin happens between the completion of the protection shell's initialization and our own code.
Any more ideas?
My company has been putting up with what sounds like the same issue you are seeing for years. The plug-in we have is a VB6 COM add-in for Outlook 2003 and it’s deployed on several hundred machines that get cycled hundreds (if not thousands) of times a day. We go through the load and unload cycles a lot.
We get a fair bit of the general errors where the plug-in is loaded but not connected and we handle that in code. (Obviously not production quality)
Dim outlook As outlook.Application
Set outlook = CreateObject("Outlook.Application")
outlook.COMAddIns("MyFancyDancyPlugin").Connect = True
Rarely, but not so rare that it isn’t an annoyance, we see the plug-in reach a state where it is loaded and we can see it in “Tools>Options>Others>Advanced Options> Com Add-Ins”, but we just can’t connect to the thing. If you try to connect you don’t get an error it just switches back to disconnected. [The equivalent of switching back to a 2 in the registry key] The COM object as far as I can tell is never created. The item is not listed in the Disabled items.
We don’t actually have to redeploy to correct this error. Removing the object through the Com Add-Ins dialogue and then re-adding it there seems to correct the issue. This is still not an acceptable solution but it does get things back and running without a reinstall.
Windows XP Professional, up-to-date
patch level
Outlook 2003
Professional, up-to-date patch level
varying versions of McAfee Virus Scan
(though disabling it has no effect -
see above)
Users are members of the
local Administrators group
This seems to fit, we don't use McAfee but the virus scanner also doesn't interact with outlook or the com add-ins. We also don't use a copy protection app.
I'm sorry I can't be of more help, but I would love to root cause this.
I am also working on Outlook Add-In and I know one reason when the Add-In gets disabled. some time when Outlook shuts down abruptly or user forcefully shut down the Outlook, add-In gets disabled. I am not sure if this is the reason in your case but it could also give you some direction to think of.
I some time use this method (closing the outlook using task manager while it is still loading) to simulate this behavior and actually I have developed a tool which scans all the machines provided to it and checks if the add-In is disabled on a machine and if yes it changes the registry value to enable it.
maybe you are a lockback policy victim. add a bypass key to registry, then it works.
modern office versions or vsto creates the key while installation. the effect is: install
a modern office too and the adddin now are also loaded in older office. please take a look
code snippet taken from NetOffice http://netoffice.codeplex.com
public static void RegisterFunction(Type type)
{
try
{
// add codebase value
Assembly thisAssembly = Assembly.GetAssembly(typeof(ExampleClassicAddin));
RegistryKey key = Registry.ClassesRoot.CreateSubKey("CLSID\\{" + type.GUID.ToString().ToUpper() + "}\\InprocServer32\\1.0.0.0");
key.SetValue("CodeBase", thisAssembly.CodeBase);
key.Close();
key = Registry.ClassesRoot.CreateSubKey("CLSID\\{" + type.GUID.ToString().ToUpper() + "}\\InprocServer32");
key.SetValue("CodeBase", thisAssembly.CodeBase);
key.Close();
// add bypass key
// http://support.microsoft.com/kb/948461
key = Registry.ClassesRoot.CreateSubKey("Interface\\{000C0601-0000-0000-C000-000000000046}");
string defaultValue = key.GetValue("") as string;
if (null == defaultValue)
key.SetValue("", "Office .NET Framework Lockback Bypass Key");
key.Close();
// add addin key
Registry.ClassesRoot.CreateSubKey(#"CLSID\{" + type.GUID.ToString().ToUpper() + #"}\Programmable");
Registry.CurrentUser.CreateSubKey(_addinRegistryKey + _prodId);
RegistryKey rk = Registry.CurrentUser.OpenSubKey(_addinRegistryKey + _prodId, true);
rk.SetValue("LoadBehavior", Convert.ToInt32(3));
rk.SetValue("FriendlyName", _addinName);
rk.SetValue("Description", "NetOffice COMAddinExample with classic UI");
rk.Close();
}
catch (Exception ex)
{
string details = string.Format("{1}{1}Details:{1}{1}{0}", ex.Message, Environment.NewLine);
MessageBox.Show("An error occured." + details, "Register " + _addinName, MessageBoxButtons.OK, MessageBoxIcon.Error);
}
}
If you have the ability for your users to run a debug program to get more information about the problem when it happens, try using Add-in Spy from Microsoft.
http://msdn.microsoft.com/en-us/library/cc984533(v=office.12).aspx
I was having an add-in fail to load and discovered that it was happening because a dependency wasn't getting preloaded. This tool should be able to tell you what specific error is happening when the Outlook add-in fails to load.
Just to close this up: The problem did eventually turn out to be caused by a bug in the third-party licensing wrapper we were using. It has been confirmed by the vendor and was fixed in more recent releases.