Revise the handling of cookie request headers so that we don't interact with them in the WebContent process. They are only needed for interaction with the server and the network process, so we should limit their scope to just the Network process. Instead, we should handle a token that represents the cookie headers in the WebContent process, which can be converted to the relevant cookie data in the network process when needed.
<rdar://problem/36785285>
Created attachment 338282 [details] Patch
Created attachment 338298 [details] Patch
Created attachment 338300 [details] Patch
Created attachment 338304 [details] Patch
Comment on attachment 338304 [details] Patch Attachment 338304 [details] did not pass mac-ews (mac): Output: http://webkit-queues.webkit.org/results/7367060 New failing tests: http/tests/websocket/tests/hybi/websocket-cookie-overwrite-behavior.html http/tests/websocket/tests/hybi/secure-cookie-insecure-connection.pl http/tests/websocket/tests/hybi/httponly-cookie.pl http/tests/websocket/tests/hybi/secure-cookie-secure-connection.pl
Created attachment 338306 [details] Archive of layout-test-results from ews102 for mac-sierra The attached test failures were seen while running run-webkit-tests on the mac-ews. Bot: ews102 Port: mac-sierra Platform: Mac OS X 10.12.6
Comment on attachment 338304 [details] Patch Attachment 338304 [details] did not pass win-ews (win): Output: http://webkit-queues.webkit.org/results/7367008 New failing tests: http/tests/websocket/tests/hybi/secure-cookie-insecure-connection.pl http/tests/websocket/tests/hybi/secure-cookie-secure-connection.pl http/tests/websocket/tests/hybi/httponly-cookie.pl
Created attachment 338307 [details] Archive of layout-test-results from ews200 for win-future The attached test failures were seen while running run-webkit-tests on the win-ews. Bot: ews200 Port: win-future Platform: CYGWIN_NT-6.1-2.9.0-0.318-5-3-x86_64-64bit
Comment on attachment 338304 [details] Patch Attachment 338304 [details] did not pass mac-wk2-ews (mac-wk2): Output: http://webkit-queues.webkit.org/results/7367087 New failing tests: http/tests/websocket/tests/hybi/websocket-cookie-overwrite-behavior.html http/tests/websocket/tests/hybi/secure-cookie-insecure-connection.pl http/tests/websocket/tests/hybi/httponly-cookie.pl http/tests/websocket/tests/hybi/simple-wss.html http/tests/websocket/tests/hybi/secure-cookie-secure-connection.pl
Created attachment 338308 [details] Archive of layout-test-results from ews107 for mac-sierra-wk2 The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews. Bot: ews107 Port: mac-sierra-wk2 Platform: Mac OS X 10.12.6
Comment on attachment 338304 [details] Patch Attachment 338304 [details] did not pass ios-sim-ews (ios-simulator-wk2): Output: http://webkit-queues.webkit.org/results/7367147 New failing tests: http/tests/websocket/tests/hybi/secure-cookie-insecure-connection.pl http/tests/websocket/tests/hybi/websocket-cookie-overwrite-behavior.html http/tests/websocket/tests/hybi/httponly-cookie.pl http/tests/websocket/tests/hybi/simple-wss.html http/tests/websocket/tests/hybi/secure-cookie-secure-connection.pl
Created attachment 338310 [details] Archive of layout-test-results from ews121 for ios-simulator-wk2 The attached test failures were seen while running run-webkit-tests on the ios-sim-ews. Bot: ews121 Port: ios-simulator-wk2 Platform: Mac OS X 10.13.4
Comment on attachment 338304 [details] Patch Attachment 338304 [details] did not pass mac-debug-ews (mac): Output: http://webkit-queues.webkit.org/results/7367213 New failing tests: http/tests/websocket/tests/hybi/websocket-cookie-overwrite-behavior.html http/tests/websocket/tests/hybi/secure-cookie-insecure-connection.pl http/tests/websocket/tests/hybi/httponly-cookie.pl http/tests/websocket/tests/hybi/secure-cookie-secure-connection.pl
Created attachment 338312 [details] Archive of layout-test-results from ews116 for mac-sierra The attached test failures were seen while running run-webkit-tests on the mac-debug-ews. Bot: ews116 Port: mac-sierra Platform: Mac OS X 10.12.6
Comment on attachment 338304 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=338304&action=review > Source/WebCore/Modules/websockets/WebSocketChannel.cpp:287 > + handle.sendData(handshakeMessage.data(), handshakeMessage.length(), WTFMove(cookieRequestHeaderFieldProxy), [this, protectedThis = makeRef(*this)] (bool success, bool didAccessSecureCookies) { I know this is preexisting code but it seems strange to use sendData to actually do the handshake. Ideally, we would follow the spec, this is probably difficult and not the right time to do it. Would it be possible instead to add a sendHandshake() or establishConnection() method that would be specialized for WK2? That might also allow moving the proxy to WK2 specific code, maybe as parameters to the related IPC message. On the reverse side, we could also add a specific message to notify that the connection was established or not and whether secure cookies were accessed. That would remove the need to modify the sendData instead of using the completion handler. > Source/WebCore/Modules/websockets/WebSocketChannel.cpp:758 > + processSecureCookies(didAccessSecureCookies); I think that didAccessSecureCookies can only be false here. So this call is probably not needed. If we add a specific message to notify that cookies are used, we would not need those modifications. > Source/WebCore/Modules/websockets/WebSocketChannel.cpp:826 > +void WebSocketChannel::sendFrame(WebSocketFrame::OpCode opCode, const char* data, size_t dataLength, WTF::Function<void(bool, bool)> completionHandler) No need for WTF:: any longer. Also we could use CompletionHandler instead of Function, modulo some modifications to the WK2 specialization. Maybe for a later patch.
Created attachment 338325 [details] Patch v2 (Does not take Youenn's comments into account)
Comment on attachment 338325 [details] Patch v2 (Does not take Youenn's comments into account) Uploading fix for EWS to confirm proper test output. Will refactor per Youenn's suggestions once I confirm the code is working properly.
Comment on attachment 338325 [details] Patch v2 (Does not take Youenn's comments into account) Attachment 338325 [details] did not pass mac-ews (mac): Output: http://webkit-queues.webkit.org/results/7372091 New failing tests: http/tests/websocket/tests/hybi/httponly-cookie.pl http/tests/websocket/tests/hybi/secure-cookie-secure-connection.pl
Created attachment 338335 [details] Archive of layout-test-results from ews102 for mac-sierra The attached test failures were seen while running run-webkit-tests on the mac-ews. Bot: ews102 Port: mac-sierra Platform: Mac OS X 10.12.6
Comment on attachment 338325 [details] Patch v2 (Does not take Youenn's comments into account) Attachment 338325 [details] did not pass ios-sim-ews (ios-simulator-wk2): Output: http://webkit-queues.webkit.org/results/7372313 New failing tests: http/tests/websocket/tests/hybi/httponly-cookie.pl http/tests/websocket/tests/hybi/simple-wss.html http/tests/websocket/tests/hybi/secure-cookie-secure-connection.pl
Created attachment 338341 [details] Archive of layout-test-results from ews122 for ios-simulator-wk2 The attached test failures were seen while running run-webkit-tests on the ios-sim-ews. Bot: ews122 Port: ios-simulator-wk2 Platform: Mac OS X 10.13.4
Comment on attachment 338325 [details] Patch v2 (Does not take Youenn's comments into account) Attachment 338325 [details] did not pass mac-debug-ews (mac): Output: http://webkit-queues.webkit.org/results/7372223 New failing tests: http/tests/websocket/tests/hybi/httponly-cookie.pl http/tests/websocket/tests/hybi/secure-cookie-secure-connection.pl
Created attachment 338342 [details] Archive of layout-test-results from ews114 for mac-sierra The attached test failures were seen while running run-webkit-tests on the mac-debug-ews. Bot: ews114 Port: mac-sierra Platform: Mac OS X 10.12.6
Comment on attachment 338325 [details] Patch v2 (Does not take Youenn's comments into account) Attachment 338325 [details] did not pass mac-wk2-ews (mac-wk2): Output: http://webkit-queues.webkit.org/results/7373019 New failing tests: http/tests/websocket/tests/hybi/secure-cookie-secure-connection.pl http/tests/websocket/tests/hybi/simple-wss.html http/tests/websocket/tests/hybi/httponly-cookie.pl
Created attachment 338346 [details] Archive of layout-test-results from ews106 for mac-sierra-wk2 The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews. Bot: ews106 Port: mac-sierra-wk2 Platform: Mac OS X 10.12.6
Created attachment 338386 [details] Patch (with Youenn's suggestions)
Attachment 338386 [details] did not pass style-queue: ERROR: Source/WebCore/platform/network/mac/CookieJarMac.mm:29: Alphabetical sorting problem. [build/include_order] [4] Total errors found: 1 in 26 files If any of these errors are false positives, please file a bug against check-webkit-style.
Created attachment 338387 [details] Patch v3 (with Youenn's comments)
Attachment 338387 [details] did not pass style-queue: ERROR: Source/WebCore/platform/network/mac/CookieJarMac.mm:29: Alphabetical sorting problem. [build/include_order] [4] Total errors found: 1 in 26 files If any of these errors are false positives, please file a bug against check-webkit-style.
Comment on attachment 338387 [details] Patch v3 (with Youenn's comments) Patch looks almost ready to me. Some comments and suggestions below. View in context: https://bugs.webkit.org/attachment.cgi?id=338387&action=review > Source/WebCore/Modules/websockets/WebSocketChannel.cpp:284 > + m_document->setSecureCookiesAccessed(); Shouldn't we check for m_document being null? Cannot we have something like socket is disconnecting and we receive just after the IPC call that we did the handshake? > Source/WebCore/loader/CookieJar.cpp:81 > + auto includeSecureCookies = (url.protocolIs("https") && !document.foundMixedContent().contains(SecurityContext::MixedContentType::Active)) ? IncludeSecureCookies::Yes : IncludeSecureCookies::No; This routine is found in other places of the code base. Maybe it would be better to have a utility function here, something like computeIncludeSecureCookies(document, url)? > Source/WebCore/loader/CookieJar.cpp:84 > + if (frame) if (auto* frame = document.frame()) maybe > Source/WebCore/loader/CookieJar.cpp:87 > + return { storageSession(document).sessionID(), document.firstPartyForCookies().isolatedCopy(), url.isolatedCopy(), std::nullopt, std::nullopt, includeSecureCookies }; Why should they be isolatedCopy()? Isn't this method always called from the main thread? > Source/WebCore/platform/network/CookieRequestHeaderFieldProxy.h:53 > + CookieRequestHeaderFieldProxy isolatedCopy() const Do we need this one? It seems that sockets on workers will not use it, right? > Source/WebCore/platform/network/CookieRequestHeaderFieldProxy.h:79 > + encoder.encodeEnum(m_includeSecureCookies); Ideally we would use encoder << m_includeSecureCookies with template<> struct EnumTraits<WebCore::IncludeSecureCookies> { using values = EnumValues< WebCore::IncludeSecureCookies, WebCore::IncludeSecureCookies::No, WebCore::IncludeSecureCookies::Yes >; }; > Source/WebCore/platform/network/SocketStreamHandle.h:57 > + void sendHandshake(CString&& handshake, std::optional<CookieRequestHeaderFieldProxy>&&, Function<void(bool, bool)>); Should be Function<>&& probably. Also Function<void(bool, bool)> is not very easy to read. Not sure whether it is worth adding some enums to improve readability. > Source/WebCore/platform/network/SocketStreamHandleImpl.cpp:80 > + data.shrink(data.size() - 2); We could add an assertion that data is terminated by '\r\n' > Source/WebCore/platform/network/SocketStreamHandleImpl.cpp:85 > + builder.appendLiteral("\r\n\r\n"); We might not need a StringBuilder here, since we create a String for a very temporary time and we do utf8 conversion. Can we try appending literal strings directly to data and utf8 convert cookieData only? Or maybe directly to m_buffer? > Source/WebKit/WebProcess/Network/WebSocketStream.cpp:109 > +void WebSocketStream::platformSendHandshake(const char* data, size_t length, const std::optional<CookieRequestHeaderFieldProxy>& headerFieldProxy, Function<void(bool, bool)>&& completionHandler) It seems we should try to migrate this code from char to uint8_t. Can we make it a const uint8_t*? > Source/WebKit/WebProcess/Network/WebSocketStream.cpp:115 > + m_sendHandshakeCallbacks.set(dataIdentifier, WTFMove(completionHandler)); s/add/set/
Comment on attachment 338387 [details] Patch v3 (with Youenn's comments) View in context: https://bugs.webkit.org/attachment.cgi?id=338387&action=review >> Source/WebCore/Modules/websockets/WebSocketChannel.cpp:284 >> + m_document->setSecureCookiesAccessed(); > > Shouldn't we check for m_document being null? > Cannot we have something like socket is disconnecting and we receive just after the IPC call that we did the handshake? Good point. If the document has gone away, there's no need to note that we accessed secure cookies, so we should just do nothing. I'll remove the assertion and do a null check. >> Source/WebCore/loader/CookieJar.cpp:81 >> + auto includeSecureCookies = (url.protocolIs("https") && !document.foundMixedContent().contains(SecurityContext::MixedContentType::Active)) ? IncludeSecureCookies::Yes : IncludeSecureCookies::No; > > This routine is found in other places of the code base. > Maybe it would be better to have a utility function here, something like computeIncludeSecureCookies(document, url)? Sure. I prefer 'shouldIncludeSecureCookies(document, url)', but otherwise totally agree. >> Source/WebCore/loader/CookieJar.cpp:84 >> + if (frame) > > if (auto* frame = document.frame()) maybe OK! >> Source/WebCore/loader/CookieJar.cpp:87 >> + return { storageSession(document).sessionID(), document.firstPartyForCookies().isolatedCopy(), url.isolatedCopy(), std::nullopt, std::nullopt, includeSecureCookies }; > > Why should they be isolatedCopy()? > Isn't this method always called from the main thread? That's true. I'll fix that. >> Source/WebCore/platform/network/CookieRequestHeaderFieldProxy.h:53 >> + CookieRequestHeaderFieldProxy isolatedCopy() const > > Do we need this one? > It seems that sockets on workers will not use it, right? Removed! >> Source/WebCore/platform/network/CookieRequestHeaderFieldProxy.h:79 >> + encoder.encodeEnum(m_includeSecureCookies); > > Ideally we would use encoder << m_includeSecureCookies with > template<> struct EnumTraits<WebCore::IncludeSecureCookies> { > using values = EnumValues< > WebCore::IncludeSecureCookies, > WebCore::IncludeSecureCookies::No, > WebCore::IncludeSecureCookies::Yes > >; > }; OK! >> Source/WebCore/platform/network/SocketStreamHandle.h:57 >> + void sendHandshake(CString&& handshake, std::optional<CookieRequestHeaderFieldProxy>&&, Function<void(bool, bool)>); > > Should be Function<>&& probably. > Also Function<void(bool, bool)> is not very easy to read. > Not sure whether it is worth adding some enums to improve readability. Should 'sendData' also be using Function<>&&? Both sites create a Function object on the stack as an argument to the 'send...' method, so that seems movable. >> Source/WebCore/platform/network/SocketStreamHandleImpl.cpp:80 >> + data.shrink(data.size() - 2); > > We could add an assertion that data is terminated by '\r\n' Good idea. >> Source/WebCore/platform/network/SocketStreamHandleImpl.cpp:85 >> + builder.appendLiteral("\r\n\r\n"); > > We might not need a StringBuilder here, since we create a String for a very temporary time and we do utf8 conversion. > Can we try appending literal strings directly to data and utf8 convert cookieData only? > Or maybe directly to m_buffer? Yeah -- I wasn't happy with this section because of all the data copying. I'll revise it to be more efficient. > Source/WebCore/platform/network/SocketStreamHandleImpl.cpp:103 > + if (auto result = platformSendInternal(data.data(), data.size())) Sadly, it looks like the entire handshake buffer needs to be sent in one shot through this method. Calling it twice (once with data, one with cookie data) causes the handshake to fail and the tests to hang. >> Source/WebKit/WebProcess/Network/WebSocketStream.cpp:109 >> +void WebSocketStream::platformSendHandshake(const char* data, size_t length, const std::optional<CookieRequestHeaderFieldProxy>& headerFieldProxy, Function<void(bool, bool)>&& completionHandler) > > It seems we should try to migrate this code from char to uint8_t. > Can we make it a const uint8_t*? Sure! >> Source/WebKit/WebProcess/Network/WebSocketStream.cpp:115 >> + m_sendHandshakeCallbacks.set(dataIdentifier, WTFMove(completionHandler)); > > s/add/set/ Done (for both places).
Created attachment 338477 [details] Patch
Attachment 338477 [details] did not pass style-queue: ERROR: Source/WebCore/platform/network/mac/CookieJarMac.mm:29: Alphabetical sorting problem. [build/include_order] [4] Total errors found: 1 in 31 files If any of these errors are false positives, please file a bug against check-webkit-style.
Comment on attachment 338477 [details] Patch r=me as long as the bots are happy. View in context: https://bugs.webkit.org/attachment.cgi?id=338477&action=review > Source/WebCore/Modules/websockets/WebSocketChannel.cpp:283 > + if (m_document) one if liner? > Source/WebCore/platform/network/SocketStreamHandleImpl.cpp:94 > + data.append('\n'); We could append a Vector of '\r\n\r\n' maybe
Created attachment 338488 [details] Patch for landing
Attachment 338488 [details] did not pass style-queue: ERROR: Source/WebCore/platform/network/mac/CookieJarMac.mm:29: Alphabetical sorting problem. [build/include_order] [4] Total errors found: 1 in 31 files If any of these errors are false positives, please file a bug against check-webkit-style.
Created attachment 338498 [details] Patch for landing
Comment on attachment 338477 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=338477&action=review >> Source/WebCore/Modules/websockets/WebSocketChannel.cpp:283 >> + if (m_document) > > one if liner? Doh! >> Source/WebCore/platform/network/SocketStreamHandleImpl.cpp:94 >> + data.append('\n'); > > We could append a Vector of '\r\n\r\n' maybe I tried this but it didn't work, so I settled on the current code. I'm not sure if it's worth creating a static Vector of this. data.append({ '\r', '\n', '\r', '\n' }); I'm bummed that didn't work. Maybe there is better syntax that would fix it... > Source/WebCore/platform/network/SocketStreamHandleImpl.cpp:95 > + I figured it out! data.appendVector(Vector<uint8_t>({ '\r', '\n', '\r', '\n' }));
Comment on attachment 338498 [details] Patch for landing Clearing flags on attachment: 338498 Committed r230875: <https://trac.webkit.org/changeset/230875>
All reviewed patches have been landed. Closing bug.
Broke the Windows build: WebKitSystemInterface.lib(WebKitSystemInterface.obj) : warning LNK4099: PDB 'WebKitSystemInterface.pdb' was not found with 'WebKitSystemInterface.lib(WebKitSystemInterface.obj)' or at 'C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking object as if no debug info [C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\Tools\TestWebKitAPI\TestWebCoreLib.vcxproj] WebKitSystemInterface.lib(MediaUI.obj) : warning LNK4099: PDB 'WebKitSystemInterface.pdb' was not found with 'WebKitSystemInterface.lib(MediaUI.obj)' or at 'C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking object as if no debug info [C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\Tools\TestWebKitAPI\TestWebCoreLib.vcxproj] WebKitSystemInterface.lib(WebKitSystemInterfacePrefix.obj) : warning LNK4099: PDB 'WebKitSystemInterface.pdb' was not found with 'WebKitSystemInterface.lib(WebKitSystemInterfacePrefix.obj)' or at 'C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking object as if no debug info [C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\Tools\TestWebKitAPI\TestWebCoreLib.vcxproj] WebKitSystemInterface.lib(SharedMediaUI.obj) : warning LNK4099: PDB 'WebKitSystemInterface.pdb' was not found with 'WebKitSystemInterface.lib(SharedMediaUI.obj)' or at 'C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking object as if no debug info [C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\Tools\TestWebKitAPI\TestWebCoreLib.vcxproj] Creating library C:/cygwin/home/buildbot/slave/win-release/build/WebKitBuild/Release/lib32/TestWebCoreLib.lib and object C:/cygwin/home/buildbot/slave/win-release/build/WebKitBuild/Release/lib32/TestWebCoreLib.exp WebCore.lib(UnifiedSource348.obj) : error LNK2019: unresolved external symbol "struct std::pair<class WTF::String,bool> __cdecl WebCore::cookieRequestHeaderFieldValue(class WebCore::NetworkStorageSession const &,struct WebCore::CookieRequestHeaderFieldProxy const &)" (?cookieRequestHeaderFieldValue@WebCore@@YA?AU?$pair@VString@WTF@@_N@std@@ABVNetworkStorageSession@1@ABUCookieRequestHeaderFieldProxy@1@@Z) referenced in function "struct std::pair<class WTF::Vector<unsigned char,0,class WTF::CrashOnOverflow,16,struct WTF::FastMalloc>,bool> __cdecl WebCore::cookieDataForHandshake(struct WebCore::CookieRequestHeaderFieldProxy const &)" (?cookieDataForHandshake@WebCore@@YA?AU?$pair@V?$Vector@E$0A@VCrashOnOverflow@WTF@@$0BA@UFastMalloc@2@@WTF@@_N@std@@ABUCookieRequestHeaderFieldProxy@1@@Z) [C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\Tools\TestWebKitAPI\TestWebCoreLib.vcxproj] C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\bin32\TestWebCoreLib.dll : fatal error LNK1120: 1 unresolved externals [C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\Tools\TestWebKitAPI\TestWebCoreLib.vcxproj] WebKitSystemInterface.lib(WebKitSystemInterface.obj) : warning LNK4099: PDB 'WebKitSystemInterface.pdb' was not found with 'WebKitSystemInterface.lib(WebKitSystemInterface.obj)' or at 'C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking object as if no debug info [C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\Source\WebKitLegacy\WebKitLegacy.vcxproj] WebKitSystemInterface.lib(MediaUI.obj) : warning LNK4099: PDB 'WebKitSystemInterface.pdb' was not found with 'WebKitSystemInterface.lib(MediaUI.obj)' or at 'C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking object as if no debug info [C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\Source\WebKitLegacy\WebKitLegacy.vcxproj] WebKitSystemInterface.lib(WebKitSystemInterfacePrefix.obj) : warning LNK4099: PDB 'WebKitSystemInterface.pdb' was not found with 'WebKitSystemInterface.lib(WebKitSystemInterfacePrefix.obj)' or at 'C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking object as if no debug info [C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\Source\WebKitLegacy\WebKitLegacy.vcxproj] WebKitSystemInterface.lib(SharedMediaUI.obj) : warning LNK4099: PDB 'WebKitSystemInterface.pdb' was not found with 'WebKitSystemInterface.lib(SharedMediaUI.obj)' or at 'C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking object as if no debug info [C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\Source\WebKitLegacy\WebKitLegacy.vcxproj] Creating library C:/cygwin/home/buildbot/slave/win-release/build/WebKitBuild/Release/lib32/WebKit.lib and object C:/cygwin/home/buildbot/slave/win-release/build/WebKitBuild/Release/lib32/WebKit.exp WebCore.lib(UnifiedSource348.obj) : error LNK2019: unresolved external symbol "struct std::pair<class WTF::String,bool> __cdecl WebCore::cookieRequestHeaderFieldValue(class WebCore::NetworkStorageSession const &,struct WebCore::CookieRequestHeaderFieldProxy const &)" (?cookieRequestHeaderFieldValue@WebCore@@YA?AU?$pair@VString@WTF@@_N@std@@ABVNetworkStorageSession@1@ABUCookieRequestHeaderFieldProxy@1@@Z) referenced in function "struct std::pair<class WTF::Vector<unsigned char,0,class WTF::CrashOnOverflow,16,struct WTF::FastMalloc>,bool> __cdecl WebCore::cookieDataForHandshake(struct WebCore::CookieRequestHeaderFieldProxy const &)" (?cookieDataForHandshake@WebCore@@YA?AU?$pair@V?$Vector@E$0A@VCrashOnOverflow@WTF@@$0BA@UFastMalloc@2@@WTF@@_N@std@@ABUCookieRequestHeaderFieldProxy@1@@Z) [C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\Source\WebKitLegacy\WebKitLegacy.vcxproj] C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\bin32\WebKit.dll : fatal error LNK1120: 1 unresolved externals [C:\cygwin\home\buildbot\slave\win-release\build\WebKitBuild\Release\Source\WebKitLegacy\WebKitLegacy.vcxproj]
(In reply to Chris Dumez from comment #42) > Broke the Windows build: > WebKitSystemInterface.lib(WebKitSystemInterface.obj) : warning LNK4099: PDB > 'WebKitSystemInterface.pdb' was not found with > 'WebKitSystemInterface.lib(WebKitSystemInterface.obj)' or at > 'C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking > object as if no debug info > [C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\Tools\TestWebKitAPI\TestWebCoreLib.vcxproj] > WebKitSystemInterface.lib(MediaUI.obj) : warning LNK4099: PDB > 'WebKitSystemInterface.pdb' was not found with > 'WebKitSystemInterface.lib(MediaUI.obj)' or at > 'C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking > object as if no debug info > [C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\Tools\TestWebKitAPI\TestWebCoreLib.vcxproj] > WebKitSystemInterface.lib(WebKitSystemInterfacePrefix.obj) : warning > LNK4099: PDB 'WebKitSystemInterface.pdb' was not found with > 'WebKitSystemInterface.lib(WebKitSystemInterfacePrefix.obj)' or at > 'C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking > object as if no debug info > [C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\Tools\TestWebKitAPI\TestWebCoreLib.vcxproj] > WebKitSystemInterface.lib(SharedMediaUI.obj) : warning LNK4099: PDB > 'WebKitSystemInterface.pdb' was not found with > 'WebKitSystemInterface.lib(SharedMediaUI.obj)' or at > 'C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking > object as if no debug info > [C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\Tools\TestWebKitAPI\TestWebCoreLib.vcxproj] > Creating library > C:/cygwin/home/buildbot/slave/win-release/build/WebKitBuild/Release/lib32/ > TestWebCoreLib.lib and object > C:/cygwin/home/buildbot/slave/win-release/build/WebKitBuild/Release/lib32/ > TestWebCoreLib.exp > WebCore.lib(UnifiedSource348.obj) : error LNK2019: unresolved external > symbol "struct std::pair<class WTF::String,bool> __cdecl > WebCore::cookieRequestHeaderFieldValue(class WebCore::NetworkStorageSession > const &,struct WebCore::CookieRequestHeaderFieldProxy const &)" > (?cookieRequestHeaderFieldValue@WebCore@@YA?AU?$pair@VString@WTF@@_N@std@@ABV > NetworkStorageSession@1@ABUCookieRequestHeaderFieldProxy@1@@Z) referenced in > function "struct std::pair<class WTF::Vector<unsigned char,0,class > WTF::CrashOnOverflow,16,struct WTF::FastMalloc>,bool> __cdecl > WebCore::cookieDataForHandshake(struct > WebCore::CookieRequestHeaderFieldProxy const &)" > (?cookieDataForHandshake@WebCore@@YA?AU?$pair@V?$Vector@E$0A@VCrashOnOverflow > @WTF@@$0BA@UFastMalloc@2@@WTF@@_N@std@@ABUCookieRequestHeaderFieldProxy@1@@Z) > [C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\Tools\TestWebKitAPI\TestWebCoreLib.vcxproj] > C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\bin32\TestWebCoreLib.dll : fatal error > LNK1120: 1 unresolved externals > [C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\Tools\TestWebKitAPI\TestWebCoreLib.vcxproj] > WebKitSystemInterface.lib(WebKitSystemInterface.obj) : warning LNK4099: PDB > 'WebKitSystemInterface.pdb' was not found with > 'WebKitSystemInterface.lib(WebKitSystemInterface.obj)' or at > 'C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking > object as if no debug info > [C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\Source\WebKitLegacy\WebKitLegacy.vcxproj] > WebKitSystemInterface.lib(MediaUI.obj) : warning LNK4099: PDB > 'WebKitSystemInterface.pdb' was not found with > 'WebKitSystemInterface.lib(MediaUI.obj)' or at > 'C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking > object as if no debug info > [C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\Source\WebKitLegacy\WebKitLegacy.vcxproj] > WebKitSystemInterface.lib(WebKitSystemInterfacePrefix.obj) : warning > LNK4099: PDB 'WebKitSystemInterface.pdb' was not found with > 'WebKitSystemInterface.lib(WebKitSystemInterfacePrefix.obj)' or at > 'C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking > object as if no debug info > [C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\Source\WebKitLegacy\WebKitLegacy.vcxproj] > WebKitSystemInterface.lib(SharedMediaUI.obj) : warning LNK4099: PDB > 'WebKitSystemInterface.pdb' was not found with > 'WebKitSystemInterface.lib(SharedMediaUI.obj)' or at > 'C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\bin32\WebKitSystemInterface.pdb'; linking > object as if no debug info > [C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\Source\WebKitLegacy\WebKitLegacy.vcxproj] > Creating library > C:/cygwin/home/buildbot/slave/win-release/build/WebKitBuild/Release/lib32/ > WebKit.lib and object > C:/cygwin/home/buildbot/slave/win-release/build/WebKitBuild/Release/lib32/ > WebKit.exp > WebCore.lib(UnifiedSource348.obj) : error LNK2019: unresolved external > symbol "struct std::pair<class WTF::String,bool> __cdecl > WebCore::cookieRequestHeaderFieldValue(class WebCore::NetworkStorageSession > const &,struct WebCore::CookieRequestHeaderFieldProxy const &)" > (?cookieRequestHeaderFieldValue@WebCore@@YA?AU?$pair@VString@WTF@@_N@std@@ABV > NetworkStorageSession@1@ABUCookieRequestHeaderFieldProxy@1@@Z) referenced in > function "struct std::pair<class WTF::Vector<unsigned char,0,class > WTF::CrashOnOverflow,16,struct WTF::FastMalloc>,bool> __cdecl > WebCore::cookieDataForHandshake(struct > WebCore::CookieRequestHeaderFieldProxy const &)" > (?cookieDataForHandshake@WebCore@@YA?AU?$pair@V?$Vector@E$0A@VCrashOnOverflow > @WTF@@$0BA@UFastMalloc@2@@WTF@@_N@std@@ABUCookieRequestHeaderFieldProxy@1@@Z) > [C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\Source\WebKitLegacy\WebKitLegacy.vcxproj] > C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\bin32\WebKit.dll : fatal error LNK1120: 1 > unresolved externals > [C:\cygwin\home\buildbot\slave\win- > release\build\WebKitBuild\Release\Source\WebKitLegacy\WebKitLegacy.vcxproj] cookieRequestHeaderFieldValue() probably needs an implementation in CookieJarCFNet.cpp
Reopen to fix window build.
Created attachment 338511 [details] Speculative Windows build fix
Created attachment 338513 [details] Speculative Windows build fix
Landed speculative Windows build fix in <https://trac.webkit.org/changeset/230880>.