summaryrefslogtreecommitdiffstats
path: root/net/socket/tcp_client_socket_pool.cc
Commit message (Collapse)AuthorAgeFilesLines
* Even more reordering the methods in headers and implementation in net/.erg@google.com2011-01-201-6/+6
| | | | | | | | | BUG=68682 TEST=compiles Review URL: http://codereview.chromium.org/6314010 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@71880 0039d316-1c4b-4281-b951-d872f2087c98
* Start deinlining non-empty virtual methods. (This will be automatically checkederg@google.com2010-12-091-0/+31
| | | | | | | | | | | for in the future.) BUG=none TEST=compiles Review URL: http://codereview.chromium.org/5574006 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@68746 0039d316-1c4b-4281-b951-d872f2087c98
* Convert implicit scoped_refptr constructor calls to explicit ones, part 2thakis@chromium.org2010-11-021-4/+4
| | | | | | | | | | | | This CL was created automatically by this clang rewriter: http://codereview.appspot.com/2826041 I then did quite a bit of manual editing to fix style issues. BUG=28083 TEST=None Review URL: http://codereview.chromium.org/4291001 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@64798 0039d316-1c4b-4281-b951-d872f2087c98
* Add actual bytes sent/received to net-internals.mmenke@chromium.org2010-10-141-2/+2
| | | | | | | | | BUG=54745 TEST=manual Review URL: http://codereview.chromium.org/3582007 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@62627 0039d316-1c4b-4281-b951-d872f2087c98
* Move Stats, histograms, and field trial into a metrics subdirectory of base andbrettw@chromium.org2010-10-141-1/+1
| | | | | | | | | put them in the base namespace. TEST=it compiles BUG=none git-svn-id: svn://svn.chromium.org/chrome/trunk/src@62510 0039d316-1c4b-4281-b951-d872f2087c98
* Add ClientSocketPoolBaseHelper support for preconnect.willchan@chromium.org2010-10-131-0/+20
| | | | | | | | | | | | | | | | | | | Adds a RequestSockets() API to ClientSocketPool interface. - no RequestPriority param, all requests default to LOWEST. - adds a |num_sockets| param to control how many sockets to try to ensure are connected. Adds an implementation for said function in ClientSocketPoolBaseHelper. Adds a new ClientSocketPoolBaseHelper::Flag type to modify socket request behavior. In this case, we bypass idle sockets. Adds a preconnect concept to ConnectJob. This lets normal requests hijack preconnect jobs. Modifies all ClientSocketPool subclasses to support new RequestSockets API(). Adds new tests. No client actually uses this API yet. We need to plumb it up to the preconnect system. BUG=54450 TEST=new tests in ClientSocketPoolBaseTest Review URL: http://codereview.chromium.org/3689004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@62365 0039d316-1c4b-4281-b951-d872f2087c98
* FBTF: Move ctors/dtors into implementation files. Adds ctors/dtors to ↵erg@google.com2010-09-301-0/+1
| | | | | | | | | | | | | | non-POD structs. Cuts ~2MB off our .a files (Debug, Linux). Also added the "virtual" keyword on a whole bunch of virtual dtors that were missing it. BUG=none TEST=compiles Review URL: http://codereview.chromium.org/3522004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@61100 0039d316-1c4b-4281-b951-d872f2087c98
* Stop refcounting ClientSocketPool.willchan@chromium.org2010-09-291-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Establishes that HttpNetworkSession owns all the socket pools. Move out all the socket pools into a ClientSocketPoolManager. This is because of the dependency tree amongst socket pools, which dictates the order in which they must be constructed and destructed. In order to better establish it, I moved them out to their own class. HttpNetworkSession owns the ClientSocketPoolManager which owns the pools. We pass the pools as raw pointers everywhere. Note that ClientSocketPoolManager owns more pools than are publicly accessible via its interface. That's because some of them are wrapped by publicly exposed pools. Also, ClientSocketPoolHistograms used to be reference counted. That's because it can be shared by multiple ClientSocketPools. But it's effectively a global as well, so I make their lifetimes persist for the length of ClientSocketPoolManager too. I also removed internal refcounting in ClientSocketPoolBase. I had refcounted it before I knew about ScopedRunnableMethodFactory back when I first started. I cleaned up the unit tests a lot. Back when I was a young padawan, I didn't really know what I was doing, so I copy/pasted a metric asston of code. Turns out most of it was stupid, so I fixed it. I also stopped the use of implementation inheritance with ClientSocketPoolTest because it's discouraged by the style guide and more importantly because it caused the ClientSocketHandles within the TestSocketRequest vector to be destroyed _after_ the pools themselves were destroyed, which is bad since the handles will call pool_->Release() which blows up. BUG=56215,56215 TEST=Existing unit tests Review URL: http://codereview.chromium.org/3389020 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@60983 0039d316-1c4b-4281-b951-d872f2087c98
* net: Fix incorrect style used in some enumerations.tfarina@chromium.org2010-09-221-15/+15
| | | | | | | | | | | | | | | | | | | Reason: They are using kConstantNaming, however the Chromium style is to use UPPER_CASE_FOR_ENUMS. Also convert some truly constants, like the following: http/http_proxy_client_socket_pool.h: enum { kDrainBodyBufferSize = 1024 }; To: static const int kDrainBodyBufferSize = 1024; BUG=55668 TEST=compiles locally. Review URL: http://codereview.chromium.org/3421026 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@60155 0039d316-1c4b-4281-b951-d872f2087c98
* Change HostResolver::RequestInfo to take a HostPortPair rather than naked ↵eroman@chromium.org2010-09-151-5/+3
| | | | | | | | host string + port. Review URL: http://codereview.chromium.org/3420001 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@59577 0039d316-1c4b-4281-b951-d872f2087c98
* Fix a regression whereby full-granularity network events were being captured ↵eroman@chromium.org2010-08-311-1/+1
| | | | | | | | | | by PassiveLogCollector. BUG=53883 TEST=Load a webpage. Now open about:net-internals and verify that the log for that URL does NOT contain the HTTP request/response headers. Review URL: http://codereview.chromium.org/3274016 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@58070 0039d316-1c4b-4281-b951-d872f2087c98
* In about:net-internals, each SOCKET created by a CONNECT_JOB holds a ↵mmenke@chromium.org2010-08-251-1/+1
| | | | | | | | | | | reference to the CONNECT_JOB that spawned it. BUG=47082 TEST=manual Review URL: http://codereview.chromium.org/3164005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@57339 0039d316-1c4b-4281-b951-d872f2087c98
* A/B experiment for re-establishing TCP connections.ziadh@chromium.org2010-08-251-1/+1
| | | | | | | | | | | | | We would like to test the impact of automatic retries when a TCP connection exceeds a certain threshold before it gets back an ACK. We are observing a fair number of sockets where the connection had been established, but the sockets were not used thereafter. r=mbelche Review URL: http://codereview.chromium.org/3191019 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@57276 0039d316-1c4b-4281-b951-d872f2087c98
* Net: Get rid of ProxyServer::host_and_port() and friends.thestig@chromium.org2010-07-221-1/+1
| | | | | | | | BUG=none TEST=none Review URL: http://codereview.chromium.org/3032017 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@53403 0039d316-1c4b-4281-b951-d872f2087c98
* Refactor how ClientSocketPoolBaseHelper avoids re-entrancy.willchan@chromium.org2010-07-151-1/+1
| | | | | | | | | Specifically, we defer asynchronous user callbacks to tasks. BUG=48861 Review URL: http://codereview.chromium.org/2994003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@52509 0039d316-1c4b-4281-b951-d872f2087c98
* Make the various SocketParams reference counted.vandebo@chromium.org2010-07-121-6/+23
| | | | | | | | | | | This is so that the SSLSocketParam can hold one of any of the existing SocketParams. BUG=30357 TEST=existing unit tests Review URL: http://codereview.chromium.org/2848029 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@52107 0039d316-1c4b-4281-b951-d872f2087c98
* Massively simplify the NetworkChangeNotifier infrastructure:pkasting@chromium.org2010-06-251-3/+1
| | | | | | | | | | | | | | * Use a process-wide object (singleton pattern) * Create/destroy this object on the main thread, make it outlive all consumers * Make observer-related functions threadsafe As a result, the notifier can now be used by any thread (eliminating things like NetworkChangeObserverProxy and NetworkChangeNotifierProxy, and expanding its usefulness); its creation and inner workings are much simplified (eliminating implementation-specific classes); and it is simpler to access (eliminating things like NetworkChangeNotifierThread and a LOT of passing pointers around). BUG=none TEST=Unittests; network changes still trigger notifications Review URL: http://codereview.chromium.org/2802015 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@50895 0039d316-1c4b-4281-b951-d872f2087c98
* A/B test for determining a value for unused socket timeout. Currently theziadh@chromium.org2010-06-211-1/+2
| | | | | | | | | | | | | | timeout defaults to 10 seconds. Having this value set too low won't allow us to take advantage of idle sockets. Setting it to too high could possibly result in more ERR_CONNECT_RESETs, requiring one RTT to receive the RST packet and possibly another RTT to re-establish the connection. r=jar Review URL: http://codereview.chromium.org/2827016 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@50364 0039d316-1c4b-4281-b951-d872f2087c98
* Reland my close on idle socket change (r43882+r44150).willchan@chromium.org2010-06-101-1/+1
| | | | | | | | | | I reverted it the first time because it was suspected of causing a hang on the IO thread. A different CL caused that. When I relanded it previously, the fix for the hang on the IO thread broke this (changed an assertion). I've now deleted that assertion. BUG=32817 Review URL: http://codereview.chromium.org/2716004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@49444 0039d316-1c4b-4281-b951-d872f2087c98
* Revert "Revert an idle sockets change to trigger reliability bot. Will ↵willchan@chromium.org2010-06-081-1/+1
| | | | | | | | revert again soon." Review URL: http://codereview.chromium.org/2761001 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@49146 0039d316-1c4b-4281-b951-d872f2087c98
* Revert an idle sockets change to trigger reliability bot. Will revert again ↵willchan@chromium.org2010-06-081-1/+1
| | | | | | | | soon. Review URL: http://codereview.chromium.org/2760001 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@49145 0039d316-1c4b-4281-b951-d872f2087c98
* Do not attempt to reuse active sockets after a socket pool flush (usually a ↵willchan@chromium.org2010-06-071-2/+7
| | | | | | | | | | | network change). Implements this functionality by adding an |id_| field to ClientSocketPoolBaseHelper that is incremented on each Flush(). BUG=45872 Review URL: http://codereview.chromium.org/2647003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@49076 0039d316-1c4b-4281-b951-d872f2087c98
* Rework the logging for sockets/connectjobs.eroman@chromium.org2010-06-031-8/+10
| | | | | | | | | | | | | | | | | | In particular, make it work better when using backup jobs / late binding (the display was very confused before because of how these asynchronous events would nest). Also changed the paradigm for how PassiveLogCollector preserves these async associations -- this fixes how it replays the events to net-internals. (Before we would collapse the event streams into the SOURCE_URL_REQUEST which lost some hiearchy.. now I keep the separate streams). Some of the particular changes to the event streams: * ConnectJobs now create their own source stream internally. * Sockets are now bounded by +SOCKET_ALIVE / -SOCKET_ALIVE events (removed the one-off SOCKET_DONE event). * The socket log streams contains +SOCKET_IN_USE / -SOCKET_IN_USE event blocks to show which URLRequest was controlling it at various points in time (this makes it much easier to understand which read/writes belonged to a particular network transaction when a socket gets re-used). * ConnectJobs are bounded by +SOCKET_POOL_CONNECT_JOB / - SOCKET_POOL_CONNECT_JOB events. * ConnectJobs log the net error they failed with. * Removed the SOCKET_BACKUP_TIMER_EXTENDED event. Review URL: http://codereview.chromium.org/2363003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@48797 0039d316-1c4b-4281-b951-d872f2087c98
* Make ClientSocketPool histograms static so that they work properly.vandebo@chromium.org2010-05-201-2/+2
| | | | | | | | | | | | Also change their names so that they appear all together on the histograms page. BUG=43375 TEST=none Review URL: http://codereview.chromium.org/2029004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@47843 0039d316-1c4b-4281-b951-d872f2087c98
* Revert r47395. Looks like it still crashes =/willchan@chromium.org2010-05-201-2/+4
| | | | | | | | BUG=40455,40457 Review URL: http://codereview.chromium.org/2104013 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@47753 0039d316-1c4b-4281-b951-d872f2087c98
* Revert r47583 for causing reliability bot crashes.willchan@chromium.org2010-05-191-1/+1
| | | | | | | | BUG=32817 Review URL: http://codereview.chromium.org/2105009 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@47616 0039d316-1c4b-4281-b951-d872f2087c98
* Reland my close on idle socket change.willchan@chromium.org2010-05-181-1/+1
| | | | | | | | | This reverts 44402 which reverted r44150 (fix for r43882) and r43882 (original change). These changes were suspected for causing the infinite loop on the IO thread and also a crash. It turned out that the backup socket was the cause. These should be able to be safely relanded. There's one minor change from the original, due to merging in the weeks of changes in between. BUG=32817 Review URL: http://codereview.chromium.org/2077004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@47583 0039d316-1c4b-4281-b951-d872f2087c98
* This is roughly a reland of r43908, although I had to rewrite a fair chunk ↵willchan@chromium.org2010-05-171-4/+2
| | | | | | | | | | since code's changed a lot. I've changed it so as not to flush the host resolver too. I suspect there might have been a bug there. I was unable to reproduce the crash here on network change, so I think either that fixed it or something else changed under the hood. In any case, I want to reland this and see if anyone encounters any problems with it. BUG=40455,40457 Review URL: http://codereview.chromium.org/2114004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@47395 0039d316-1c4b-4281-b951-d872f2087c98
* More cleanup to address TODOs in net_log.h.eroman@chromium.org2010-04-281-5/+6
| | | | | | | | | | | * Removes 9 methods: AddEventWithParameters, BeginEventWithParameters, EndEventWithParameters, BeginEventWithString, BeginEventWithInteger, AddEventWithString, AddEventWithInteger, EndEventWithParameters, EndEventWithInteger. This was becoming ridiculous, instead made the EventParameters* a required parameter. * Moves CapturingBoundNetLog / CapturingNetLog to its own file. BUG=37421 Review URL: http://codereview.chromium.org/1746012 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@45843 0039d316-1c4b-4281-b951-d872f2087c98
* Cleanup: Address some of the todos in net_log.heroman@chromium.org2010-04-271-3/+7
| | | | | | | | | | | - Get rid of the AddString() and AddStringLiteral() methods. - Make EventParameters able to serialize to JSON, instead of a string. BUG=37421 Review URL: http://codereview.chromium.org/1716007 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@45750 0039d316-1c4b-4281-b951-d872f2087c98
* Add net log entries that summarize transmit and receive byte counts.vandebo@chromium.org2010-04-231-2/+3
| | | | | | | | | | | | | | | | | | | | | Tx/Rx summaries are integrated into the net log at the last point that bytes were transmitted or received. Hopefully this will help resolve http://crbug.com/37729 by showing if we've received bytes over the network when we hit the "Waiting for cache" bug. This change also modernizes the use of NetLog: - ClientSocket now has a net_log() accessor - ClientSocket::Connect no longer takes a NetLog, instead the TCPClientSocket constructor takes one, others use their transport socket's NetLog - TCPClientSocket creates a new source id with source type SOCKET Also updates PassiveLogCollector infrastructure: - The LiveRequestsObserver lets a RequestTracker update a RequestInfo just before it is displayed. This allows ConnectJobs to be associated with URLRequests while connecting and then reassociated if they are late-bound to a different request. BUG=37729 TEST=tx/rx lines show up in chrome://net-internals/ Review URL: http://codereview.chromium.org/1696005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@45449 0039d316-1c4b-4281-b951-d872f2087c98
* Fix crash on IP address change.willchan@chromium.org2010-04-191-2/+4
| | | | | | | | | | Revert r43908: "Flush socket pools and SPDY session pool properly on explicit requests and network changes." This change was reverted on the 375 branch, but not on trunk. BUG=41190 Review URL: http://codereview.chromium.org/1604045 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@44973 0039d316-1c4b-4281-b951-d872f2087c98
* Revert r44150 and r43882.willchan@chromium.org2010-04-131-1/+1
| | | | | | | | | r44150 is a fix for r43882. r43882 is a bugfix to include the idle sockets in the socket limit and close them when we hit the limit. Including the idle sockets in the socket limit check make us hit the socket limit more often, which exposed some bugs in the existing code. These two patches may also have introduced some separate crashes of their own (it's unclear from the stacktraces). I'm reverting them for now to mitigate the crashes. BUG=32817,41228 Review URL: http://codereview.chromium.org/1657001 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@44402 0039d316-1c4b-4281-b951-d872f2087c98
* Fix crash when we delete the group we're requesting a socket for.willchan@chromium.org2010-04-091-1/+1
| | | | | | | | BUG=40952 Review URL: http://codereview.chromium.org/1618011 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@44150 0039d316-1c4b-4281-b951-d872f2087c98
* Flush socket pools and SPDY session pool properly on explicit requests and ↵willchan@chromium.org2010-04-081-4/+2
| | | | | | | | | | network changes. BUG=40455,40457 Review URL: http://codereview.chromium.org/1615005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@43908 0039d316-1c4b-4281-b951-d872f2087c98
* Implement SOCKSClientSocketPoolvandebo@chromium.org2010-03-221-5/+10
| | | | | | | | | | | | | | This is the first layered pool, so there are several infrastructure changes in this change as well. Add a ConnectionTimeout method to pools so that layered pools can timeout each phase. Add a name method to pools to support per pool UMA histograms. Change SOCKS sockets to take a ClientSocketHandle instead of a ClientSocket BUG=30357 (blocks an SSL Pool) TEST=existing unit tests Review URL: http://codereview.chromium.org/668097 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@42231 0039d316-1c4b-4281-b951-d872f2087c98
* Disable backup connect jobs for all pools except TCPClientSocketPoolvandebo@chromium.org2010-03-201-1/+3
| | | | | | | | | BUG=36629 TEST=net_unittests Review URL: http://codereview.chromium.org/1120007 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@42179 0039d316-1c4b-4281-b951-d872f2087c98
* Generalize the net module's LoadLog facility from a passive container, to an ↵eroman@chromium.org2010-03-161-15/+13
| | | | | | | | | | | | | | | | event stream (NetLog). This makes it possible to associate a single NetLog with a URLRequestContext, and then attach observers to that log to watch the stream of events. This changelist attempts to do the most direct translation, so there will be subsequent iterations to clean up. The user-visible behavior should remain unchanged. BUG=37421 Review URL: http://codereview.chromium.org/848006 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@41689 0039d316-1c4b-4281-b951-d872f2087c98
* Make a proper TCPSocketParamsvandebo@chromium.org2010-02-271-11/+12
| | | | | | | | | BUG=none TEST=existing unit tests Review URL: http://codereview.chromium.org/661194 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@40182 0039d316-1c4b-4281-b951-d872f2087c98
* Pass the NetworkChangeNotifier to HostResolver.willchan@chromium.org2010-02-041-1/+1
| | | | | | | | | | | | | | | This requires the following refactors: (1) NetworkChangeNotifier moves out of HttpNetworkSession into IOThread. (2) HostResolver gets initialized with NetworkChangeNotifier. (3) NetworkChangeNotifier needs to get passed into HttpCache and HttpNetworkSession (required updating a lot of files). (4) NetworkChangeNotifier is no longer reference counted. It is owned by IOThread. (5) IOThread gains a new struct: Globals. It can only be used on the io thread. (6) ChromeURLRequestContextFactory uses IOThread::Globals to initialize ChromeURLRequest objects with the host resolver and network change notifier. BUG=26159 Review URL: http://codereview.chromium.org/552117 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@38052 0039d316-1c4b-4281-b951-d872f2087c98
* Switch on socket late binding - Take 2.willchan@chromium.org2010-01-271-3/+2
| | | | | | | | | | | | | | | Re-enable socket late binding. The mac valgrind errors happened due to threading bugs in test_shell_tests. The ui thread would TearDown() the test object, which deleted the TestURLRequestContext, which eventually deletes the TCPClientSocketPool, which deletes its ConnectJobs. However, those ConnectJobs might be running simultaneously on the io thread. Therefore, we have a race condition. This change fixes that. Histograms for the 4.0.266.0 dev channel release indicate the following changes for late binding: (a) Net.TCPSocketType shows a decrease (from 41.85% to 39.29%) in used of newly connected sockets. Part of this decrease is due to using previously used sockets more often (increase from 58.15% to 58.53%), but is primarily due to being able to use sockets that were connected, but not immediately handed over to a socket request (increased from 0 [not supported without late binding] to 2.18%). (b) Net.SocketIdleTimeBeforeNextUse_ReusedSocket indicates that reused sockets are getting used more quickly than before, with a decrease of mean idle time from 11.65 seconds to 11.34 seconds. (c) Net.Transaction_Connected_Under_10 indicates shows that the mean for time until the first byte of the transaction response decreased from 1585ms to 1481ms. The code change deletes the old non socket late binding code paths, cleaning up the code significantly. It also deletes duplicated tests in ClientSocketPoolBase which covered both pathways. A TCPClientSocketPool test had to be updated as well. BUG=http://crbug.com/30354. Review URL: http://codereview.chromium.org/549093 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@37311 0039d316-1c4b-4281-b951-d872f2087c98
* Add some extra request information to LoadLog:eroman@chromium.org2010-01-211-0/+10
| | | | | | | | | | | | | | * the error code that URLRequest Start() failed with. * which peer the TCP connection is being made to [Only in full logging mode] (for example, when using a proxy, this may differ from the host:port of the URL). * the type of redirect that was received, and its URL [Only in full logging mode] BUG=NONE TEST=NONE Review URL: http://codereview.chromium.org/553043 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@36813 0039d316-1c4b-4281-b951-d872f2087c98
* Revert 36230 - Switch on socket late binding.willchan@chromium.org2010-01-141-2/+3
| | | | | | | | | | | | | | | | | Broke mac valgrind for test_shell_tests. Various errors. I also missed fixing a bug that eroman pointed out in review. Total brainfart. Histograms for the 4.0.266.0 dev channel release indicate the following changes for late binding: (a) Net.TCPSocketType shows a decrease (from 41.85% to 39.29%) in used of newly connected sockets. Part of this decrease is due to using previously used sockets more often (increase from 58.15% to 58.53%), but is primarily due to being able to use sockets that were connected, but not immediately handed over to a socket request (increased from 0 [not supported without late binding] to 2.18%). (b) Net.SocketIdleTimeBeforeNextUse_ReusedSocket indicates that reused sockets are getting used more quickly than before, with a decrease of mean idle time from 11.65 seconds to 11.34 seconds. (c) Net.Transaction_Connected_Under_10 indicates shows that the mean for time until the first byte of the transaction response decreased from 1585ms to 1481ms. The code change deletes the old non socket late binding code paths, cleaning up the code significantly. It also deletes duplicated tests in ClientSocketPoolBase which covered both pathways. A TCPClientSocketPool test had to be updated as well. BUG=http://crbug.com/30354. Review URL: http://codereview.chromium.org/543051 TBR=willchan@chromium.org Review URL: http://codereview.chromium.org/549047 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@36231 0039d316-1c4b-4281-b951-d872f2087c98
* Switch on socket late binding.willchan@chromium.org2010-01-141-3/+2
| | | | | | | | | | | | | Histograms for the 4.0.266.0 dev channel release indicate the following changes for late binding: (a) Net.TCPSocketType shows a decrease (from 41.85% to 39.29%) in used of newly connected sockets. Part of this decrease is due to using previously used sockets more often (increase from 58.15% to 58.53%), but is primarily due to being able to use sockets that were connected, but not immediately handed over to a socket request (increased from 0 [not supported without late binding] to 2.18%). (b) Net.SocketIdleTimeBeforeNextUse_ReusedSocket indicates that reused sockets are getting used more quickly than before, with a decrease of mean idle time from 11.65 seconds to 11.34 seconds. (c) Net.Transaction_Connected_Under_10 indicates shows that the mean for time until the first byte of the transaction response decreased from 1585ms to 1481ms. The code change deletes the old non socket late binding code paths, cleaning up the code significantly. It also deletes duplicated tests in ClientSocketPoolBase which covered both pathways. A TCPClientSocketPool test had to be updated as well. BUG=http://crbug.com/30354. Review URL: http://codereview.chromium.org/543051 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@36230 0039d316-1c4b-4281-b951-d872f2087c98
* Fixup some connection time measurements.mbelshe@google.com2009-12-281-3/+12
| | | | | | | | | | | | | | | | - Net.TCP_Connection_Latency Time to connect only - Net.DNS_Resolution_And_TCP_Connection_Latency2 Time to do DNS and Connect - Net.HttpConnectionLatency Time for HTTP to connect (includes all delays, DNS, Queue, TCP, etc) BUG=none TEST=none Review URL: http://codereview.chromium.org/517014 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@35316 0039d316-1c4b-4281-b951-d872f2087c98
* Detects network changes. Only for Mac OS X so far. Hooks up ↵willchan@chromium.org2009-12-211-2/+4
| | | | | | | | | | | TCPClientSocketPool to flush idle sockets on IP address change. BUG=http://crbug.com/26156 TEST=Run chrome with both network cable and wireless on. Go to www.google.com, twice. Verify second time via chrome://net-internals that the second request did not need a TCP_CONNECT_JOB, since we reused idle sockets. Unplug network cable. This should flush idle sockets. Go back to www.google.com. Check chrome://net-internals. Verify that there is a TCP_CONNECT_JOB for that request, because there was no idle socket to reuse. Review URL: http://codereview.chromium.org/460149 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@35107 0039d316-1c4b-4281-b951-d872f2087c98
* Update network priorities to support better granularitymbelshe@google.com2009-12-021-1/+1
| | | | | | | | | | | | | | of resource loading from WebKit into the network stack. In order to fully make these work, webkit changes are needed as well. BUG=none TEST=none Review URL: http://codereview.chromium.org/452033 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@33546 0039d316-1c4b-4281-b951-d872f2087c98
* Add LoadLog to ClientSocket::Connect().willchan@chromium.org2009-11-021-2/+2
| | | | | | | | TODO: Use LoadLog in FLIP code. Review URL: http://codereview.chromium.org/344026 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@30765 0039d316-1c4b-4281-b951-d872f2087c98
* Increase the timeout for TcpConnectJob from 60 seconds to 240 seconds.eroman@chromium.org2009-10-061-1/+8
| | | | | | | | BUG=23364 Review URL: http://codereview.chromium.org/255085 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@28087 0039d316-1c4b-4281-b951-d872f2087c98
* Make GetLoadState virtual in ConnectJob().willchan@chromium.org2009-09-011-4/+14
| | | | | | | | | This will be necessary for SSLClientSocketPool since the SSLConnectJob will contain a ClientSocketHandle for the TCPClientSocketPool. SSLConnectJob::GetLoadState() will forward to ClientSocketHandle::GetLoadState() if the SSLConnectJob is waiting on requesting a TCP socket. BUG=http://crbug.com/13289. Review URL: http://codereview.chromium.org/176024 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@25032 0039d316-1c4b-4281-b951-d872f2087c98
generated by cgit v1.1 at 2025-05-20 16:04:19 +0000