185746 2018-05-17 14:35:33 -0700 [Curl] Enable HTTP/2 Multiplexing 2018-05-17 17:12:21 -0700 1 1 1 Unclassified WebKit Platform WebKit Nightly Build Unspecified Unspecified RESOLVED FIXED InRadar P2 Normal --- 1 don.olmstead webkit-unassigned bfulgham commit-queue ews-watchlist galpeter pvollan webkit-bug-importer youennf oldest_to_newest 1424921 0 don.olmstead 2018-05-17 14:35:33 -0700 libcurl requires setup to enable HTTP/2 multiplexing in a multi handle. 1424941 1 340654 don.olmstead 2018-05-17 14:55:26 -0700 Created attachment 340654 Patch 1424958 2 340654 youennf 2018-05-17 15:13:18 -0700 Comment on attachment 340654 Patch View in context: https://bugs.webkit.org/attachment.cgi?id=340654&action=review > Source/WebCore/ChangeLog:3 > + [Curl] Enable HTTP/2 Multiplexing Nice! > Source/WebCore/ChangeLog:8 > + Covered by existing tests. I do not think WebKit has HTTP/2 test infra so I guess you have your own internal tests? > Source/WebCore/platform/network/curl/CurlContext.cpp:202 > + curl_multi_setopt(m_multiHandle, CURLMOPT_PIPELINING, CURLPIPE_MULTIPLEX); Will the handle support both HTTP/2 and HTTP/1.1? 1424990 3 don.olmstead 2018-05-17 16:26:20 -0700 (In reply to youenn fablet from comment #2) > Comment on attachment 340654 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=340654&action=review > > > Source/WebCore/ChangeLog:3 > > + [Curl] Enable HTTP/2 Multiplexing > > Nice! Getting closer to having it working all the way but this flag is needed for us to keep digging deeper into it. This particular flag on curl_multi got lost in us dividing up patches to land. > > > Source/WebCore/ChangeLog:8 > > + Covered by existing tests. > > I do not think WebKit has HTTP/2 test infra so I guess you have your own > internal tests? Oh my mistake. I figured there was something in there to test that HTTP/2 is working properly. > > > Source/WebCore/platform/network/curl/CurlContext.cpp:202 > > + curl_multi_setopt(m_multiHandle, CURLMOPT_PIPELINING, CURLPIPE_MULTIPLEX); > > Will the handle support both HTTP/2 and HTTP/1.1? Yea internally curl makes that decision. 1425004 4 340654 commit-queue 2018-05-17 17:05:05 -0700 Comment on attachment 340654 Patch Clearing flags on attachment: 340654 Committed r231932: <https://trac.webkit.org/changeset/231932> 1425005 5 commit-queue 2018-05-17 17:05:06 -0700 All reviewed patches have been landed. Closing bug. 1425006 6 webkit-bug-importer 2018-05-17 17:12:21 -0700 <rdar://problem/40350788> 340654 2018-05-17 14:55:26 -0700 2018-05-17 17:05:05 -0700 Patch bug-185746.diff text/plain 1231 don.olmstead ZGlmZiAtLWdpdCBhL1NvdXJjZS9XZWJDb3JlL0NoYW5nZUxvZyBiL1NvdXJjZS9XZWJDb3JlL0No YW5nZUxvZwppbmRleCAxY2JlY2EwMmYxZC4uNDU2NTJjZWU0MTAgMTAwNjQ0Ci0tLSBhL1NvdXJj ZS9XZWJDb3JlL0NoYW5nZUxvZworKysgYi9Tb3VyY2UvV2ViQ29yZS9DaGFuZ2VMb2cKQEAgLTEs MyArMSwxNSBAQAorMjAxOC0wNS0xNyAgRG9uIE9sbXN0ZWFkICA8ZG9uLm9sbXN0ZWFkQHNvbnku Y29tPgorCisgICAgICAgIFtDdXJsXSBFbmFibGUgSFRUUC8yIE11bHRpcGxleGluZworICAgICAg ICBodHRwczovL2J1Z3Mud2Via2l0Lm9yZy9zaG93X2J1Zy5jZ2k/aWQ9MTg1NzQ2CisKKyAgICAg ICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChPT1BTISkuCisKKyAgICAgICAgQ292ZXJlZCBieSBleGlz dGluZyB0ZXN0cy4KKworICAgICAgICAqIHBsYXRmb3JtL25ldHdvcmsvY3VybC9DdXJsQ29udGV4 dC5jcHA6CisgICAgICAgIChXZWJDb3JlOjpDdXJsTXVsdGlIYW5kbGU6OkN1cmxNdWx0aUhhbmRs ZSk6CisKIDIwMTgtMDUtMTcgIEplciBOb2JsZSAgPGplci5ub2JsZUBhcHBsZS5jb20+CiAKICAg ICAgICAgQ1JBU0ggaW4gSW1hZ2VEZWNvZGVyQVZGT2JqQzo6c2FtcGxlQXRJbmRleCgpCmRpZmYg LS1naXQgYS9Tb3VyY2UvV2ViQ29yZS9wbGF0Zm9ybS9uZXR3b3JrL2N1cmwvQ3VybENvbnRleHQu Y3BwIGIvU291cmNlL1dlYkNvcmUvcGxhdGZvcm0vbmV0d29yay9jdXJsL0N1cmxDb250ZXh0LmNw cAppbmRleCBlZjEzMjk3NGFlYi4uMTYzOTQ5NmEzMzkgMTAwNjQ0Ci0tLSBhL1NvdXJjZS9XZWJD b3JlL3BsYXRmb3JtL25ldHdvcmsvY3VybC9DdXJsQ29udGV4dC5jcHAKKysrIGIvU291cmNlL1dl YkNvcmUvcGxhdGZvcm0vbmV0d29yay9jdXJsL0N1cmxDb250ZXh0LmNwcApAQCAtMTk3LDYgKzE5 Nyw5IEBAIExvY2sqIEN1cmxTaGFyZUhhbmRsZTo6bXV0ZXhGb3IoY3VybF9sb2NrX2RhdGEgZGF0 YSkKIEN1cmxNdWx0aUhhbmRsZTo6Q3VybE11bHRpSGFuZGxlKCkKIHsKICAgICBtX211bHRpSGFu ZGxlID0gY3VybF9tdWx0aV9pbml0KCk7CisKKyAgICBpZiAoQ3VybENvbnRleHQ6OnNpbmdsZXRv bigpLmlzSHR0cDJFbmFibGVkKCkpCisgICAgICAgIGN1cmxfbXVsdGlfc2V0b3B0KG1fbXVsdGlI YW5kbGUsIENVUkxNT1BUX1BJUEVMSU5JTkcsIENVUkxQSVBFX01VTFRJUExFWCk7CiB9CiAKIEN1 cmxNdWx0aUhhbmRsZTo6fkN1cmxNdWx0aUhhbmRsZSgpCg==