From mboxrd@z Thu Jan 1 00:00:00 1970 From: Doug Ledford Subject: Re: [PATCH V1 FIX for-3.19] IB/ipoib: Fix broken multicast flow Date: Tue, 13 Jan 2015 16:01:00 -0500 Message-ID: <1421182860.43839.214.camel@redhat.com> References: <1420643066-3599-1-git-send-email-ogerlitz@mellanox.com> <1421167544.43839.190.camel@redhat.com> Mime-Version: 1.0 Content-Type: multipart/signed; micalg="pgp-sha1"; protocol="application/pgp-signature"; boundary="=-U0Grz3J/tS3PMqYO0wIA" Return-path: In-Reply-To: Sender: linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Or Gerlitz Cc: Roland Dreier , "linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org" , Amir Vadai , Eyal Perry , Erez Shitrit List-Id: linux-rdma@vger.kernel.org --=-U0Grz3J/tS3PMqYO0wIA Content-Type: multipart/mixed; boundary="=-0TbpH9TgXhteuE4LhfY4" --=-0TbpH9TgXhteuE4LhfY4 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable On Tue, 2015-01-13 at 22:13 +0200, Or Gerlitz wrote: > On Tue, Jan 13, 2015 at 6:45 PM, Doug Ledford wrote= : > > On Fri, 2015-01-09 at 10:32 +0200, Or Gerlitz wrote: > >> On Wed, Jan 7, 2015 at 5:04 PM, Or Gerlitz wro= te: > >> > From: Erez Shitrit > >> > > >> > Following commit 016d9fb25cd9 "IPoIB: fix MCAST_FLAG_BUSY usage" > >> > both IPv6 traffic and for the most cases all IPv4 multicast traffic > >> > aren't working. > >> > >> Doug, can you ack the breakage introduced by your commit and the fix? > > > > I haven't double checked the breakage, I'll take your word for it >=20 > just try ping6 or iperf multicast and see it for yourself, please. I have. I have them working now. >=20 > > (at the time I did my work, I had multicast debugging on and I verified= the > > join/leave process, but I had assumed that the process would work the > > same for optional multicast groups as it does for the IPoIB broadcast > > group and other default IPoIB groups, so I didn't specifically test > > additional multicast groups above and beyond the broadcast/etc groups). > > > > However, the fix is not workable. In particular, as soon as this patch > > is added to the kernel, you will start getting messages like this: > > > > mlx4_ib0: ipoib_mcast_leave on an in-flight join >=20 >=20 > I don't see it on my systems, is that upstream you're running? what entit= y does > ,;x4_ib0: prefixed prints and under what settings, is that the IPoIB driv= er? No, that's my internal rhel7 tree, but it's so close to bare upstream when it comes to IPoIB that there really shouldn't be any difference. And mlx4_ib0 is just the name that I renamed ib0 to (An internal standard in my test cluster is that all ib interfaces are named based upon the hardware they are tied to, so mlx4_ib0, mlx5_ib0, qib_ib0, etc. Makes my life in verifying code coverage easier) I can test again with an upstream kernel later today. But, for now, suffice it to say the problem is not resolved with the patch in this thread, but with the much simpler patch I've attached to this email (note, this is made against rhel7, I'm only attaching it so you can try it yourself, assuming it even applies, and once I've tested it on an upstream kernel myself and verified that it works properly, I'll submit the final upstream version under a new thread). --=20 Doug Ledford GPG KeyID: 0E572FDD --=-0TbpH9TgXhteuE4LhfY4 Content-Disposition: attachment; filename="ipoib-mcast-test.patch" Content-Transfer-Encoding: base64 Content-Type: text/x-patch; name="ipoib-mcast-test.patch"; charset="UTF-8" Y29tbWl0IGRjMmM3MTg5NmVkYzNiNTQ5ZTc5YTUxY2Y4M2RiODEzZDYwMTJhMzQKQXV0aG9yOiBE b3VnIExlZGZvcmQgPGRsZWRmb3JkQHJlZGhhdC5jb20+CkRhdGU6ICAgVHVlIEphbiAxMyAxMzoz NTo0MCAyMDE1IC0wNTAwCgogICAgSUIvaXBvaWI6IEZpeCBmYWlsZWQgbXVsdGljYXN0IGpvaW5z L3NlbmRzCiAgICAKICAgIFRoZSB1c2FnZSBvZiBJUE9JQl9NQ0FTVF9SVU4gYXMgYSBmbGFnIGlz IGluY29uc2lzdGVudC4gIEluIHNvbWUgcGxhY2VzCiAgICBpdCBpcyB1c2VkIHRvIG1lYW4gIm91 ciBkZXZpY2UgaXMgYWRtaW5pc3RyYXRpdmVseSBhbGxvd2VkIHRvIHNlbmQKICAgIG11bHRpY2Fz dCBqb2lucy9sZWF2ZXMvcGFja2V0cyIgYW5kIGluIG90aGVyIHBsYWNlcyBpdCBtZWFucyAib3Vy CiAgICBtdWx0aWNhc3Qgam9pbiB0YXNrIHRocmVhZCBpcyBjdXJyZW50bHkgcnVubmluZyBhbmQg d2lsbCBwcm9jZXNzIHlvdXIKICAgIHJlcXVlc3QgaWYgeW91IHB1dCBpdCBvbiB0aGUgcXVldWUi LiAgSG93ZXZlciwgdGhpcyBsYXR0ZXIgbWVhbmluZyBpcyBpbgogICAgZmFjdCBmbGF3ZWQgYXMg dGhlcmUgaXMgYSByYWNlIGNvbmRpdGlvbiBiZXR3ZWVuIHRoZSBqb2luIHRhc2sgdGVzdGluZwog ICAgdGhlIG1jYXN0IGxpc3QgYW5kIGZpbmRpbmcgaXQgZW1wdHkgb2YgcmVtYWluaW5nIHdvcmss IGRyb3BwaW5nIHRoZQogICAgbWNhc3QgbXV0ZXggYW5kIGFsc28gdGhlIHByaXYtPmxvY2sgc3Bp bmxvY2ssIGFuZCBjbGVhcmluZyB0aGUKICAgIElQT0lCX01DQVNUX1JVTiBmbGFnLiAgRnVydGhl ciwgdGhlcmUgYXJlIG51bWVyb3VzIGxvY2F0aW9ucyB0aGF0IHVzZQogICAgdGhlIGZsYWcgaW4g dGhlIGZvcm1lciBmYXNoaW9uLCBhbmQgd2hlbiBhbGwgdGFza3MgY29tcGxldGUgYW5kIHRoZSB0 YXNrCiAgICB0aHJlYWQgY2xlYXJzIHRoZSBSVU4gZmxhZywgYWxsIG9mIHRob3NlIG90aGVyIGxv Y2F0aW9ucyB3aWxsIGZhaWwgdG8KICAgIGV2ZXIgYWdhaW4gcXVldWUgYW55IHdvcmsuICBUaGlz IHJlc3VsdHMgaW4gdGhlIGludGVyZmFjZSBjb21pbmcgdXAgZmluZQogICAgaW5pdGlhbGx5LCBi dXQgaGF2aW5nIHByb2JsZW1zIGFkZGluZyBuZXcgbXVsdGljYXN0IGdyb3VwcyBhZnRlciB0aGUK ICAgIGZpcnN0IHJvdW5kIG9mIGdyb3VwcyBoYXZlIGFsbCBiZWVuIGFkZGVkIGFuZCB0aGUgUlVO IGZsYWcgaXMgY2xlYXJlZCBieQogICAgdGhlIGpvaW4gdGFzayB0aHJlYWQgd2hlbiBpdCB0aGlu a3MgaXQgaXMgZG9uZS4gIFRvIHJlc29sdmUgdGhpcyBpc3N1ZSwKICAgIGNvbnZlcnQgYWxsIGxv Y2F0aW9ucyBpbiB0aGUgY29kZSB0byB0cmVhdCB0aGUgUlVOIGZsYWcgYXMgYW4gaW5kaWNhdG9y CiAgICB0aGF0IHRoZSBtdWx0aWNhc3QgcG9ydGlvbiBvZiB0aGlzIGludGVyZmFjZSBpcyBpbiBm YWN0IGFkbWluaXN0cmF0aXZlbHkKICAgIHVwIGFuZCBqb2lucy9sZWF2ZXMvc2VuZHMgY2FuIGJl IHBlcmZvcm1lZC4gIFRoZXJlIGlzIG5vIGhhcm0gKG90aGVyCiAgICB0aGFuIGEgc2xpZ2h0IHBl cmZvcm1hbmNlIHBlbmFsdHkpIHRvIG5ldmVyIGNsZWFyaW5nIHRoaXMgZmxhZyBhbmQgdXNpbmcK ICAgIGl0IGluIHRoaXMgZmFzaGlvbiBhcyBpdCBzaW1wbHkgbWVhbnMgdGhhdCBhIGZldyBwbGFj ZXMgdGhhdCB1c2VkIHRvCiAgICBtaWNyby1vcHRpbWl6ZSBob3cgb2Z0ZW4gdGhpcyB0YXNrIHdh cyBxdWV1ZWQgb24gYSB3b3JrIHF1ZXVlIHdpbGwgbm93CiAgICBxdWV1ZSB0aGUgdGFzayBhIGZl dyBleHRyYSB0aW1lcy4gIFdlIGNhbiBhZGRyZXNzIHRoYXQgc3Vib3B0aW1hbAogICAgYmVoYXZp b3IgaW4gZnV0dXJlIHBhdGNoZXMuCiAgICAKICAgIFNpZ25lZC1vZmYtYnk6IERvdWcgTGVkZm9y ZCA8ZGxlZGZvcmRAcmVkaGF0LmNvbT4KCmRpZmYgLS1naXQgYS9kcml2ZXJzL2luZmluaWJhbmQv dWxwL2lwb2liL2lwb2liX211bHRpY2FzdC5jIGIvZHJpdmVycy9pbmZpbmliYW5kL3VscC9pcG9p Yi9pcG9pYl9tdWx0aWNhc3QuYwppbmRleCA4YTUzOGMwMTBiOS4uY2JhNmUxNjBkZjIgMTAwNjQ0 Ci0tLSBhL2RyaXZlcnMvaW5maW5pYmFuZC91bHAvaXBvaWIvaXBvaWJfbXVsdGljYXN0LmMKKysr IGIvZHJpdmVycy9pbmZpbmliYW5kL3VscC9pcG9pYi9pcG9pYl9tdWx0aWNhc3QuYwpAQCAtNjM4 LDggKzYzOCw2IEBAIHZvaWQgaXBvaWJfbWNhc3Rfam9pbl90YXNrKHN0cnVjdCB3b3JrX3N0cnVj dCAqd29yaykKIAl9CiAKIAlpcG9pYl9kYmdfbWNhc3QocHJpdiwgInN1Y2Nlc3NmdWxseSBqb2lu ZWQgYWxsIG11bHRpY2FzdCBncm91cHNcbiIpOwotCi0JY2xlYXJfYml0KElQT0lCX01DQVNUX1JV TiwgJnByaXYtPmZsYWdzKTsKIH0KIAogaW50IGlwb2liX21jYXN0X3N0YXJ0X3RocmVhZChzdHJ1 Y3QgbmV0X2RldmljZSAqZGV2KQpAQCAtNjQ5LDggKzY0Nyw4IEBAIGludCBpcG9pYl9tY2FzdF9z dGFydF90aHJlYWQoc3RydWN0IG5ldF9kZXZpY2UgKmRldikKIAlpcG9pYl9kYmdfbWNhc3QocHJp diwgInN0YXJ0aW5nIG11bHRpY2FzdCB0aHJlYWRcbiIpOwogCiAJbXV0ZXhfbG9jaygmbWNhc3Rf bXV0ZXgpOwotCWlmICghdGVzdF9hbmRfc2V0X2JpdChJUE9JQl9NQ0FTVF9SVU4sICZwcml2LT5m bGFncykpCi0JCXF1ZXVlX2RlbGF5ZWRfd29yayhwcml2LT53cSwgJnByaXYtPm1jYXN0X3Rhc2ss IDApOworCXNldF9iaXQoSVBPSUJfTUNBU1RfUlVOLCAmcHJpdi0+ZmxhZ3MpOworCXF1ZXVlX2Rl bGF5ZWRfd29yayhwcml2LT53cSwgJnByaXYtPm1jYXN0X3Rhc2ssIDApOwogCW11dGV4X3VubG9j aygmbWNhc3RfbXV0ZXgpOwogCiAJcmV0dXJuIDA7CkBAIC03MzMsNyArNzMxLDcgQEAgdm9pZCBp cG9pYl9tY2FzdF9zZW5kKHN0cnVjdCBuZXRfZGV2aWNlICpkZXYsIHU4ICpkYWRkciwgc3RydWN0 IHNrX2J1ZmYgKnNrYikKIAkJbWVtY3B5KG1jYXN0LT5tY21lbWJlci5tZ2lkLnJhdywgbWdpZCwg c2l6ZW9mICh1bmlvbiBpYl9naWQpKTsKIAkJX19pcG9pYl9tY2FzdF9hZGQoZGV2LCBtY2FzdCk7 CiAJCWxpc3RfYWRkX3RhaWwoJm1jYXN0LT5saXN0LCAmcHJpdi0+bXVsdGljYXN0X2xpc3QpOwot CQlpZiAoIXRlc3RfYW5kX3NldF9iaXQoSVBPSUJfTUNBU1RfUlVOLCAmcHJpdi0+ZmxhZ3MpKQor CQlpZiAodGVzdF9iaXQoSVBPSUJfTUNBU1RfUlVOLCAmcHJpdi0+ZmxhZ3MpKQogCQkJcXVldWVf ZGVsYXllZF93b3JrKHByaXYtPndxLCAmcHJpdi0+bWNhc3RfdGFzaywgMCk7CiAJfQogCkBAIC05 NTksNyArOTU3LDggQEAgdm9pZCBpcG9pYl9tY2FzdF9yZXN0YXJ0X3Rhc2soc3RydWN0IHdvcmtf c3RydWN0ICp3b3JrKQogCS8qCiAJICogUmVzdGFydCBvdXIgam9pbiB0YXNrIGlmIG5lZWRlZAog CSAqLwotCWlwb2liX21jYXN0X3N0YXJ0X3RocmVhZChkZXYpOworCWlmICh0ZXN0X2JpdChJUE9J Ql9NQ0FTVF9SVU4sICZwcml2LT5mbGFncykpCisJCXF1ZXVlX2RlbGF5ZWRfd29yayhwcml2LT53 cSwgJnByaXYtPm1jYXN0X3Rhc2ssIDApOwogCXJ0bmxfdW5sb2NrKCk7CiB9CiAK --=-0TbpH9TgXhteuE4LhfY4-- --=-U0Grz3J/tS3PMqYO0wIA Content-Type: application/pgp-signature; name="signature.asc" Content-Description: This is a digitally signed message part Content-Transfer-Encoding: 7bit -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAABAgAGBQJUtYeNAAoJELgmozMOVy/dESIP/07CV72EjFij4t9+gybewiXq YvKNAAHzo7cvdsxaK8Atdup0s8qD5KX8YUjtv2eu+6swkEYxpmWZ1L26xRkNkcmm bcyKfjXwT2Nl95Hu07eVOMWD/oqM2Ym5mOnNmBTIekJJ5wGyEfH5kLz2Lpkzb1gc sSuImkek03XpvyFoNG/i9GNXNovVHx2795ji+erCrP8NZOp/nr34IYG3asuaWGbo tFLYkIxfmaTh9EUFsUZHHkbDX16yL0N1CIMuEg48UcePnL7nW9CvkEXzzptBwuUT kaEEkwkzUkZJED3hSYV9+6VteAhL9LKzPRrV7kXJfv9pN01Ij3Eb8Bh3HfALiaqQ CpD4O1ZMqr3/k1JqEjVQjGCsh6RUsFr3/eUEIUwLvNHP+TD/AkRfqjYRP2QC+xX8 G0BZozXYBukBUisJELpflGEBEYRwAt6SyFbc3g0ECFzwLMMlWk0v69OZwox37A/O Y+CrY70MZPIW1uFtI1HSSIV+RuJjkcuBVjttbzM979hwVYE6YCmixas4gNU6LgVm GdiQbC06iXLcmu5ypX07cSZpWewT7bdrWoRUr3Z39lYoOWBjdxuEOKARO+BM/0RZ oPCa2l6aM8XMYbUcekbE9810MB2PNXwM7TzA4OQJJSlZTNU7YCXNvQJFdaIJPYfU gGGh1vG2FJDGdbjMZdjQ =163I -----END PGP SIGNATURE----- --=-U0Grz3J/tS3PMqYO0wIA-- -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html