| Message ID | 1505885986-5873-1-git-send-email-Michal.Kalderon@cavium.com (mailing list archive) |
|---|---|
| State | Accepted |
| Headers | show |
On Wed, 2017-09-20 at 08:39 +0300, Michal Kalderon wrote: > Initialize the rdma_type (iWARP or RoCE) which is set according to > device configuration in qed. > > Fixes: e6a38c54faf ("RDMA/qedr: Add support for registering an iWARP > device") > > Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> > Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com> > > --- > This patch applies cleanly on k.o/for-next-merged > > Changes from V0: > - Added Fixes in commit comment. Sorry, I was wrong in my email to the previous posting of this patch. I had pulled this v1 patch from patchworks, not the original patch, and so the Fixes: line was yours. Patchworks did not add the Fixes: line to the v0 patch when I downloaded it just now. That being said, I still can't build with this patch ;-)
From: Doug Ledford <dledford@redhat.com> Sent: Wednesday, September 27, 2017 4:27 PM >On Wed, 2017-09-20 at 08:39 +0300, Michal Kalderon wrote: >> Initialize the rdma_type (iWARP or RoCE) which is set according to >> device configuration in qed. >> >> Fixes: e6a38c54faf ("RDMA/qedr: Add support for registering an iWARP >> device") >> >> Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> >> Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com> >> >> --- >> This patch applies cleanly on k.o/for-next-merged >> >> Changes from V0: >> - Added Fixes in commit comment. > >Sorry, I was wrong in my email to the previous posting of this patch. >I had pulled this v1 patch from patchworks, not the original patch, and >so the Fixes: line was yours. Patchworks did not add the Fixes: line >to the v0 patch when I downloaded it just now. >That being said, I still can't build with this patch ;-) Are you building against the rdma-next branch ? The patch requires the iWARP series, originally this was tested against for-next-merged, what's the error you're getting ? thanks, Michal -- Doug Ledford <dledford@redhat.com> GPG KeyID: B826A3330E572FDD Key fingerprint = AE6B 1BDA 122B 23B4 265B 1274 B826 A333 0E57 2FDD -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Sep 27, 2017 at 04:18:50PM +0000, Kalderon, Michal wrote: > From: Doug Ledford <dledford@redhat.com> > Sent: Wednesday, September 27, 2017 4:27 PM > > >On Wed, 2017-09-20 at 08:39 +0300, Michal Kalderon wrote: > >> Initialize the rdma_type (iWARP or RoCE) which is set according to > >> device configuration in qed. > >> > >> Fixes: e6a38c54faf ("RDMA/qedr: Add support for registering an iWARP > >> device") > >> > >> Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> > >> Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com> > >> > >> --- > >> This patch applies cleanly on k.o/for-next-merged > >> > >> Changes from V0: > >> - Added Fixes in commit comment. > > > >Sorry, I was wrong in my email to the previous posting of this patch. > >I had pulled this v1 patch from patchworks, not the original patch, and > >so the Fixes: line was yours. Patchworks did not add the Fixes: line > >to the v0 patch when I downloaded it just now. > > >That being said, I still can't build with this patch ;-) > Are you building against the rdma-next branch ? > The patch requires the iWARP series, originally this was tested against for-next-merged, > what's the error you're getting ? On rdma-next: drivers/infiniband/hw/qedr/main.c: In function ‘qedr_add’: drivers/infiniband/hw/qedr/main.c:780:7: error: ‘struct qedr_dev’ has no member named ‘rdma_type’; did you mean ‘rdma_ctx’? dev->rdma_type = dev_info.rdma_type; ^~~~~~~~~ rdma_ctx make[4]: *** [scripts/Makefile.build:311: drivers/infiniband/hw/qedr/main.o] Error 1 > > thanks, > Michal > > -- > Doug Ledford <dledford@redhat.com> > GPG KeyID: B826A3330E572FDD > Key fingerprint = AE6B 1BDA 122B 23B4 265B 1274 B826 A333 0E57 2FDD >
On Wed, Sep 27, 2017 at 07:53:39PM +0300, Leon Romanovsky wrote: > On Wed, Sep 27, 2017 at 04:18:50PM +0000, Kalderon, Michal wrote: > > From: Doug Ledford <dledford@redhat.com> > > Sent: Wednesday, September 27, 2017 4:27 PM > > > > >On Wed, 2017-09-20 at 08:39 +0300, Michal Kalderon wrote: > > >> Initialize the rdma_type (iWARP or RoCE) which is set according to > > >> device configuration in qed. > > >> > > >> Fixes: e6a38c54faf ("RDMA/qedr: Add support for registering an iWARP > > >> device") > > >> > > >> Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> > > >> Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com> > > >> > > >> --- > > >> This patch applies cleanly on k.o/for-next-merged > > >> > > >> Changes from V0: > > >> - Added Fixes in commit comment. > > > > > >Sorry, I was wrong in my email to the previous posting of this patch. > > >I had pulled this v1 patch from patchworks, not the original patch, and > > >so the Fixes: line was yours. Patchworks did not add the Fixes: line > > >to the v0 patch when I downloaded it just now. > > > > >That being said, I still can't build with this patch ;-) > > Are you building against the rdma-next branch ? > > The patch requires the iWARP series, originally this was tested against for-next-merged, > > what's the error you're getting ? > > On rdma-next: > drivers/infiniband/hw/qedr/main.c: In function ‘qedr_add’: > drivers/infiniband/hw/qedr/main.c:780:7: error: ‘struct qedr_dev’ has no member named ‘rdma_type’; did you mean ‘rdma_ctx’? > dev->rdma_type = dev_info.rdma_type; > ^~~~~~~~~ > rdma_ctx > make[4]: *** [scripts/Makefile.build:311: drivers/infiniband/hw/qedr/main.o] Error 1 My config: ➜ linux-rdma git:(rdma-next) grep QED .config CONFIG_QED=m CONFIG_QED_LL2=y CONFIG_QED_SRIOV=y CONFIG_QEDE=m CONFIG_QED_RDMA=y CONFIG_INFINIBAND_QEDR=m > > > > > > thanks, > > Michal > > > > -- > > Doug Ledford <dledford@redhat.com> > > GPG KeyID: B826A3330E572FDD > > Key fingerprint = AE6B 1BDA 122B 23B4 265B 1274 B826 A333 0E57 2FDD > >
From: Leon Romanovsky <leon@kernel.org> Sent: Wednesday, September 27, 2017 7:56 PM On Wed, Sep 27, 2017 at 07:53:39PM +0300, Leon Romanovsky wrote: >> On Wed, Sep 27, 2017 at 04:18:50PM +0000, Kalderon, Michal wrote: >> > From: Doug Ledford <dledford@redhat.com> >> > Sent: Wednesday, September 27, 2017 4:27 PM >> > >> > >On Wed, 2017-09-20 at 08:39 +0300, Michal Kalderon wrote: >> > >> Initialize the rdma_type (iWARP or RoCE) which is set according to >> > >> device configuration in qed. >> > >> >> > >> Fixes: e6a38c54faf ("RDMA/qedr: Add support for registering an iWARP >> > >> device") >> > >> >> > >> Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> >> > >> Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com> >> > >> >> > >> --- >> > >> This patch applies cleanly on k.o/for-next-merged >> > >> >> > >> Changes from V0: >> > >> - Added Fixes in commit comment. >> > > >> > >Sorry, I was wrong in my email to the previous posting of this patch. >> > >I had pulled this v1 patch from patchworks, not the original patch, and >> > >so the Fixes: line was yours. Patchworks did not add the Fixes: line >> > >to the v0 patch when I downloaded it just now. >> > >> > >That being said, I still can't build with this patch ;-) >> > Are you building against the rdma-next branch ? >> > The patch requires the iWARP series, originally this was tested against for-next-merged, >> > what's the error you're getting ? >> >> On rdma-next: >> drivers/infiniband/hw/qedr/main.c: In function ‘qedr_add’: >> drivers/infiniband/hw/qedr/main.c:780:7: error: ‘struct qedr_dev’ has no member named ‘rdma_type’; did you mean ‘rdma_ctx’? >> dev->rdma_type = dev_info.rdma_type; >> ^~~~~~~~~ >> rdma_ctx >> make[4]: *** [scripts/Makefile.build:311: drivers/infiniband/hw/qedr/main.o] Error 1 > >My config: > >➜ linux-rdma git:(rdma-next) grep QED .config >CONFIG_QED=m >CONFIG_QED_LL2=y >CONFIG_QED_SRIOV=y >CONFIG_QEDE=m >CONFIG_QED_RDMA=y >CONFIG_INFINIBAND_QEDR=m Thanks Leon for helping out with this. Unfortunately i'm not seeing this. The rdma_type field was added with the following patch: https://kernel.googlesource.com/pub/scm/linux/kernel/git/dledford/rdma/+/e6a38c54faf38498170e227c82ea25cb8bc1ae71 When running git log on the k.o/for-next I see the patch (last in list below), applying this patch on top of rdma-next works for me and compiles cleanly. 2e2f2d0 RDMA/qedr: Fix rdma_type initialization 1848757 Merge branches 'hns' and 'misc' into k.o/for-next 931bc0d IB: Move PCI dependency from root KConfig to HW's KConfigs 7ae6f2a Merge branch 'qedr' into k.o/for-next bd491d2 RDMA/qedr: fix build error without ipv6 89fd257 RDMA/qedr: Missing error code in qedr_init_user_queue() 847cb1a RDMA/qedr: fix spelling mistake: "invlaid" -> "invalid" 69ad0e7 RDMA/qedr: Add support for iWARP in user space e411e05 RDMA/qedr: Add iWARP connection management functions de0089e RDMA/qedr: Add iWARP connection management qp related callbacks fb1a22b RDMA/qedr: Add support for read with invalidate, supported in iWARP f5b1b17 RDMA/qedr: Add iWARP support in existing verbs e6a38c5 RDMA/qedr: Add support for registering an iWARP device am i missing something ? thanks, Michal
On Wed, Sep 27, 2017 at 08:33:38PM +0000, Kalderon, Michal wrote: > From: Leon Romanovsky <leon@kernel.org> > Sent: Wednesday, September 27, 2017 7:56 PM > On Wed, Sep 27, 2017 at 07:53:39PM +0300, Leon Romanovsky wrote: > >> On Wed, Sep 27, 2017 at 04:18:50PM +0000, Kalderon, Michal wrote: > >> > From: Doug Ledford <dledford@redhat.com> > >> > Sent: Wednesday, September 27, 2017 4:27 PM > >> > > >> > >On Wed, 2017-09-20 at 08:39 +0300, Michal Kalderon wrote: > >> > >> Initialize the rdma_type (iWARP or RoCE) which is set according to > >> > >> device configuration in qed. > >> > >> > >> > >> Fixes: e6a38c54faf ("RDMA/qedr: Add support for registering an iWARP > >> > >> device") > >> > >> > >> > >> Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> > >> > >> Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com> > >> > >> > >> > >> --- > >> > >> This patch applies cleanly on k.o/for-next-merged > >> > >> > >> > >> Changes from V0: > >> > >> - Added Fixes in commit comment. > >> > > > >> > >Sorry, I was wrong in my email to the previous posting of this patch. > >> > >I had pulled this v1 patch from patchworks, not the original patch, and > >> > >so the Fixes: line was yours. Patchworks did not add the Fixes: line > >> > >to the v0 patch when I downloaded it just now. > >> > > >> > >That being said, I still can't build with this patch ;-) > >> > Are you building against the rdma-next branch ? > >> > The patch requires the iWARP series, originally this was tested against for-next-merged, > >> > what's the error you're getting ? > >> > >> On rdma-next: > >> drivers/infiniband/hw/qedr/main.c: In function ‘qedr_add’: > >> drivers/infiniband/hw/qedr/main.c:780:7: error: ‘struct qedr_dev’ has no member named ‘rdma_type’; did you mean ‘rdma_ctx’? > >> dev->rdma_type = dev_info.rdma_type; > >> ^~~~~~~~~ > >> rdma_ctx > >> make[4]: *** [scripts/Makefile.build:311: drivers/infiniband/hw/qedr/main.o] Error 1 > > > >My config: > > > >➜ linux-rdma git:(rdma-next) grep QED .config > >CONFIG_QED=m > >CONFIG_QED_LL2=y > >CONFIG_QED_SRIOV=y > >CONFIG_QEDE=m > >CONFIG_QED_RDMA=y > >CONFIG_INFINIBAND_QEDR=m > > Thanks Leon for helping out with this. Unfortunately i'm not seeing this. > The rdma_type field was added with the following patch: > https://kernel.googlesource.com/pub/scm/linux/kernel/git/dledford/rdma/+/e6a38c54faf38498170e227c82ea25cb8bc1ae71 > > When running git log on the k.o/for-next I see the patch (last in list below), > applying this patch on top of rdma-next works for me and compiles cleanly. > > 2e2f2d0 RDMA/qedr: Fix rdma_type initialization > 1848757 Merge branches 'hns' and 'misc' into k.o/for-next > 931bc0d IB: Move PCI dependency from root KConfig to HW's KConfigs > 7ae6f2a Merge branch 'qedr' into k.o/for-next > bd491d2 RDMA/qedr: fix build error without ipv6 > 89fd257 RDMA/qedr: Missing error code in qedr_init_user_queue() > 847cb1a RDMA/qedr: fix spelling mistake: "invlaid" -> "invalid" > 69ad0e7 RDMA/qedr: Add support for iWARP in user space > e411e05 RDMA/qedr: Add iWARP connection management functions > de0089e RDMA/qedr: Add iWARP connection management qp related callbacks > fb1a22b RDMA/qedr: Add support for read with invalidate, supported in iWARP > f5b1b17 RDMA/qedr: Add iWARP support in existing verbs > e6a38c5 RDMA/qedr: Add support for registering an iWARP device > > am i missing something ? I refreshed my branches now and it worked, but your git log history looks strange. Where did you see the commit 1848757 and 931bc0d? https://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma.git/log/?h=k.o/for-next Age Commit message (Expand) Author Files Lines 3 days Merge branch 'qedr' into k.o/for-nextfor-nextk.o/for-next Doug Ledford 11 -91/+1208 3 days Merge tag 'v4.14-rc2' into k.o/for-next Doug Ledford 12044 -388475/+631018 3 days Linux 4.14-rc2HEADmaster Linus Torvalds 1 -1/+1 ➜ linux-rdma git:(iwarp-compile) git log --oneline f3506957237b (HEAD -> iwarp-compile) RDMA/qedr: Fix rdma_type initialization 7ae6f2a3d552 (rdma/k.o/for-next, grdma/k.o/for-next) Merge branch 'qedr' into k.o/for-next 0d9c2ff1c9f7 Merge tag 'v4.14-rc2' into k.o/for-next e19b205be43d (tag: v4.14-rc2, rdma/master, origin/master, ml/master, grdma/master, grdma/hns, backup/master, master) Linux 4.14-rc2 Thanks > thanks, > Michal >
PiBGcm9tOiBMZW9uIFJvbWFub3Zza3kgW21haWx0bzpsZW9uQGtlcm5lbC5vcmddDQo+IFNlbnQ6 IFRodXJzZGF5LCBTZXB0ZW1iZXIgMjgsIDIwMTcgNjoyNyBBTQ0KPiBPbiBXZWQsIFNlcCAyNywg MjAxNyBhdCAwODozMzozOFBNICswMDAwLCBLYWxkZXJvbiwgTWljaGFsIHdyb3RlOg0KPiA+IEZy b206IExlb24gUm9tYW5vdnNreSA8bGVvbkBrZXJuZWwub3JnPg0KPiA+IFNlbnQ6IFdlZG5lc2Rh eSwgU2VwdGVtYmVyIDI3LCAyMDE3IDc6NTYgUE0gT24gV2VkLCBTZXAgMjcsIDIwMTcgYXQNCj4g PiAwNzo1MzozOVBNICswMzAwLCBMZW9uIFJvbWFub3Zza3kgd3JvdGU6DQo+ID4gPj4gT24gV2Vk LCBTZXAgMjcsIDIwMTcgYXQgMDQ6MTg6NTBQTSArMDAwMCwgS2FsZGVyb24sIE1pY2hhbCB3cm90 ZToNCj4gPiA+PiA+IEZyb206IERvdWcgTGVkZm9yZCA8ZGxlZGZvcmRAcmVkaGF0LmNvbT4NCj4g PiA+PiA+IFNlbnQ6IFdlZG5lc2RheSwgU2VwdGVtYmVyIDI3LCAyMDE3IDQ6MjcgUE0NCj4gPiA+ PiA+DQo+ID4gPj4gPiA+T24gV2VkLCAyMDE3LTA5LTIwIGF0IDA4OjM5ICswMzAwLCBNaWNoYWwg S2FsZGVyb24gd3JvdGU6DQo+ID4gPj4gPiA+PiBJbml0aWFsaXplIHRoZSByZG1hX3R5cGUgKGlX QVJQIG9yIFJvQ0UpIHdoaWNoIGlzIHNldA0KPiA+ID4+ID4gPj4gYWNjb3JkaW5nIHRvIGRldmlj ZSBjb25maWd1cmF0aW9uIGluIHFlZC4NCj4gPiA+PiA+ID4+DQo+ID4gPj4gPiA+PiBGaXhlczog ZTZhMzhjNTRmYWYgKCJSRE1BL3FlZHI6IEFkZCBzdXBwb3J0IGZvciByZWdpc3RlcmluZyBhbg0K PiA+ID4+ID4gPj4gaVdBUlANCj4gPiA+PiA+ID4+IGRldmljZSIpDQo+ID4gPj4gPiA+Pg0KPiA+ ID4+ID4gPj4gU2lnbmVkLW9mZi1ieTogTWljaGFsIEthbGRlcm9uIDxNaWNoYWwuS2FsZGVyb25A Y2F2aXVtLmNvbT4NCj4gPiA+PiA+ID4+IFNpZ25lZC1vZmYtYnk6IEFyaWVsIEVsaW9yIDxBcmll bC5FbGlvckBjYXZpdW0uY29tPg0KPiA+ID4+ID4gPj4NCj4gPiA+PiA+ID4+IC0tLQ0KPiA+ID4+ ID4gPj4gVGhpcyBwYXRjaCBhcHBsaWVzIGNsZWFubHkgb24gay5vL2Zvci1uZXh0LW1lcmdlZA0K PiA+ID4+ID4gPj4NCj4gPiA+PiA+ID4+IENoYW5nZXMgZnJvbSBWMDoNCj4gPiA+PiA+ID4+IC0g QWRkZWQgRml4ZXMgaW4gY29tbWl0IGNvbW1lbnQuDQo+ID4gPj4gPiA+DQo+ID4gPj4gPiA+U29y cnksIEkgd2FzIHdyb25nIGluIG15IGVtYWlsIHRvIHRoZSBwcmV2aW91cyBwb3N0aW5nIG9mIHRo aXMgcGF0Y2guDQo+ID4gPj4gPiA+SSBoYWQgcHVsbGVkIHRoaXMgdjEgcGF0Y2ggZnJvbSBwYXRj aHdvcmtzLCBub3QgdGhlIG9yaWdpbmFsDQo+ID4gPj4gPiA+cGF0Y2gsIGFuZCBzbyB0aGUgRml4 ZXM6IGxpbmUgd2FzIHlvdXJzLiAgUGF0Y2h3b3JrcyBkaWQgbm90IGFkZA0KPiA+ID4+ID4gPnRo ZSBGaXhlczogbGluZSB0byB0aGUgdjAgcGF0Y2ggd2hlbiBJIGRvd25sb2FkZWQgaXQganVzdCBu b3cuDQo+ID4gPj4gPg0KPiA+ID4+ID4gPlRoYXQgYmVpbmcgc2FpZCwgSSBzdGlsbCBjYW4ndCBi dWlsZCB3aXRoIHRoaXMgcGF0Y2ggOy0pDQo+ID4gPj4gPiBBcmUgeW91IGJ1aWxkaW5nIGFnYWlu c3QgdGhlIHJkbWEtbmV4dCBicmFuY2ggPw0KPiA+ID4+ID4gVGhlIHBhdGNoIHJlcXVpcmVzIHRo ZSBpV0FSUCBzZXJpZXMsIG9yaWdpbmFsbHkgdGhpcyB3YXMgdGVzdGVkDQo+ID4gPj4gPiBhZ2Fp bnN0IGZvci1uZXh0LW1lcmdlZCwgd2hhdCdzIHRoZSBlcnJvciB5b3UncmUgZ2V0dGluZyA/DQo+ ID4gPj4NCj4gPiA+PiBPbiByZG1hLW5leHQ6DQo+ID4gPj4gZHJpdmVycy9pbmZpbmliYW5kL2h3 L3FlZHIvbWFpbi5jOiBJbiBmdW5jdGlvbiDigJhxZWRyX2FkZOKAmToNCj4gPiA+PiBkcml2ZXJz L2luZmluaWJhbmQvaHcvcWVkci9tYWluLmM6NzgwOjc6IGVycm9yOiDigJhzdHJ1Y3QgcWVkcl9k ZXbigJkgaGFzDQo+IG5vIG1lbWJlciBuYW1lZCDigJhyZG1hX3R5cGXigJk7IGRpZCB5b3UgbWVh biDigJhyZG1hX2N0eOKAmT8NCj4gPiA+PiAgIGRldi0+cmRtYV90eXBlID0gZGV2X2luZm8ucmRt YV90eXBlOw0KPiA+ID4+ICAgICAgICBefn5+fn5+fn4NCj4gPiA+PiAgICAgICAgcmRtYV9jdHgN Cj4gPiA+PiBtYWtlWzRdOiAqKiogW3NjcmlwdHMvTWFrZWZpbGUuYnVpbGQ6MzExOg0KPiA+ID4+ IGRyaXZlcnMvaW5maW5pYmFuZC9ody9xZWRyL21haW4ub10gRXJyb3IgMQ0KPiA+ID4NCj4gPiA+ TXkgY29uZmlnOg0KPiA+ID4NCj4gPiA+4p6cICBsaW51eC1yZG1hIGdpdDoocmRtYS1uZXh0KSBn cmVwIFFFRCAuY29uZmlnIENPTkZJR19RRUQ9bQ0KPiA+ID5DT05GSUdfUUVEX0xMMj15IENPTkZJ R19RRURfU1JJT1Y9eSBDT05GSUdfUUVERT1tDQo+IENPTkZJR19RRURfUkRNQT15DQo+ID4gPkNP TkZJR19JTkZJTklCQU5EX1FFRFI9bQ0KPiA+DQo+ID4gVGhhbmtzIExlb24gZm9yIGhlbHBpbmcg b3V0IHdpdGggdGhpcy4gVW5mb3J0dW5hdGVseSBpJ20gbm90IHNlZWluZyB0aGlzLg0KPiA+IFRo ZSByZG1hX3R5cGUgZmllbGQgd2FzIGFkZGVkIHdpdGggdGhlIGZvbGxvd2luZyBwYXRjaDoNCj4g Pg0KPiBodHRwczovL2tlcm5lbC5nb29nbGVzb3VyY2UuY29tL3B1Yi9zY20vbGludXgva2VybmVs L2dpdC9kbGVkZm9yZC9yZG1hDQo+ID4gLysvZTZhMzhjNTRmYWYzODQ5ODE3MGUyMjdjODJlYTI1 Y2I4YmMxYWU3MQ0KPiA+DQo+ID4gV2hlbiBydW5uaW5nIGdpdCBsb2cgb24gdGhlIGsuby9mb3It bmV4dCBJIHNlZSB0aGUgcGF0Y2ggKGxhc3QgaW4gbGlzdA0KPiA+IGJlbG93KSwgYXBwbHlpbmcg dGhpcyBwYXRjaCBvbiB0b3Agb2YgIHJkbWEtbmV4dCB3b3JrcyBmb3IgbWUgYW5kDQo+IGNvbXBp bGVzIGNsZWFubHkuDQo+ID4NCj4gPiAyZTJmMmQwIFJETUEvcWVkcjogRml4IHJkbWFfdHlwZSBp bml0aWFsaXphdGlvbg0KPiA+IDE4NDg3NTcgTWVyZ2UgYnJhbmNoZXMgJ2hucycgYW5kICdtaXNj JyBpbnRvIGsuby9mb3ItbmV4dCA5MzFiYzBkIElCOg0KPiA+IE1vdmUgUENJIGRlcGVuZGVuY3kg ZnJvbSByb290IEtDb25maWcgdG8gSFcncyBLQ29uZmlncyA3YWU2ZjJhIE1lcmdlDQo+ID4gYnJh bmNoICdxZWRyJyBpbnRvIGsuby9mb3ItbmV4dA0KPiA+IGJkNDkxZDIgUkRNQS9xZWRyOiBmaXgg YnVpbGQgZXJyb3Igd2l0aG91dCBpcHY2DQo+ID4gODlmZDI1NyBSRE1BL3FlZHI6IE1pc3Npbmcg ZXJyb3IgY29kZSBpbiBxZWRyX2luaXRfdXNlcl9xdWV1ZSgpDQo+ID4gODQ3Y2IxYSBSRE1BL3Fl ZHI6IGZpeCBzcGVsbGluZyBtaXN0YWtlOiAiaW52bGFpZCIgLT4gImludmFsaWQiDQo+ID4gNjlh ZDBlNyBSRE1BL3FlZHI6IEFkZCBzdXBwb3J0IGZvciBpV0FSUCBpbiB1c2VyIHNwYWNlDQo+ID4g ZTQxMWUwNSBSRE1BL3FlZHI6IEFkZCBpV0FSUCBjb25uZWN0aW9uIG1hbmFnZW1lbnQgZnVuY3Rp b25zDQo+IGRlMDA4OWUNCj4gPiBSRE1BL3FlZHI6IEFkZCBpV0FSUCBjb25uZWN0aW9uIG1hbmFn ZW1lbnQgcXAgcmVsYXRlZCBjYWxsYmFja3MNCj4gPiBmYjFhMjJiIFJETUEvcWVkcjogQWRkIHN1 cHBvcnQgZm9yIHJlYWQgd2l0aCBpbnZhbGlkYXRlLCBzdXBwb3J0ZWQgaW4NCj4gPiBpV0FSUA0K PiA+IGY1YjFiMTcgUkRNQS9xZWRyOiBBZGQgaVdBUlAgc3VwcG9ydCBpbiBleGlzdGluZyB2ZXJi cw0KPiA+IGU2YTM4YzUgUkRNQS9xZWRyOiBBZGQgc3VwcG9ydCBmb3IgcmVnaXN0ZXJpbmcgYW4g aVdBUlAgZGV2aWNlDQo+ID4NCj4gPiBhbSBpIG1pc3Npbmcgc29tZXRoaW5nID8NCj4gDQo+IEkg cmVmcmVzaGVkIG15IGJyYW5jaGVzIG5vdyBhbmQgaXQgd29ya2VkLCBidXQgeW91ciBnaXQgbG9n IGhpc3RvcnkgbG9va3MNCj4gc3RyYW5nZS4gV2hlcmUgZGlkIHlvdSBzZWUgdGhlIGNvbW1pdCAx ODQ4NzU3IGFuZCA5MzFiYzBkPw0KDQpnaXQ6Ly9naXRodWIuY29tL2RsZWRmb3JkL2xpbnV4Lmdp dA0KZ2l0IGxvZyAtLW9uZWxpbmUgLS0gZHJpdmVycy9pbmZpbmliYW5kL2h3L3FlZHINCmh0dHBz Oi8vZ2l0aHViLmNvbS9kbGVkZm9yZC9saW51eC9jb21taXQvMTg0ODc1N2MzZDhiZmYzYWRkZjg4 MWM4YmNjMmI4NWUyNWNjZjI1ZQ0KDQpJIHdhcyB1bmRlciB0aGUgYXNzdW1wdGlvbiBhIHBhdGNo IGdvZXMgdG8gZ2l0aHViIGJlZm9yZSBrLm8gLSBpcyB0aGlzIG5vdCB0aGUgY2FzZSA/IA0KVGhh bmtzIGFnYWluLCANCg0KRG91ZyAtIGFyZSB5b3Ugc3RpbGwgc2VlaW5nIGlzc3VlcyB3aXRoIHRo aXMgcGF0Y2ggPyANClRoYW5rcywNCk1pY2hhbA0KDQo+IA0KPiBodHRwczovL2dpdC5rZXJuZWwu b3JnL3B1Yi9zY20vbGludXgva2VybmVsL2dpdC9kbGVkZm9yZC9yZG1hLmdpdC9sb2cvP2g9aw0K PiAuby9mb3ItbmV4dA0KPiBBZ2UJQ29tbWl0IG1lc3NhZ2UgKEV4cGFuZCkJQXV0aG9yCUZpbGVz CUxpbmVzDQo+IDMgZGF5cwlNZXJnZSBicmFuY2ggJ3FlZHInIGludG8gay5vL2Zvci1uZXh0Zm9y LW5leHRrLm8vZm9yLW5leHQJIERvdWcNCj4gTGVkZm9yZAkxMQktOTEvKzEyMDgNCj4gMyBkYXlz CU1lcmdlIHRhZyAndjQuMTQtcmMyJyBpbnRvIGsuby9mb3ItbmV4dAkgRG91ZyBMZWRmb3JkCTEy MDQ0CS0NCj4gMzg4NDc1Lys2MzEwMTgNCj4gMyBkYXlzCUxpbnV4IDQuMTQtcmMySEVBRG1hc3Rl cgkgTGludXMgVG9ydmFsZHMJMQktMS8rMQ0KPiANCj4g4p6cICBsaW51eC1yZG1hIGdpdDooaXdh cnAtY29tcGlsZSkgZ2l0IGxvZyAtLW9uZWxpbmUgZjM1MDY5NTcyMzdiIChIRUFEIC0+DQo+IGl3 YXJwLWNvbXBpbGUpIFJETUEvcWVkcjogRml4IHJkbWFfdHlwZSBpbml0aWFsaXphdGlvbg0KPiA3 YWU2ZjJhM2Q1NTIgKHJkbWEvay5vL2Zvci1uZXh0LCBncmRtYS9rLm8vZm9yLW5leHQpIE1lcmdl IGJyYW5jaCAncWVkcicNCj4gaW50byBrLm8vZm9yLW5leHQNCj4gMGQ5YzJmZjFjOWY3IE1lcmdl IHRhZyAndjQuMTQtcmMyJyBpbnRvIGsuby9mb3ItbmV4dCBlMTliMjA1YmU0M2QgKHRhZzoNCj4g djQuMTQtcmMyLCByZG1hL21hc3Rlciwgb3JpZ2luL21hc3RlciwgbWwvbWFzdGVyLCBncmRtYS9t YXN0ZXIsDQo+IGdyZG1hL2hucywgYmFja3VwL21hc3RlciwgbWFzdGVyKSBMaW51eCA0LjE0LXJj Mg0KPiANCj4gVGhhbmtzDQo+IA0KPiA+IHRoYW5rcywNCj4gPiBNaWNoYWwNCj4gPg0K -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Thu, Sep 28, 2017 at 06:22:59AM +0000, Kalderon, Michal wrote: > > > > > > When running git log on the k.o/for-next I see the patch (last in list > > > below), applying this patch on top of rdma-next works for me and > > compiles cleanly. > > > > > > 2e2f2d0 RDMA/qedr: Fix rdma_type initialization > > > 1848757 Merge branches 'hns' and 'misc' into k.o/for-next 931bc0d IB: > > > Move PCI dependency from root KConfig to HW's KConfigs 7ae6f2a Merge > > > branch 'qedr' into k.o/for-next > > > bd491d2 RDMA/qedr: fix build error without ipv6 > > > 89fd257 RDMA/qedr: Missing error code in qedr_init_user_queue() > > > 847cb1a RDMA/qedr: fix spelling mistake: "invlaid" -> "invalid" > > > 69ad0e7 RDMA/qedr: Add support for iWARP in user space > > > e411e05 RDMA/qedr: Add iWARP connection management functions > > de0089e > > > RDMA/qedr: Add iWARP connection management qp related callbacks > > > fb1a22b RDMA/qedr: Add support for read with invalidate, supported in > > > iWARP > > > f5b1b17 RDMA/qedr: Add iWARP support in existing verbs > > > e6a38c5 RDMA/qedr: Add support for registering an iWARP device > > > > > > am i missing something ? > > > > I refreshed my branches now and it worked, but your git log history looks > > strange. Where did you see the commit 1848757 and 931bc0d? > > git://github.com/dledford/linux.git > git log --oneline -- drivers/infiniband/hw/qedr > https://github.com/dledford/linux/commit/1848757c3d8bff3addf881c8bcc2b85e25ccf25e > > I was under the assumption a patch goes to github before k.o - is this not the case ? Repo on github is Doug's private repository, while it is much safer to stick with one official on kernel.org. Thanks
> From: Leon Romanovsky [mailto:leon@kernel.org] > Sent: Thursday, September 28, 2017 9:50 AM > > On Thu, Sep 28, 2017 at 06:22:59AM +0000, Kalderon, Michal wrote: > > > > > > > > When running git log on the k.o/for-next I see the patch (last in > > > > list below), applying this patch on top of rdma-next works for me > > > > and > > > compiles cleanly. > > > > > > > > 2e2f2d0 RDMA/qedr: Fix rdma_type initialization > > > > 1848757 Merge branches 'hns' and 'misc' into k.o/for-next 931bc0d IB: > > > > Move PCI dependency from root KConfig to HW's KConfigs 7ae6f2a > > > > Merge branch 'qedr' into k.o/for-next > > > > bd491d2 RDMA/qedr: fix build error without ipv6 > > > > 89fd257 RDMA/qedr: Missing error code in qedr_init_user_queue() > > > > 847cb1a RDMA/qedr: fix spelling mistake: "invlaid" -> "invalid" > > > > 69ad0e7 RDMA/qedr: Add support for iWARP in user space > > > > e411e05 RDMA/qedr: Add iWARP connection management functions > > > de0089e > > > > RDMA/qedr: Add iWARP connection management qp related callbacks > > > > fb1a22b RDMA/qedr: Add support for read with invalidate, supported > > > > in iWARP > > > > f5b1b17 RDMA/qedr: Add iWARP support in existing verbs > > > > e6a38c5 RDMA/qedr: Add support for registering an iWARP device > > > > > > > > am i missing something ? > > > > > > I refreshed my branches now and it worked, but your git log history > > > looks strange. Where did you see the commit 1848757 and 931bc0d? > > > > git://github.com/dledford/linux.git > > git log --oneline -- drivers/infiniband/hw/qedr > > > https://github.com/dledford/linux/commit/1848757c3d8bff3addf881c8bcc2b > > 85e25ccf25e > > > > I was under the assumption a patch goes to github before k.o - is this not > the case ? > > Repo on github is Doug's private repository, while it is much safer to stick > with one official on kernel.org. > > Thanks Ok, thanks. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 9/28/2017 2:22 AM, Kalderon, Michal wrote: >> From: Leon Romanovsky [mailto:leon@kernel.org] >> Sent: Thursday, September 28, 2017 6:27 AM >> On Wed, Sep 27, 2017 at 08:33:38PM +0000, Kalderon, Michal wrote: >>> From: Leon Romanovsky <leon@kernel.org> >>> Sent: Wednesday, September 27, 2017 7:56 PM On Wed, Sep 27, 2017 at >>> 07:53:39PM +0300, Leon Romanovsky wrote: >>>>> On Wed, Sep 27, 2017 at 04:18:50PM +0000, Kalderon, Michal wrote: >>>>>> From: Doug Ledford <dledford@redhat.com> >>>>>> Sent: Wednesday, September 27, 2017 4:27 PM >>>>>> >>>>>>> On Wed, 2017-09-20 at 08:39 +0300, Michal Kalderon wrote: >>>>>>>> Initialize the rdma_type (iWARP or RoCE) which is set >>>>>>>> according to device configuration in qed. >>>>>>>> >>>>>>>> Fixes: e6a38c54faf ("RDMA/qedr: Add support for registering an >>>>>>>> iWARP >>>>>>>> device") >>>>>>>> >>>>>>>> Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> >>>>>>>> Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com> >>>>>>>> >>>>>>>> --- >>>>>>>> This patch applies cleanly on k.o/for-next-merged >>>>>>>> >>>>>>>> Changes from V0: >>>>>>>> - Added Fixes in commit comment. >>>>>>> >>>>>>> Sorry, I was wrong in my email to the previous posting of this patch. >>>>>>> I had pulled this v1 patch from patchworks, not the original >>>>>>> patch, and so the Fixes: line was yours. Patchworks did not add >>>>>>> the Fixes: line to the v0 patch when I downloaded it just now. >>>>>> >>>>>>> That being said, I still can't build with this patch ;-) >>>>>> Are you building against the rdma-next branch ? >>>>>> The patch requires the iWARP series, originally this was tested >>>>>> against for-next-merged, what's the error you're getting ? >>>>> >>>>> On rdma-next: >>>>> drivers/infiniband/hw/qedr/main.c: In function ‘qedr_add’: >>>>> drivers/infiniband/hw/qedr/main.c:780:7: error: ‘struct qedr_dev’ has >> no member named ‘rdma_type’; did you mean ‘rdma_ctx’? >>>>> dev->rdma_type = dev_info.rdma_type; >>>>> ^~~~~~~~~ >>>>> rdma_ctx >>>>> make[4]: *** [scripts/Makefile.build:311: >>>>> drivers/infiniband/hw/qedr/main.o] Error 1 >>>> >>>> My config: >>>> >>>> ➜ linux-rdma git:(rdma-next) grep QED .config CONFIG_QED=m >>>> CONFIG_QED_LL2=y CONFIG_QED_SRIOV=y CONFIG_QEDE=m >> CONFIG_QED_RDMA=y >>>> CONFIG_INFINIBAND_QEDR=m >>> >>> Thanks Leon for helping out with this. Unfortunately i'm not seeing this. >>> The rdma_type field was added with the following patch: >>> >> https://kernel.googlesource.com/pub/scm/linux/kernel/git/dledford/rdma >>> /+/e6a38c54faf38498170e227c82ea25cb8bc1ae71 >>> >>> When running git log on the k.o/for-next I see the patch (last in list >>> below), applying this patch on top of rdma-next works for me and >> compiles cleanly. >>> >>> 2e2f2d0 RDMA/qedr: Fix rdma_type initialization >>> 1848757 Merge branches 'hns' and 'misc' into k.o/for-next 931bc0d IB: >>> Move PCI dependency from root KConfig to HW's KConfigs 7ae6f2a Merge >>> branch 'qedr' into k.o/for-next >>> bd491d2 RDMA/qedr: fix build error without ipv6 >>> 89fd257 RDMA/qedr: Missing error code in qedr_init_user_queue() >>> 847cb1a RDMA/qedr: fix spelling mistake: "invlaid" -> "invalid" >>> 69ad0e7 RDMA/qedr: Add support for iWARP in user space >>> e411e05 RDMA/qedr: Add iWARP connection management functions >> de0089e >>> RDMA/qedr: Add iWARP connection management qp related callbacks >>> fb1a22b RDMA/qedr: Add support for read with invalidate, supported in >>> iWARP >>> f5b1b17 RDMA/qedr: Add iWARP support in existing verbs >>> e6a38c5 RDMA/qedr: Add support for registering an iWARP device >>> >>> am i missing something ? >> >> I refreshed my branches now and it worked, but your git log history looks >> strange. Where did you see the commit 1848757 and 931bc0d? > > git://github.com/dledford/linux.git > git log --oneline -- drivers/infiniband/hw/qedr > https://github.com/dledford/linux/commit/1848757c3d8bff3addf881c8bcc2b85e25ccf25e > > I was under the assumption a patch goes to github before k.o - is this not the case ? It does. I always send things to github first, wait for 0day testing results, then send it to k.o if it passes. However, what is important here, is that if it *doesn't* pass 0day testing, then I will rebase the branch to fix the broken patch in some cases. Once a patch series lands on k.o, if it's broken, we will either revert or incrementally patch things. > Thanks again, > > Doug - are you still seeing issues with this patch ? I'll recheck today. > Thanks, > Michal > >> >> https://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma.git/log/?h=k >> .o/for-next >> Age Commit message (Expand) Author Files Lines >> 3 days Merge branch 'qedr' into k.o/for-nextfor-nextk.o/for-next Doug >> Ledford 11 -91/+1208 >> 3 days Merge tag 'v4.14-rc2' into k.o/for-next Doug Ledford 12044 - >> 388475/+631018 >> 3 days Linux 4.14-rc2HEADmaster Linus Torvalds 1 -1/+1 >> >> ➜ linux-rdma git:(iwarp-compile) git log --oneline f3506957237b (HEAD -> >> iwarp-compile) RDMA/qedr: Fix rdma_type initialization >> 7ae6f2a3d552 (rdma/k.o/for-next, grdma/k.o/for-next) Merge branch 'qedr' >> into k.o/for-next >> 0d9c2ff1c9f7 Merge tag 'v4.14-rc2' into k.o/for-next e19b205be43d (tag: >> v4.14-rc2, rdma/master, origin/master, ml/master, grdma/master, >> grdma/hns, backup/master, master) Linux 4.14-rc2 >> >> Thanks >> >>> thanks, >>> Michal >>>
On Wed, 2017-09-27 at 09:27 -0400, Doug Ledford wrote: > On Wed, 2017-09-20 at 08:39 +0300, Michal Kalderon wrote: > > Initialize the rdma_type (iWARP or RoCE) which is set according to > > device configuration in qed. > > > > Fixes: e6a38c54faf ("RDMA/qedr: Add support for registering an > > iWARP > > device") > > > > Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> > > Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com> > > > > --- > > This patch applies cleanly on k.o/for-next-merged > > > > Changes from V0: > > - Added Fixes in commit comment. > > Sorry, I was wrong in my email to the previous posting of this > patch. > I had pulled this v1 patch from patchworks, not the original patch, > and > so the Fixes: line was yours. Patchworks did not add the Fixes: line > to the v0 patch when I downloaded it just now. > > That being said, I still can't build with this patch ;-) OK, the build failure was my fault. As this was a minor patch, I had lumped it into a bundle I labeled as "misc". That by itself is fine, but I then applied that to a clean for-next head of v4.14-rc2, without the merge of the qedr branch, and so the build failed. Putting this into a branch with the qedr patches already merged resolves the issue. So, thanks, applied.
diff --git a/drivers/infiniband/hw/qedr/main.c b/drivers/infiniband/hw/qedr/main.c index ca93171..1afc267 100644 --- a/drivers/infiniband/hw/qedr/main.c +++ b/drivers/infiniband/hw/qedr/main.c @@ -870,6 +870,7 @@ static struct qedr_dev *qedr_add(struct qed_dev *cdev, struct pci_dev *pdev, goto init_err; dev->user_dpm_enabled = dev_info.user_dpm_enabled; + dev->rdma_type = dev_info.rdma_type; dev->num_hwfns = dev_info.common.num_hwfns; dev->rdma_ctx = dev->ops->rdma_get_rdma_ctx(cdev);
Initialize the rdma_type (iWARP or RoCE) which is set according to device configuration in qed. Fixes: e6a38c54faf ("RDMA/qedr: Add support for registering an iWARP device") Signed-off-by: Michal Kalderon <Michal.Kalderon@cavium.com> Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com> --- This patch applies cleanly on k.o/for-next-merged Changes from V0: - Added Fixes in commit comment. --- drivers/infiniband/hw/qedr/main.c | 1 + 1 file changed, 1 insertion(+)