| Message ID | 20220421071747.1892-13-chengyou@linux.alibaba.com (mailing list archive) |
|---|---|
| State | Superseded |
| Delegated to: | Jason Gunthorpe |
| Headers | show |
| Series | Elastic RDMA Adapter (ERDMA) driver | expand |
On Thu, Apr 21, 2022 at 03:17:47PM +0800, Cheng Xu wrote: > Signed-off-by: Cheng Xu <chengyou@linux.alibaba.com> > diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig > index 33d3ce9c888e..cc6a7ff88ff3 100644 > +++ b/drivers/infiniband/Kconfig > @@ -92,6 +92,7 @@ source "drivers/infiniband/hw/hns/Kconfig" > source "drivers/infiniband/hw/bnxt_re/Kconfig" > source "drivers/infiniband/hw/hfi1/Kconfig" > source "drivers/infiniband/hw/qedr/Kconfig" > +source "drivers/infiniband/hw/erdma/Kconfig" > source "drivers/infiniband/sw/rdmavt/Kconfig" > source "drivers/infiniband/sw/rxe/Kconfig" > source "drivers/infiniband/sw/siw/Kconfig" keep sorted > diff --git a/drivers/infiniband/hw/Makefile b/drivers/infiniband/hw/Makefile > index fba0b3be903e..6b3a88046125 100644 > +++ b/drivers/infiniband/hw/Makefile > @@ -13,3 +13,4 @@ obj-$(CONFIG_INFINIBAND_HFI1) += hfi1/ > obj-$(CONFIG_INFINIBAND_HNS) += hns/ > obj-$(CONFIG_INFINIBAND_QEDR) += qedr/ > obj-$(CONFIG_INFINIBAND_BNXT_RE) += bnxt_re/ > +obj-$(CONFIG_INFINIBAND_ERDMA) += erdma/ > diff --git a/drivers/infiniband/hw/erdma/Kconfig b/drivers/infiniband/hw/erdma/Kconfig > new file mode 100644 > index 000000000000..c90f2be1ea63 > +++ b/drivers/infiniband/hw/erdma/Kconfig > @@ -0,0 +1,12 @@ > +# SPDX-License-Identifier: GPL-2.0-only > +config INFINIBAND_ERDMA > + tristate "Alibaba Elastic RDMA Adapter (ERDMA) support" > + depends on PCI_MSI && 64BIT && !CPU_BIG_ENDIAN Why !CPU_BIG_ENDIAN? That is usually not OK. Did you run sparse on this? Jason
On 5/10/22 9:18 PM, Jason Gunthorpe wrote: > On Thu, Apr 21, 2022 at 03:17:47PM +0800, Cheng Xu wrote: >> Signed-off-by: Cheng Xu <chengyou@linux.alibaba.com> >> diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig >> index 33d3ce9c888e..cc6a7ff88ff3 100644 >> +++ b/drivers/infiniband/Kconfig >> @@ -92,6 +92,7 @@ source "drivers/infiniband/hw/hns/Kconfig" >> source "drivers/infiniband/hw/bnxt_re/Kconfig" >> source "drivers/infiniband/hw/hfi1/Kconfig" >> source "drivers/infiniband/hw/qedr/Kconfig" >> +source "drivers/infiniband/hw/erdma/Kconfig" >> source "drivers/infiniband/sw/rdmavt/Kconfig" >> source "drivers/infiniband/sw/rxe/Kconfig" >> source "drivers/infiniband/sw/siw/Kconfig" > > keep sorted > OK, I thought the order follows the rule: "comes first, appears first", so I add our drivers to the tail of "drivers/infiniband/hw/..." Since the current list is not sorted, and if it should be, I will fix this in next version. >> diff --git a/drivers/infiniband/hw/Makefile b/drivers/infiniband/hw/Makefile >> index fba0b3be903e..6b3a88046125 100644 >> +++ b/drivers/infiniband/hw/Makefile >> @@ -13,3 +13,4 @@ obj-$(CONFIG_INFINIBAND_HFI1) += hfi1/ >> obj-$(CONFIG_INFINIBAND_HNS) += hns/ >> obj-$(CONFIG_INFINIBAND_QEDR) += qedr/ >> obj-$(CONFIG_INFINIBAND_BNXT_RE) += bnxt_re/ >> +obj-$(CONFIG_INFINIBAND_ERDMA) += erdma/ >> diff --git a/drivers/infiniband/hw/erdma/Kconfig b/drivers/infiniband/hw/erdma/Kconfig >> new file mode 100644 >> index 000000000000..c90f2be1ea63 >> +++ b/drivers/infiniband/hw/erdma/Kconfig >> @@ -0,0 +1,12 @@ >> +# SPDX-License-Identifier: GPL-2.0-only >> +config INFINIBAND_ERDMA >> + tristate "Alibaba Elastic RDMA Adapter (ERDMA) support" >> + depends on PCI_MSI && 64BIT && !CPU_BIG_ENDIAN > > Why !CPU_BIG_ENDIAN? That is usually not OK. we want use !CPU_BIG_ENDIAN to disable the erdma compilation on big endian machine, because we only have little endian machines, and don't support big endian machines. I have no idea why it is usually not OK, could you explain it any more? Thanks > Did you run sparse on this? > No, I will check this. I think this may be OK, because EFA also use !CPU_BIG_ENDIAN in the same way. Thanks, Cheng Xu > Jason
On 5/16/22 11:40 AM, Cheng Xu wrote: > > > On 5/10/22 9:18 PM, Jason Gunthorpe wrote: >> On Thu, Apr 21, 2022 at 03:17:47PM +0800, Cheng Xu wrote: >>> Signed-off-by: Cheng Xu <chengyou@linux.alibaba.com> >>> diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig >>> index 33d3ce9c888e..cc6a7ff88ff3 100644 >>> +++ b/drivers/infiniband/Kconfig >>> @@ -92,6 +92,7 @@ source "drivers/infiniband/hw/hns/Kconfig" >>> source "drivers/infiniband/hw/bnxt_re/Kconfig" >>> source "drivers/infiniband/hw/hfi1/Kconfig" >>> source "drivers/infiniband/hw/qedr/Kconfig" >>> +source "drivers/infiniband/hw/erdma/Kconfig" >>> source "drivers/infiniband/sw/rdmavt/Kconfig" >>> source "drivers/infiniband/sw/rxe/Kconfig" >>> source "drivers/infiniband/sw/siw/Kconfig" >> >> keep sorted >> > > OK, I thought the order follows the rule: "comes first, appears first", > so I add our drivers to the tail of "drivers/infiniband/hw/..." > > Since the current list is not sorted, and if it should be, I will fix > this in next version. > >>> diff --git a/drivers/infiniband/hw/Makefile >>> b/drivers/infiniband/hw/Makefile >>> index fba0b3be903e..6b3a88046125 100644 >>> +++ b/drivers/infiniband/hw/Makefile >>> @@ -13,3 +13,4 @@ obj-$(CONFIG_INFINIBAND_HFI1) += hfi1/ >>> obj-$(CONFIG_INFINIBAND_HNS) += hns/ >>> obj-$(CONFIG_INFINIBAND_QEDR) += qedr/ >>> obj-$(CONFIG_INFINIBAND_BNXT_RE) += bnxt_re/ >>> +obj-$(CONFIG_INFINIBAND_ERDMA) += erdma/ >>> diff --git a/drivers/infiniband/hw/erdma/Kconfig >>> b/drivers/infiniband/hw/erdma/Kconfig >>> new file mode 100644 >>> index 000000000000..c90f2be1ea63 >>> +++ b/drivers/infiniband/hw/erdma/Kconfig >>> @@ -0,0 +1,12 @@ >>> +# SPDX-License-Identifier: GPL-2.0-only >>> +config INFINIBAND_ERDMA >>> + tristate "Alibaba Elastic RDMA Adapter (ERDMA) support" >>> + depends on PCI_MSI && 64BIT && !CPU_BIG_ENDIAN >> >> Why !CPU_BIG_ENDIAN? That is usually not OK. > > we want use !CPU_BIG_ENDIAN to disable the erdma compilation on big > endian machine, because we only have little endian machines, and > don't support big endian machines. I have no idea why it is usually not > OK, could you explain it any more? > > Thanks > >> Did you run sparse on this? >> > > No, I will check this. I think this may be OK, because EFA also use > !CPU_BIG_ENDIAN in the same way. > I use the command 'make C=2' as the document says to check the code, It seems ok, except some warnings. I reviewed and checked the warnings in erdma part. Fix them or not fix, is all OK for functionality. Is cleaning the warnings necessary for upstream? If so, I will work on it. Thanks Cheng Xu > Thanks, > Cheng Xu > >> Jason
On 5/16/22 3:11 PM, Cheng Xu wrote: > > > On 5/16/22 11:40 AM, Cheng Xu wrote: >> >> >> On 5/10/22 9:18 PM, Jason Gunthorpe wrote: >>> On Thu, Apr 21, 2022 at 03:17:47PM +0800, Cheng Xu wrote: >>>> Signed-off-by: Cheng Xu <chengyou@linux.alibaba.com> >>>> diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig >>>> index 33d3ce9c888e..cc6a7ff88ff3 100644 >>>> +++ b/drivers/infiniband/Kconfig >>>> @@ -92,6 +92,7 @@ source "drivers/infiniband/hw/hns/Kconfig" >>>> source "drivers/infiniband/hw/bnxt_re/Kconfig" >>>> source "drivers/infiniband/hw/hfi1/Kconfig" >>>> source "drivers/infiniband/hw/qedr/Kconfig" >>>> +source "drivers/infiniband/hw/erdma/Kconfig" >>>> source "drivers/infiniband/sw/rdmavt/Kconfig" >>>> source "drivers/infiniband/sw/rxe/Kconfig" >>>> source "drivers/infiniband/sw/siw/Kconfig" >>> >>> keep sorted >>> >> >> OK, I thought the order follows the rule: "comes first, appears first", >> so I add our drivers to the tail of "drivers/infiniband/hw/..." >> >> Since the current list is not sorted, and if it should be, I will fix >> this in next version. >> >>>> diff --git a/drivers/infiniband/hw/Makefile >>>> b/drivers/infiniband/hw/Makefile >>>> index fba0b3be903e..6b3a88046125 100644 >>>> +++ b/drivers/infiniband/hw/Makefile >>>> @@ -13,3 +13,4 @@ obj-$(CONFIG_INFINIBAND_HFI1) += hfi1/ >>>> obj-$(CONFIG_INFINIBAND_HNS) += hns/ >>>> obj-$(CONFIG_INFINIBAND_QEDR) += qedr/ >>>> obj-$(CONFIG_INFINIBAND_BNXT_RE) += bnxt_re/ >>>> +obj-$(CONFIG_INFINIBAND_ERDMA) += erdma/ >>>> diff --git a/drivers/infiniband/hw/erdma/Kconfig >>>> b/drivers/infiniband/hw/erdma/Kconfig >>>> new file mode 100644 >>>> index 000000000000..c90f2be1ea63 >>>> +++ b/drivers/infiniband/hw/erdma/Kconfig >>>> @@ -0,0 +1,12 @@ >>>> +# SPDX-License-Identifier: GPL-2.0-only >>>> +config INFINIBAND_ERDMA >>>> + tristate "Alibaba Elastic RDMA Adapter (ERDMA) support" >>>> + depends on PCI_MSI && 64BIT && !CPU_BIG_ENDIAN >>> >>> Why !CPU_BIG_ENDIAN? That is usually not OK. >> >> we want use !CPU_BIG_ENDIAN to disable the erdma compilation on big >> endian machine, because we only have little endian machines, and >> don't support big endian machines. I have no idea why it is usually not >> OK, could you explain it any more? >> >> Thanks >> >>> Did you run sparse on this? >>> >> >> No, I will check this. I think this may be OK, because EFA also use >> !CPU_BIG_ENDIAN in the same way. >> > > ...... Is > cleaning the warnings necessary for upstream? If so, I will work on it. Please ignore this, it seems easy to clean the warnings, I will fix them any way. Thanks, Cheng Xu
On Mon, May 16, 2022 at 11:40:46AM +0800, Cheng Xu wrote: > > > diff --git a/drivers/infiniband/hw/Makefile b/drivers/infiniband/hw/Makefile > > > index fba0b3be903e..6b3a88046125 100644 > > > +++ b/drivers/infiniband/hw/Makefile > > > @@ -13,3 +13,4 @@ obj-$(CONFIG_INFINIBAND_HFI1) += hfi1/ > > > obj-$(CONFIG_INFINIBAND_HNS) += hns/ > > > obj-$(CONFIG_INFINIBAND_QEDR) += qedr/ > > > obj-$(CONFIG_INFINIBAND_BNXT_RE) += bnxt_re/ > > > +obj-$(CONFIG_INFINIBAND_ERDMA) += erdma/ > > > diff --git a/drivers/infiniband/hw/erdma/Kconfig b/drivers/infiniband/hw/erdma/Kconfig > > > new file mode 100644 > > > index 000000000000..c90f2be1ea63 > > > +++ b/drivers/infiniband/hw/erdma/Kconfig > > > @@ -0,0 +1,12 @@ > > > +# SPDX-License-Identifier: GPL-2.0-only > > > +config INFINIBAND_ERDMA > > > + tristate "Alibaba Elastic RDMA Adapter (ERDMA) support" > > > + depends on PCI_MSI && 64BIT && !CPU_BIG_ENDIAN > > > > Why !CPU_BIG_ENDIAN? That is usually not OK. > > we want use !CPU_BIG_ENDIAN to disable the erdma compilation on big > endian machine Do not copy this > because we only have little endian machines, and don't support big > endian machines. I have no idea why it is usually not OK, could you > explain it any more? It is considered bad coding practice in the kernel. Write the required endian swaps. Jason
On 5/16/22 10:13 PM, Jason Gunthorpe wrote: > On Mon, May 16, 2022 at 11:40:46AM +0800, Cheng Xu wrote: > >>>> diff --git a/drivers/infiniband/hw/Makefile b/drivers/infiniband/hw/Makefile >>>> index fba0b3be903e..6b3a88046125 100644 >>>> +++ b/drivers/infiniband/hw/Makefile >>>> @@ -13,3 +13,4 @@ obj-$(CONFIG_INFINIBAND_HFI1) += hfi1/ >>>> obj-$(CONFIG_INFINIBAND_HNS) += hns/ >>>> obj-$(CONFIG_INFINIBAND_QEDR) += qedr/ >>>> obj-$(CONFIG_INFINIBAND_BNXT_RE) += bnxt_re/ >>>> +obj-$(CONFIG_INFINIBAND_ERDMA) += erdma/ >>>> diff --git a/drivers/infiniband/hw/erdma/Kconfig b/drivers/infiniband/hw/erdma/Kconfig >>>> new file mode 100644 >>>> index 000000000000..c90f2be1ea63 >>>> +++ b/drivers/infiniband/hw/erdma/Kconfig >>>> @@ -0,0 +1,12 @@ >>>> +# SPDX-License-Identifier: GPL-2.0-only >>>> +config INFINIBAND_ERDMA >>>> + tristate "Alibaba Elastic RDMA Adapter (ERDMA) support" >>>> + depends on PCI_MSI && 64BIT && !CPU_BIG_ENDIAN >>> >>> Why !CPU_BIG_ENDIAN? That is usually not OK. >> >> we want use !CPU_BIG_ENDIAN to disable the erdma compilation on big >> endian machine > > Do not copy this > >> because we only have little endian machines, and don't support big >> endian machines. I have no idea why it is usually not OK, could you >> explain it any more? > > It is considered bad coding practice in the kernel. Write the required endian > swaps. Get it. Thanks. Cheng Xu
diff --git a/MAINTAINERS b/MAINTAINERS index fd768d43e048..7aa57ef45a27 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -732,6 +732,14 @@ S: Maintained F: Documentation/i2c/busses/i2c-ali1563.rst F: drivers/i2c/busses/i2c-ali1563.c +ALIBABA ELASTIC RDMA DRIVER +M: Cheng Xu <chengyou@linux.alibaba.com> +M: Kai Shen <kaishen@linux.alibaba.com> +L: linux-rdma@vger.kernel.org +S: Supported +F: drivers/infiniband/hw/erdma +F: include/uapi/rdma/erdma-abi.h + ALIENWARE WMI DRIVER L: Dell.Client.Kernel@dell.com S: Maintained diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig index 33d3ce9c888e..cc6a7ff88ff3 100644 --- a/drivers/infiniband/Kconfig +++ b/drivers/infiniband/Kconfig @@ -92,6 +92,7 @@ source "drivers/infiniband/hw/hns/Kconfig" source "drivers/infiniband/hw/bnxt_re/Kconfig" source "drivers/infiniband/hw/hfi1/Kconfig" source "drivers/infiniband/hw/qedr/Kconfig" +source "drivers/infiniband/hw/erdma/Kconfig" source "drivers/infiniband/sw/rdmavt/Kconfig" source "drivers/infiniband/sw/rxe/Kconfig" source "drivers/infiniband/sw/siw/Kconfig" diff --git a/drivers/infiniband/hw/Makefile b/drivers/infiniband/hw/Makefile index fba0b3be903e..6b3a88046125 100644 --- a/drivers/infiniband/hw/Makefile +++ b/drivers/infiniband/hw/Makefile @@ -13,3 +13,4 @@ obj-$(CONFIG_INFINIBAND_HFI1) += hfi1/ obj-$(CONFIG_INFINIBAND_HNS) += hns/ obj-$(CONFIG_INFINIBAND_QEDR) += qedr/ obj-$(CONFIG_INFINIBAND_BNXT_RE) += bnxt_re/ +obj-$(CONFIG_INFINIBAND_ERDMA) += erdma/ diff --git a/drivers/infiniband/hw/erdma/Kconfig b/drivers/infiniband/hw/erdma/Kconfig new file mode 100644 index 000000000000..c90f2be1ea63 --- /dev/null +++ b/drivers/infiniband/hw/erdma/Kconfig @@ -0,0 +1,12 @@ +# SPDX-License-Identifier: GPL-2.0-only +config INFINIBAND_ERDMA + tristate "Alibaba Elastic RDMA Adapter (ERDMA) support" + depends on PCI_MSI && 64BIT && !CPU_BIG_ENDIAN + depends on INFINIBAND_ADDR_TRANS + depends on INFINIBAND_USER_ACCESS + help + This is a RDMA/iWarp driver for Alibaba Elastic RDMA Adapter(ERDMA), + which supports RDMA features in Alibaba cloud environment. + + To compile this driver as module, choose M here. The module will be + called erdma. diff --git a/drivers/infiniband/hw/erdma/Makefile b/drivers/infiniband/hw/erdma/Makefile new file mode 100644 index 000000000000..51d2ef91905a --- /dev/null +++ b/drivers/infiniband/hw/erdma/Makefile @@ -0,0 +1,4 @@ +# SPDX-License-Identifier: GPL-2.0 +obj-$(CONFIG_INFINIBAND_ERDMA) := erdma.o + +erdma-y := erdma_cm.o erdma_main.o erdma_cmdq.o erdma_cq.o erdma_verbs.o erdma_qp.o erdma_eq.o
Signed-off-by: Cheng Xu <chengyou@linux.alibaba.com> --- MAINTAINERS | 8 ++++++++ drivers/infiniband/Kconfig | 1 + drivers/infiniband/hw/Makefile | 1 + drivers/infiniband/hw/erdma/Kconfig | 12 ++++++++++++ drivers/infiniband/hw/erdma/Makefile | 4 ++++ 5 files changed, 26 insertions(+) create mode 100644 drivers/infiniband/hw/erdma/Kconfig create mode 100644 drivers/infiniband/hw/erdma/Makefile