[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <40F65EF2B5E2254199711F58E3ACB84D7825D318@MX104CL02.corp.emc.com>
Date: Wed, 20 May 2015 21:26:52 +0000
From: "Hubbe, Allen" <Allen.Hubbe@....com>
To: "Jiang, Dave" <dave.jiang@...el.com>,
"bhelgaas@...gle.com" <bhelgaas@...gle.com>
CC: "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"linux-pci@...r.kernel.org" <linux-pci@...r.kernel.org>,
"jdmason@...zu.us" <jdmason@...zu.us>,
"linux-ntb@...glegroups.com" <linux-ntb@...glegroups.com>
Subject: RE: [PATCH 04/16] Check the DID for certain workaround error flags
to be set.
From: linux-ntb@...glegroups.com [mailto:linux-ntb@...glegroups.com] On Behalf Of Jiang, Dave
> On Wed, 2015-05-20 at 16:11 -0500, Bjorn Helgaas wrote:
> > On Wed, May 20, 2015 at 10:41 AM, Allen Hubbe <Allen.Hubbe@....com>
> wrote:
> > > From: Dave Jiang <dave.jiang@...el.com>
> > >
> > > Signed-off-by: Dave Jiang <dave.jiang@...el.com>
> >
> > Needs a topic in the subject line and a changelog.
> >
> > It also seems to do a lot more than just checking device ID (I assume
> > that's what "DID" means), so this should probably be split into
> > several patches that each do one thing. I see at least:
> >
> > - cosmetic code restructuring
> > - work_struct/tasklet_struct changes
> > - new #defines and bar2_off() changes
>
> I think this patch got mangled with couple other patches. Allen?
Yes, there will be a v2 to fix the mangling. The code restructuring is meant for a different patch.
This is all that is meant to change in this patch:
> > > @@ -1747,29 +1722,68 @@ static int snb_init_dev(struct intel_ntb_dev
> *ndev)
> > > u8 ppd;
> > > int rc, mem;
> > >
> > > + pdev = ndev_pdev(ndev);
> > > +
> > > + switch (pdev->device) {
> > > /* There is a Xeon hardware errata related to writes to
> SDOORBELL or
> > > * B2BDOORBELL in conjunction with inbound access to NTB
> MMIO Space,
> > > * which may hang the system. To workaround this use the
> second memory
> > > * window to access the interrupt and scratch pad registers
> on the
> > > * remote system.
> > > */
> > > - ndev->hwerr_flags |= NTB_HWERR_SDOORBELL_LOCKUP;
> > > + case PCI_DEVICE_ID_INTEL_NTB_SS_JSF:
> > > + case PCI_DEVICE_ID_INTEL_NTB_PS_JSF:
> > > + case PCI_DEVICE_ID_INTEL_NTB_B2B_JSF:
> > > + case PCI_DEVICE_ID_INTEL_NTB_SS_SNB:
> > > + case PCI_DEVICE_ID_INTEL_NTB_PS_SNB:
> > > + case PCI_DEVICE_ID_INTEL_NTB_B2B_SNB:
> > > + case PCI_DEVICE_ID_INTEL_NTB_SS_IVT:
> > > + case PCI_DEVICE_ID_INTEL_NTB_PS_IVT:
> > > + case PCI_DEVICE_ID_INTEL_NTB_B2B_IVT:
> > > + case PCI_DEVICE_ID_INTEL_NTB_SS_HSX:
> > > + case PCI_DEVICE_ID_INTEL_NTB_PS_HSX:
> > > + case PCI_DEVICE_ID_INTEL_NTB_B2B_HSX:
> > > + ndev->hwerr_flags |= NTB_HWERR_SDOORBELL_LOCKUP;
> > > + break;
> > > + }
> > >
> > > + switch (pdev->device) {
> > > /* There is a hardware errata related to accessing any
> register in
> > > * SB01BASE in the presence of bidirectional traffic
> crossing the NTB.
> > > */
> > > - ndev->hwerr_flags |= NTB_HWERR_SB01BASE_LOCKUP;
> > > + case PCI_DEVICE_ID_INTEL_NTB_SS_IVT:
> > > + case PCI_DEVICE_ID_INTEL_NTB_PS_IVT:
> > > + case PCI_DEVICE_ID_INTEL_NTB_B2B_IVT:
> > > + case PCI_DEVICE_ID_INTEL_NTB_SS_HSX:
> > > + case PCI_DEVICE_ID_INTEL_NTB_PS_HSX:
> > > + case PCI_DEVICE_ID_INTEL_NTB_B2B_HSX:
> > > + ndev->hwerr_flags |= NTB_HWERR_SB01BASE_LOCKUP;
> > > + break;
> > > + }
> > >
> > > + switch (pdev->device) {
> > > /* HW Errata on bit 14 of b2bdoorbell register. Writes will
> not be
> > > * mirrored to the remote system. Shrink the number of bits
> by one,
> > > * since bit 14 is the last bit.
> > > */
> > > - ndev->hwerr_flags |= NTB_HWERR_B2BDOORBELL_BIT14;
> > > + case PCI_DEVICE_ID_INTEL_NTB_SS_JSF:
> > > + case PCI_DEVICE_ID_INTEL_NTB_PS_JSF:
> > > + case PCI_DEVICE_ID_INTEL_NTB_B2B_JSF:
> > > + case PCI_DEVICE_ID_INTEL_NTB_SS_SNB:
> > > + case PCI_DEVICE_ID_INTEL_NTB_PS_SNB:
> > > + case PCI_DEVICE_ID_INTEL_NTB_B2B_SNB:
> > > + case PCI_DEVICE_ID_INTEL_NTB_SS_IVT:
> > > + case PCI_DEVICE_ID_INTEL_NTB_PS_IVT:
> > > + case PCI_DEVICE_ID_INTEL_NTB_B2B_IVT:
> > > + case PCI_DEVICE_ID_INTEL_NTB_SS_HSX:
> > > + case PCI_DEVICE_ID_INTEL_NTB_PS_HSX:
> > > + case PCI_DEVICE_ID_INTEL_NTB_B2B_HSX:
> > > + ndev->hwerr_flags |= NTB_HWERR_B2BDOORBELL_BIT14;
> > > + break;
> > > + }
> > >
> > > ndev->reg = &snb_reg;
> > >
> > > - pdev = ndev_pdev(ndev);
> > > -
> > > rc = pci_read_config_byte(pdev, SNB_PPD_OFFSET, &ppd);
> > > if (rc)
> > > return -EIO;
Powered by blists - more mailing lists