* [PATCH v3] staging/rdma/hfi1: set Gen3 half-swing for integrated devices
@ 2015-11-16 22:03 ira.weiny-ral2JQCrhuEAvxtiuMwx3w
[not found] ` <1447711429-10389-1-git-send-email-ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
0 siblings, 1 reply; 5+ messages in thread
From: ira.weiny-ral2JQCrhuEAvxtiuMwx3w @ 2015-11-16 22:03 UTC (permalink / raw)
To: gregkh-hQyY1W1yCW8ekmWlsbkhG0B+6BGkLq7r,
devel-gWbeCf7V1WCQmaza687I9mD2FQJk+8+b
Cc: dledford-H+wXaHxf7aLQT0dZR+AlfA,
linux-rdma-u79uwXL29TY76Z2rM5mHXA,
mike.marciniszyn-ral2JQCrhuEAvxtiuMwx3w, Dean Luick, Ira Weiny
From: Dean Luick <dean.luick-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Correctly set half-swing for integrated devices. A0 needs all fields set for
CcePcieCtrl. B0 and later only need a few fields set.
Reviewed-by: Stuart Summers <john.s.summers-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Dean Luick <dean.luick-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Ira Weiny <ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
Changes from V1:
Add comments concerning the very long names.
Changes from V2:
Remove PC Macro and define short names to be used in the code.
drivers/staging/rdma/hfi1/chip_registers.h | 11 ++++
drivers/staging/rdma/hfi1/pcie.c | 82 ++++++++++++++++++++++++++++--
2 files changed, 89 insertions(+), 4 deletions(-)
diff --git a/drivers/staging/rdma/hfi1/chip_registers.h b/drivers/staging/rdma/hfi1/chip_registers.h
index bf45de29d8bd..d0deb2278635 100644
--- a/drivers/staging/rdma/hfi1/chip_registers.h
+++ b/drivers/staging/rdma/hfi1/chip_registers.h
@@ -549,6 +549,17 @@
#define CCE_MSIX_TABLE_UPPER (CCE + 0x000000100008)
#define CCE_MSIX_TABLE_UPPER_RESETCSR 0x0000000100000000ull
#define CCE_MSIX_VEC_CLR_WITHOUT_INT (CCE + 0x000000110400)
+#define CCE_PCIE_CTRL (CCE + 0x0000000000C0)
+#define CCE_PCIE_CTRL_PCIE_LANE_BUNDLE_MASK 0x3ull
+#define CCE_PCIE_CTRL_PCIE_LANE_BUNDLE_SHIFT 0
+#define CCE_PCIE_CTRL_PCIE_LANE_DELAY_MASK 0xFull
+#define CCE_PCIE_CTRL_PCIE_LANE_DELAY_SHIFT 2
+#define CCE_PCIE_CTRL_XMT_MARGIN_OVERWRITE_ENABLE_SHIFT 8
+#define CCE_PCIE_CTRL_XMT_MARGIN_SHIFT 9
+#define CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_OVERWRITE_ENABLE_MASK 0x1ull
+#define CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_OVERWRITE_ENABLE_SHIFT 12
+#define CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_MASK 0x7ull
+#define CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_SHIFT 13
#define CCE_REVISION (CCE + 0x000000000000)
#define CCE_REVISION2 (CCE + 0x000000000008)
#define CCE_REVISION2_HFI_ID_MASK 0x1ull
diff --git a/drivers/staging/rdma/hfi1/pcie.c b/drivers/staging/rdma/hfi1/pcie.c
index 0b7eafb0fc70..a92ad4577cc8 100644
--- a/drivers/staging/rdma/hfi1/pcie.c
+++ b/drivers/staging/rdma/hfi1/pcie.c
@@ -867,6 +867,83 @@ static void arm_gasket_logic(struct hfi1_devdata *dd)
}
/*
+ * CCE_PCIE_CTRL long name helpers
+ * We redefine these shorter macros to use in the code while leaving
+ * chip_registers.h to be autogenerated from the hardware spec.
+ */
+#define LANE_BUNDLE_MASK CCE_PCIE_CTRL_PCIE_LANE_BUNDLE_MASK
+#define LANE_BUNDLE_SHIFT CCE_PCIE_CTRL_PCIE_LANE_BUNDLE_SHIFT
+#define LANE_DELAY_MASK CCE_PCIE_CTRL_PCIE_LANE_DELAY_MASK
+#define LANE_DELAY_SHIFT CCE_PCIE_CTRL_PCIE_LANE_DELAY_SHIFT
+#define MARGIN_OVERWRITE_ENABLE_SHIFT CCE_PCIE_CTRL_XMT_MARGIN_OVERWRITE_ENABLE_SHIFT
+#define MARGIN_SHIFT CCE_PCIE_CTRL_XMT_MARGIN_SHIFT
+#define MARGIN_G1_G2_OVERWRITE_MASK CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_OVERWRITE_ENABLE_MASK
+#define MARGIN_G1_G2_OVERWRITE_SHIFT CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_OVERWRITE_ENABLE_SHIFT
+#define MARGIN_GEN1_GEN2_MASK CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_MASK
+#define MARGIN_GEN1_GEN2_SHIFT CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_SHIFT
+
+ /*
+ * Write xmt_margin for full-swing (WFR-B) or half-swing (WFR-C).
+ */
+static void write_xmt_margin(struct hfi1_devdata *dd, const char *fname)
+{
+ u64 pcie_ctrl;
+ u64 xmt_margin;
+ u64 xmt_margin_oe;
+ u64 lane_delay;
+ u64 lane_bundle;
+
+ pcie_ctrl = read_csr(dd, CCE_PCIE_CTRL);
+
+ /*
+ * For Discrete, use full-swing.
+ * - PCIe TX defaults to full-swing.
+ * Leave this register as default.
+ * For Integrated, use half-swing
+ * - Copy xmt_margin and xmt_margin_oe
+ * from Gen1/Gen2 to Gen3.
+ */
+ if (dd->pcidev->device == PCI_DEVICE_ID_INTEL1) { /* integrated */
+ /* extract initial fields */
+ xmt_margin = (pcie_ctrl >> MARGIN_GEN1_GEN2_SHIFT)
+ & MARGIN_GEN1_GEN2_MASK;
+ xmt_margin_oe = (pcie_ctrl >> MARGIN_G1_G2_OVERWRITE_SHIFT)
+ & MARGIN_G1_G2_OVERWRITE_MASK;
+ lane_delay = (pcie_ctrl >> LANE_DELAY_SHIFT) & LANE_DELAY_MASK;
+ lane_bundle = (pcie_ctrl >> LANE_BUNDLE_SHIFT)
+ & LANE_BUNDLE_MASK;
+
+ /*
+ * For A0, EFUSE values are not set. Override with the
+ * correct values.
+ */
+ if (is_a0(dd)) {
+ /*
+ * xmt_margin and OverwiteEnabel should be the
+ * same for Gen1/Gen2 and Gen3
+ */
+ xmt_margin = 0x5;
+ xmt_margin_oe = 0x1;
+ lane_delay = 0xF; /* Delay 240ns. */
+ lane_bundle = 0x0; /* Set to 1 lane. */
+ }
+
+ /* overwrite existing values */
+ pcie_ctrl = (xmt_margin << MARGIN_GEN1_GEN2_SHIFT)
+ | (xmt_margin_oe << MARGIN_G1_G2_OVERWRITE_SHIFT)
+ | (xmt_margin << MARGIN_SHIFT)
+ | (xmt_margin_oe << MARGIN_OVERWRITE_ENABLE_SHIFT)
+ | (lane_delay << LANE_DELAY_SHIFT)
+ | (lane_bundle << LANE_BUNDLE_SHIFT);
+
+ write_csr(dd, CCE_PCIE_CTRL, pcie_ctrl);
+ }
+
+ dd_dev_info(dd, "%s: program XMT margin, CcePcieCtrl 0x%llx\n",
+ fname, pcie_ctrl);
+}
+
+/*
* Do all the steps needed to transition the PCIe link to Gen3 speed.
*/
int do_pcie_gen3_transition(struct hfi1_devdata *dd)
@@ -1064,11 +1141,8 @@ retry:
/*
* step 5d: program XMT margin
- * Right now, leave the default alone. To change, do a
- * read-modify-write of:
- * CcePcieCtrl.XmtMargin
- * CcePcieCtrl.XmitMarginOverwriteEnable
*/
+ write_xmt_margin(dd, __func__);
/* step 5e: disable active state power management (ASPM) */
dd_dev_info(dd, "%s: clearing ASPM\n", __func__);
--
1.8.2
--
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
^ permalink raw reply related [flat|nested] 5+ messages in thread[parent not found: <1447711429-10389-1-git-send-email-ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>]
* Re: [PATCH v3] staging/rdma/hfi1: set Gen3 half-swing for integrated devices [not found] ` <1447711429-10389-1-git-send-email-ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org> @ 2015-11-20 1:12 ` Greg KH 2015-11-20 15:14 ` Marciniszyn, Mike 0 siblings, 1 reply; 5+ messages in thread From: Greg KH @ 2015-11-20 1:12 UTC (permalink / raw) To: ira.weiny-ral2JQCrhuEAvxtiuMwx3w Cc: devel-gWbeCf7V1WCQmaza687I9mD2FQJk+8+b, dledford-H+wXaHxf7aLQT0dZR+AlfA, linux-rdma-u79uwXL29TY76Z2rM5mHXA, mike.marciniszyn-ral2JQCrhuEAvxtiuMwx3w, Dean Luick On Mon, Nov 16, 2015 at 05:03:49PM -0500, ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org wrote: > From: Dean Luick <dean.luick-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org> > > Correctly set half-swing for integrated devices. A0 needs all fields set for > CcePcieCtrl. B0 and later only need a few fields set. > > Reviewed-by: Stuart Summers <john.s.summers-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org> > Signed-off-by: Dean Luick <dean.luick-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org> > Signed-off-by: Ira Weiny <ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org> > > --- > Changes from V1: > Add comments concerning the very long names. > > Changes from V2: > Remove PC Macro and define short names to be used in the code. > > drivers/staging/rdma/hfi1/chip_registers.h | 11 ++++ > drivers/staging/rdma/hfi1/pcie.c | 82 ++++++++++++++++++++++++++++-- > 2 files changed, 89 insertions(+), 4 deletions(-) > > diff --git a/drivers/staging/rdma/hfi1/chip_registers.h b/drivers/staging/rdma/hfi1/chip_registers.h > index bf45de29d8bd..d0deb2278635 100644 > --- a/drivers/staging/rdma/hfi1/chip_registers.h > +++ b/drivers/staging/rdma/hfi1/chip_registers.h > @@ -549,6 +549,17 @@ > #define CCE_MSIX_TABLE_UPPER (CCE + 0x000000100008) > #define CCE_MSIX_TABLE_UPPER_RESETCSR 0x0000000100000000ull > #define CCE_MSIX_VEC_CLR_WITHOUT_INT (CCE + 0x000000110400) > +#define CCE_PCIE_CTRL (CCE + 0x0000000000C0) > +#define CCE_PCIE_CTRL_PCIE_LANE_BUNDLE_MASK 0x3ull > +#define CCE_PCIE_CTRL_PCIE_LANE_BUNDLE_SHIFT 0 > +#define CCE_PCIE_CTRL_PCIE_LANE_DELAY_MASK 0xFull > +#define CCE_PCIE_CTRL_PCIE_LANE_DELAY_SHIFT 2 > +#define CCE_PCIE_CTRL_XMT_MARGIN_OVERWRITE_ENABLE_SHIFT 8 > +#define CCE_PCIE_CTRL_XMT_MARGIN_SHIFT 9 > +#define CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_OVERWRITE_ENABLE_MASK 0x1ull > +#define CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_OVERWRITE_ENABLE_SHIFT 12 > +#define CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_MASK 0x7ull > +#define CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_SHIFT 13 > #define CCE_REVISION (CCE + 0x000000000000) > #define CCE_REVISION2 (CCE + 0x000000000008) > #define CCE_REVISION2_HFI_ID_MASK 0x1ull > diff --git a/drivers/staging/rdma/hfi1/pcie.c b/drivers/staging/rdma/hfi1/pcie.c > index 0b7eafb0fc70..a92ad4577cc8 100644 > --- a/drivers/staging/rdma/hfi1/pcie.c > +++ b/drivers/staging/rdma/hfi1/pcie.c > @@ -867,6 +867,83 @@ static void arm_gasket_logic(struct hfi1_devdata *dd) > } > > /* > + * CCE_PCIE_CTRL long name helpers > + * We redefine these shorter macros to use in the code while leaving > + * chip_registers.h to be autogenerated from the hardware spec. > + */ > +#define LANE_BUNDLE_MASK CCE_PCIE_CTRL_PCIE_LANE_BUNDLE_MASK > +#define LANE_BUNDLE_SHIFT CCE_PCIE_CTRL_PCIE_LANE_BUNDLE_SHIFT > +#define LANE_DELAY_MASK CCE_PCIE_CTRL_PCIE_LANE_DELAY_MASK > +#define LANE_DELAY_SHIFT CCE_PCIE_CTRL_PCIE_LANE_DELAY_SHIFT > +#define MARGIN_OVERWRITE_ENABLE_SHIFT CCE_PCIE_CTRL_XMT_MARGIN_OVERWRITE_ENABLE_SHIFT > +#define MARGIN_SHIFT CCE_PCIE_CTRL_XMT_MARGIN_SHIFT > +#define MARGIN_G1_G2_OVERWRITE_MASK CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_OVERWRITE_ENABLE_MASK > +#define MARGIN_G1_G2_OVERWRITE_SHIFT CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_OVERWRITE_ENABLE_SHIFT > +#define MARGIN_GEN1_GEN2_MASK CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_MASK > +#define MARGIN_GEN1_GEN2_SHIFT CCE_PCIE_CTRL_XMT_MARGIN_GEN1_GEN2_SHIFT > + > + /* > + * Write xmt_margin for full-swing (WFR-B) or half-swing (WFR-C). > + */ > +static void write_xmt_margin(struct hfi1_devdata *dd, const char *fname) > +{ > + u64 pcie_ctrl; > + u64 xmt_margin; > + u64 xmt_margin_oe; > + u64 lane_delay; > + u64 lane_bundle; > + > + pcie_ctrl = read_csr(dd, CCE_PCIE_CTRL); > + > + /* > + * For Discrete, use full-swing. > + * - PCIe TX defaults to full-swing. > + * Leave this register as default. > + * For Integrated, use half-swing > + * - Copy xmt_margin and xmt_margin_oe > + * from Gen1/Gen2 to Gen3. > + */ > + if (dd->pcidev->device == PCI_DEVICE_ID_INTEL1) { /* integrated */ > + /* extract initial fields */ > + xmt_margin = (pcie_ctrl >> MARGIN_GEN1_GEN2_SHIFT) > + & MARGIN_GEN1_GEN2_MASK; > + xmt_margin_oe = (pcie_ctrl >> MARGIN_G1_G2_OVERWRITE_SHIFT) > + & MARGIN_G1_G2_OVERWRITE_MASK; > + lane_delay = (pcie_ctrl >> LANE_DELAY_SHIFT) & LANE_DELAY_MASK; > + lane_bundle = (pcie_ctrl >> LANE_BUNDLE_SHIFT) > + & LANE_BUNDLE_MASK; > + > + /* > + * For A0, EFUSE values are not set. Override with the > + * correct values. > + */ > + if (is_a0(dd)) { > + /* > + * xmt_margin and OverwiteEnabel should be the > + * same for Gen1/Gen2 and Gen3 > + */ > + xmt_margin = 0x5; > + xmt_margin_oe = 0x1; > + lane_delay = 0xF; /* Delay 240ns. */ > + lane_bundle = 0x0; /* Set to 1 lane. */ > + } > + > + /* overwrite existing values */ > + pcie_ctrl = (xmt_margin << MARGIN_GEN1_GEN2_SHIFT) > + | (xmt_margin_oe << MARGIN_G1_G2_OVERWRITE_SHIFT) > + | (xmt_margin << MARGIN_SHIFT) > + | (xmt_margin_oe << MARGIN_OVERWRITE_ENABLE_SHIFT) > + | (lane_delay << LANE_DELAY_SHIFT) > + | (lane_bundle << LANE_BUNDLE_SHIFT); > + > + write_csr(dd, CCE_PCIE_CTRL, pcie_ctrl); > + } > + > + dd_dev_info(dd, "%s: program XMT margin, CcePcieCtrl 0x%llx\n", > + fname, pcie_ctrl); Why spam the log with this all the time? Shouldn't this be a debug line instead? -- 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 ^ permalink raw reply [flat|nested] 5+ messages in thread
* RE: [PATCH v3] staging/rdma/hfi1: set Gen3 half-swing for integrated devices 2015-11-20 1:12 ` Greg KH @ 2015-11-20 15:14 ` Marciniszyn, Mike [not found] ` <32E1700B9017364D9B60AED9960492BC259BD75C-RjuIdWtd+YbTXloPLtfHfbfspsVTdybXVpNB7YpNyf8@public.gmane.org> 0 siblings, 1 reply; 5+ messages in thread From: Marciniszyn, Mike @ 2015-11-20 15:14 UTC (permalink / raw) To: Greg KH, Weiny, Ira Cc: devel@driverdev.osuosl.org, linux-rdma@vger.kernel.org, dledford@redhat.com, Luick, Dean > > + dd_dev_info(dd, "%s: program XMT margin, CcePcieCtrl 0x%llx\n", > > + fname, pcie_ctrl); > > Why spam the log with this all the time? Shouldn't this be a debug line > instead? The current implementation is layered on dev_info(). Are you saying dev_dbg() is a better choice? Mike ^ permalink raw reply [flat|nested] 5+ messages in thread
[parent not found: <32E1700B9017364D9B60AED9960492BC259BD75C-RjuIdWtd+YbTXloPLtfHfbfspsVTdybXVpNB7YpNyf8@public.gmane.org>]
* Re: [PATCH v3] staging/rdma/hfi1: set Gen3 half-swing for integrated devices [not found] ` <32E1700B9017364D9B60AED9960492BC259BD75C-RjuIdWtd+YbTXloPLtfHfbfspsVTdybXVpNB7YpNyf8@public.gmane.org> @ 2015-11-20 15:21 ` Greg KH [not found] ` <20151120152111.GA13281-U8xfFu+wG4EAvxtiuMwx3w@public.gmane.org> 0 siblings, 1 reply; 5+ messages in thread From: Greg KH @ 2015-11-20 15:21 UTC (permalink / raw) To: Marciniszyn, Mike Cc: Weiny, Ira, devel-gWbeCf7V1WCQmaza687I9mD2FQJk+8+b@public.gmane.org, dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org, linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Luick, Dean On Fri, Nov 20, 2015 at 03:14:19PM +0000, Marciniszyn, Mike wrote: > > > + dd_dev_info(dd, "%s: program XMT margin, CcePcieCtrl 0x%llx\n", > > > + fname, pcie_ctrl); > > > > Why spam the log with this all the time? Shouldn't this be a debug line > > instead? > > The current implementation is layered on dev_info(). > > Are you saying dev_dbg() is a better choice? Yes, why would a user ever want to see that above line? What can they do with it? -- 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 ^ permalink raw reply [flat|nested] 5+ messages in thread
[parent not found: <20151120152111.GA13281-U8xfFu+wG4EAvxtiuMwx3w@public.gmane.org>]
* RE: [PATCH v3] staging/rdma/hfi1: set Gen3 half-swing for integrated devices [not found] ` <20151120152111.GA13281-U8xfFu+wG4EAvxtiuMwx3w@public.gmane.org> @ 2015-11-20 15:54 ` Marciniszyn, Mike 0 siblings, 0 replies; 5+ messages in thread From: Marciniszyn, Mike @ 2015-11-20 15:54 UTC (permalink / raw) To: Greg KH Cc: Weiny, Ira, devel-gWbeCf7V1WCQmaza687I9mD2FQJk+8+b@public.gmane.org, dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org, linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Luick, Dean > > Are you saying dev_dbg() is a better choice? > > Yes, why would a user ever want to see that above line? What can they do > with it? Will do. Mike -- 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 ^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2015-11-20 15:54 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2015-11-16 22:03 [PATCH v3] staging/rdma/hfi1: set Gen3 half-swing for integrated devices ira.weiny-ral2JQCrhuEAvxtiuMwx3w
[not found] ` <1447711429-10389-1-git-send-email-ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
2015-11-20 1:12 ` Greg KH
2015-11-20 15:14 ` Marciniszyn, Mike
[not found] ` <32E1700B9017364D9B60AED9960492BC259BD75C-RjuIdWtd+YbTXloPLtfHfbfspsVTdybXVpNB7YpNyf8@public.gmane.org>
2015-11-20 15:21 ` Greg KH
[not found] ` <20151120152111.GA13281-U8xfFu+wG4EAvxtiuMwx3w@public.gmane.org>
2015-11-20 15:54 ` Marciniszyn, Mike
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox