[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20240105112547.7301-5-ilpo.jarvinen@linux.intel.com>
Date: Fri, 5 Jan 2024 13:25:43 +0200
From: Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>
To: linux-pci@...r.kernel.org,
Bjorn Helgaas <bhelgaas@...gle.com>,
Lorenzo Pieralisi <lorenzo.pieralisi@....com>,
Rob Herring <robh@...nel.org>,
Krzysztof Wilczyński <kw@...ux.com>,
Lukas Wunner <lukas@...ner.de>,
Alexandru Gagniuc <mr.nuke.me@...il.com>,
Krishna chaitanya chundru <quic_krichai@...cinc.com>,
Srinivas Pandruvada <srinivas.pandruvada@...ux.intel.com>,
"Rafael J. Wysocki" <rafael@...nel.org>,
linux-pm@...r.kernel.org,
Dennis Dalessandro <dennis.dalessandro@...nelisnetworks.com>,
Jason Gunthorpe <jgg@...pe.ca>,
Leon Romanovsky <leon@...nel.org>,
linux-rdma@...r.kernel.org,
linux-kernel@...r.kernel.org
Cc: Alex Deucher <alexdeucher@...il.com>,
Daniel Lezcano <daniel.lezcano@...aro.org>,
Amit Kucheria <amitk@...nel.org>,
Zhang Rui <rui.zhang@...el.com>,
Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>,
Dean Luick <dean.luick@...nelisnetworks.com>
Subject: [PATCH v4 4/8] RDMA/hfi1: Use RMW accessors for changing LNKCTL2
Don't assume that only the driver would be accessing LNKCTL2. In the
case of upstream (parent), the driver does not even own the device it's
changing the registers for.
Use RMW capability accessors which do proper locking to avoid losing
concurrent updates to the register value. This change is also useful as
a cleanup.
Suggested-by: Lukas Wunner <lukas@...ner.de>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>
Reviewed-by: Dean Luick <dean.luick@...nelisnetworks.com>
---
drivers/infiniband/hw/hfi1/pcie.c | 30 ++++++++----------------------
1 file changed, 8 insertions(+), 22 deletions(-)
diff --git a/drivers/infiniband/hw/hfi1/pcie.c b/drivers/infiniband/hw/hfi1/pcie.c
index 119ec2f1382b..7133964749f8 100644
--- a/drivers/infiniband/hw/hfi1/pcie.c
+++ b/drivers/infiniband/hw/hfi1/pcie.c
@@ -1207,14 +1207,11 @@ int do_pcie_gen3_transition(struct hfi1_devdata *dd)
(u32)lnkctl2);
/* only write to parent if target is not as high as ours */
if ((lnkctl2 & PCI_EXP_LNKCTL2_TLS) < target_vector) {
- lnkctl2 &= ~PCI_EXP_LNKCTL2_TLS;
- lnkctl2 |= target_vector;
- dd_dev_info(dd, "%s: ..new link control2: 0x%x\n", __func__,
- (u32)lnkctl2);
- ret = pcie_capability_write_word(parent,
- PCI_EXP_LNKCTL2, lnkctl2);
+ ret = pcie_capability_clear_and_set_word(parent, PCI_EXP_LNKCTL2,
+ PCI_EXP_LNKCTL2_TLS,
+ target_vector);
if (ret) {
- dd_dev_err(dd, "Unable to write to PCI config\n");
+ dd_dev_err(dd, "Unable to change parent PCI target speed\n");
return_error = 1;
goto done;
}
@@ -1223,22 +1220,11 @@ int do_pcie_gen3_transition(struct hfi1_devdata *dd)
}
dd_dev_info(dd, "%s: setting target link speed\n", __func__);
- ret = pcie_capability_read_word(dd->pcidev, PCI_EXP_LNKCTL2, &lnkctl2);
+ ret = pcie_capability_clear_and_set_word(dd->pcidev, PCI_EXP_LNKCTL2,
+ PCI_EXP_LNKCTL2_TLS,
+ target_vector);
if (ret) {
- dd_dev_err(dd, "Unable to read from PCI config\n");
- return_error = 1;
- goto done;
- }
-
- dd_dev_info(dd, "%s: ..old link control2: 0x%x\n", __func__,
- (u32)lnkctl2);
- lnkctl2 &= ~PCI_EXP_LNKCTL2_TLS;
- lnkctl2 |= target_vector;
- dd_dev_info(dd, "%s: ..new link control2: 0x%x\n", __func__,
- (u32)lnkctl2);
- ret = pcie_capability_write_word(dd->pcidev, PCI_EXP_LNKCTL2, lnkctl2);
- if (ret) {
- dd_dev_err(dd, "Unable to write to PCI config\n");
+ dd_dev_err(dd, "Unable to change device PCI target speed\n");
return_error = 1;
goto done;
}
--
2.39.2
Powered by blists - more mailing lists