In the subject: PCI: mediatek: Assert PERST# for 100ms for power and clock to stabilize On Tue, Dec 07, 2021 at 02:05:50PM +0800, qizhong cheng wrote: > Described in PCIe CEM specification setctions 2.2 (PERST# Signal) and > 2.2.1 (Initial Power-Up (G3 to S0)). The deassertion of PERST# should > be delayed 100ms (TPVPERL) for the power and clock to become stable. Please fix the typos above and in the comment below: "setctions" should be "sections". I mentioned this yesterday: https://lore.kernel.org/all/20211207015323.GA26237@bhelgaas/ But maybe the "s/setctions/sections/" notation wasn't obvious. This is used in sed, vi, and other tools, and it is a command that searches for the first regular expression ("setctions") and replaces it with the second ("sections"). > Signed-off-by: qizhong cheng <qizhong.cheng@xxxxxxxxxxxx> > Change-Id: Ia9abe1e763564a5bad1d045fd268c38e76e2ae95 No need for the "Change-Id". I assume this came from a gerrit server somewhere, but unless there's something to tell us *which* server that is, and unless that server is public, this isn't really useful. > --- > drivers/pci/controller/pcie-mediatek.c | 7 +++++++ > 1 file changed, 7 insertions(+) > > diff --git a/drivers/pci/controller/pcie-mediatek.c b/drivers/pci/controller/pcie-mediatek.c > index 2f3f974977a3..a61ea3940471 100644 > --- a/drivers/pci/controller/pcie-mediatek.c > +++ b/drivers/pci/controller/pcie-mediatek.c > @@ -702,6 +702,13 @@ static int mtk_pcie_startup_port_v2(struct mtk_pcie_port *port) > */ > writel(PCIE_LINKDOWN_RST_EN, port->base + PCIE_RST_CTRL); > > + /* > + * Described in PCIe CEM specification setctions 2.2 (PERST# Signal) and > + * 2.2.1 (Initial Power-Up (G3 to S0)). The deassertion of PERST# should > + * be delayed 100ms (TPVPERL) for the power and clock to become stable. > + */ > + msleep(100); > + > /* De-assert PHY, PE, PIPE, MAC and configuration reset */ > val = readl(port->base + PCIE_RST_CTRL); > val |= PCIE_PHY_RSTB | PCIE_PERSTB | PCIE_PIPE_SRSTB | > -- > 2.25.1 >