diff options
author | Pali Rohár <pali@kernel.org> | 2020-07-21 14:11:08 +0200 |
---|---|---|
committer | Hauke Mehrtens <hauke@hauke-m.de> | 2020-10-11 16:16:24 +0200 |
commit | eb777539fd18fec624aa461feec20f1f92624620 (patch) | |
tree | 62bf7ceb2cceaee9184b99356f8561a3af8dc073 /target/linux/mvebu/patches-5.4/402-PCI-aardvark-Wait-for-endpoint-to-be-ready-before-tr.patch | |
parent | 5b3ff53ebd6846a0f03bfa29d9aa5d22a6a05bfb (diff) | |
download | upstream-eb777539fd18fec624aa461feec20f1f92624620.tar.gz upstream-eb777539fd18fec624aa461feec20f1f92624620.tar.bz2 upstream-eb777539fd18fec624aa461feec20f1f92624620.zip |
mvebu: armada-37xx: Backport PCI aardvark patches
This commit contains patches for PCI aardvark driver and relevant DTS
changes for Turris MOX and EspressoBin backported from mainline kernel.
It fixes support for old ATF, various wifi cards, mainly Compex WLE900VX.
Signed-off-by: Pali Rohár <pali@kernel.org>
Diffstat (limited to 'target/linux/mvebu/patches-5.4/402-PCI-aardvark-Wait-for-endpoint-to-be-ready-before-tr.patch')
-rw-r--r-- | target/linux/mvebu/patches-5.4/402-PCI-aardvark-Wait-for-endpoint-to-be-ready-before-tr.patch | 50 |
1 files changed, 50 insertions, 0 deletions
diff --git a/target/linux/mvebu/patches-5.4/402-PCI-aardvark-Wait-for-endpoint-to-be-ready-before-tr.patch b/target/linux/mvebu/patches-5.4/402-PCI-aardvark-Wait-for-endpoint-to-be-ready-before-tr.patch new file mode 100644 index 0000000000..95abbce002 --- /dev/null +++ b/target/linux/mvebu/patches-5.4/402-PCI-aardvark-Wait-for-endpoint-to-be-ready-before-tr.patch @@ -0,0 +1,50 @@ +From f4c7d053d7f77cd5c1a1ba7c7ce085ddba13d1d7 Mon Sep 17 00:00:00 2001 +From: Remi Pommarel <repk@triplefau.lt> +Date: Wed, 22 May 2019 23:33:50 +0200 +Subject: [PATCH] PCI: aardvark: Wait for endpoint to be ready before training + link + +When configuring pcie reset pin from gpio (e.g. initially set by +u-boot) to pcie function this pin goes low for a brief moment +asserting the PERST# signal. Thus connected device enters fundamental +reset process and link configuration can only begin after a minimal +100ms delay (see [1]). + +Because the pin configuration comes from the "default" pinctrl it is +implicitly configured before the probe callback is called: + +driver_probe_device() + really_probe() + ... + pinctrl_bind_pins() /* Here pin goes from gpio to PCIE reset + function and PERST# is asserted */ + ... + drv->probe() + +[1] "PCI Express Base Specification", REV. 4.0 + PCI Express, February 19 2014, 6.6.1 Conventional Reset + +Signed-off-by: Remi Pommarel <repk@triplefau.lt> +Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com> +Acked-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com> +--- + drivers/pci/controller/pci-aardvark.c | 8 ++++++++ + 1 file changed, 8 insertions(+) + +--- a/drivers/pci/controller/pci-aardvark.c ++++ b/drivers/pci/controller/pci-aardvark.c +@@ -337,6 +337,14 @@ static void advk_pcie_setup_hw(struct ad + reg |= PIO_CTRL_ADDR_WIN_DISABLE; + advk_writel(pcie, reg, PIO_CTRL); + ++ /* ++ * PERST# signal could have been asserted by pinctrl subsystem before ++ * probe() callback has been called, making the endpoint going into ++ * fundamental reset. As required by PCI Express spec a delay for at ++ * least 100ms after such a reset before link training is needed. ++ */ ++ msleep(PCI_PM_D3COLD_WAIT); ++ + /* Start link training */ + reg = advk_readl(pcie, PCIE_CORE_LINK_CTRL_STAT_REG); + reg |= PCIE_CORE_LINK_TRAINING; |