Re: [PATCH v2] PCI: apple: Follow the PCIe specifications when resetting the port

Date: Tue, 23 Nov 2021 12:24:10 +0000
From: Marc Zyngier <maz@kernel.org>

On Mon, 22 Nov 2021 21:50:48 +0000,
Mark Kettenis [off-list ref] wrote:

quoted

quoted

Date: Mon, 22 Nov 2021 14:43:37 +0000
From: Marc Zyngier <maz@kernel.org>

On Mon, 22 Nov 2021 12:03:47 +0000,
Pali Rohár [off-list ref] wrote:

quoted

On Monday 22 November 2021 10:41:56 Marc Zyngier wrote:

quoted

While the Apple PCIe driver works correctly when directly booted
from the firmware, it fails to initialise when the kernel is booted
from a bootloader using PCIe such as u-boot.

That's beacuse we're missing a proper reset of the port (we only
clear the reset, but never assert it).

The PCIe spec requirements are two-fold:

- #PERST must be asserted before setting up the clocks, and
  stay asserted for at least 100us (Tperst-clk).

- Once #PERST is deasserted, the OS must wait for at least 100ms
  "from the end of a Conventional Reset" before we can start talking
  to the devices

Implementing this results in a booting system.

Fixes: 1e33888fbe44 ("PCI: apple: Add initial hardware bring-up")
Signed-off-by: Marc Zyngier <maz@kernel.org>
Cc: Alyssa Rosenzweig <redacted>
Cc: Lorenzo Pieralisi <redacted>
Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Pali Rohár <pali@kernel.org>

Looks good, but see comment below.

Acked-by: Pali Rohár <pali@kernel.org>

Thanks for that.

quoted

quoted

---
 drivers/pci/controller/pcie-apple.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/drivers/pci/controller/pcie-apple.c b/drivers/pci/controller/pcie-apple.c
index 1bf4d75b61be..957960a733c4 100644
--- a/drivers/pci/controller/pcie-apple.c
+++ b/drivers/pci/controller/pcie-apple.c

@@ -539,13 +539,23 @@ static int apple_pcie_setup_port(struct apple_pcie *pcie,
 
 	rmw_set(PORT_APPCLK_EN, port->base + PORT_APPCLK);
 
+	/* Engage #PERST before setting up the clock */
+	gpiod_set_value(reset, 0);
+
 	ret = apple_pcie_setup_refclk(pcie, port);
 	if (ret < 0)
 		return ret;
 
+	/* The minimal Tperst-clk value is 100us (PCIe CMS r2.0, 2.6.2) */
+	usleep_range(100, 200);
+
+	/* Deassert #PERST */
 	rmw_set(PORT_PERST_OFF, port->base + PORT_PERST);
 	gpiod_set_value(reset, 1);

+ Luca

Just one comment. PERST# (PCIe Reset) is active-low signal. De-asserting
means to really set value to 1.

But there was a discussion that de-asserting should be done by call:
  gpiod_set_value(reset, 0);

https://lore.kernel.org/linux-pci/51be082a-ff10-8a19-5648-f279aabcac51@lucaceresoli.net/ (local)

Could we make this new pcie-apple.c driver to use gpiod_set_value(reset, 0)
for de-asserting, like in other drivers?

I guess it depends whether you care about the assertion or the signal
itself. I think we may have a bug in the way the GPIOs are handled at
the moment, as it makes no difference whether I register the GPIO are
active high or active low...

That's unfortunate.  But maybe that's an opportunity to fix the
devicetree to use GPIO_ACTIVE_LOW for these GPIOs?

Indeed. The following hack does the right thing, and I can then
reverse the polarity of the reset in the Linux driver. Of course, it
breaks u-boot at the same time (and I suspect OpenBSD would be equally
affected).

As I said in my other reply (that clearly crossed this message) the
U-Boot PCIe driver has not been upstreamed yet.  OpenBSD (and also
NetBSD) rely on the configuration done by U-Boot and don't mess with
the reset GPIO at this moment.

So if we are going down that road, we may need a flag day where all
the moving parts change. I don't really mind not being able to boot
older kernels, but this goes beyond Linux at this point.

So I don't think this is needed.  The earlier we fix this the better!

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel