diff mbox series

[U-Boot] drivers: pci: imx: fix imx_pcie_remove function

Message ID 1507204002-14559-1-git-send-email-fabio.estevam@nxp.com
State Accepted
Commit b2915ba25e350caa79a13a1cd3513eb82709e1f9
Delegated to: Stefano Babic
Headers show
Series [U-Boot] drivers: pci: imx: fix imx_pcie_remove function | expand

Commit Message

Fabio Estevam Oct. 5, 2017, 11:46 a.m. UTC 
From: Sven-Ola Tuecke <sven-ola.tuecke@numberfour.eu>

We have at least a minor count of boards, that failed to re-initialize
PCI express in the Linux kernel. Typical failure rate is 20% on affected
boards. This is mitigated by commit 6ecbe1375671 ("drivers: pci: imx:
add imx_pcie_remove function").

However, at least on some i.MX6 custom boards, when calling
assert_core_reset() as part of the first-time PCIe init, read access
to PCIE_PL_PFLR simply hangs. Surround this readl() with 
imx_pcie_fix_dabt_handler() does not help. For this reason, the forced
LTSSM detection is only used on the second assert_core_reset() that is
called shorly before starting the Linux kernel.

Signed-off-by: Sven-Ola Tuecke <sven-ola.tuecke@numberfour.eu>
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
---
 drivers/pci/pcie_imx.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

Comments

Fabio Estevam Oct. 10, 2017, 12:12 p.m. UTC | #1 
Hi David,

On Thu, Oct 5, 2017 at 8:46 AM, Fabio Estevam <fabio.estevam@nxp.com> wrote:
> From: Sven-Ola Tuecke <sven-ola.tuecke@numberfour.eu>
>
> We have at least a minor count of boards, that failed to re-initialize
> PCI express in the Linux kernel. Typical failure rate is 20% on affected
> boards. This is mitigated by commit 6ecbe1375671 ("drivers: pci: imx:
> add imx_pcie_remove function").
>
> However, at least on some i.MX6 custom boards, when calling
> assert_core_reset() as part of the first-time PCIe init, read access
> to PCIE_PL_PFLR simply hangs. Surround this readl() with
> imx_pcie_fix_dabt_handler() does not help. For this reason, the forced
> LTSSM detection is only used on the second assert_core_reset() that is
> called shorly before starting the Linux kernel.
>
> Signed-off-by: Sven-Ola Tuecke <sven-ola.tuecke@numberfour.eu>
> Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>

Does this patch solve the PCIE_PL_PFLR hang you reported at
https://lists.denx.de/pipermail/u-boot/2017-September/307191.html ?
David Müller (ELSOFT AG) Oct. 11, 2017, 8:18 a.m. UTC | #2 
Hi Fabio

Fabio Estevam wrote:
> From: Sven-Ola Tuecke <sven-ola.tuecke@numberfour.eu>
> 
> However, at least on some i.MX6 custom boards, when calling

By "some", I assume that not all i.MX6 systems are showing this
behaviour. Does anyone have an idea which systems are affected and which
are not?

> assert_core_reset() as part of the first-time PCIe init, read access
> to PCIE_PL_PFLR simply hangs. Surround this readl() with 
> imx_pcie_fix_dabt_handler() does not help. For this reason, the forced
> LTSSM detection is only used on the second assert_core_reset() that is
> called shorly before starting the Linux kernel.
> 
> Signed-off-by: Sven-Ola Tuecke <sven-ola.tuecke@numberfour.eu>
> Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>

Seems to work on our board therefore

Tested-by: David Müller <d.mueller@elsoft.ch>
Stefano Babic Oct. 12, 2017, 4:24 p.m. UTC | #3 
On 05/10/2017 13:46, Fabio Estevam wrote:
> From: Sven-Ola Tuecke <sven-ola.tuecke@numberfour.eu>
> 
> We have at least a minor count of boards, that failed to re-initialize
> PCI express in the Linux kernel. Typical failure rate is 20% on affected
> boards. This is mitigated by commit 6ecbe1375671 ("drivers: pci: imx:
> add imx_pcie_remove function").
> 
> However, at least on some i.MX6 custom boards, when calling
> assert_core_reset() as part of the first-time PCIe init, read access
> to PCIE_PL_PFLR simply hangs. Surround this readl() with 
> imx_pcie_fix_dabt_handler() does not help. For this reason, the forced
> LTSSM detection is only used on the second assert_core_reset() that is
> called shorly before starting the Linux kernel.
> 
> Signed-off-by: Sven-Ola Tuecke <sven-ola.tuecke@numberfour.eu>
> Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
> ---


Applied to u-boot-imx, thanks !

Best regards,
Stefano Babic
Fabio Estevam Oct. 13, 2017, 12:18 a.m. UTC | #4 
Hi David,

On Wed, Oct 11, 2017 at 5:18 AM, David Müller (ELSOFT AG)
<d.mueller@elsoft.ch> wrote:
> Hi Fabio
>
> Fabio Estevam wrote:
>> From: Sven-Ola Tuecke <sven-ola.tuecke@numberfour.eu>
>>
>> However, at least on some i.MX6 custom boards, when calling
>
> By "some", I assume that not all i.MX6 systems are showing this
> behaviour. Does anyone have an idea which systems are affected and which
> are not?

Yes, the mx6 board I have access to do not show this issue.
Not sure why it fails on your board and on Sven's.

>> assert_core_reset() as part of the first-time PCIe init, read access
>> to PCIE_PL_PFLR simply hangs. Surround this readl() with
>> imx_pcie_fix_dabt_handler() does not help. For this reason, the forced
>> LTSSM detection is only used on the second assert_core_reset() that is
>> called shorly before starting the Linux kernel.
>>
>> Signed-off-by: Sven-Ola Tuecke <sven-ola.tuecke@numberfour.eu>
>> Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
>
> Seems to work on our board therefore
>
> Tested-by: David Müller <d.mueller@elsoft.ch>

Thanks for testing and glad it works for both of you.
diff mbox series

Patch

diff --git a/drivers/pci/pcie_imx.c b/drivers/pci/pcie_imx.c
index eab0a2b..2900c8d 100644
--- a/drivers/pci/pcie_imx.c
+++ b/drivers/pci/pcie_imx.c
@@ -431,7 +431,7 @@  static int imx_pcie_write_config(struct pci_controller *hose, pci_dev_t d,
 /*
  * Initial bus setup
  */
-static int imx6_pcie_assert_core_reset(void)
+static int imx6_pcie_assert_core_reset(bool prepare_for_boot)
 {
 	struct iomuxc *iomuxc_regs = (struct iomuxc *)IOMUXC_BASE_ADDR;
 
@@ -459,7 +459,7 @@  static int imx6_pcie_assert_core_reset(void)
 	 * If both LTSSM_ENABLE and REF_SSP_ENABLE are active we have a strong
 	 * indication that the bootloader activated the link.
 	 */
-	if (is_mx6dq()) {
+	if (is_mx6dq() && prepare_for_boot) {
 		u32 val, gpr1, gpr12;
 
 		gpr1 = readl(&iomuxc_regs->gpr[1]);
@@ -605,7 +605,7 @@  static int imx_pcie_link_up(void)
 	uint32_t tmp;
 	int count = 0;
 
-	imx6_pcie_assert_core_reset();
+	imx6_pcie_assert_core_reset(false);
 	imx6_pcie_init_phy();
 	imx6_pcie_deassert_core_reset();
 
@@ -687,7 +687,7 @@  void imx_pcie_init(void)
 
 void imx_pcie_remove(void)
 {
-	imx6_pcie_assert_core_reset();
+	imx6_pcie_assert_core_reset(true);
 }
 
 /* Probe function. */