linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Lucas Stach <l.stach@pengutronix.de>
To: Stefan Agner <stefan@agner.ch>, hongxing.zhu@nxp.com
Cc: robin.murphy@arm.com, tpiepho@impinj.com, linux@armlinux.org.uk,
	leonard.crestez@nxp.com, andrew.smirnov@gmail.com,
	festevam@gmail.com, lorenzo.pieralisi@arm.com,
	bhelgaas@google.com, linux-pci@vger.kernel.org,
	linux-arm-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 2/2] pci: imx6: support kernels built in Thumb-2 mode
Date: Mon, 11 Feb 2019 17:57:53 +0100	[thread overview]
Message-ID: <1549904273.2546.16.camel@pengutronix.de> (raw)
In-Reply-To: <20181204132733.14422-2-stefan@agner.ch>

Am Dienstag, den 04.12.2018, 14:27 +0100 schrieb Stefan Agner:
> Add a fault handler which handles immediate reads in Thumb-2
> mode. Install the appropriate handler depending on which mode
> the kernel has been built. This avoids an "Unhandled fault:
> external abort on non-linefetch (0x1008) at 0xf0a80000"
> during boot on a device with a PCIe switch connected.
> 
> Link: https://lore.kernel.org/linux-pci/20181126161645.8177-1-stefan@agner.ch/
> Signed-off-by: Stefan Agner <stefan@agner.ch>

Acked-by: Lucas Stach <l.stach@pengutronix.de>

> ---
> Changes since v1:
> - Added Thumb-2 32-bit instruction support (tested by inserting .w
>   instructions in arch/arm/include/asm/io.h)
> - Avoid dereferencing if fault happened in user mode
> 
>  drivers/pci/controller/dwc/pci-imx6.c | 59 ++++++++++++++++++++++++++-
>  1 file changed, 58 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/pci/controller/dwc/pci-imx6.c b/drivers/pci/controller/dwc/pci-imx6.c
> index 54a29e441303..a9bbbf176c4a 100644
> --- a/drivers/pci/controller/dwc/pci-imx6.c
> +++ b/drivers/pci/controller/dwc/pci-imx6.c
> @@ -29,6 +29,7 @@
>  #include <linux/reset.h>
>  #include <linux/pm_domain.h>
>  #include <linux/pm_runtime.h>
> +#include <asm/opcodes.h>
>  
>  #include "pcie-designware.h"
>  
> @@ -305,6 +306,59 @@ static int imx6q_pcie_abort_handler(unsigned long addr,
>  	return 1;
>  }
>  
> +static int imx6q_pcie_abort_handler_thumb2(unsigned long addr,
> +		unsigned int fsr, struct pt_regs *regs)
> +{
> +	unsigned long pc = instruction_pointer(regs);
> +	unsigned long instr;
> +
> +	if (user_mode(regs))
> +		return 1;
> +
> +	instr = __mem_to_opcode_thumb32(*(unsigned long *)pc);
> +
> +	if (__opcode_is_thumb32(instr)) {
> +		/* Load word/byte and halfword immediate offset */
> +		if ((instr & 0xff100000UL) == 0xf8100000UL) {
> +			int reg = (instr >> 12) & 0xf;
> +			unsigned long val;
> +
> +			if ((instr & 0x00700000UL) == 0x00100000UL)
> +				val = 0xff;
> +			else if ((instr & 0x00700000UL) == 0x00300000UL)
> +				val = 0xffff;
> +			else
> +				val = 0xffffffffUL;
> +
> +			regs->uregs[reg] = val;
> +			regs->ARM_pc += 4;
> +			return 0;
> +		}
> +	} else {
> +		instr = __mem_to_opcode_thumb16(*(unsigned long *)pc);
> +
> +		/* Load word/byte and halfword immediate offset */
> +		if (((instr & 0xe800) == 0x6800) ||
> +		    ((instr & 0xf800) == 0x8800)) {
> +			int reg = instr & 0x7;
> +			unsigned long val;
> +
> +			if (instr & 0x1000)
> +				val = 0xff;
> +			else if (instr & 0x8000)
> +				val = 0xffff;
> +			else
> +				val = 0xffffffffUL;
> +
> +			regs->uregs[reg] = val;
> +			regs->ARM_pc += 2;
> +			return 0;
> +		}
> +	}
> +
> +	return 1;
> +}
> +
>  static int imx6_pcie_attach_pd(struct device *dev)
>  {
>  	struct imx6_pcie *imx6_pcie = dev_get_drvdata(dev);
> @@ -1075,6 +1129,8 @@ static struct platform_driver imx6_pcie_driver = {
>  
>  static int __init imx6_pcie_init(void)
>  {
> +	bool thumb2 = IS_ENABLED(CONFIG_THUMB2_KERNEL);
> +
>  	/*
>  	 * Since probe() can be deferred we need to make sure that
>  	 * hook_fault_code is not called after __init memory is freed
> @@ -1082,7 +1138,8 @@ static int __init imx6_pcie_init(void)
>  	 * we can install the handler here without risking it
>  	 * accessing some uninitialized driver state.
>  	 */
> -	hook_fault_code(8, imx6q_pcie_abort_handler, SIGBUS, 0,
> +	hook_fault_code(8, thumb2 ? imx6q_pcie_abort_handler_thumb2 :
> +			imx6q_pcie_abort_handler, SIGBUS, 0,
>  			"external abort on non-linefetch");
>  
>  	return platform_driver_register(&imx6_pcie_driver);

  reply	other threads:[~2019-02-11 16:58 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-12-04 13:27 [PATCH v2 1/2] pci: imx6: avoid dereferencing program counter from user mode Stefan Agner
2018-12-04 13:27 ` [PATCH v2 2/2] pci: imx6: support kernels built in Thumb-2 mode Stefan Agner
2019-02-11 16:57   ` Lucas Stach [this message]
2019-02-08 12:13 ` [PATCH v2 1/2] pci: imx6: avoid dereferencing program counter from user mode Lorenzo Pieralisi
2019-02-11 16:56 ` Lucas Stach

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1549904273.2546.16.camel@pengutronix.de \
    --to=l.stach@pengutronix.de \
    --cc=andrew.smirnov@gmail.com \
    --cc=bhelgaas@google.com \
    --cc=festevam@gmail.com \
    --cc=hongxing.zhu@nxp.com \
    --cc=leonard.crestez@nxp.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=linux@armlinux.org.uk \
    --cc=lorenzo.pieralisi@arm.com \
    --cc=robin.murphy@arm.com \
    --cc=stefan@agner.ch \
    --cc=tpiepho@impinj.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).