IronPort-SDR: L28AYcyUoWBwk8Eei0uJhXpK1KjZLFoJ5Ep00NIDpMAMy3DbVxngpkJH0WdGZhEgfxZm1grBtM
 t3Z4GSJiqRr8wEsyhA0XTq8YBc39QHGja5aGn+a5zbcE/puEgQm253x8J1JY02fm2vTM4Rm0LU
 Mk1iuF1ACKQWaftMTBCZUFGMOqvV2IVs2LqTRIVaJityWZhoP8HyFnQrPe6o7TZNzF9u9taDrK
 1SJyVN0VDwhh7TLJsvrDveLw5YZYW0AiKQgzeO7DRustExvSK4oJ7hlnYtaCep2njUUTWU+oNp
 Je4=
IronPort-SDR: w4FoTf2sadxcBbIPYPjhMFNDV0oGSs45KF+PKKrC5bLSd0rWLhvXs3CrrIzsL1HGTY9XRyPC5R
 C4ohbtrcDHU8JMBmB1dL/s+BA844T7A1m2jw7iBAr/ya3KMwq+/Dg/IOEudJKxk/el23lw+5k7
 yF6su9sobN/XIceasOaGcAOIQY/vy61Se64wZ/lfuYAvQFUg8TPH9c697vvTfOpELcNLSQdXLn
 KwMewT7BhhaQykDa52g+oHjGne80JY7aWa1GQ97ZD5/GNr1fKNxZteGqyUk1RXLkL0wu5oIgiw
 /ww=
WDCIronportException: Internal
Message-ID: <67d08dce-4610-5e34-853a-1d09c5389607@opensource.wdc.com>
Date:   Wed, 15 Feb 2023 10:26:52 +0900
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.7.1
Subject: Re: [PATCH v2 7/9] PCI: rockchip: Fix legacy IRQ generation for
 RK3399 PCIe endpoint core
Content-Language: en-US
To:     Rick Wertenbroek <rick.wertenbroek@gmail.com>,
        alberto.dassatti@heig-vd.ch
Cc:     xxm@rock-chips.com, rick.wertenbroek@heig-vd.ch,
        stable@vger.kernel.org, Rob Herring <robh+dt@kernel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>,
        Heiko Stuebner <heiko@sntech.de>,
        Shawn Lin <shawn.lin@rock-chips.com>,
        Lorenzo Pieralisi <lpieralisi@kernel.org>,
        =?UTF-8?Q?Krzysztof_Wilczy=c5=84ski?= <kw@linux.com>,
        Bjorn Helgaas <bhelgaas@google.com>,
        Jani Nikula <jani.nikula@intel.com>,
        Mikko Kovanen <mikko.kovanen@aavamobile.com>,
        Rodrigo Vivi <rodrigo.vivi@intel.com>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        devicetree@vger.kernel.org, linux-arm-kernel@lists.infradead.org,
        linux-rockchip@lists.infradead.org, linux-kernel@vger.kernel.org,
        linux-pci@vger.kernel.org
References: <20230214140858.1133292-1-rick.wertenbroek@gmail.com>
 <20230214140858.1133292-8-rick.wertenbroek@gmail.com>
From:   Damien Le Moal <damien.lemoal@opensource.wdc.com>
Organization: Western Digital Research
In-Reply-To: <20230214140858.1133292-8-rick.wertenbroek@gmail.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 2/14/23 23:08, Rick Wertenbroek wrote:
> Fix legacy IRQ generation for RK3399 PCIe endpoint core according to
> the technical reference manual (TRM). Assert and deassert legacy
> interrupt (INTx) through the legacy interrupt control register
> ("PCIE_CLIENT_LEGACY_INT_CTRL") instead of manually generating a PCIe
> message. The generation of the legacy interrupt was tested and validated
> with the PCIe endpoint test driver.
> 
> Fixes: cf590b078391 ("PCI: rockchip: Add EP driver for Rockchip PCIe controller")
> Cc: stable@vger.kernel.org
> Signed-off-by: Rick Wertenbroek <rick.wertenbroek@gmail.com>

Some nits below. But otherwise works fine for me.

> ---
>  drivers/pci/controller/pcie-rockchip-ep.c | 38 +++++------------------
>  drivers/pci/controller/pcie-rockchip.h    |  6 ++++
>  2 files changed, 14 insertions(+), 30 deletions(-)
> 
> diff --git a/drivers/pci/controller/pcie-rockchip-ep.c b/drivers/pci/controller/pcie-rockchip-ep.c
> index cbc281a6a..ca5b363ba 100644
> --- a/drivers/pci/controller/pcie-rockchip-ep.c
> +++ b/drivers/pci/controller/pcie-rockchip-ep.c
> @@ -328,45 +328,23 @@ static void rockchip_pcie_ep_assert_intx(struct rockchip_pcie_ep *ep, u8 fn,
>  					 u8 intx, bool is_asserted)
>  {
>  	struct rockchip_pcie *rockchip = &ep->rockchip;
> -	u32 r = ep->max_regions - 1;
> -	u32 offset;
> -	u32 status;
> -	u8 msg_code;
> -
> -	if (unlikely(ep->irq_pci_addr != ROCKCHIP_PCIE_EP_PCI_LEGACY_IRQ_ADDR ||
> -		     ep->irq_pci_fn != fn)) {
> -		rockchip_pcie_prog_ep_ob_atu(rockchip, fn, r,
> -					     AXI_WRAPPER_NOR_MSG,
> -					     ep->irq_phys_addr, 0, 0);
> -		ep->irq_pci_addr = ROCKCHIP_PCIE_EP_PCI_LEGACY_IRQ_ADDR;
> -		ep->irq_pci_fn = fn;
> -	}
>  
>  	intx &= 3;
>  	if (is_asserted) {
>  		ep->irq_pending |= BIT(intx);
> -		msg_code = ROCKCHIP_PCIE_MSG_CODE_ASSERT_INTA + intx;
>  	} else {
>  		ep->irq_pending &= ~BIT(intx);
> -		msg_code = ROCKCHIP_PCIE_MSG_CODE_DEASSERT_INTA + intx;
>  	}
>  
> -	status = rockchip_pcie_read(rockchip,
> -				    ROCKCHIP_PCIE_EP_FUNC_BASE(fn) +
> -				    ROCKCHIP_PCIE_EP_CMD_STATUS);
> -	status &= ROCKCHIP_PCIE_EP_CMD_STATUS_IS;
> -
> -	if ((status != 0) ^ (ep->irq_pending != 0)) {
> -		status ^= ROCKCHIP_PCIE_EP_CMD_STATUS_IS;
> -		rockchip_pcie_write(rockchip, status,
> -				    ROCKCHIP_PCIE_EP_FUNC_BASE(fn) +
> -				    ROCKCHIP_PCIE_EP_CMD_STATUS);
> +	if (is_asserted) {
> +		rockchip_pcie_write(rockchip,
> +			PCIE_CLIENT_INT_IN_ASSERT | PCIE_CLIENT_INT_PEND_ST_PEND,
> +			PCIE_CLIENT_LEGACY_INT_CTRL);
> +	} else {
> +		rockchip_pcie_write(rockchip,
> +			PCIE_CLIENT_INT_IN_DEASSERT | PCIE_CLIENT_INT_PEND_ST_NORMAL,
> +			PCIE_CLIENT_LEGACY_INT_CTRL);
>  	}

With this change, you have now twice "if (is_asserted) {", which is not
necessary. You can simplify the code a bit:

static void rockchip_pcie_ep_assert_intx(struct rockchip_pcie_ep *ep,
					 u8 fn, u8 intx, bool do_assert)
{

        u8 msg_code;


        intx &= 3;

        if (do_assert) {

                ep->irq_pending |= BIT(intx);

                msg_code = ROCKCHIP_PCIE_MSG_CODE_ASSERT_INTA + intx;

                rockchip_pcie_write(&ep->rockchip,

                        PCIE_CLIENT_INT_IN_ASSERT |

                        PCIE_CLIENT_INT_PEND_ST_PEND,

                        PCIE_CLIENT_LEGACY_INT_CTRL);

                return;

        }


        ep->irq_pending &= ~BIT(intx);

        msg_code = ROCKCHIP_PCIE_MSG_CODE_DEASSERT_INTA + intx;

        rockchip_pcie_write(&ep->rockchip,

                            PCIE_CLIENT_INT_IN_DEASSERT |

                            PCIE_CLIENT_INT_PEND_ST_NORMAL,

                            PCIE_CLIENT_LEGACY_INT_CTRL);

}

Note also the renaming of the argument "is_asserted" to "do_assert". The
name is_asserted is badly misleading considering the english meaning given
that it is true when we *must* do the assert and false when we must
deassert. So do_assert as a name better match the use of that argument I
think.

> -
> -	offset =
> -	   ROCKCHIP_PCIE_MSG_ROUTING(ROCKCHIP_PCIE_MSG_ROUTING_LOCAL_INTX) |
> -	   ROCKCHIP_PCIE_MSG_CODE(msg_code) | ROCKCHIP_PCIE_MSG_NO_DATA;
> -	writel(0, ep->irq_cpu_addr + offset);
>  }
>  
>  static int rockchip_pcie_ep_send_legacy_irq(struct rockchip_pcie_ep *ep, u8 fn,
> diff --git a/drivers/pci/controller/pcie-rockchip.h b/drivers/pci/controller/pcie-rockchip.h
> index 72e427a0f..e90c2a2b8 100644
> --- a/drivers/pci/controller/pcie-rockchip.h
> +++ b/drivers/pci/controller/pcie-rockchip.h
> @@ -39,6 +39,12 @@
>  #define   PCIE_CLIENT_GEN_SEL_1		  HIWORD_UPDATE(0x0080, 0)
>  #define   PCIE_CLIENT_GEN_SEL_2		  HIWORD_UPDATE_BIT(0x0080)
>  #define PCIE_CLIENT_SIDE_BAND_STATUS	(PCIE_CLIENT_BASE + 0x20)
> +#define PCIE_CLIENT_LEGACY_INT_CTRL		(PCIE_CLIENT_BASE + 0x0c)
> +#define   PCIE_CLIENT_INT_IN_ASSERT		HIWORD_UPDATE_BIT(0x0002)
> +#define   PCIE_CLIENT_INT_IN_DEASSERT	HIWORD_UPDATE(0x0002, 0)
> +#define   PCIE_CLIENT_INT_PEND_ST_PEND	HIWORD_UPDATE_BIT(0x0001)
> +#define   PCIE_CLIENT_INT_PEND_ST_NORMAL	HIWORD_UPDATE(0x0001, 0)
> +#define PCIE_CLIENT_SIDE_BAND_STATUS	(PCIE_CLIENT_BASE + 0x20)
>  #define   PCIE_CLIENT_PHY_ST			BIT(12)
>  #define PCIE_CLIENT_DEBUG_OUT_0		(PCIE_CLIENT_BASE + 0x3c)
>  #define   PCIE_CLIENT_DEBUG_LTSSM_MASK		GENMASK(5, 0)

-- 
Damien Le Moal
Western Digital Research