lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <21d17701-e40b-49c5-bf16-5e70a6daa2c4@collabora.com>
Date: Thu, 15 Feb 2024 11:40:52 +0100
From: AngeloGioacchino Del Regno <angelogioacchino.delregno@...labora.com>
To: Chun-Kuang Hu <chunkuang.hu@...nel.org>,
 Matthias Brugger <matthias.bgg@...il.com>,
 Mauro Carvalho Chehab <mchehab@...nel.org>,
 linux-arm-kernel@...ts.infradead.org, linux-mediatek@...ts.infradead.org,
 linux-kernel@...r.kernel.org, linux-media@...r.kernel.org
Subject: Re: [PATCH 2/9] soc: mediatek: cmdq: Add parameter shift_pa to
 cmdq_pkt_jump()

Il 15/02/24 01:49, Chun-Kuang Hu ha scritto:
> In original design, cmdq_pkt_jump() call cmdq_get_shift_pa() every
> time to get shift_pa. But the shift_pa is constant value for each
> SoC, so client driver just need to call cmdq_get_shift_pa() once
> and pass shift_pa to cmdq_pkt_jump() to prevent frequent function
> call.
> 

As far as I understand, the CMDQ supports both relative and absolute jumps, right?

Here's my proposal:
  - Add a new function cmdq_pkt_jump_rel() or cmdq_pkt_jump_relative()
    * note: I prefer "rel", as maybe in a future we'll get a jump_abs function? :-)
  - Don't touch the cmdq_pkt_jump() function for one cycle
    - Migrate mediatek-drm to use cmdq_pkt_jump_rel()
  - Remove cmdq_pkt_jump() in the next cycle.

What do you think?

Regards,
Angelo

> Signed-off-by: Chun-Kuang Hu <chunkuang.hu@...nel.org>
> ---
>   drivers/soc/mediatek/mtk-cmdq-helper.c | 5 ++---
>   include/linux/soc/mediatek/mtk-cmdq.h  | 6 ++++--
>   2 files changed, 6 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/soc/mediatek/mtk-cmdq-helper.c b/drivers/soc/mediatek/mtk-cmdq-helper.c
> index 67e17974d1e6..ed4ef95adf5b 100644
> --- a/drivers/soc/mediatek/mtk-cmdq-helper.c
> +++ b/drivers/soc/mediatek/mtk-cmdq-helper.c
> @@ -348,14 +348,13 @@ int cmdq_pkt_assign(struct cmdq_pkt *pkt, u16 reg_idx, u32 value)
>   }
>   EXPORT_SYMBOL(cmdq_pkt_assign);
>   
> -int cmdq_pkt_jump(struct cmdq_pkt *pkt, dma_addr_t addr)
> +int cmdq_pkt_jump(struct cmdq_pkt *pkt, dma_addr_t addr, u8 shift_pa)
>   {
>   	struct cmdq_instruction inst = {};
>   
>   	inst.op = CMDQ_CODE_JUMP;
>   	inst.offset = CMDQ_JUMP_RELATIVE;
> -	inst.value = addr >>
> -		cmdq_get_shift_pa(((struct cmdq_client *)pkt->cl)->chan);
> +	inst.value = addr >> shift_pa;
>   	return cmdq_pkt_append_command(pkt, inst);
>   }
>   EXPORT_SYMBOL(cmdq_pkt_jump);
> diff --git a/include/linux/soc/mediatek/mtk-cmdq.h b/include/linux/soc/mediatek/mtk-cmdq.h
> index 6c42d817d368..6215191a328d 100644
> --- a/include/linux/soc/mediatek/mtk-cmdq.h
> +++ b/include/linux/soc/mediatek/mtk-cmdq.h
> @@ -238,10 +238,12 @@ int cmdq_pkt_assign(struct cmdq_pkt *pkt, u16 reg_idx, u32 value);
>    *		     a physical address which should contains more instruction.
>    * @pkt:        the CMDQ packet
>    * @addr:       physical address of target instruction buffer
> + * @shift_pa:	shift bits of physical address in CMDQ instruction. This value
> + *		is got by cmdq_get_shift_pa().
>    *
>    * Return: 0 for success; else the error code is returned
>    */
> -int cmdq_pkt_jump(struct cmdq_pkt *pkt, dma_addr_t addr);
> +int cmdq_pkt_jump(struct cmdq_pkt *pkt, dma_addr_t addr, u8 shift_pa);
>   
>   /**
>    * cmdq_pkt_finalize() - Append EOC and jump command to pkt.
> @@ -339,7 +341,7 @@ static inline int cmdq_pkt_assign(struct cmdq_pkt *pkt, u16 reg_idx, u32 value)
>   	return -EINVAL;
>   }
>   
> -static inline int cmdq_pkt_jump(struct cmdq_pkt *pkt, dma_addr_t addr)
> +static inline int cmdq_pkt_jump(struct cmdq_pkt *pkt, dma_addr_t addr, u8 shift_pa)
>   {
>   	return -EINVAL;
>   }




Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ