linux-kernel - Re: [PATCH] [v2] dmaengine: dw-edma: reduce stack usage after debugfs rework

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: <20230131064237.GA5548@thinkpad>
Date:   Tue, 31 Jan 2023 12:12:37 +0530
From:   Manivannan Sadhasivam <mani@...nel.org>
To:     Arnd Bergmann <arnd@...nel.org>
Cc:     Gustavo Pimentel <gustavo.pimentel@...opsys.com>,
        Vinod Koul <vkoul@...nel.org>,
        Serge Semin <Sergey.Semin@...kalelectronics.ru>,
        Arnd Bergmann <arnd@...db.de>,
        Bjorn Helgaas <bhelgaas@...gle.com>,
        Lorenzo Pieralisi <lpieralisi@...nel.org>,
        Frank Li <Frank.Li@....com>, dmaengine@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] [v2] dmaengine: dw-edma: reduce stack usage after
 debugfs rework

On Mon, Jan 30, 2023 at 07:50:42PM +0100, Arnd Bergmann wrote:
> From: Arnd Bergmann <arnd@...db.de>
> 
> After the dw_edma_debugfs_entry arrays are no longer compile-time

This sentence doesn't look complete to me.

> constant, they take up space on the stack, which exceeds the warning
> limit after inlining:
> 
> drivers/dma/dw-edma/dw-edma-v0-debugfs.c:280:6: error: stack frame size (1784) exceeds limit (1400) in 'dw_edma_v0_debugfs_on' [-Werror,-Wframe-larger-than]
> void dw_edma_v0_debugfs_on(struct dw_edma *dw)
> 
> Work around this by preventing dw_edma_debugfs_regs_{wr,rd} from both
> being inlined together, which cuts the stack frame size in half and
> makes it fit below the warning limit.
> 
> Fixes: 5c0373eafd83 ("dmaengine: dw-edma: Move eDMA data pointer to debugfs node descriptor")
> Signed-off-by: Arnd Bergmann <arnd@...db.de>

Small style comment below, with that:

Reviewed-by: Manivannan Sadhasivam <mani@...nel.org>

> ---
> v2: rebase on top of dmaengine tree
> ---
>  drivers/dma/dw-edma/dw-edma-v0-debugfs.c | 6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/dma/dw-edma/dw-edma-v0-debugfs.c b/drivers/dma/dw-edma/dw-edma-v0-debugfs.c
> index 42380bf64a70..6542060bd01a 100644
> --- a/drivers/dma/dw-edma/dw-edma-v0-debugfs.c
> +++ b/drivers/dma/dw-edma/dw-edma-v0-debugfs.c
> @@ -136,7 +136,8 @@ static void dw_edma_debugfs_regs_ch(struct dw_edma *dw, enum dw_edma_dir dir,
>  	dw_edma_debugfs_create_x32(dw, debugfs_regs, nr_entries, dent);
>  }
>  
> -static void dw_edma_debugfs_regs_wr(struct dw_edma *dw, struct dentry *dent)
> +static noinline_for_stack void
> +dw_edma_debugfs_regs_wr(struct dw_edma *dw, struct dentry *dent)

For consistency, the function arguments could be wrapped:

static noinline_for_stack void dw_edma_debugfs_regs_wr(struct dw_edma *dw,
						       struct dentry *dent)

I could see that with the addition of "noinline_for_stack", the arguments are
shifted towards the end of the line. But this style matches rest of the
functions in this file.

Thanks,
Mani

>  {
>  	const struct dw_edma_debugfs_entry debugfs_regs[] = {
>  		/* eDMA global registers */
> @@ -197,7 +198,8 @@ static void dw_edma_debugfs_regs_wr(struct dw_edma *dw, struct dentry *dent)
>  	}
>  }
>  
> -static void dw_edma_debugfs_regs_rd(struct dw_edma *dw, struct dentry *dent)
> +static noinline_for_stack void
> +dw_edma_debugfs_regs_rd(struct dw_edma *dw, struct dentry *dent)
>  {
>  	const struct dw_edma_debugfs_entry debugfs_regs[] = {
>  		/* eDMA global registers */
> -- 
> 2.39.0
> 

-- 
மணிவண்ணன் சதாசிவம்