netdev - Re: [net-next PATCH] octeontx2-pf: Add support for adaptive interrupt coalescing

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: <20220512165842.4f0ed0f8@kernel.org>
Date:   Thu, 12 May 2022 16:58:42 -0700
From:   Jakub Kicinski <kuba@...nel.org>
To:     Suman Ghosh <sumang@...vell.com>
Cc:     <davem@...emloft.net>, <edumazet@...gle.com>, <pabeni@...hat.com>,
        <sgoutham@...vell.com>, <sbhatta@...vell.com>,
        <gakula@...vell.com>, <Sunil.Goutham@...ium.com>,
        <hkelam@...vell.com>, <colin.king@...el.com>,
        <netdev@...r.kernel.org>
Subject: Re: [net-next PATCH] octeontx2-pf: Add support for adaptive
 interrupt coalescing

On Thu, 12 May 2022 12:49:12 +0530 Suman Ghosh wrote:
> Added support for adaptive IRQ coalescing. It uses net_dim
> algorithm to find the suitable delay/IRQ count based on the
> current packet rate.
> 
> Signed-off-by: Suman Ghosh <sumang@...vell.com>
> Reviewed-by: Sunil Kovvuri Goutham <sgoutham@...vell.com>

You still claim this is tested? How can it work if you're not changing
.supported_coalesce_params? Do we have a bug in the core?

> +	/* Check and update coalesce status */
> +	if ((pfvf->flags & OTX2_FLAG_ADPTV_INT_COAL_ENABLED) ==
> +			OTX2_FLAG_ADPTV_INT_COAL_ENABLED) {
> +		priv_coalesce_status = 1;
> +		if (!ec->use_adaptive_rx_coalesce || !ec->use_adaptive_tx_coalesce)
> +			pfvf->flags &= ~OTX2_FLAG_ADPTV_INT_COAL_ENABLED;
> +	} else {
> +		priv_coalesce_status = 0;
> +		if (ec->use_adaptive_rx_coalesce || ec->use_adaptive_tx_coalesce)
> +			pfvf->flags |= OTX2_FLAG_ADPTV_INT_COAL_ENABLED;
> +	}

Won't this flip-flop the flag? You should probably reject 
adaptive_rx != adaptive_tx.

>  		/* Re-enable interrupts */
>  		otx2_write64(pfvf, NIX_LF_CINTX_ENA_W1S(cq_poll->cint_idx),
> -			     BIT_ULL(0));
> +				BIT_ULL(0));

Unrelated and possibly misalinged?