netdev - Re: [net-next,V0.5,4/5] eth: fbnic: Remove retry support

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: <20260114033814.1033266-1-kuba@kernel.org>
Date: Tue, 13 Jan 2026 19:38:14 -0800
From: Jakub Kicinski <kuba@...nel.org>
To: mohsin.bashr@...il.com
Cc: Jakub Kicinski <kuba@...nel.org>,
	netdev@...r.kernel.org,
	jacob.e.keller@...el.com,
	davem@...emloft.net,
	sanman.p211993@...il.com,
	pabeni@...hat.com,
	andrew+netdev@...n.ch,
	horms@...nel.org,
	alexanderduyck@...com,
	kernel-team@...a.com,
	edumazet@...gle.com,
	lee@...ger.us
Subject: Re: [net-next,V0.5,4/5] eth: fbnic: Remove retry support

This is an AI-generated review of your patch. The human sending this
email has considered the AI review valid, or at least pausible.

For local reproduction steps see https://netdev-ai.bots.linux.dev/ai-local.html
---
    eth: fbnic: Update RX mbox timeout value

This commit unifies timeout values for firmware mailbox completion waits
by introducing a wrapper function fbnic_mbx_wait_for_cmpl(). However,
there appear to be two regressions where the condition logic was
accidentally inverted.

> diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_fw.h b/drivers/net/ethernet/meta/fbnic/fbnic_fw.h
> index 1ecd777aaada..6b3fb163d381 100644
> --- a/drivers/net/ethernet/meta/fbnic/fbnic_fw.h
> +++ b/drivers/net/ethernet/meta/fbnic/fbnic_fw.h

[ ... ]

> +static inline bool fbnic_mbx_wait_for_cmpl(struct fbnic_fw_completion *cmpl)
> +{
> +	return wait_for_completion_timeout(&cmpl->done,
> +					   FBNIC_MBX_RX_TO_SEC * HZ);
> +}

The wrapper returns the same value as wait_for_completion_timeout():
non-zero on success, zero on timeout.

> diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_devlink.c b/drivers/net/ethernet/meta/fbnic/fbnic_devlink.c
> index b62b1d5b1453..193f554717b3 100644
> --- a/drivers/net/ethernet/meta/fbnic/fbnic_devlink.c
> +++ b/drivers/net/ethernet/meta/fbnic/fbnic_devlink.c
> @@ -178,7 +178,7 @@ fbnic_flash_start(struct fbnic_dev *fbd, struct pldmfw_component *component)
>  		goto cmpl_free;
>
>  	/* Wait for firmware to ack firmware upgrade start */
> -	if (wait_for_completion_timeout(&cmpl->done, 10 * HZ))
> +	if (!fbnic_mbx_wait_for_cmpl(cmpl))
>  		err = cmpl->result;
>  	else
>  		err = -ETIMEDOUT;

Was the negation here intentional? The original code sets err to
cmpl->result when the completion succeeds (wait returns non-zero). With
the added negation, it now sets err to cmpl->result when the wait times
out (returns zero), and sets err to -ETIMEDOUT when the wait succeeds.
This appears to invert the success and timeout handling.

[ ... ]

> @@ -447,7 +447,7 @@ static int fbnic_fw_reporter_dump(struct devlink_health_reporter *reporter,
>  				goto cmpl_cleanup;
>  		}
>
> -		if (wait_for_completion_timeout(&fw_cmpl->done, 2 * HZ)) {
> +		if (!fbnic_mbx_wait_for_cmpl(fw_cmpl)) {
>  			reinit_completion(&fw_cmpl->done);
>  		} else {
>  			NL_SET_ERR_MSG_FMT_MOD(extack,
>  					       "Timed out waiting on core dump (%d/%d)",

Similarly here, the original code calls reinit_completion() when the wait
succeeds (to prepare for the next iteration), and reports "Timed out" in
the else branch when it fails. With the negation added, reinit_completion()
is now called on timeout, and the "Timed out" error message is reported
when the wait actually succeeds. Should the negation be removed from this
call site as well?
-- 
pw-bot: cr