netdev - Re: [PATCH v5 3/9] net: stmmac: Add Loongson DWGMAC definitions

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]

Date: Fri, 24 Nov 2023 19:44:03 +0300
From: Serge Semin <fancer.lancer@...il.com>
To: Andrew Lunn <andrew@...n.ch>
Cc: Yanteng Si <siyanteng@...ngson.cn>, hkallweit1@...il.com, 
	peppe.cavallaro@...com, alexandre.torgue@...s.st.com, joabreu@...opsys.com, 
	Jose.Abreu@...opsys.com, chenhuacai@...ngson.cn, linux@...linux.org.uk, 
	dongbiao@...ngson.cn, guyinggang@...ngson.cn, netdev@...r.kernel.org, 
	loongarch@...ts.linux.dev, chris.chenfeiyang@...il.com
Subject: Re: [PATCH v5 3/9] net: stmmac: Add Loongson DWGMAC definitions

On Fri, Nov 24, 2023 at 03:51:08PM +0100, Andrew Lunn wrote:
> > In general, we split one into two.
> > 
> > the details are as follows：
> > 
> > DMA_INTR_ENA_NIE = DMA_INTR_ENA_NIE_LOONGSON= DMA_INTR_ENA_TX_NIE +
> > DMA_INTR_ENA_RX_NIE
> 
> What does the documentation from Synopsys say about the bit you have
> used for DMA_INTR_ENA_NIE_LOONGSON? Is it marked as being usable by IP
> integrators for whatever they want, or is it marked as reserved?
> 
> I'm just wondering if we are heading towards a problem when the next
> version of this IP assigns the bit to mean something else.

That's what I started to figure out in my initial message:
Link: https://lore.kernel.org/netdev/gxods3yclaqkrud6jxhvcjm67vfp5zmuoxjlr6llcddny7zwsr@473g74uk36xn/
but Yanteng for some reason ignored all my comments.

Anyway AFAICS this Loongson GMAC has NIE and AIE flags defined differently:
DW GMAC: NIE - BIT(16) - all non-fatal Tx and Rx errors,
         AIE - BIT(15) - all fatal Tx, Rx and Bus errors.
Loongson GMAC: NIE - BIT(18) | BIT(17) - one flag for Tx and another for Rx errors.
               AIE - BIT(16) | BIT(15) - Tx, Rx and don't know about the Bus errors.
So the Loongson GMAC has not only NIE/AIE flags re-defined, but
also get to occupy two reserved in the generic DW GMAC bits: BIT(18) and BIT(17).

Moreover Yanteng in his patch defines DMA_INTR_NORMAL as a combination
of both _generic_ DW and Loongson-specific NIE flags and
DMA_INTR_ABNORMAL as a combination of both _generic_ DW and
Loongson-specific AIE flags. At the very least it doesn't look
correct, since _generic_ DW GMAC NIE flag BIT(16) is defined as a part
of the Loongson GMAC AIE flags set.

-Serge(y)

> 
> 	Andrew