linux-kernel - Re: [PATCH v4 01/24] EDAC/amd64: Define Data Fabric operations

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: <YikfPWACGyWPIxKP@yaz-ubuntu>
Date:   Wed, 9 Mar 2022 21:42:21 +0000
From:   Yazen Ghannam <yazen.ghannam@....com>
To:     Borislav Petkov <bp@...en8.de>
Cc:     linux-edac@...r.kernel.org, linux-kernel@...r.kernel.org,
        mchehab@...nel.org, tony.luck@...el.com, james.morse@....com,
        rric@...nel.org, Smita.KoralahalliChannabasappa@....com
Subject: Re: [PATCH v4 01/24] EDAC/amd64: Define Data Fabric operations

On Fri, Feb 11, 2022 at 08:10:43PM +0100, Borislav Petkov wrote:
> > +/*
> > + * Glossary of acronyms used in address translation for Zen-based systems
> > + *
> > + * DF          =       Data Fabric
> 
> Yeah, "DF: Data Fabric" is probably easier to parse, without that weird
> spacing and equals sign.
> 

Okay, will change.

> > + */
> > +
> >  /* Protect the PCI config register pairs used for DF indirect access. */
> >  static DEFINE_MUTEX(df_indirect_mutex);
> >  
> > @@ -1058,6 +1064,14 @@ struct addr_ctx {
> >  	u8 inst_id;
> >  };
> >  
> > +struct data_fabric_ops {
> > +};
> 
> I know that this is not the only example but we have struct definitions
> interspersed with functions in the .c file while former should be all in
> the header. It is a lot cleaner to have definitions and inline functions
> in the header and the actual functionality in the C file but I leave it
> up to you to decide what you prefer.
>

Makes sense. I'll make the change.
 
> > +
> > +struct data_fabric_ops df2_ops = {
> > +};
> > +
> > +struct data_fabric_ops *df_ops;
> > +
> >  static int umc_normaddr_to_sysaddr(u64 norm_addr, u16 nid, u8 umc, u64 *sys_addr)
> >  {
> >  	u64 dram_base_addr, dram_limit_addr, dram_hole_base;
> > @@ -1072,6 +1086,11 @@ static int umc_normaddr_to_sysaddr(u64 norm_addr, u16 nid, u8 umc, u64 *sys_addr
> >  
> >  	struct addr_ctx ctx;
> >  
> > +	if (!df_ops) {
> > +		pr_debug("Data Fabric Operations not set");
> 
> That probably wants to be a WARN_ON_ONCE() so that it is loud and
> prominent when it happens...
>

I'd like to keep this a debug message because address translation updates will
almost certainly lag general EDAC enablement. For example, Family 19h Model
10h is enabled now, but df_ops won't be set. The translation code will return
early here, and the EDAC message will say that the tranlsation failed. But
this isn't a bug that needs a WARNing. Rather it's just that the feature isn't
enabled yet for new systems.

Thanks,
Yazen