答复: [????] Re: [PATCH] PCI: sysfs: Suppress FW_BUG warning when NUMA node already matches

["Li,Rongqing(ACG CCN)" <lirongqing@baidu.com>]

> 
> Hello,
> 
> > The numa_node sysfs interface allows users to manually override a PCI
> > device's NUMA node assignment. Currently, every write triggers a
> > FW_BUG warning and taints the kernel, even when writing the same value
> > that is already set.
> 
> So, this works as intended, then?
> 

I don't know if this is intended, but it feels unreasonable―writing the same value still triggers a FW_BUG warning.

> What makes multiple writes to this sysfs attribute, if you don't mind me
> asking?  Do you have some tool that does this?  Some automation?
> 

I have a tool that triggered this warning, and I will fix my tool.

> Especially, that you seem to be writing the same value over and over.
> 
Not , but a user is allowed to write to this file,  and subsequently will generate a massive amount of kernel logs―is this considered unreasonable? Should we add some rate limit?


> >  	if (node != NUMA_NO_NODE && !node_online(node))
> >  		return -EINVAL;
> >
> > -	add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
> > -	pci_alert(pdev, FW_BUG "Overriding NUMA node to %d.  Contact your
> vendor for updates.",
> > -		  node);
> > +	if (node != dev->numa_node) {
> > +		add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
> > +		pci_alert(pdev, FW_BUG "Overriding NUMA node to %d.  Contact
> your vendor for updates.",
> > +			  node);
> > +		dev->numa_node = node;
> > +	}
> 
> You could invert the check and make it an early return where you just return
> count.  Would save on the new indent level.
> 

Good idle, I will fix this in v2

Thanks

[Li,Rongqing] 

> Thank you!
> 
> 	Krzysztof