linux-kernel - Re: [PATCH] mm, memory hotplug: print more failure information for online

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:	Thu, 25 Feb 2016 09:07:26 +0800
From:	Chen Yucong <slaoub@...il.com>
To:	David Rientjes <rientjes@...gle.com>
Cc:	akpm@...ux-foundation.org, vbabka@...e.cz, linux-mm@...ck.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] mm, memory hotplug: print more failure information for
 online_pages

On Wed, 2016-02-24 at 13:33 -0800, David Rientjes wrote:
> On Wed, 24 Feb 2016, Chen Yucong wrote:
> 
> > diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> > index c832ef3..e4b6dec3 100644
> > --- a/mm/memory_hotplug.c
> > +++ b/mm/memory_hotplug.c
> > @@ -1059,10 +1059,9 @@ int __ref online_pages(unsigned long pfn, unsigned long nr_pages, int online_typ
> >  
> >  	ret = memory_notify(MEM_GOING_ONLINE, &arg);
> >  	ret = notifier_to_errno(ret);
> > -	if (ret) {
> > -		memory_notify(MEM_CANCEL_ONLINE, &arg);
> > -		return ret;
> > -	}
> > +	if (ret)
> > +		goto failed_addition;
> > +
> >  	/*
> >  	 * If this zone is not populated, then it is not in zonelist.
> >  	 * This means the page allocator ignores this zone.
> > @@ -1080,12 +1079,7 @@ int __ref online_pages(unsigned long pfn, unsigned long nr_pages, int online_typ
> >  		if (need_zonelists_rebuild)
> >  			zone_pcp_reset(zone);
> >  		mutex_unlock(&zonelists_mutex);
> > -		printk(KERN_DEBUG "online_pages [mem %#010llx-%#010llx] failed\n",
> > -		       (unsigned long long) pfn << PAGE_SHIFT,
> > -		       (((unsigned long long) pfn + nr_pages)
> > -			    << PAGE_SHIFT) - 1);
> > -		memory_notify(MEM_CANCEL_ONLINE, &arg);
> > -		return ret;
> > +		goto failed_addition;
> >  	}
> >  
> >  	zone->present_pages += onlined_pages;
> > @@ -1118,6 +1112,13 @@ int __ref online_pages(unsigned long pfn, unsigned long nr_pages, int online_typ
> >  	if (onlined_pages)
> >  		memory_notify(MEM_ONLINE, &arg);
> >  	return 0;
> > +
> > +failed_addition:
> > +	pr_info("online_pages [mem %#010llx-%#010llx] failed\n",
> > +		(unsigned long long) pfn << PAGE_SHIFT,
> > +		(((unsigned long long) pfn + nr_pages) << PAGE_SHIFT) - 1);
> > +	memory_notify(MEM_CANCEL_ONLINE, &arg);
> > +	return ret;
> >  }
> >  #endif /* CONFIG_MEMORY_HOTPLUG_SPARSE */
> >  
> 
> Please explain how the conversion from KERN_DEBUG to KERN_INFO level is 
> better?

Like __offline_pages(), printk() in online_pages() is used for reporting
an failed addition rather than debug information.
Another reason is that pr_debug() is not an exact equivalent of 
printk(KERN_DEBUG ...)

/* If you are writing a driver, please use dev_dbg instead */
#if defined(CONFIG_DYNAMIC_DEBUG)
/* dynamic_pr_debug() uses pr_fmt() internally so we don't need it here
*/
#define pr_debug(fmt, ...) \
        dynamic_pr_debug(fmt, ##__VA_ARGS__)
#elif defined(DEBUG)
#define pr_debug(fmt, ...) \
        printk(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__)
#else
#define pr_debug(fmt, ...) \
        no_printk(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__)
#endif
 

> If the onlining returns an error value, which it will, why do we need to 
> leave an artifact behind in the kernel log that it failed?

In __offline_pages(), we can find the following snippet:

...
        ret = memory_notify(MEM_GOING_OFFLINE, &arg);
        ret = notifier_to_errno(ret);
        if (ret)
                goto failed_removal;
...
        offlined_pages = check_pages_isolated(start_pfn, end_pfn);
        if (offlined_pages < 0) {
                ret = -EBUSY;
                goto failed_removal;
        }
...
failed_removal:
        printk(KERN_INFO "memory offlining [mem %#010llx-%#010llx] 
...

Similarly, there's no single cause for failed online_pages operation.
So if memory_notify(MEM_GOING_ONLINE, &arg) returns an error
value, the result of online_pages is also ""online_pages [mem %#010llx-%
#010llx] failed\n".

thx!
    cyc