Re: [RFC PATCH 2/3] x86, MCE: Avoid potential deadlock in MCE context

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Mon, Jul 21, 2014 at 05:14:06PM +0000, Luck, Tony wrote:
> We've evolved a bunch of mechanisms:
> 
> 1) mce_ring: to pass pfn for AO errors from MCE context to a work thread
> 2) mce_info: to pass pfn for AR errors from MCE context to same process running in process context
> 3) mce_log: to pass entire "mce" structures from any context (MCE, CMCI, or init-time) to /dev/mcelog
> 
> I was actually wondering about going in the other direction. Make the
> /dev/mcelog code register a notifier on x86_mce_decoder_chain (and
> perhaps move all the /dev/mcelog functions out of mce.c into an actual
> driver file).  Then use Chen Gong's NMI safe code to just unconditionally
> make safe copies of anything that gets passed to mce_log() and run all
> the notifiers from his do_mce_irqwork().
> 
> -Tony

OK, I can cook some patches based on Tony's suggestion:
patch 1: add a generic lock-less memory pool to save error records
patch 2: remove mce_info (Tony has done a draft)
patch 3: remove mce_ring
patch 4: remove mce log buffer
patch 5: move all mce log related logic into a separate file lke mcelog.c
         under the same directory with mce.c

Attachment: signature.asc
Description: Digital signature


[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux]     [Linux OMAP]     [Linux MIPS]     [ECOS]     [Asterisk Internet PBX]     [Linux API]