[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <5C4C569E8A4B9B42A84A977CF070A35B2C57B4FA66@USINDEVS01.corp.hds.com>
Date: Fri, 18 Nov 2011 18:04:40 -0500
From: Seiji Aguchi <seiji.aguchi@....com>
To: Huang Ying <ying.huang@...el.com>
CC: "linux-acpi@...r.kernel.org" <linux-acpi@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"lenb@...nel.org" <lenb@...nel.org>,
"Luck, Tony" <tony.luck@...el.com>,
"Chen, Gong" <gong.chen@...el.com>,
Matthew Garrett <mjg@...hat.com>,
"dle-develop@...ts.sourceforge.net"
<dle-develop@...ts.sourceforge.net>,
Satoru Moriya <satoru.moriya@....com>
Subject: RE: [RFC][PATCH]ACPI,APEI,ERST, back end driver for NVRAM
Hi,
>According to ACPI 4.0a spec:
>
>"""
>17.5.2.4.1 Error Log Address Range Resides in NVRAM
>
Thank you for giving me the information.
Let me clarify one thing.
If the busy bit can be cleared immediately, we don't need to
check busy bit with CHECK_BUSY_STATUS.
So, we should simply execute OSPM operations as follows.
- Writing
1. BEGIN_WRITE
2. SET_RECORD_OFFSET
3. EXECUTE_OPERATION
4. END
- Reading
1. BEGIN_READ
2. SET_RECORD_OFFSET
3. SET_RECORD_ID
4. EXECUTE_OPERATION
(END operation is not needed because OSPM requires no platform
support to read.)
- Clearing
1. BEGIN_CLEAR
2. SET_RECORD_ID
3. EXECUTE_OPERATION
(END operation is not needed because OSPM requires no platform
support to clear.)
Is this what you expected?
Seiji
Powered by blists - more mailing lists