Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 8BIT
Subject: bug in drivers/edac/mpc85xx_edac.c:mpc85xx_mc_check() ?
Date: Wed, 8 Apr 2009 14:57:42 -0700
Message-ID: <57AC2FA1761300418C7AB8F3EA493C9702C5F200@HQ-EXCH-5.corp.brocade.com>
Thread-Topic: bug in drivers/edac/mpc85xx_edac.c:mpc85xx_mc_check() ?
Thread-Index: Acm4lQp/520fvHxRSN2NDaRoPxya+Q==
From: "Jeff Haran" <jharan@Brocade.COM>
To: <linux-kernel@vger.kernel.org>
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1213
Lines: 42

Hi,

Recent versions of this function start off with:

static void mpc85xx_mc_check(struct mem_ctl_info *mci)
{
    struct mpc85xx_mc_pdata *pdata = mci->pvt_info;
    ...

    err_detect = in_be32(pdata->mc_vbase + MPC85XX_MC_ERR_DETECT);
    if (err_detect)
        return;

    ...
}

My reading of the Freescale 8548E Manual leads me to conclude that the
Memory Error Detect register (ERR_DETECT) will have various bits set if
the memory controller has detected an error since the last time it was
cleared. If no memory error has occurred, the register will contain 0.

Perhaps I am missing something very basic, but it seem to me that the
above "if" should be:

    if (!err_detect)
        return;

as the existing code would seem to read "if any errors have occurred,
ignore them", though perhaps testing has demonstrated that the Freescale
manual is in error.

Please include this email address in responses as I do not subscribe.

Thanks,

Jeff Haran
Brocade
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/