* [PATCH 3/4] 44x/fsp2: tvsense workaround for dd1
@ 2017-11-02 13:07 Ivan Mikhaylov
2017-11-27 5:50 ` Alistair Popple
0 siblings, 1 reply; 2+ messages in thread
From: Ivan Mikhaylov @ 2017-11-02 13:07 UTC (permalink / raw)
To: Alistair Popple, Matt Porter
Cc: linuxppc-dev, linux-kernel, Benjamin Herrenschmidt, Joel Stanley,
Paul Mackerras, Michael Ellerman
* tvsense(temperature and voltage sensors) may provide
erratic sense values which may result in parity errors
on CMU.
Signed-off-by: Ivan Mikhaylov <ivan@de.ibm.com>
---
arch/powerpc/platforms/44x/fsp2.c | 17 +++++++++++++++++
1 files changed, 17 insertions(+), 0 deletions(-)
diff --git a/arch/powerpc/platforms/44x/fsp2.c b/arch/powerpc/platforms/44x/fsp2.c
index 9585725..4e12490 100644
--- a/arch/powerpc/platforms/44x/fsp2.c
+++ b/arch/powerpc/platforms/44x/fsp2.c
@@ -59,6 +59,23 @@ static int __init fsp2_probe(void)
mtdcr(DCRN_PLB6_HD, 0xffff0000);
mtdcr(DCRN_PLB6_SHD, 0xffff0000);
+ /* TVSENSE reset is blocked (clock gated) by the POR default of the TVS
+ * sleep config bit. As a consequence, TVSENSE will provide erratic
+ * sensor values, which may result in spurious (parity) errors
+ * recorded in the CMU FIR and leading to erroneous interrupt requests
+ * once the CMU interrupt is unmasked.
+ */
+
+ /* 1. set TVS1[UNDOZE] */
+ val = mfcmu(CMUN_TVS1);
+ val |= 0x4;
+ mtcmu(CMUN_TVS1, val);
+
+ /* 2. clear FIR[TVS] and FIR[TVSPAR] */
+ val = mfcmu(CMUN_FIR0);
+ val |= 0x30000000;
+ mtcmu(CMUN_FIR0, val);
+
/* L2 machine checks */
mtl2(L2PLBMCKEN0, 0xffffffff);
mtl2(L2PLBMCKEN1, 0x0000ffff);
--
1.7.1
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH 3/4] 44x/fsp2: tvsense workaround for dd1
2017-11-02 13:07 [PATCH 3/4] 44x/fsp2: tvsense workaround for dd1 Ivan Mikhaylov
@ 2017-11-27 5:50 ` Alistair Popple
0 siblings, 0 replies; 2+ messages in thread
From: Alistair Popple @ 2017-11-27 5:50 UTC (permalink / raw)
To: Ivan Mikhaylov
Cc: Matt Porter, linuxppc-dev, linux-kernel, Benjamin Herrenschmidt,
Joel Stanley, Paul Mackerras, Michael Ellerman
Again, I'm not familiar with this specific hardware so can only provide general
comments.
On Thursday, 2 November 2017 4:07:05 PM AEDT Ivan Mikhaylov wrote:
> * tvsense(temperature and voltage sensors) may provide
> erratic sense values which may result in parity errors
> on CMU.
It would be good if you could provide a more detailed description of the problem
(like in the comment below) and what this patch does to address it.
Feel free to add my reviewed by as well.
- Alistair
> Signed-off-by: Ivan Mikhaylov <ivan@de.ibm.com>
> ---
> arch/powerpc/platforms/44x/fsp2.c | 17 +++++++++++++++++
> 1 files changed, 17 insertions(+), 0 deletions(-)
>
> diff --git a/arch/powerpc/platforms/44x/fsp2.c b/arch/powerpc/platforms/44x/fsp2.c
> index 9585725..4e12490 100644
> --- a/arch/powerpc/platforms/44x/fsp2.c
> +++ b/arch/powerpc/platforms/44x/fsp2.c
> @@ -59,6 +59,23 @@ static int __init fsp2_probe(void)
> mtdcr(DCRN_PLB6_HD, 0xffff0000);
> mtdcr(DCRN_PLB6_SHD, 0xffff0000);
>
> + /* TVSENSE reset is blocked (clock gated) by the POR default of the TVS
> + * sleep config bit. As a consequence, TVSENSE will provide erratic
> + * sensor values, which may result in spurious (parity) errors
> + * recorded in the CMU FIR and leading to erroneous interrupt requests
> + * once the CMU interrupt is unmasked.
> + */
> +
> + /* 1. set TVS1[UNDOZE] */
> + val = mfcmu(CMUN_TVS1);
> + val |= 0x4;
> + mtcmu(CMUN_TVS1, val);
> +
> + /* 2. clear FIR[TVS] and FIR[TVSPAR] */
> + val = mfcmu(CMUN_FIR0);
> + val |= 0x30000000;
> + mtcmu(CMUN_FIR0, val);
> +
> /* L2 machine checks */
> mtl2(L2PLBMCKEN0, 0xffffffff);
> mtl2(L2PLBMCKEN1, 0x0000ffff);
>
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2017-11-27 6:24 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2017-11-02 13:07 [PATCH 3/4] 44x/fsp2: tvsense workaround for dd1 Ivan Mikhaylov
2017-11-27 5:50 ` Alistair Popple
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).