Re: [PATCHv2] EDAC, altera: Move Stratix10 SDRAM ECC to peripheral

From: Thor Thayer
Date: Thu Jul 25 2019 - 19:27:03 EST

Hi James,

On 7/25/19 7:46 AM, James Morse wrote:
Hi Thor,

On 12/07/2019 19:28, thor.thayer@xxxxxxxxxxxxxxx wrote:
From: Thor Thayer <thor.thayer@xxxxxxxxxxxxxxx>

ARM32 SoCFPGAs had separate IRQs for SDRAM. ARM64 SoCFPGAs
send all DBEs to SError so filtering by source is necessary.

The Stratix10 SDRAM ECC is a better match with the generic
Altera peripheral ECC framework because the linked list can
be searched to find the ECC block offset and printout
the DBE Address.

diff --git a/drivers/edac/altera_edac.c b/drivers/edac/altera_edac.c
index c2e693e34d43..09a80b53acea 100644
--- a/drivers/edac/altera_edac.c
+++ b/drivers/edac/altera_edac.c

@@ -2231,13 +2275,15 @@ static int altr_edac_a10_probe(struct platform_device *pdev)
of_device_is_compatible(child, "altr,socfpga-dma-ecc") ||
of_device_is_compatible(child, "altr,socfpga-usb-ecc") ||
of_device_is_compatible(child, "altr,socfpga-qspi-ecc") ||
+ of_device_is_compatible(child, "altr,sdram-edac-s10") ||
of_device_is_compatible(child, "altr,socfpga-sdmmc-ecc"))

I'm just curious: This list looks suspiciously like the altr_edac_a10_device_of_match[]
list. Is there a reason it can't use of_match_device() here?

Good point. Yes, it does look very much like the match list. Although of_match_device() doesn't fit with the available parameters (device pointer), your question prompted me to look closer and I noticed of_match_node() is perfect.

I'll create a version 3 with this change.

Thanks for reviewing!

altr_edac_a10_device_add(edac, child);
- else if ((of_device_is_compatible(child, "altr,sdram-edac-a10")) ||
- (of_device_is_compatible(child, "altr,sdram-edac-s10")))
+ else if (of_device_is_compatible(child, "altr,sdram-edac-a10"))
NULL, &pdev->dev);

Acked-by: James Morse <james.morse@xxxxxxx>