On Wed, Jan 25, 2017 at 12:04:04PM +0000, David Binderman wrote: > You'll have a very long wait to get a linux patch from me. > > I am happy for someone else to invent a patch.
Ah ok, I thought you wanted to give it a try and would want me to help you with it. :-) Anyway, here it is: --- From: Borislav Petkov <[email protected]> Date: Wed, 25 Jan 2017 16:08:27 +0100 Subject: [PATCH] EDAC, i7300: Test for the second channel properly REDMEMB[17] is the ECC_Locator bit, which, when set, identifies the CS[3:2] as the simbols in error. And thus the second channel. The macro computing it was wrong so get rid of it (it was used at one place only) and get rid of the conditional too. Generates better code this way anyway. Signed-off-by: Borislav Petkov <[email protected]> Reported-by: David Binderman <[email protected]> --- drivers/edac/i7300_edac.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/edac/i7300_edac.c b/drivers/edac/i7300_edac.c index 0a912bf6de00..e391f5a716be 100644 --- a/drivers/edac/i7300_edac.c +++ b/drivers/edac/i7300_edac.c @@ -304,7 +304,6 @@ static const char *ferr_global_lo_name[] = { #define REDMEMA 0xdc #define REDMEMB 0x7c - #define IS_SECOND_CH(v) ((v) * (1 << 17)) #define RECMEMA 0xe0 #define RECMEMA_BANK(v) (((v) >> 12) & 7) @@ -483,8 +482,9 @@ static void i7300_process_fbd_error(struct mem_ctl_info *mci) pci_read_config_dword(pvt->pci_dev_16_1_fsb_addr_map, REDMEMB, &value); channel = (branch << 1); - if (IS_SECOND_CH(value)) - channel++; + + /* Second channel ? */ + channel += !!(value & BIT(17)); /* Clear the error bit */ pci_write_config_dword(pvt->pci_dev_16_1_fsb_addr_map, -- 2.11.0 -- Regards/Gruss, Boris. Good mailing practices for 400: avoid top-posting and trim the reply.
