[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1248106385-27514-2-git-send-email-borislav.petkov@amd.com>
Date: Mon, 20 Jul 2009 18:12:52 +0200
From: Borislav Petkov <borislav.petkov@....com>
To: <mingo@...e.hu>, <hpa@...or.com>, <tglx@...utronix.de>,
<norsk5@...oo.com>, <aris@...hat.com>
CC: <linux-kernel@...r.kernel.org>, <x86@...nel.org>
Subject: [PATCH 01/14] amd64_edac: simplify error type bits extractors
Teach the error code macros to generate the error description strings
and use them in the error type decoders directly which removes a bunch
of code and makes the decoding functions much more readable. Also, fix
strings and shorten macro names.
Signed-off-by: Borislav Petkov <borislav.petkov@....com>
---
drivers/edac/amd64_edac.c | 72 ++++++++++------------------------
drivers/edac/amd64_edac.h | 28 +++++++-------
drivers/edac/amd64_edac_err_types.c | 55 ++++++++++++---------------
3 files changed, 59 insertions(+), 96 deletions(-)
diff --git a/drivers/edac/amd64_edac.c b/drivers/edac/amd64_edac.c
index 858fe60..93b119b 100644
--- a/drivers/edac/amd64_edac.c
+++ b/drivers/edac/amd64_edac.c
@@ -1099,8 +1099,8 @@ static void k8_map_sysaddr_to_csrow(struct mem_ctl_info *mci,
u32 page, offset;
/* Extract the syndrome parts and form a 16-bit syndrome */
- syndrome = EXTRACT_HIGH_SYNDROME(info->nbsl) << 8;
- syndrome |= EXTRACT_LOW_SYNDROME(info->nbsh);
+ syndrome = HIGH_SYNDROME(info->nbsl) << 8;
+ syndrome |= LOW_SYNDROME(info->nbsh);
/* CHIPKILL enabled */
if (info->nbcfg & K8_NBCFG_CHIPKILL) {
@@ -1699,8 +1699,8 @@ static void f10_map_sysaddr_to_csrow(struct mem_ctl_info *mci,
if (csrow >= 0) {
error_address_to_page_and_offset(sys_addr, &page, &offset);
- syndrome = EXTRACT_HIGH_SYNDROME(info->nbsl) << 8;
- syndrome |= EXTRACT_LOW_SYNDROME(info->nbsh);
+ syndrome = HIGH_SYNDROME(info->nbsl) << 8;
+ syndrome |= LOW_SYNDROME(info->nbsh);
/*
* Is CHIPKILL on? If so, then we can attempt to use the
@@ -2153,36 +2153,22 @@ static int amd64_get_error_info(struct mem_ctl_info *mci,
static inline void amd64_decode_gart_tlb_error(struct mem_ctl_info *mci,
struct amd64_error_info_regs *info)
{
- u32 err_code;
- u32 ec_tt; /* error code transaction type (2b) */
- u32 ec_ll; /* error code cache level (2b) */
-
- err_code = EXTRACT_ERROR_CODE(info->nbsl);
- ec_ll = EXTRACT_LL_CODE(err_code);
- ec_tt = EXTRACT_TT_CODE(err_code);
+ u32 ec = ERROR_CODE(info->nbsl);
amd64_mc_printk(mci, KERN_ERR,
"GART TLB event: transaction type(%s), "
- "cache level(%s)\n", tt_msgs[ec_tt], ll_msgs[ec_ll]);
+ "cache level(%s)\n", TT(ec), LL(ec));
}
static inline void amd64_decode_mem_cache_error(struct mem_ctl_info *mci,
struct amd64_error_info_regs *info)
{
- u32 err_code;
- u32 ec_rrrr; /* error code memory transaction (4b) */
- u32 ec_tt; /* error code transaction type (2b) */
- u32 ec_ll; /* error code cache level (2b) */
-
- err_code = EXTRACT_ERROR_CODE(info->nbsl);
- ec_ll = EXTRACT_LL_CODE(err_code);
- ec_tt = EXTRACT_TT_CODE(err_code);
- ec_rrrr = EXTRACT_RRRR_CODE(err_code);
+ u32 ec = ERROR_CODE(info->nbsl);
amd64_mc_printk(mci, KERN_ERR,
"cache hierarchy error: memory transaction type(%s), "
"transaction type(%s), cache level(%s)\n",
- rrrr_msgs[ec_rrrr], tt_msgs[ec_tt], ll_msgs[ec_ll]);
+ RRRR(ec), TT(ec), LL(ec));
}
@@ -2262,21 +2248,8 @@ static void amd64_handle_ue(struct mem_ctl_info *mci,
static void amd64_decode_bus_error(struct mem_ctl_info *mci,
struct amd64_error_info_regs *info)
{
- u32 err_code, ext_ec;
- u32 ec_pp; /* error code participating processor (2p) */
- u32 ec_to; /* error code timed out (1b) */
- u32 ec_rrrr; /* error code memory transaction (4b) */
- u32 ec_ii; /* error code memory or I/O (2b) */
- u32 ec_ll; /* error code cache level (2b) */
-
- ext_ec = EXTRACT_EXT_ERROR_CODE(info->nbsl);
- err_code = EXTRACT_ERROR_CODE(info->nbsl);
-
- ec_ll = EXTRACT_LL_CODE(err_code);
- ec_ii = EXTRACT_II_CODE(err_code);
- ec_rrrr = EXTRACT_RRRR_CODE(err_code);
- ec_to = EXTRACT_TO_CODE(err_code);
- ec_pp = EXTRACT_PP_CODE(err_code);
+ u32 ec = ERROR_CODE(info->nbsl);
+ u32 xec = EXT_ERROR_CODE(info->nbsl);
amd64_mc_printk(mci, KERN_ERR,
"BUS ERROR:\n"
@@ -2284,20 +2257,17 @@ static void amd64_decode_bus_error(struct mem_ctl_info *mci,
" participating processor(%s)\n"
" memory transaction type(%s)\n"
" cache level(%s) Error Found by: %s\n",
- to_msgs[ec_to],
- ii_msgs[ec_ii],
- pp_msgs[ec_pp],
- rrrr_msgs[ec_rrrr],
- ll_msgs[ec_ll],
+ TO(ec), II(ec), PP(ec), RRRR(ec), LL(ec),
(info->nbsh & K8_NBSH_ERR_SCRUBER) ?
"Scrubber" : "Normal Operation");
- /* If this was an 'observed' error, early out */
- if (ec_pp == K8_NBSL_PP_OBS)
- return; /* We aren't the node involved */
+
+ /* Bail early out if this was an 'observed' error */
+ if (((ec >> 9) & 0x3) == K8_NBSL_PP_OBS)
+ return;
/* Parse out the extended error code for ECC events */
- switch (ext_ec) {
+ switch (xec) {
/* F10 changed to one Extended ECC error code */
case F10_NBSL_EXT_ERR_RES: /* Reserved field */
case F10_NBSL_EXT_ERR_ECC: /* F10 ECC ext err code */
@@ -2377,7 +2347,7 @@ int amd64_process_error_info(struct mem_ctl_info *mci,
(regs->nbsh & K8_NBSH_CORE3) ? "True" : "False");
- err_code = EXTRACT_ERROR_CODE(regs->nbsl);
+ err_code = ERROR_CODE(regs->nbsl);
/* Determine which error type:
* 1) GART errors - non-fatal, developmental events
@@ -2385,7 +2355,7 @@ int amd64_process_error_info(struct mem_ctl_info *mci,
* 3) BUS errors
* 4) Unknown error
*/
- if (TEST_TLB_ERROR(err_code)) {
+ if (TLB_ERROR(err_code)) {
/*
* GART errors are intended to help graphics driver developers
* to detect bad GART PTEs. It is recommended by AMD to disable
@@ -2409,10 +2379,10 @@ int amd64_process_error_info(struct mem_ctl_info *mci,
debugf1("GART TLB error\n");
amd64_decode_gart_tlb_error(mci, info);
- } else if (TEST_MEM_ERROR(err_code)) {
+ } else if (MEM_ERROR(err_code)) {
debugf1("Memory/Cache error\n");
amd64_decode_mem_cache_error(mci, info);
- } else if (TEST_BUS_ERROR(err_code)) {
+ } else if (BUS_ERROR(err_code)) {
debugf1("Bus (Link/DRAM) error\n");
amd64_decode_bus_error(mci, info);
} else {
@@ -2422,7 +2392,7 @@ int amd64_process_error_info(struct mem_ctl_info *mci,
err_code);
}
- ext_ec = EXTRACT_EXT_ERROR_CODE(regs->nbsl);
+ ext_ec = EXT_ERROR_CODE(regs->nbsl);
amd64_mc_printk(mci, KERN_ERR,
"ExtErr=(0x%x) %s\n", ext_ec, ext_msgs[ext_ec]);
diff --git a/drivers/edac/amd64_edac.h b/drivers/edac/amd64_edac.h
index ba73015..e050a92 100644
--- a/drivers/edac/amd64_edac.h
+++ b/drivers/edac/amd64_edac.h
@@ -303,9 +303,6 @@ enum {
#define K8_NBSL 0x48
-#define EXTRACT_HIGH_SYNDROME(x) (((x) >> 24) & 0xff)
-#define EXTRACT_EXT_ERROR_CODE(x) (((x) >> 16) & 0x1f)
-
/* Family F10h: Normalized Extended Error Codes */
#define F10_NBSL_EXT_ERR_RES 0x0
#define F10_NBSL_EXT_ERR_CRC 0x1
@@ -348,16 +345,20 @@ enum {
#define K8_NBSL_EXT_ERR_CHIPKILL_ECC 0x8
#define K8_NBSL_EXT_ERR_DRAM_PARITY 0xD
-#define EXTRACT_ERROR_CODE(x) ((x) & 0xffff)
-#define TEST_TLB_ERROR(x) (((x) & 0xFFF0) == 0x0010)
-#define TEST_MEM_ERROR(x) (((x) & 0xFF00) == 0x0100)
-#define TEST_BUS_ERROR(x) (((x) & 0xF800) == 0x0800)
-#define EXTRACT_TT_CODE(x) (((x) >> 2) & 0x3)
-#define EXTRACT_II_CODE(x) (((x) >> 2) & 0x3)
-#define EXTRACT_LL_CODE(x) (((x) >> 0) & 0x3)
-#define EXTRACT_RRRR_CODE(x) (((x) >> 4) & 0xf)
-#define EXTRACT_TO_CODE(x) (((x) >> 8) & 0x1)
-#define EXTRACT_PP_CODE(x) (((x) >> 9) & 0x3)
+#define ERROR_CODE(x) ((x) & 0xffff)
+#define EXT_ERROR_CODE(x) (((x) >> 16) & 0x1f)
+#define LOW_SYNDROME(x) (((x) >> 15) & 0xff)
+#define HIGH_SYNDROME(x) (((x) >> 24) & 0xff)
+
+#define TLB_ERROR(x) (((x) & 0xFFF0) == 0x0010)
+#define MEM_ERROR(x) (((x) & 0xFF00) == 0x0100)
+#define BUS_ERROR(x) (((x) & 0xF800) == 0x0800)
+#define TT(x) tt_msgs[(((x) >> 2) & 0x3)]
+#define II(x) ii_msgs[(((x) >> 2) & 0x3)]
+#define LL(x) ll_msgs[(((x) >> 0) & 0x3)]
+#define RRRR(x) rrrr_msgs[(((x) >> 4) & 0xf)]
+#define TO(x) to_msgs[(((x) >> 8) & 0x1)]
+#define PP(x) pp_msgs[(((x) >> 9) & 0x3)]
/*
* The following are for BUS type errors AFTER values have been normalized by
@@ -388,7 +389,6 @@ enum {
#define EXTRACT_LDT_LINK(x) (((x) >> 4) & 0x7)
#define EXTRACT_ERR_CPU_MAP(x) ((x) & 0xF)
-#define EXTRACT_LOW_SYNDROME(x) (((x) >> 15) & 0xff)
#define K8_NBEAL 0x50
diff --git a/drivers/edac/amd64_edac_err_types.c b/drivers/edac/amd64_edac_err_types.c
index f212ff1..cacdd54 100644
--- a/drivers/edac/amd64_edac_err_types.c
+++ b/drivers/edac/amd64_edac_err_types.c
@@ -62,55 +62,48 @@ struct scrubrate scrubrates[] = {
* or MSR0000_0411.
*/
const char *tt_msgs[] = { /* transaction type */
- "instruction",
- "data",
- "generic",
- "reserved"
+ "Instruction",
+ "Data",
+ "Generic",
+ "Reserved"
};
const char *ll_msgs[] = { /* cache level */
- "L0",
+ "Reserved",
"L1",
"L2",
- "L3/generic"
+ "L3/Generic"
};
const char *rrrr_msgs[] = {
- "generic",
- "generic read",
- "generic write",
- "data read",
- "data write",
- "inst fetch",
- "prefetch",
- "evict",
- "snoop",
- "reserved RRRR= 9",
- "reserved RRRR= 10",
- "reserved RRRR= 11",
- "reserved RRRR= 12",
- "reserved RRRR= 13",
- "reserved RRRR= 14",
- "reserved RRRR= 15"
+ "Generic",
+ "Generic Read",
+ "Generic Write",
+ "Data Read",
+ "Data Write",
+ "Instruction Fetch",
+ "Prefetch",
+ "Evict",
+ "Snoop (Probe)"
};
const char *pp_msgs[] = { /* participating processor */
- "local node originated (SRC)",
- "local node responded to request (RES)",
- "local node observed as 3rd party (OBS)",
+ "local node originated the request",
+ "local node responded to request",
+ "local node observed error as 3rd party",
"generic"
};
const char *to_msgs[] = {
- "no timeout",
- "timed out"
+ "No timeout",
+ "Timed out"
};
const char *ii_msgs[] = { /* memory or i/o */
- "mem access",
- "reserved",
- "i/o access",
- "generic"
+ "Memory Access",
+ "Reserved",
+ "IO Access",
+ "Generic"
};
/* Map the 5 bits of Extended Error code to the string table. */
--
1.6.3.3
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists