* [PATCH v3 01/10] EDAC/mc: Split edac_mc_alloc() into smaller functions
2020-01-23 9:02 [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Robert Richter
@ 2020-01-23 9:02 ` Robert Richter
2020-01-23 9:02 ` [PATCH v3 02/10] EDAC/mc: Reorder functions edac_mc_alloc*() Robert Richter
` (9 subsequent siblings)
10 siblings, 0 replies; 17+ messages in thread
From: Robert Richter @ 2020-01-23 9:02 UTC (permalink / raw)
To: Borislav Petkov, Mauro Carvalho Chehab, Tony Luck
Cc: James Morse, Aristeu Rozanski, Robert Richter, linux-edac,
linux-kernel, Mauro Carvalho Chehab
edac_mc_alloc() is huge. Factor out code by moving it to the two new
functions edac_mc_alloc_csrows() and edac_mc_alloc_dimms(). Do not
move code yet for better review.
Signed-off-by: Robert Richter <rrichter@marvell.com>
Reviewed-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
Acked-by: Aristeu Rozanski <aris@redhat.com>
---
drivers/edac/edac_mc.c | 105 +++++++++++++++++++++++++++--------------
1 file changed, 70 insertions(+), 35 deletions(-)
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index 058efcd9032e..d71006a4f07b 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -311,6 +311,9 @@ static void mci_release(struct device *dev)
kfree(mci);
}
+static int edac_mc_alloc_csrows(struct mem_ctl_info *mci);
+static int edac_mc_alloc_dimms(struct mem_ctl_info *mci);
+
struct mem_ctl_info *edac_mc_alloc(unsigned int mc_num,
unsigned int n_layers,
struct edac_mc_layer *layers,
@@ -318,15 +321,11 @@ struct mem_ctl_info *edac_mc_alloc(unsigned int mc_num,
{
struct mem_ctl_info *mci;
struct edac_mc_layer *layer;
- struct csrow_info *csr;
- struct rank_info *chan;
- struct dimm_info *dimm;
u32 *ce_per_layer[EDAC_MAX_LAYERS], *ue_per_layer[EDAC_MAX_LAYERS];
- unsigned int pos[EDAC_MAX_LAYERS];
unsigned int idx, size, tot_dimms = 1, count = 1;
unsigned int tot_csrows = 1, tot_channels = 1, tot_errcount = 0;
- void *pvt, *p, *ptr = NULL;
- int i, j, row, chn, n, len;
+ void *pvt, *ptr = NULL;
+ int i;
bool per_rank = false;
if (WARN_ON(n_layers > EDAC_MAX_LAYERS || n_layers == 0))
@@ -401,16 +400,43 @@ struct mem_ctl_info *edac_mc_alloc(unsigned int mc_num,
mci->num_cschannel = tot_channels;
mci->csbased = per_rank;
+ if (edac_mc_alloc_csrows(mci))
+ goto error;
+
+ if (edac_mc_alloc_dimms(mci))
+ goto error;
+
+ mci->op_state = OP_ALLOC;
+
+ return mci;
+
+error:
+ _edac_mc_free(mci);
+
+ return NULL;
+}
+EXPORT_SYMBOL_GPL(edac_mc_alloc);
+
+static int edac_mc_alloc_csrows(struct mem_ctl_info *mci)
+{
+ unsigned int tot_csrows = mci->nr_csrows;
+ unsigned int tot_channels = mci->num_cschannel;
+ unsigned int row, chn;
+
/*
* Alocate and fill the csrow/channels structs
*/
mci->csrows = kcalloc(tot_csrows, sizeof(*mci->csrows), GFP_KERNEL);
if (!mci->csrows)
- goto error;
+ return -ENOMEM;
+
for (row = 0; row < tot_csrows; row++) {
+ struct csrow_info *csr;
+
csr = kzalloc(sizeof(**mci->csrows), GFP_KERNEL);
if (!csr)
- goto error;
+ return -ENOMEM;
+
mci->csrows[row] = csr;
csr->csrow_idx = row;
csr->mci = mci;
@@ -418,34 +444,51 @@ struct mem_ctl_info *edac_mc_alloc(unsigned int mc_num,
csr->channels = kcalloc(tot_channels, sizeof(*csr->channels),
GFP_KERNEL);
if (!csr->channels)
- goto error;
+ return -ENOMEM;
for (chn = 0; chn < tot_channels; chn++) {
+ struct rank_info *chan;
+
chan = kzalloc(sizeof(**csr->channels), GFP_KERNEL);
if (!chan)
- goto error;
+ return -ENOMEM;
+
csr->channels[chn] = chan;
chan->chan_idx = chn;
chan->csrow = csr;
}
}
+ return 0;
+}
+
+static int edac_mc_alloc_dimms(struct mem_ctl_info *mci)
+{
+ void *p;
+ unsigned int pos[EDAC_MAX_LAYERS];
+ unsigned int row, chn, idx;
+ int layer;
+
/*
* Allocate and fill the dimm structs
*/
- mci->dimms = kcalloc(tot_dimms, sizeof(*mci->dimms), GFP_KERNEL);
+ mci->dimms = kcalloc(mci->tot_dimms, sizeof(*mci->dimms), GFP_KERNEL);
if (!mci->dimms)
- goto error;
+ return -ENOMEM;
memset(&pos, 0, sizeof(pos));
row = 0;
chn = 0;
- for (idx = 0; idx < tot_dimms; idx++) {
+ for (idx = 0; idx < mci->tot_dimms; idx++) {
+ struct dimm_info *dimm;
+ struct rank_info *chan;
+ int n, len;
+
chan = mci->csrows[row]->channels[chn];
dimm = kzalloc(sizeof(**mci->dimms), GFP_KERNEL);
if (!dimm)
- goto error;
+ return -ENOMEM;
mci->dimms[idx] = dimm;
dimm->mci = mci;
dimm->idx = idx;
@@ -455,16 +498,16 @@ struct mem_ctl_info *edac_mc_alloc(unsigned int mc_num,
*/
len = sizeof(dimm->label);
p = dimm->label;
- n = snprintf(p, len, "mc#%u", mc_num);
+ n = snprintf(p, len, "mc#%u", mci->mc_idx);
p += n;
len -= n;
- for (j = 0; j < n_layers; j++) {
+ for (layer = 0; layer < mci->n_layers; layer++) {
n = snprintf(p, len, "%s#%u",
- edac_layer_name[layers[j].type],
- pos[j]);
+ edac_layer_name[mci->layers[layer].type],
+ pos[layer]);
p += n;
len -= n;
- dimm->location[j] = pos[j];
+ dimm->location[layer] = pos[layer];
if (len <= 0)
break;
@@ -476,39 +519,31 @@ struct mem_ctl_info *edac_mc_alloc(unsigned int mc_num,
dimm->cschannel = chn;
/* Increment csrow location */
- if (layers[0].is_virt_csrow) {
+ if (mci->layers[0].is_virt_csrow) {
chn++;
- if (chn == tot_channels) {
+ if (chn == mci->num_cschannel) {
chn = 0;
row++;
}
} else {
row++;
- if (row == tot_csrows) {
+ if (row == mci->nr_csrows) {
row = 0;
chn++;
}
}
/* Increment dimm location */
- for (j = n_layers - 1; j >= 0; j--) {
- pos[j]++;
- if (pos[j] < layers[j].size)
+ for (layer = mci->n_layers - 1; layer >= 0; layer--) {
+ pos[layer]++;
+ if (pos[layer] < mci->layers[layer].size)
break;
- pos[j] = 0;
+ pos[layer] = 0;
}
}
- mci->op_state = OP_ALLOC;
-
- return mci;
-
-error:
- _edac_mc_free(mci);
-
- return NULL;
+ return 0;
}
-EXPORT_SYMBOL_GPL(edac_mc_alloc);
void edac_mc_free(struct mem_ctl_info *mci)
{
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH v3 02/10] EDAC/mc: Reorder functions edac_mc_alloc*()
2020-01-23 9:02 [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Robert Richter
2020-01-23 9:02 ` [PATCH v3 01/10] EDAC/mc: Split edac_mc_alloc() into smaller functions Robert Richter
@ 2020-01-23 9:02 ` Robert Richter
2020-01-23 19:02 ` Matthias Brugger
2020-01-23 9:02 ` [PATCH v3 03/10] EDAC: Store error type in struct edac_raw_error_desc Robert Richter
` (8 subsequent siblings)
10 siblings, 1 reply; 17+ messages in thread
From: Robert Richter @ 2020-01-23 9:02 UTC (permalink / raw)
To: Borislav Petkov, Mauro Carvalho Chehab, Tony Luck
Cc: James Morse, Aristeu Rozanski, Robert Richter, linux-edac,
linux-kernel, Mauro Carvalho Chehab
Reorder the new created functions edac_mc_alloc_csrows() and
edac_mc_alloc_dimms() and move them before edac_mc_alloc(). No further
code changes.
Signed-off-by: Robert Richter <rrichter@marvell.com>
Reviewed-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
Acked-by: Aristeu Rozanski <aris@redhat.com>
---
drivers/edac/edac_mc.c | 209 ++++++++++++++++++++---------------------
1 file changed, 103 insertions(+), 106 deletions(-)
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index d71006a4f07b..1e227e69e216 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -311,112 +311,6 @@ static void mci_release(struct device *dev)
kfree(mci);
}
-static int edac_mc_alloc_csrows(struct mem_ctl_info *mci);
-static int edac_mc_alloc_dimms(struct mem_ctl_info *mci);
-
-struct mem_ctl_info *edac_mc_alloc(unsigned int mc_num,
- unsigned int n_layers,
- struct edac_mc_layer *layers,
- unsigned int sz_pvt)
-{
- struct mem_ctl_info *mci;
- struct edac_mc_layer *layer;
- u32 *ce_per_layer[EDAC_MAX_LAYERS], *ue_per_layer[EDAC_MAX_LAYERS];
- unsigned int idx, size, tot_dimms = 1, count = 1;
- unsigned int tot_csrows = 1, tot_channels = 1, tot_errcount = 0;
- void *pvt, *ptr = NULL;
- int i;
- bool per_rank = false;
-
- if (WARN_ON(n_layers > EDAC_MAX_LAYERS || n_layers == 0))
- return NULL;
-
- /*
- * Calculate the total amount of dimms and csrows/cschannels while
- * in the old API emulation mode
- */
- for (idx = 0; idx < n_layers; idx++) {
- tot_dimms *= layers[idx].size;
-
- if (layers[idx].is_virt_csrow)
- tot_csrows *= layers[idx].size;
- else
- tot_channels *= layers[idx].size;
-
- if (layers[idx].type == EDAC_MC_LAYER_CHIP_SELECT)
- per_rank = true;
- }
-
- /* Figure out the offsets of the various items from the start of an mc
- * structure. We want the alignment of each item to be at least as
- * stringent as what the compiler would provide if we could simply
- * hardcode everything into a single struct.
- */
- mci = edac_align_ptr(&ptr, sizeof(*mci), 1);
- layer = edac_align_ptr(&ptr, sizeof(*layer), n_layers);
- for (i = 0; i < n_layers; i++) {
- count *= layers[i].size;
- edac_dbg(4, "errcount layer %d size %d\n", i, count);
- ce_per_layer[i] = edac_align_ptr(&ptr, sizeof(u32), count);
- ue_per_layer[i] = edac_align_ptr(&ptr, sizeof(u32), count);
- tot_errcount += 2 * count;
- }
-
- edac_dbg(4, "allocating %d error counters\n", tot_errcount);
- pvt = edac_align_ptr(&ptr, sz_pvt, 1);
- size = ((unsigned long)pvt) + sz_pvt;
-
- edac_dbg(1, "allocating %u bytes for mci data (%d %s, %d csrows/channels)\n",
- size,
- tot_dimms,
- per_rank ? "ranks" : "dimms",
- tot_csrows * tot_channels);
-
- mci = kzalloc(size, GFP_KERNEL);
- if (mci == NULL)
- return NULL;
-
- mci->dev.release = mci_release;
- device_initialize(&mci->dev);
-
- /* Adjust pointers so they point within the memory we just allocated
- * rather than an imaginary chunk of memory located at address 0.
- */
- layer = (struct edac_mc_layer *)(((char *)mci) + ((unsigned long)layer));
- for (i = 0; i < n_layers; i++) {
- mci->ce_per_layer[i] = (u32 *)((char *)mci + ((unsigned long)ce_per_layer[i]));
- mci->ue_per_layer[i] = (u32 *)((char *)mci + ((unsigned long)ue_per_layer[i]));
- }
- pvt = sz_pvt ? (((char *)mci) + ((unsigned long)pvt)) : NULL;
-
- /* setup index and various internal pointers */
- mci->mc_idx = mc_num;
- mci->tot_dimms = tot_dimms;
- mci->pvt_info = pvt;
- mci->n_layers = n_layers;
- mci->layers = layer;
- memcpy(mci->layers, layers, sizeof(*layer) * n_layers);
- mci->nr_csrows = tot_csrows;
- mci->num_cschannel = tot_channels;
- mci->csbased = per_rank;
-
- if (edac_mc_alloc_csrows(mci))
- goto error;
-
- if (edac_mc_alloc_dimms(mci))
- goto error;
-
- mci->op_state = OP_ALLOC;
-
- return mci;
-
-error:
- _edac_mc_free(mci);
-
- return NULL;
-}
-EXPORT_SYMBOL_GPL(edac_mc_alloc);
-
static int edac_mc_alloc_csrows(struct mem_ctl_info *mci)
{
unsigned int tot_csrows = mci->nr_csrows;
@@ -545,6 +439,109 @@ static int edac_mc_alloc_dimms(struct mem_ctl_info *mci)
return 0;
}
+struct mem_ctl_info *edac_mc_alloc(unsigned int mc_num,
+ unsigned int n_layers,
+ struct edac_mc_layer *layers,
+ unsigned int sz_pvt)
+{
+ struct mem_ctl_info *mci;
+ struct edac_mc_layer *layer;
+ u32 *ce_per_layer[EDAC_MAX_LAYERS], *ue_per_layer[EDAC_MAX_LAYERS];
+ unsigned int idx, size, tot_dimms = 1, count = 1;
+ unsigned int tot_csrows = 1, tot_channels = 1, tot_errcount = 0;
+ void *pvt, *ptr = NULL;
+ int i;
+ bool per_rank = false;
+
+ if (WARN_ON(n_layers > EDAC_MAX_LAYERS || n_layers == 0))
+ return NULL;
+
+ /*
+ * Calculate the total amount of dimms and csrows/cschannels while
+ * in the old API emulation mode
+ */
+ for (idx = 0; idx < n_layers; idx++) {
+ tot_dimms *= layers[idx].size;
+
+ if (layers[idx].is_virt_csrow)
+ tot_csrows *= layers[idx].size;
+ else
+ tot_channels *= layers[idx].size;
+
+ if (layers[idx].type == EDAC_MC_LAYER_CHIP_SELECT)
+ per_rank = true;
+ }
+
+ /* Figure out the offsets of the various items from the start of an mc
+ * structure. We want the alignment of each item to be at least as
+ * stringent as what the compiler would provide if we could simply
+ * hardcode everything into a single struct.
+ */
+ mci = edac_align_ptr(&ptr, sizeof(*mci), 1);
+ layer = edac_align_ptr(&ptr, sizeof(*layer), n_layers);
+ for (i = 0; i < n_layers; i++) {
+ count *= layers[i].size;
+ edac_dbg(4, "errcount layer %d size %d\n", i, count);
+ ce_per_layer[i] = edac_align_ptr(&ptr, sizeof(u32), count);
+ ue_per_layer[i] = edac_align_ptr(&ptr, sizeof(u32), count);
+ tot_errcount += 2 * count;
+ }
+
+ edac_dbg(4, "allocating %d error counters\n", tot_errcount);
+ pvt = edac_align_ptr(&ptr, sz_pvt, 1);
+ size = ((unsigned long)pvt) + sz_pvt;
+
+ edac_dbg(1, "allocating %u bytes for mci data (%d %s, %d csrows/channels)\n",
+ size,
+ tot_dimms,
+ per_rank ? "ranks" : "dimms",
+ tot_csrows * tot_channels);
+
+ mci = kzalloc(size, GFP_KERNEL);
+ if (mci == NULL)
+ return NULL;
+
+ mci->dev.release = mci_release;
+ device_initialize(&mci->dev);
+
+ /* Adjust pointers so they point within the memory we just allocated
+ * rather than an imaginary chunk of memory located at address 0.
+ */
+ layer = (struct edac_mc_layer *)(((char *)mci) + ((unsigned long)layer));
+ for (i = 0; i < n_layers; i++) {
+ mci->ce_per_layer[i] = (u32 *)((char *)mci + ((unsigned long)ce_per_layer[i]));
+ mci->ue_per_layer[i] = (u32 *)((char *)mci + ((unsigned long)ue_per_layer[i]));
+ }
+ pvt = sz_pvt ? (((char *)mci) + ((unsigned long)pvt)) : NULL;
+
+ /* setup index and various internal pointers */
+ mci->mc_idx = mc_num;
+ mci->tot_dimms = tot_dimms;
+ mci->pvt_info = pvt;
+ mci->n_layers = n_layers;
+ mci->layers = layer;
+ memcpy(mci->layers, layers, sizeof(*layer) * n_layers);
+ mci->nr_csrows = tot_csrows;
+ mci->num_cschannel = tot_channels;
+ mci->csbased = per_rank;
+
+ if (edac_mc_alloc_csrows(mci))
+ goto error;
+
+ if (edac_mc_alloc_dimms(mci))
+ goto error;
+
+ mci->op_state = OP_ALLOC;
+
+ return mci;
+
+error:
+ _edac_mc_free(mci);
+
+ return NULL;
+}
+EXPORT_SYMBOL_GPL(edac_mc_alloc);
+
void edac_mc_free(struct mem_ctl_info *mci)
{
edac_dbg(1, "\n");
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* Re: [PATCH v3 02/10] EDAC/mc: Reorder functions edac_mc_alloc*()
2020-01-23 9:02 ` [PATCH v3 02/10] EDAC/mc: Reorder functions edac_mc_alloc*() Robert Richter
@ 2020-01-23 19:02 ` Matthias Brugger
2020-01-30 18:49 ` Robert Richter
0 siblings, 1 reply; 17+ messages in thread
From: Matthias Brugger @ 2020-01-23 19:02 UTC (permalink / raw)
To: Robert Richter, Borislav Petkov, Mauro Carvalho Chehab, Tony Luck
Cc: James Morse, Aristeu Rozanski, linux-edac, linux-kernel,
Mauro Carvalho Chehab
On 23/01/2020 10:02, Robert Richter wrote:
> Reorder the new created functions edac_mc_alloc_csrows() and
> edac_mc_alloc_dimms() and move them before edac_mc_alloc(). No further
> code changes.
>
> Signed-off-by: Robert Richter <rrichter@marvell.com>
> Reviewed-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
> Acked-by: Aristeu Rozanski <aris@redhat.com>
> ---
> drivers/edac/edac_mc.c | 209 ++++++++++++++++++++---------------------
> 1 file changed, 103 insertions(+), 106 deletions(-)
>
> diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
> index d71006a4f07b..1e227e69e216 100644
> --- a/drivers/edac/edac_mc.c
> +++ b/drivers/edac/edac_mc.c
> @@ -311,112 +311,6 @@ static void mci_release(struct device *dev)
> kfree(mci);
> }
>
> -static int edac_mc_alloc_csrows(struct mem_ctl_info *mci);
> -static int edac_mc_alloc_dimms(struct mem_ctl_info *mci);
> -
> -struct mem_ctl_info *edac_mc_alloc(unsigned int mc_num,
> - unsigned int n_layers,
> - struct edac_mc_layer *layers,
> - unsigned int sz_pvt)
> -{
[...]
> - edac_dbg(1, "allocating %u bytes for mci data (%d %s, %d csrows/channels)\n",
> - size,
> - tot_dimms,
> - per_rank ? "ranks" : "dimms",
> - tot_csrows * tot_channels);
> -
> - mci = kzalloc(size, GFP_KERNEL);
> - if (mci == NULL)
> - return NULL;
> -
> - mci->dev.release = mci_release;
> - device_initialize(&mci->dev);
Seems like some leftovers in your working tree. Are you sure you created the
patches against a clean upstream tree?
Regards,
Matthias
^ permalink raw reply [flat|nested] 17+ messages in thread* Re: [PATCH v3 02/10] EDAC/mc: Reorder functions edac_mc_alloc*()
2020-01-23 19:02 ` Matthias Brugger
@ 2020-01-30 18:49 ` Robert Richter
0 siblings, 0 replies; 17+ messages in thread
From: Robert Richter @ 2020-01-30 18:49 UTC (permalink / raw)
To: Matthias Brugger
Cc: Borislav Petkov, Mauro Carvalho Chehab, Tony Luck, James Morse,
Aristeu Rozanski, linux-edac, linux-kernel,
Mauro Carvalho Chehab
Matthias,
On 23.01.20 20:02:00, Matthias Brugger wrote:
> On 23/01/2020 10:02, Robert Richter wrote:
> > Reorder the new created functions edac_mc_alloc_csrows() and
> > edac_mc_alloc_dimms() and move them before edac_mc_alloc(). No further
> > code changes.
> >
> > Signed-off-by: Robert Richter <rrichter@marvell.com>
> > Reviewed-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
> > Acked-by: Aristeu Rozanski <aris@redhat.com>
> > ---
> > drivers/edac/edac_mc.c | 209 ++++++++++++++++++++---------------------
> > 1 file changed, 103 insertions(+), 106 deletions(-)
> >
> > diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
> > index d71006a4f07b..1e227e69e216 100644
Sorry, I missed that mail.
I have rebased v3 series on top of edac-for-next *plus* use-after-
free-fix, see here:
https://lore.kernel.org/patchwork/patch/1169444/
So the patch above must be applied before the series.
Sorry for the confusion and not being more clear here.
-Robert
^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH v3 03/10] EDAC: Store error type in struct edac_raw_error_desc
2020-01-23 9:02 [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Robert Richter
2020-01-23 9:02 ` [PATCH v3 01/10] EDAC/mc: Split edac_mc_alloc() into smaller functions Robert Richter
2020-01-23 9:02 ` [PATCH v3 02/10] EDAC/mc: Reorder functions edac_mc_alloc*() Robert Richter
@ 2020-01-23 9:02 ` Robert Richter
2020-01-23 9:02 ` [PATCH v3 04/10] EDAC/mc: Determine mci pointer from the error descriptor Robert Richter
` (7 subsequent siblings)
10 siblings, 0 replies; 17+ messages in thread
From: Robert Richter @ 2020-01-23 9:02 UTC (permalink / raw)
To: Borislav Petkov, Mauro Carvalho Chehab, Tony Luck
Cc: James Morse, Aristeu Rozanski, Robert Richter, linux-edac,
linux-kernel, kbuild test robot
Store the error type in struct edac_raw_error_desc. This makes the
type parameter of edac_raw_mc_handle_error() obsolete.
Reported-by: kbuild test robot <lkp@intel.com>
Signed-off-by: Robert Richter <rrichter@marvell.com>
Reviewed-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Acked-by: Aristeu Rozanski <aris@redhat.com>
---
drivers/edac/edac_mc.c | 10 +++++-----
drivers/edac/edac_mc.h | 4 +---
drivers/edac/ghes_edac.c | 11 +++++------
include/linux/edac.h | 2 ++
4 files changed, 13 insertions(+), 14 deletions(-)
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index 1e227e69e216..8ef69d24297d 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -1086,8 +1086,7 @@ static void edac_ue_error(struct mem_ctl_info *mci,
edac_inc_ue_error(mci, enable_per_layer_report, pos, error_count);
}
-void edac_raw_mc_handle_error(const enum hw_event_mc_err_type type,
- struct mem_ctl_info *mci,
+void edac_raw_mc_handle_error(struct mem_ctl_info *mci,
struct edac_raw_error_desc *e)
{
char detail[80];
@@ -1102,14 +1101,14 @@ void edac_raw_mc_handle_error(const enum hw_event_mc_err_type type,
/* Report the error via the trace interface */
if (IS_ENABLED(CONFIG_RAS))
- trace_mc_event(type, e->msg, e->label, e->error_count,
+ trace_mc_event(e->type, e->msg, e->label, e->error_count,
mci->mc_idx, e->top_layer, e->mid_layer,
e->low_layer,
(e->page_frame_number << PAGE_SHIFT) | e->offset_in_page,
grain_bits, e->syndrome, e->other_detail);
/* Memory type dependent details about the error */
- if (type == HW_EVENT_ERR_CORRECTED) {
+ if (e->type == HW_EVENT_ERR_CORRECTED) {
snprintf(detail, sizeof(detail),
"page:0x%lx offset:0x%lx grain:%ld syndrome:0x%lx",
e->page_frame_number, e->offset_in_page,
@@ -1154,6 +1153,7 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
/* Fills the error report buffer */
memset(e, 0, sizeof (*e));
e->error_count = error_count;
+ e->type = type;
e->top_layer = top_layer;
e->mid_layer = mid_layer;
e->low_layer = low_layer;
@@ -1284,6 +1284,6 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
if (p > e->location)
*(p - 1) = '\0';
- edac_raw_mc_handle_error(type, mci, e);
+ edac_raw_mc_handle_error(mci, e);
}
EXPORT_SYMBOL_GPL(edac_mc_handle_error);
diff --git a/drivers/edac/edac_mc.h b/drivers/edac/edac_mc.h
index 02aac5c61d00..5d78be774f9e 100644
--- a/drivers/edac/edac_mc.h
+++ b/drivers/edac/edac_mc.h
@@ -212,7 +212,6 @@ extern int edac_mc_find_csrow_by_page(struct mem_ctl_info *mci,
* edac_raw_mc_handle_error() - Reports a memory event to userspace without
* doing anything to discover the error location.
*
- * @type: severity of the error (CE/UE/Fatal)
* @mci: a struct mem_ctl_info pointer
* @e: error description
*
@@ -220,8 +219,7 @@ extern int edac_mc_find_csrow_by_page(struct mem_ctl_info *mci,
* only be called directly when the hardware error come directly from BIOS,
* like in the case of APEI GHES driver.
*/
-void edac_raw_mc_handle_error(const enum hw_event_mc_err_type type,
- struct mem_ctl_info *mci,
+void edac_raw_mc_handle_error(struct mem_ctl_info *mci,
struct edac_raw_error_desc *e);
/**
diff --git a/drivers/edac/ghes_edac.c b/drivers/edac/ghes_edac.c
index b99080d8a10c..7c3e5264a41e 100644
--- a/drivers/edac/ghes_edac.c
+++ b/drivers/edac/ghes_edac.c
@@ -201,7 +201,6 @@ static void ghes_edac_dmidecode(const struct dmi_header *dh, void *arg)
void ghes_edac_report_mem_error(int sev, struct cper_sec_mem_err *mem_err)
{
- enum hw_event_mc_err_type type;
struct edac_raw_error_desc *e;
struct mem_ctl_info *mci;
struct ghes_edac_pvt *pvt;
@@ -240,17 +239,17 @@ void ghes_edac_report_mem_error(int sev, struct cper_sec_mem_err *mem_err)
switch (sev) {
case GHES_SEV_CORRECTED:
- type = HW_EVENT_ERR_CORRECTED;
+ e->type = HW_EVENT_ERR_CORRECTED;
break;
case GHES_SEV_RECOVERABLE:
- type = HW_EVENT_ERR_UNCORRECTED;
+ e->type = HW_EVENT_ERR_UNCORRECTED;
break;
case GHES_SEV_PANIC:
- type = HW_EVENT_ERR_FATAL;
+ e->type = HW_EVENT_ERR_FATAL;
break;
default:
case GHES_SEV_NO:
- type = HW_EVENT_ERR_INFO;
+ e->type = HW_EVENT_ERR_INFO;
}
edac_dbg(1, "error validation_bits: 0x%08llx\n",
@@ -442,7 +441,7 @@ void ghes_edac_report_mem_error(int sev, struct cper_sec_mem_err *mem_err)
if (p > pvt->other_detail)
*(p - 1) = '\0';
- edac_raw_mc_handle_error(type, mci, e);
+ edac_raw_mc_handle_error(mci, e);
unlock:
spin_unlock_irqrestore(&ghes_lock, flags);
diff --git a/include/linux/edac.h b/include/linux/edac.h
index cc31b9742684..6703eb492cd2 100644
--- a/include/linux/edac.h
+++ b/include/linux/edac.h
@@ -442,6 +442,7 @@ struct errcount_attribute_data {
* struct edac_raw_error_desc - Raw error report structure
* @grain: minimum granularity for an error report, in bytes
* @error_count: number of errors of the same type
+ * @type: severity of the error (CE/UE/Fatal)
* @top_layer: top layer of the error (layer[0])
* @mid_layer: middle layer of the error (layer[1])
* @low_layer: low layer of the error (layer[2])
@@ -462,6 +463,7 @@ struct edac_raw_error_desc {
long grain;
u16 error_count;
+ enum hw_event_mc_err_type type;
int top_layer;
int mid_layer;
int low_layer;
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH v3 04/10] EDAC/mc: Determine mci pointer from the error descriptor
2020-01-23 9:02 [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Robert Richter
` (2 preceding siblings ...)
2020-01-23 9:02 ` [PATCH v3 03/10] EDAC: Store error type in struct edac_raw_error_desc Robert Richter
@ 2020-01-23 9:02 ` Robert Richter
2020-01-23 9:02 ` [PATCH v3 05/10] EDAC/mc: Create new function edac_inc_csrow() Robert Richter
` (6 subsequent siblings)
10 siblings, 0 replies; 17+ messages in thread
From: Robert Richter @ 2020-01-23 9:02 UTC (permalink / raw)
To: Borislav Petkov, Mauro Carvalho Chehab, Tony Luck
Cc: James Morse, Aristeu Rozanski, Robert Richter, linux-edac,
linux-kernel, Mauro Carvalho Chehab
Each struct mci has its own error descriptor. Create a function
error_desc_to_mci() to determine the corresponding mci from an error
descriptor. This eases the parameter list of edac_raw_mc_handle_
error() as the mci pointer do not need to be passed any longer.
Signed-off-by: Robert Richter <rrichter@marvell.com>
Reviewed-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
Acked-by: Aristeu Rozanski <aris@redhat.com>
---
drivers/edac/edac_mc.c | 11 ++++++++---
drivers/edac/edac_mc.h | 4 +---
drivers/edac/ghes_edac.c | 2 +-
3 files changed, 10 insertions(+), 7 deletions(-)
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index 8ef69d24297d..3c00c046acc9 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -55,6 +55,11 @@ static LIST_HEAD(mc_devices);
*/
static const char *edac_mc_owner;
+static struct mem_ctl_info *error_desc_to_mci(struct edac_raw_error_desc *e)
+{
+ return container_of(e, struct mem_ctl_info, error_desc);
+}
+
int edac_get_report_status(void)
{
return edac_report;
@@ -1086,9 +1091,9 @@ static void edac_ue_error(struct mem_ctl_info *mci,
edac_inc_ue_error(mci, enable_per_layer_report, pos, error_count);
}
-void edac_raw_mc_handle_error(struct mem_ctl_info *mci,
- struct edac_raw_error_desc *e)
+void edac_raw_mc_handle_error(struct edac_raw_error_desc *e)
{
+ struct mem_ctl_info *mci = error_desc_to_mci(e);
char detail[80];
int pos[EDAC_MAX_LAYERS] = { e->top_layer, e->mid_layer, e->low_layer };
u8 grain_bits;
@@ -1284,6 +1289,6 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
if (p > e->location)
*(p - 1) = '\0';
- edac_raw_mc_handle_error(mci, e);
+ edac_raw_mc_handle_error(e);
}
EXPORT_SYMBOL_GPL(edac_mc_handle_error);
diff --git a/drivers/edac/edac_mc.h b/drivers/edac/edac_mc.h
index 5d78be774f9e..881b00eadf7a 100644
--- a/drivers/edac/edac_mc.h
+++ b/drivers/edac/edac_mc.h
@@ -212,15 +212,13 @@ extern int edac_mc_find_csrow_by_page(struct mem_ctl_info *mci,
* edac_raw_mc_handle_error() - Reports a memory event to userspace without
* doing anything to discover the error location.
*
- * @mci: a struct mem_ctl_info pointer
* @e: error description
*
* This raw function is used internally by edac_mc_handle_error(). It should
* only be called directly when the hardware error come directly from BIOS,
* like in the case of APEI GHES driver.
*/
-void edac_raw_mc_handle_error(struct mem_ctl_info *mci,
- struct edac_raw_error_desc *e);
+void edac_raw_mc_handle_error(struct edac_raw_error_desc *e);
/**
* edac_mc_handle_error() - Reports a memory event to userspace.
diff --git a/drivers/edac/ghes_edac.c b/drivers/edac/ghes_edac.c
index 7c3e5264a41e..bef8a428c429 100644
--- a/drivers/edac/ghes_edac.c
+++ b/drivers/edac/ghes_edac.c
@@ -441,7 +441,7 @@ void ghes_edac_report_mem_error(int sev, struct cper_sec_mem_err *mem_err)
if (p > pvt->other_detail)
*(p - 1) = '\0';
- edac_raw_mc_handle_error(mci, e);
+ edac_raw_mc_handle_error(e);
unlock:
spin_unlock_irqrestore(&ghes_lock, flags);
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH v3 05/10] EDAC/mc: Create new function edac_inc_csrow()
2020-01-23 9:02 [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Robert Richter
` (3 preceding siblings ...)
2020-01-23 9:02 ` [PATCH v3 04/10] EDAC/mc: Determine mci pointer from the error descriptor Robert Richter
@ 2020-01-23 9:02 ` Robert Richter
2020-02-13 12:47 ` Borislav Petkov
2020-02-14 14:17 ` [PATCH v2] " Robert Richter
2020-01-23 9:03 ` [PATCH v3 06/10] EDAC/mc: Report "unknown memory" on too many DIMM labels found Robert Richter
` (5 subsequent siblings)
10 siblings, 2 replies; 17+ messages in thread
From: Robert Richter @ 2020-01-23 9:02 UTC (permalink / raw)
To: Borislav Petkov, Mauro Carvalho Chehab, Tony Luck
Cc: James Morse, Aristeu Rozanski, Robert Richter, linux-edac, linux-kernel
Have a separate function to count errors in csrow/channel. This better
separates code and reduces the indentation level. No functional
changes.
Signed-off-by: Robert Richter <rrichter@marvell.com>
Reviewed-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Acked-by: Aristeu Rozanski <aris@redhat.com>
---
drivers/edac/edac_mc.c | 40 +++++++++++++++++++++++++---------------
1 file changed, 25 insertions(+), 15 deletions(-)
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index 3c00c046acc9..e75cb7a9c454 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -1091,6 +1091,26 @@ static void edac_ue_error(struct mem_ctl_info *mci,
edac_inc_ue_error(mci, enable_per_layer_report, pos, error_count);
}
+static void edac_inc_csrow(struct edac_raw_error_desc *e, int row, int chan)
+{
+ struct mem_ctl_info *mci = error_desc_to_mci(e);
+ u16 count = e->error_count;
+ enum hw_event_mc_err_type type = e->type;
+
+ if (row < 0)
+ return;
+
+ edac_dbg(4, "csrow/channel to increment: (%d,%d)\n", row, chan);
+
+ if (type == HW_EVENT_ERR_CORRECTED) {
+ mci->csrows[row]->ce_count += count;
+ if (chan >= 0)
+ mci->csrows[row]->channels[chan]->ce_count += count;
+ } else {
+ mci->csrows[row]->ue_count += count;
+ }
+}
+
void edac_raw_mc_handle_error(struct edac_raw_error_desc *e)
{
struct mem_ctl_info *mci = error_desc_to_mci(e);
@@ -1258,22 +1278,12 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
chan = -2;
}
- if (!e->enable_per_layer_report) {
+ if (!e->enable_per_layer_report)
strcpy(e->label, "any memory");
- } else {
- edac_dbg(4, "csrow/channel to increment: (%d,%d)\n", row, chan);
- if (p == e->label)
- strcpy(e->label, "unknown memory");
- if (type == HW_EVENT_ERR_CORRECTED) {
- if (row >= 0) {
- mci->csrows[row]->ce_count += error_count;
- if (chan >= 0)
- mci->csrows[row]->channels[chan]->ce_count += error_count;
- }
- } else
- if (row >= 0)
- mci->csrows[row]->ue_count += error_count;
- }
+ else if (!*e->label)
+ strcpy(e->label, "unknown memory");
+
+ edac_inc_csrow(e, row, chan);
/* Fill the RAM location data */
p = e->location;
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* Re: [PATCH v3 05/10] EDAC/mc: Create new function edac_inc_csrow()
2020-01-23 9:02 ` [PATCH v3 05/10] EDAC/mc: Create new function edac_inc_csrow() Robert Richter
@ 2020-02-13 12:47 ` Borislav Petkov
2020-02-14 12:40 ` Robert Richter
2020-02-14 14:17 ` [PATCH v2] " Robert Richter
1 sibling, 1 reply; 17+ messages in thread
From: Borislav Petkov @ 2020-02-13 12:47 UTC (permalink / raw)
To: Robert Richter
Cc: Mauro Carvalho Chehab, Tony Luck, James Morse, Aristeu Rozanski,
linux-edac, linux-kernel
On Thu, Jan 23, 2020 at 09:02:58AM +0000, Robert Richter wrote:
> Have a separate function to count errors in csrow/channel. This better
> separates code and reduces the indentation level. No functional
> changes.
>
> Signed-off-by: Robert Richter <rrichter@marvell.com>
> Reviewed-by: Mauro Carvalho Chehab <mchehab@kernel.org>
> Acked-by: Aristeu Rozanski <aris@redhat.com>
> ---
> drivers/edac/edac_mc.c | 40 +++++++++++++++++++++++++---------------
> 1 file changed, 25 insertions(+), 15 deletions(-)
>
> diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
> index 3c00c046acc9..e75cb7a9c454 100644
> --- a/drivers/edac/edac_mc.c
> +++ b/drivers/edac/edac_mc.c
> @@ -1091,6 +1091,26 @@ static void edac_ue_error(struct mem_ctl_info *mci,
> edac_inc_ue_error(mci, enable_per_layer_report, pos, error_count);
> }
>
> +static void edac_inc_csrow(struct edac_raw_error_desc *e, int row, int chan)
> +{
> + struct mem_ctl_info *mci = error_desc_to_mci(e);
> + u16 count = e->error_count;
> + enum hw_event_mc_err_type type = e->type;
Please sort function local variables declaration in a reverse christmas
tree order:
<type A> longest_variable_name;
<type B> shorter_var_name;
<type C> even_shorter;
<type D> i;
> +
> + if (row < 0)
> + return;
> +
> + edac_dbg(4, "csrow/channel to increment: (%d,%d)\n", row, chan);
> +
> + if (type == HW_EVENT_ERR_CORRECTED) {
> + mci->csrows[row]->ce_count += count;
> + if (chan >= 0)
> + mci->csrows[row]->channels[chan]->ce_count += count;
> + } else {
> + mci->csrows[row]->ue_count += count;
> + }
> +}
> +
> void edac_raw_mc_handle_error(struct edac_raw_error_desc *e)
> {
> struct mem_ctl_info *mci = error_desc_to_mci(e);
> @@ -1258,22 +1278,12 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
> chan = -2;
> }
>
> - if (!e->enable_per_layer_report) {
> + if (!e->enable_per_layer_report)
> strcpy(e->label, "any memory");
> - } else {
> - edac_dbg(4, "csrow/channel to increment: (%d,%d)\n", row, chan);
> - if (p == e->label)
> - strcpy(e->label, "unknown memory");
> - if (type == HW_EVENT_ERR_CORRECTED) {
> - if (row >= 0) {
> - mci->csrows[row]->ce_count += error_count;
> - if (chan >= 0)
> - mci->csrows[row]->channels[chan]->ce_count += error_count;
> - }
> - } else
> - if (row >= 0)
> - mci->csrows[row]->ue_count += error_count;
> - }
> + else if (!*e->label)
> + strcpy(e->label, "unknown memory");
> +
> + edac_inc_csrow(e, row, chan);
Err, but this has functional changes: the !e->enable_per_layer_report
case sets only the e->label and the else branch only does increment
->ce_count.
Your change gets ->ce_count incremented in both cases.
Why?
--
Regards/Gruss,
Boris.
https://people.kernel.org/tglx/notes-about-netiquette
^ permalink raw reply [flat|nested] 17+ messages in thread* Re: [PATCH v3 05/10] EDAC/mc: Create new function edac_inc_csrow()
2020-02-13 12:47 ` Borislav Petkov
@ 2020-02-14 12:40 ` Robert Richter
0 siblings, 0 replies; 17+ messages in thread
From: Robert Richter @ 2020-02-14 12:40 UTC (permalink / raw)
To: Borislav Petkov
Cc: Mauro Carvalho Chehab, Tony Luck, James Morse, Aristeu Rozanski,
linux-edac, linux-kernel
On 13.02.20 13:47:08, Borislav Petkov wrote:
> On Thu, Jan 23, 2020 at 09:02:58AM +0000, Robert Richter wrote:
> > Have a separate function to count errors in csrow/channel. This better
> > separates code and reduces the indentation level. No functional
> > changes.
> >
> > Signed-off-by: Robert Richter <rrichter@marvell.com>
> > Reviewed-by: Mauro Carvalho Chehab <mchehab@kernel.org>
> > Acked-by: Aristeu Rozanski <aris@redhat.com>
> > ---
> > drivers/edac/edac_mc.c | 40 +++++++++++++++++++++++++---------------
> > 1 file changed, 25 insertions(+), 15 deletions(-)
> >
> > diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
> > index 3c00c046acc9..e75cb7a9c454 100644
> > --- a/drivers/edac/edac_mc.c
> > +++ b/drivers/edac/edac_mc.c
> > @@ -1091,6 +1091,26 @@ static void edac_ue_error(struct mem_ctl_info *mci,
> > edac_inc_ue_error(mci, enable_per_layer_report, pos, error_count);
> > }
> >
> > +static void edac_inc_csrow(struct edac_raw_error_desc *e, int row, int chan)
> > +{
> > + struct mem_ctl_info *mci = error_desc_to_mci(e);
> > + u16 count = e->error_count;
> > + enum hw_event_mc_err_type type = e->type;
>
> Please sort function local variables declaration in a reverse christmas
> tree order:
>
> <type A> longest_variable_name;
> <type B> shorter_var_name;
> <type C> even_shorter;
> <type D> i;
I can change this. Does variable name length include the assignment?
Generally I prefer to sort it by size_of() to avoid holes due to
padding, though the compiler uses probably registers here anyway.
It's just a flavor.
If it's just this change, could you edit the patch to avoid respin?
>
> > +
> > + if (row < 0)
> > + return;
> > +
> > + edac_dbg(4, "csrow/channel to increment: (%d,%d)\n", row, chan);
> > +
> > + if (type == HW_EVENT_ERR_CORRECTED) {
> > + mci->csrows[row]->ce_count += count;
> > + if (chan >= 0)
> > + mci->csrows[row]->channels[chan]->ce_count += count;
> > + } else {
> > + mci->csrows[row]->ue_count += count;
> > + }
> > +}
> > +
> > void edac_raw_mc_handle_error(struct edac_raw_error_desc *e)
> > {
> > struct mem_ctl_info *mci = error_desc_to_mci(e);
> > @@ -1258,22 +1278,12 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
> > chan = -2;
> > }
> >
> > - if (!e->enable_per_layer_report) {
> > + if (!e->enable_per_layer_report)
> > strcpy(e->label, "any memory");
> > - } else {
> > - edac_dbg(4, "csrow/channel to increment: (%d,%d)\n", row, chan);
> > - if (p == e->label)
> > - strcpy(e->label, "unknown memory");
> > - if (type == HW_EVENT_ERR_CORRECTED) {
> > - if (row >= 0) {
> > - mci->csrows[row]->ce_count += error_count;
> > - if (chan >= 0)
> > - mci->csrows[row]->channels[chan]->ce_count += error_count;
> > - }
> > - } else
> > - if (row >= 0)
> > - mci->csrows[row]->ue_count += error_count;
> > - }
> > + else if (!*e->label)
> > + strcpy(e->label, "unknown memory");
> > +
> > + edac_inc_csrow(e, row, chan);
>
> Err, but this has functional changes: the !e->enable_per_layer_report
> case sets only the e->label and the else branch only does increment
> ->ce_count.
>
> Your change gets ->ce_count incremented in both cases.
No, there is a check in edac_inc_csrow(): if (row < 0) ... In the case
of "any memory", row is also < 0, so nothing is counted. This is
reasonable since no dimm is found and row/channel is still set to the
setup value of -1.
-Robert
^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH v2] EDAC/mc: Create new function edac_inc_csrow()
2020-01-23 9:02 ` [PATCH v3 05/10] EDAC/mc: Create new function edac_inc_csrow() Robert Richter
2020-02-13 12:47 ` Borislav Petkov
@ 2020-02-14 14:17 ` Robert Richter
1 sibling, 0 replies; 17+ messages in thread
From: Robert Richter @ 2020-02-14 14:17 UTC (permalink / raw)
To: Borislav Petkov, Mauro Carvalho Chehab, Tony Luck
Cc: James Morse, Aristeu Rozanski, Robert Richter, linux-edac, linux-kernel
Have a separate function to count errors in csrow/channel. This better
separates code and reduces the indentation level.
Implementation note: Function edac_inc_csrow() counts the same as
before, ->ce_count is only incremented if row >= 0. This is esp. true
for the case of (!e->enable_per_layer_report). Here, a DIMM was not
found, variable row still has a value of -1 and ->ce_count is not
incremented.
Signed-off-by: Robert Richter <rrichter@marvell.com>
Reviewed-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Acked-by: Aristeu Rozanski <aris@redhat.com>
---
v2:
* updated patch description to address the case fir
(!e->enable_per_layer_report),
* reordered variable declarations
---
drivers/edac/edac_mc.c | 40 +++++++++++++++++++++++++---------------
1 file changed, 25 insertions(+), 15 deletions(-)
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index e817a710739f..31ba988359d2 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -1089,6 +1089,26 @@ static void edac_ue_error(struct mem_ctl_info *mci,
edac_inc_ue_error(mci, enable_per_layer_report, pos, error_count);
}
+static void edac_inc_csrow(struct edac_raw_error_desc *e, int row, int chan)
+{
+ struct mem_ctl_info *mci = error_desc_to_mci(e);
+ enum hw_event_mc_err_type type = e->type;
+ u16 count = e->error_count;
+
+ if (row < 0)
+ return;
+
+ edac_dbg(4, "csrow/channel to increment: (%d,%d)\n", row, chan);
+
+ if (type == HW_EVENT_ERR_CORRECTED) {
+ mci->csrows[row]->ce_count += count;
+ if (chan >= 0)
+ mci->csrows[row]->channels[chan]->ce_count += count;
+ } else {
+ mci->csrows[row]->ue_count += count;
+ }
+}
+
void edac_raw_mc_handle_error(struct edac_raw_error_desc *e)
{
struct mem_ctl_info *mci = error_desc_to_mci(e);
@@ -1256,22 +1276,12 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
chan = -2;
}
- if (!e->enable_per_layer_report) {
+ if (!e->enable_per_layer_report)
strcpy(e->label, "any memory");
- } else {
- edac_dbg(4, "csrow/channel to increment: (%d,%d)\n", row, chan);
- if (p == e->label)
- strcpy(e->label, "unknown memory");
- if (type == HW_EVENT_ERR_CORRECTED) {
- if (row >= 0) {
- mci->csrows[row]->ce_count += error_count;
- if (chan >= 0)
- mci->csrows[row]->channels[chan]->ce_count += error_count;
- }
- } else
- if (row >= 0)
- mci->csrows[row]->ue_count += error_count;
- }
+ else if (!*e->label)
+ strcpy(e->label, "unknown memory");
+
+ edac_inc_csrow(e, row, chan);
/* Fill the RAM location data */
p = e->location;
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH v3 06/10] EDAC/mc: Report "unknown memory" on too many DIMM labels found
2020-01-23 9:02 [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Robert Richter
` (4 preceding siblings ...)
2020-01-23 9:02 ` [PATCH v3 05/10] EDAC/mc: Create new function edac_inc_csrow() Robert Richter
@ 2020-01-23 9:03 ` Robert Richter
2020-01-23 9:03 ` [PATCH v3 07/10] EDAC/mc: Remove enable_per_layer_report function arguments Robert Richter
` (4 subsequent siblings)
10 siblings, 0 replies; 17+ messages in thread
From: Robert Richter @ 2020-01-23 9:03 UTC (permalink / raw)
To: Borislav Petkov, Mauro Carvalho Chehab, Tony Luck
Cc: James Morse, Aristeu Rozanski, Robert Richter, linux-edac, linux-kernel
There is a limitation to report only EDAC_MAX_LABELS in e->label of
the error descriptor. This is to prevent a possible string overflow.
Current implementation falls back to "any memory" in this case and
also stops all further processing to find a unique row and channel of
the possible error location. Reporting "any memory" is wrong as the
memory controller reported an error location for one of the layers.
Instead, report "unknown memory" and also do not break early in the
loop to further check row and channel for uniqueness.
Signed-off-by: Robert Richter <rrichter@marvell.com>
Acked-by: Aristeu Rozanski <aris@redhat.com>
---
drivers/edac/edac_mc.c | 21 +++++++++++----------
1 file changed, 11 insertions(+), 10 deletions(-)
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index e75cb7a9c454..aa94152777fe 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -1245,20 +1245,21 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
* channel/memory controller/... may be affected.
* Also, don't show errors for empty DIMM slots.
*/
- if (!e->enable_per_layer_report || !dimm->nr_pages)
+ if (!dimm->nr_pages)
continue;
- if (n_labels >= EDAC_MAX_LABELS) {
- e->enable_per_layer_report = false;
- break;
- }
n_labels++;
- if (p != e->label) {
- strcpy(p, OTHER_LABEL);
- p += strlen(OTHER_LABEL);
+ if (n_labels > EDAC_MAX_LABELS) {
+ p = e->label;
+ *p = '\0';
+ } else {
+ if (p != e->label) {
+ strcpy(p, OTHER_LABEL);
+ p += strlen(OTHER_LABEL);
+ }
+ strcpy(p, dimm->label);
+ p += strlen(p);
}
- strcpy(p, dimm->label);
- p += strlen(p);
/*
* get csrow/channel of the DIMM, in order to allow
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH v3 07/10] EDAC/mc: Remove enable_per_layer_report function arguments
2020-01-23 9:02 [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Robert Richter
` (5 preceding siblings ...)
2020-01-23 9:03 ` [PATCH v3 06/10] EDAC/mc: Report "unknown memory" on too many DIMM labels found Robert Richter
@ 2020-01-23 9:03 ` Robert Richter
2020-01-23 9:03 ` [PATCH v3 08/10] EDAC/mc: Pass the error descriptor to error reporting functions Robert Richter
` (3 subsequent siblings)
10 siblings, 0 replies; 17+ messages in thread
From: Robert Richter @ 2020-01-23 9:03 UTC (permalink / raw)
To: Borislav Petkov, Mauro Carvalho Chehab, Tony Luck
Cc: James Morse, Aristeu Rozanski, Robert Richter, linux-edac, linux-kernel
Many functions carry the enable_per_layer_report argument. This is a
bool value indicating the error information contains some location
data where the error occurred. This can easily being determined by
checking the pos[] array for values. Negative values indicate there is
no location available. So if the top layer is negative, the error
location is unknown.
Just check if the top layer is negative and remove
enable_per_layer_report as function argument and also from struct
edac_raw_error_desc.
Signed-off-by: Robert Richter <rrichter@marvell.com>
Acked-by: Aristeu Rozanski <aris@redhat.com>
---
drivers/edac/edac_mc.c | 42 +++++++++++++++++++---------------------
drivers/edac/ghes_edac.c | 5 +----
include/linux/edac.h | 3 ---
3 files changed, 21 insertions(+), 29 deletions(-)
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index aa94152777fe..35e427f89949 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -948,7 +948,6 @@ const char *edac_layer_name[] = {
EXPORT_SYMBOL_GPL(edac_layer_name);
static void edac_inc_ce_error(struct mem_ctl_info *mci,
- bool enable_per_layer_report,
const int pos[EDAC_MAX_LAYERS],
const u16 count)
{
@@ -956,7 +955,7 @@ static void edac_inc_ce_error(struct mem_ctl_info *mci,
mci->ce_mc += count;
- if (!enable_per_layer_report) {
+ if (pos[0] < 0) {
mci->ce_noinfo_count += count;
return;
}
@@ -973,7 +972,6 @@ static void edac_inc_ce_error(struct mem_ctl_info *mci,
}
static void edac_inc_ue_error(struct mem_ctl_info *mci,
- bool enable_per_layer_report,
const int pos[EDAC_MAX_LAYERS],
const u16 count)
{
@@ -981,7 +979,7 @@ static void edac_inc_ue_error(struct mem_ctl_info *mci,
mci->ue_mc += count;
- if (!enable_per_layer_report) {
+ if (pos[0] < 0) {
mci->ue_noinfo_count += count;
return;
}
@@ -1005,7 +1003,6 @@ static void edac_ce_error(struct mem_ctl_info *mci,
const char *label,
const char *detail,
const char *other_detail,
- const bool enable_per_layer_report,
const unsigned long page_frame_number,
const unsigned long offset_in_page,
long grain)
@@ -1028,7 +1025,7 @@ static void edac_ce_error(struct mem_ctl_info *mci,
error_count, msg, msg_aux, label,
location, detail);
}
- edac_inc_ce_error(mci, enable_per_layer_report, pos, error_count);
+ edac_inc_ce_error(mci, pos, error_count);
if (mci->scrub_mode == SCRUB_SW_SRC) {
/*
@@ -1058,8 +1055,7 @@ static void edac_ue_error(struct mem_ctl_info *mci,
const char *location,
const char *label,
const char *detail,
- const char *other_detail,
- const bool enable_per_layer_report)
+ const char *other_detail)
{
char *msg_aux = "";
@@ -1088,7 +1084,7 @@ static void edac_ue_error(struct mem_ctl_info *mci,
msg, msg_aux, label, location, detail);
}
- edac_inc_ue_error(mci, enable_per_layer_report, pos, error_count);
+ edac_inc_ue_error(mci, pos, error_count);
}
static void edac_inc_csrow(struct edac_raw_error_desc *e, int row, int chan)
@@ -1138,16 +1134,16 @@ void edac_raw_mc_handle_error(struct edac_raw_error_desc *e)
"page:0x%lx offset:0x%lx grain:%ld syndrome:0x%lx",
e->page_frame_number, e->offset_in_page,
e->grain, e->syndrome);
- edac_ce_error(mci, e->error_count, pos, e->msg, e->location, e->label,
- detail, e->other_detail, e->enable_per_layer_report,
+ edac_ce_error(mci, e->error_count, pos, e->msg, e->location,
+ e->label, detail, e->other_detail,
e->page_frame_number, e->offset_in_page, e->grain);
} else {
snprintf(detail, sizeof(detail),
"page:0x%lx offset:0x%lx grain:%ld",
e->page_frame_number, e->offset_in_page, e->grain);
- edac_ue_error(mci, e->error_count, pos, e->msg, e->location, e->label,
- detail, e->other_detail, e->enable_per_layer_report);
+ edac_ue_error(mci, e->error_count, pos, e->msg, e->location,
+ e->label, detail, e->other_detail);
}
@@ -1172,6 +1168,7 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
int pos[EDAC_MAX_LAYERS] = { top_layer, mid_layer, low_layer };
int i, n_labels = 0;
struct edac_raw_error_desc *e = &mci->error_desc;
+ bool any_memory = true;
edac_dbg(3, "MC%d\n", mci->mc_idx);
@@ -1190,9 +1187,9 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
/*
* Check if the event report is consistent and if the memory
- * location is known. If it is known, enable_per_layer_report will be
- * true, the DIMM(s) label info will be filled and the per-layer
- * error counters will be incremented.
+ * location is known. If it is known, the DIMM(s) label info
+ * will be filled and the per-layer error counters will be
+ * incremented.
*/
for (i = 0; i < mci->n_layers; i++) {
if (pos[i] >= (int)mci->layers[i].size) {
@@ -1210,7 +1207,7 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
pos[i] = -1;
}
if (pos[i] >= 0)
- e->enable_per_layer_report = true;
+ any_memory = false;
}
/*
@@ -1240,10 +1237,11 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
e->grain = dimm->grain;
/*
- * If the error is memory-controller wide, there's no need to
- * seek for the affected DIMMs because the whole
- * channel/memory controller/... may be affected.
- * Also, don't show errors for empty DIMM slots.
+ * If the error is memory-controller wide, there's no
+ * need to seek for the affected DIMMs because the
+ * whole channel/memory controller/... may be
+ * affected. Also, don't show errors for empty DIMM
+ * slots.
*/
if (!dimm->nr_pages)
continue;
@@ -1279,7 +1277,7 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
chan = -2;
}
- if (!e->enable_per_layer_report)
+ if (any_memory)
strcpy(e->label, "any memory");
else if (!*e->label)
strcpy(e->label, "unknown memory");
diff --git a/drivers/edac/ghes_edac.c b/drivers/edac/ghes_edac.c
index bef8a428c429..cb3dab56a875 100644
--- a/drivers/edac/ghes_edac.c
+++ b/drivers/edac/ghes_edac.c
@@ -355,11 +355,8 @@ void ghes_edac_report_mem_error(int sev, struct cper_sec_mem_err *mem_err)
mem_err->mem_dev_handle);
index = get_dimm_smbios_index(mci, mem_err->mem_dev_handle);
- if (index >= 0) {
+ if (index >= 0)
e->top_layer = index;
- e->enable_per_layer_report = true;
- }
-
}
if (p > e->location)
*(p - 1) = '\0';
diff --git a/include/linux/edac.h b/include/linux/edac.h
index 6703eb492cd2..815f246e0abd 100644
--- a/include/linux/edac.h
+++ b/include/linux/edac.h
@@ -454,8 +454,6 @@ struct errcount_attribute_data {
* @location: location of the error
* @label: label of the affected DIMM(s)
* @other_detail: other driver-specific detail about the error
- * @enable_per_layer_report: if false, the error affects all layers
- * (typically, a memory controller error)
*/
struct edac_raw_error_desc {
char location[LOCATION_SIZE];
@@ -472,7 +470,6 @@ struct edac_raw_error_desc {
unsigned long syndrome;
const char *msg;
const char *other_detail;
- bool enable_per_layer_report;
};
/* MEMORY controller information structure
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH v3 08/10] EDAC/mc: Pass the error descriptor to error reporting functions
2020-01-23 9:02 [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Robert Richter
` (6 preceding siblings ...)
2020-01-23 9:03 ` [PATCH v3 07/10] EDAC/mc: Remove enable_per_layer_report function arguments Robert Richter
@ 2020-01-23 9:03 ` Robert Richter
2020-01-23 9:03 ` [PATCH v3 09/10] EDAC/mc: Remove detail[] string and cleanup error string generation Robert Richter
` (2 subsequent siblings)
10 siblings, 0 replies; 17+ messages in thread
From: Robert Richter @ 2020-01-23 9:03 UTC (permalink / raw)
To: Borislav Petkov, Mauro Carvalho Chehab, Tony Luck
Cc: James Morse, Aristeu Rozanski, Robert Richter, linux-edac, linux-kernel
Most arguments of error reporting functions are already stored in
struct edac_raw_error_desc error descriptor. Pass the error descriptor
to the functions and reduce the functions' arg list.
Signed-off-by: Robert Richter <rrichter@marvell.com>
Acked-by: Aristeu Rozanski <aris@redhat.com>
---
drivers/edac/edac_mc.c | 100 +++++++++++++++++------------------------
1 file changed, 42 insertions(+), 58 deletions(-)
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index 35e427f89949..c71af845ef16 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -947,16 +947,16 @@ const char *edac_layer_name[] = {
};
EXPORT_SYMBOL_GPL(edac_layer_name);
-static void edac_inc_ce_error(struct mem_ctl_info *mci,
- const int pos[EDAC_MAX_LAYERS],
- const u16 count)
+static void edac_inc_ce_error(struct edac_raw_error_desc *e)
{
+ struct mem_ctl_info *mci = error_desc_to_mci(e);
+ int pos[EDAC_MAX_LAYERS] = { e->top_layer, e->mid_layer, e->low_layer };
int i, index = 0;
- mci->ce_mc += count;
+ mci->ce_mc += e->error_count;
if (pos[0] < 0) {
- mci->ce_noinfo_count += count;
+ mci->ce_noinfo_count += e->error_count;
return;
}
@@ -964,23 +964,23 @@ static void edac_inc_ce_error(struct mem_ctl_info *mci,
if (pos[i] < 0)
break;
index += pos[i];
- mci->ce_per_layer[i][index] += count;
+ mci->ce_per_layer[i][index] += e->error_count;
if (i < mci->n_layers - 1)
index *= mci->layers[i + 1].size;
}
}
-static void edac_inc_ue_error(struct mem_ctl_info *mci,
- const int pos[EDAC_MAX_LAYERS],
- const u16 count)
+static void edac_inc_ue_error(struct edac_raw_error_desc *e)
{
+ struct mem_ctl_info *mci = error_desc_to_mci(e);
+ int pos[EDAC_MAX_LAYERS] = { e->top_layer, e->mid_layer, e->low_layer };
int i, index = 0;
- mci->ue_mc += count;
+ mci->ue_mc += e->error_count;
if (pos[0] < 0) {
- mci->ue_noinfo_count += count;
+ mci->ue_noinfo_count += e->error_count;
return;
}
@@ -988,44 +988,37 @@ static void edac_inc_ue_error(struct mem_ctl_info *mci,
if (pos[i] < 0)
break;
index += pos[i];
- mci->ue_per_layer[i][index] += count;
+ mci->ue_per_layer[i][index] += e->error_count;
if (i < mci->n_layers - 1)
index *= mci->layers[i + 1].size;
}
}
-static void edac_ce_error(struct mem_ctl_info *mci,
- const u16 error_count,
- const int pos[EDAC_MAX_LAYERS],
- const char *msg,
- const char *location,
- const char *label,
- const char *detail,
- const char *other_detail,
- const unsigned long page_frame_number,
- const unsigned long offset_in_page,
- long grain)
+static void edac_ce_error(struct edac_raw_error_desc *e,
+ const char *detail)
{
+ struct mem_ctl_info *mci = error_desc_to_mci(e);
unsigned long remapped_page;
char *msg_aux = "";
- if (*msg)
+ if (*e->msg)
msg_aux = " ";
if (edac_mc_get_log_ce()) {
- if (other_detail && *other_detail)
+ if (e->other_detail && *e->other_detail)
edac_mc_printk(mci, KERN_WARNING,
"%d CE %s%son %s (%s %s - %s)\n",
- error_count, msg, msg_aux, label,
- location, detail, other_detail);
+ e->error_count, e->msg, msg_aux, e->label,
+ e->location, detail, e->other_detail);
else
edac_mc_printk(mci, KERN_WARNING,
"%d CE %s%son %s (%s %s)\n",
- error_count, msg, msg_aux, label,
- location, detail);
+ e->error_count, e->msg, msg_aux, e->label,
+ e->location, detail);
}
- edac_inc_ce_error(mci, pos, error_count);
+
+ edac_inc_ce_error(e);
if (mci->scrub_mode == SCRUB_SW_SRC) {
/*
@@ -1040,51 +1033,46 @@ static void edac_ce_error(struct mem_ctl_info *mci,
* be scrubbed.
*/
remapped_page = mci->ctl_page_to_phys ?
- mci->ctl_page_to_phys(mci, page_frame_number) :
- page_frame_number;
+ mci->ctl_page_to_phys(mci, e->page_frame_number) :
+ e->page_frame_number;
- edac_mc_scrub_block(remapped_page,
- offset_in_page, grain);
+ edac_mc_scrub_block(remapped_page, e->offset_in_page, e->grain);
}
}
-static void edac_ue_error(struct mem_ctl_info *mci,
- const u16 error_count,
- const int pos[EDAC_MAX_LAYERS],
- const char *msg,
- const char *location,
- const char *label,
- const char *detail,
- const char *other_detail)
+static void edac_ue_error(struct edac_raw_error_desc *e,
+ const char *detail)
{
+ struct mem_ctl_info *mci = error_desc_to_mci(e);
char *msg_aux = "";
- if (*msg)
+ if (*e->msg)
msg_aux = " ";
if (edac_mc_get_log_ue()) {
- if (other_detail && *other_detail)
+ if (e->other_detail && *e->other_detail)
edac_mc_printk(mci, KERN_WARNING,
"%d UE %s%son %s (%s %s - %s)\n",
- error_count, msg, msg_aux, label,
- location, detail, other_detail);
+ e->error_count, e->msg, msg_aux, e->label,
+ e->location, detail, e->other_detail);
else
edac_mc_printk(mci, KERN_WARNING,
"%d UE %s%son %s (%s %s)\n",
- error_count, msg, msg_aux, label,
- location, detail);
+ e->error_count, e->msg, msg_aux, e->label,
+ e->location, detail);
}
if (edac_mc_get_panic_on_ue()) {
- if (other_detail && *other_detail)
+ if (e->other_detail && *e->other_detail)
panic("UE %s%son %s (%s%s - %s)\n",
- msg, msg_aux, label, location, detail, other_detail);
+ e->msg, msg_aux, e->label, e->location, detail,
+ e->other_detail);
else
panic("UE %s%son %s (%s%s)\n",
- msg, msg_aux, label, location, detail);
+ e->msg, msg_aux, e->label, e->location, detail);
}
- edac_inc_ue_error(mci, pos, error_count);
+ edac_inc_ue_error(e);
}
static void edac_inc_csrow(struct edac_raw_error_desc *e, int row, int chan)
@@ -1111,7 +1099,6 @@ void edac_raw_mc_handle_error(struct edac_raw_error_desc *e)
{
struct mem_ctl_info *mci = error_desc_to_mci(e);
char detail[80];
- int pos[EDAC_MAX_LAYERS] = { e->top_layer, e->mid_layer, e->low_layer };
u8 grain_bits;
/* Sanity-check driver-supplied grain value. */
@@ -1134,16 +1121,13 @@ void edac_raw_mc_handle_error(struct edac_raw_error_desc *e)
"page:0x%lx offset:0x%lx grain:%ld syndrome:0x%lx",
e->page_frame_number, e->offset_in_page,
e->grain, e->syndrome);
- edac_ce_error(mci, e->error_count, pos, e->msg, e->location,
- e->label, detail, e->other_detail,
- e->page_frame_number, e->offset_in_page, e->grain);
+ edac_ce_error(e, detail);
} else {
snprintf(detail, sizeof(detail),
"page:0x%lx offset:0x%lx grain:%ld",
e->page_frame_number, e->offset_in_page, e->grain);
- edac_ue_error(mci, e->error_count, pos, e->msg, e->location,
- e->label, detail, e->other_detail);
+ edac_ue_error(e, detail);
}
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH v3 09/10] EDAC/mc: Remove detail[] string and cleanup error string generation
2020-01-23 9:02 [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Robert Richter
` (7 preceding siblings ...)
2020-01-23 9:03 ` [PATCH v3 08/10] EDAC/mc: Pass the error descriptor to error reporting functions Robert Richter
@ 2020-01-23 9:03 ` Robert Richter
2020-01-23 9:03 ` [PATCH v3 10/10] EDAC/mc: Remove per layer counters Robert Richter
2020-02-17 15:49 ` [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Borislav Petkov
10 siblings, 0 replies; 17+ messages in thread
From: Robert Richter @ 2020-01-23 9:03 UTC (permalink / raw)
To: Borislav Petkov, Mauro Carvalho Chehab, Tony Luck
Cc: James Morse, Aristeu Rozanski, Robert Richter, linux-edac, linux-kernel
The error descriptor is passed to the error reporting functions, so
the error details can be directly generated there. Move string
generation from edac_raw_mc_handle_error() to edac_ce_error() and
edac_ue_error(). The intermediate detail[] string can be removed then.
Also, cleanup the string generation by switching to a single variant
only using the ternary operator.
Signed-off-by: Robert Richter <rrichter@marvell.com>
Acked-by: Aristeu Rozanski <aris@redhat.com>
---
drivers/edac/edac_mc.c | 83 +++++++++++++-----------------------------
1 file changed, 25 insertions(+), 58 deletions(-)
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index c71af845ef16..60639def8697 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -995,27 +995,18 @@ static void edac_inc_ue_error(struct edac_raw_error_desc *e)
}
}
-static void edac_ce_error(struct edac_raw_error_desc *e,
- const char *detail)
+static void edac_ce_error(struct edac_raw_error_desc *e)
{
struct mem_ctl_info *mci = error_desc_to_mci(e);
unsigned long remapped_page;
- char *msg_aux = "";
-
- if (*e->msg)
- msg_aux = " ";
if (edac_mc_get_log_ce()) {
- if (e->other_detail && *e->other_detail)
- edac_mc_printk(mci, KERN_WARNING,
- "%d CE %s%son %s (%s %s - %s)\n",
- e->error_count, e->msg, msg_aux, e->label,
- e->location, detail, e->other_detail);
- else
- edac_mc_printk(mci, KERN_WARNING,
- "%d CE %s%son %s (%s %s)\n",
- e->error_count, e->msg, msg_aux, e->label,
- e->location, detail);
+ edac_mc_printk(mci, KERN_WARNING,
+ "%d CE %s%son %s (%s page:0x%lx offset:0x%lx grain:%ld syndrome:0x%lx%s%s)\n",
+ e->error_count, e->msg, *e->msg ? " " : "", e->label,
+ e->location, e->page_frame_number, e->offset_in_page,
+ e->grain, e->syndrome, *e->other_detail ? " - " : "",
+ e->other_detail);
}
edac_inc_ce_error(e);
@@ -1040,36 +1031,24 @@ static void edac_ce_error(struct edac_raw_error_desc *e,
}
}
-static void edac_ue_error(struct edac_raw_error_desc *e,
- const char *detail)
+static void edac_ue_error(struct edac_raw_error_desc *e)
{
struct mem_ctl_info *mci = error_desc_to_mci(e);
- char *msg_aux = "";
-
- if (*e->msg)
- msg_aux = " ";
if (edac_mc_get_log_ue()) {
- if (e->other_detail && *e->other_detail)
- edac_mc_printk(mci, KERN_WARNING,
- "%d UE %s%son %s (%s %s - %s)\n",
- e->error_count, e->msg, msg_aux, e->label,
- e->location, detail, e->other_detail);
- else
- edac_mc_printk(mci, KERN_WARNING,
- "%d UE %s%son %s (%s %s)\n",
- e->error_count, e->msg, msg_aux, e->label,
- e->location, detail);
+ edac_mc_printk(mci, KERN_WARNING,
+ "%d UE %s%son %s (%s page:0x%lx offset:0x%lx grain:%ld%s%s)\n",
+ e->error_count, e->msg, *e->msg ? " " : "", e->label,
+ e->location, e->page_frame_number, e->offset_in_page,
+ e->grain, *e->other_detail ? " - " : "",
+ e->other_detail);
}
if (edac_mc_get_panic_on_ue()) {
- if (e->other_detail && *e->other_detail)
- panic("UE %s%son %s (%s%s - %s)\n",
- e->msg, msg_aux, e->label, e->location, detail,
- e->other_detail);
- else
- panic("UE %s%son %s (%s%s)\n",
- e->msg, msg_aux, e->label, e->location, detail);
+ panic("UE %s%son %s (%s page:0x%lx offset:0x%lx grain:%ld%s%s)\n",
+ e->msg, *e->msg ? " " : "", e->label, e->location,
+ e->page_frame_number, e->offset_in_page, e->grain,
+ *e->other_detail ? " - " : "", e->other_detail);
}
edac_inc_ue_error(e);
@@ -1098,7 +1077,6 @@ static void edac_inc_csrow(struct edac_raw_error_desc *e, int row, int chan)
void edac_raw_mc_handle_error(struct edac_raw_error_desc *e)
{
struct mem_ctl_info *mci = error_desc_to_mci(e);
- char detail[80];
u8 grain_bits;
/* Sanity-check driver-supplied grain value. */
@@ -1115,22 +1093,10 @@ void edac_raw_mc_handle_error(struct edac_raw_error_desc *e)
(e->page_frame_number << PAGE_SHIFT) | e->offset_in_page,
grain_bits, e->syndrome, e->other_detail);
- /* Memory type dependent details about the error */
- if (e->type == HW_EVENT_ERR_CORRECTED) {
- snprintf(detail, sizeof(detail),
- "page:0x%lx offset:0x%lx grain:%ld syndrome:0x%lx",
- e->page_frame_number, e->offset_in_page,
- e->grain, e->syndrome);
- edac_ce_error(e, detail);
- } else {
- snprintf(detail, sizeof(detail),
- "page:0x%lx offset:0x%lx grain:%ld",
- e->page_frame_number, e->offset_in_page, e->grain);
-
- edac_ue_error(e, detail);
- }
-
-
+ if (e->type == HW_EVENT_ERR_CORRECTED)
+ edac_ce_error(e);
+ else
+ edac_ue_error(e);
}
EXPORT_SYMBOL_GPL(edac_raw_mc_handle_error);
@@ -1166,8 +1132,9 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
e->page_frame_number = page_frame_number;
e->offset_in_page = offset_in_page;
e->syndrome = syndrome;
- e->msg = msg;
- e->other_detail = other_detail;
+ /* need valid strings here for both: */
+ e->msg = msg ? msg : "";
+ e->other_detail = other_detail ? other_detail : "";
/*
* Check if the event report is consistent and if the memory
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH v3 10/10] EDAC/mc: Remove per layer counters
2020-01-23 9:02 [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Robert Richter
` (8 preceding siblings ...)
2020-01-23 9:03 ` [PATCH v3 09/10] EDAC/mc: Remove detail[] string and cleanup error string generation Robert Richter
@ 2020-01-23 9:03 ` Robert Richter
2020-02-17 15:49 ` [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Borislav Petkov
10 siblings, 0 replies; 17+ messages in thread
From: Robert Richter @ 2020-01-23 9:03 UTC (permalink / raw)
To: Borislav Petkov, Mauro Carvalho Chehab, Tony Luck
Cc: James Morse, Aristeu Rozanski, Robert Richter, linux-edac, linux-kernel
Looking at how mci->{ue,ce}_per_layer[EDAC_MAX_LAYERS] is used, it
turns out that only the leaves in the memory hierarchy are consumed
(in sysfs), but not the intermediate layers, e.g.:
count = dimm->mci->ce_per_layer[dimm->mci->n_layers-1][dimm->idx];
These unused counters only add complexity, remove them. The error
counter values are directly stored in struct dimm_info now.
Signed-off-by: Robert Richter <rrichter@marvell.com>
Acked-by: Aristeu Rozanski <aris@redhat.com>
---
drivers/edac/edac_mc.c | 65 +++++++++---------------------------
drivers/edac/edac_mc_sysfs.c | 20 +++++------
include/linux/edac.h | 4 ++-
3 files changed, 26 insertions(+), 63 deletions(-)
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index 60639def8697..fbd9faa5c0f9 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -451,11 +451,9 @@ struct mem_ctl_info *edac_mc_alloc(unsigned int mc_num,
{
struct mem_ctl_info *mci;
struct edac_mc_layer *layer;
- u32 *ce_per_layer[EDAC_MAX_LAYERS], *ue_per_layer[EDAC_MAX_LAYERS];
- unsigned int idx, size, tot_dimms = 1, count = 1;
- unsigned int tot_csrows = 1, tot_channels = 1, tot_errcount = 0;
+ unsigned int idx, size, tot_dimms = 1;
+ unsigned int tot_csrows = 1, tot_channels = 1;
void *pvt, *ptr = NULL;
- int i;
bool per_rank = false;
if (WARN_ON(n_layers > EDAC_MAX_LAYERS || n_layers == 0))
@@ -482,19 +480,10 @@ struct mem_ctl_info *edac_mc_alloc(unsigned int mc_num,
* stringent as what the compiler would provide if we could simply
* hardcode everything into a single struct.
*/
- mci = edac_align_ptr(&ptr, sizeof(*mci), 1);
- layer = edac_align_ptr(&ptr, sizeof(*layer), n_layers);
- for (i = 0; i < n_layers; i++) {
- count *= layers[i].size;
- edac_dbg(4, "errcount layer %d size %d\n", i, count);
- ce_per_layer[i] = edac_align_ptr(&ptr, sizeof(u32), count);
- ue_per_layer[i] = edac_align_ptr(&ptr, sizeof(u32), count);
- tot_errcount += 2 * count;
- }
-
- edac_dbg(4, "allocating %d error counters\n", tot_errcount);
- pvt = edac_align_ptr(&ptr, sz_pvt, 1);
- size = ((unsigned long)pvt) + sz_pvt;
+ mci = edac_align_ptr(&ptr, sizeof(*mci), 1);
+ layer = edac_align_ptr(&ptr, sizeof(*layer), n_layers);
+ pvt = edac_align_ptr(&ptr, sz_pvt, 1);
+ size = ((unsigned long)pvt) + sz_pvt;
edac_dbg(1, "allocating %u bytes for mci data (%d %s, %d csrows/channels)\n",
size,
@@ -513,10 +502,6 @@ struct mem_ctl_info *edac_mc_alloc(unsigned int mc_num,
* rather than an imaginary chunk of memory located at address 0.
*/
layer = (struct edac_mc_layer *)(((char *)mci) + ((unsigned long)layer));
- for (i = 0; i < n_layers; i++) {
- mci->ce_per_layer[i] = (u32 *)((char *)mci + ((unsigned long)ce_per_layer[i]));
- mci->ue_per_layer[i] = (u32 *)((char *)mci + ((unsigned long)ue_per_layer[i]));
- }
pvt = sz_pvt ? (((char *)mci) + ((unsigned long)pvt)) : NULL;
/* setup index and various internal pointers */
@@ -951,48 +936,28 @@ static void edac_inc_ce_error(struct edac_raw_error_desc *e)
{
struct mem_ctl_info *mci = error_desc_to_mci(e);
int pos[EDAC_MAX_LAYERS] = { e->top_layer, e->mid_layer, e->low_layer };
- int i, index = 0;
+ struct dimm_info *dimm = edac_get_dimm(mci, pos[0], pos[1], pos[2]);
mci->ce_mc += e->error_count;
- if (pos[0] < 0) {
+ if (dimm)
+ dimm->ce_count += e->error_count;
+ else
mci->ce_noinfo_count += e->error_count;
- return;
- }
-
- for (i = 0; i < mci->n_layers; i++) {
- if (pos[i] < 0)
- break;
- index += pos[i];
- mci->ce_per_layer[i][index] += e->error_count;
-
- if (i < mci->n_layers - 1)
- index *= mci->layers[i + 1].size;
- }
}
static void edac_inc_ue_error(struct edac_raw_error_desc *e)
{
struct mem_ctl_info *mci = error_desc_to_mci(e);
int pos[EDAC_MAX_LAYERS] = { e->top_layer, e->mid_layer, e->low_layer };
- int i, index = 0;
+ struct dimm_info *dimm = edac_get_dimm(mci, pos[0], pos[1], pos[2]);
mci->ue_mc += e->error_count;
- if (pos[0] < 0) {
+ if (dimm)
+ dimm->ue_count += e->error_count;
+ else
mci->ue_noinfo_count += e->error_count;
- return;
- }
-
- for (i = 0; i < mci->n_layers; i++) {
- if (pos[i] < 0)
- break;
- index += pos[i];
- mci->ue_per_layer[i][index] += e->error_count;
-
- if (i < mci->n_layers - 1)
- index *= mci->layers[i + 1].size;
- }
}
static void edac_ce_error(struct edac_raw_error_desc *e)
@@ -1139,7 +1104,7 @@ void edac_mc_handle_error(const enum hw_event_mc_err_type type,
/*
* Check if the event report is consistent and if the memory
* location is known. If it is known, the DIMM(s) label info
- * will be filled and the per-layer error counters will be
+ * will be filled and the DIMM's error counters will be
* incremented.
*/
for (i = 0; i < mci->n_layers; i++) {
diff --git a/drivers/edac/edac_mc_sysfs.c b/drivers/edac/edac_mc_sysfs.c
index 408bace699dc..20657530a108 100644
--- a/drivers/edac/edac_mc_sysfs.c
+++ b/drivers/edac/edac_mc_sysfs.c
@@ -551,10 +551,8 @@ static ssize_t dimmdev_ce_count_show(struct device *dev,
char *data)
{
struct dimm_info *dimm = to_dimm(dev);
- u32 count;
- count = dimm->mci->ce_per_layer[dimm->mci->n_layers-1][dimm->idx];
- return sprintf(data, "%u\n", count);
+ return sprintf(data, "%u\n", dimm->ce_count);
}
static ssize_t dimmdev_ue_count_show(struct device *dev,
@@ -562,10 +560,8 @@ static ssize_t dimmdev_ue_count_show(struct device *dev,
char *data)
{
struct dimm_info *dimm = to_dimm(dev);
- u32 count;
- count = dimm->mci->ue_per_layer[dimm->mci->n_layers-1][dimm->idx];
- return sprintf(data, "%u\n", count);
+ return sprintf(data, "%u\n", dimm->ue_count);
}
/* dimm/rank attribute files */
@@ -661,7 +657,9 @@ static ssize_t mci_reset_counters_store(struct device *dev,
const char *data, size_t count)
{
struct mem_ctl_info *mci = to_mci(dev);
- int cnt, row, chan, i;
+ struct dimm_info *dimm;
+ int row, chan;
+
mci->ue_mc = 0;
mci->ce_mc = 0;
mci->ue_noinfo_count = 0;
@@ -677,11 +675,9 @@ static ssize_t mci_reset_counters_store(struct device *dev,
ri->channels[chan]->ce_count = 0;
}
- cnt = 1;
- for (i = 0; i < mci->n_layers; i++) {
- cnt *= mci->layers[i].size;
- memset(mci->ce_per_layer[i], 0, cnt * sizeof(u32));
- memset(mci->ue_per_layer[i], 0, cnt * sizeof(u32));
+ mci_for_each_dimm(mci, dimm) {
+ dimm->ue_count = 0;
+ dimm->ce_count = 0;
}
mci->start_time = jiffies;
diff --git a/include/linux/edac.h b/include/linux/edac.h
index 815f246e0abd..0f20b986b0ab 100644
--- a/include/linux/edac.h
+++ b/include/linux/edac.h
@@ -383,6 +383,9 @@ struct dimm_info {
unsigned int csrow, cschannel; /* Points to the old API data */
u16 smbios_handle; /* Handle for SMBIOS type 17 */
+
+ u32 ce_count;
+ u32 ue_count;
};
/**
@@ -559,7 +562,6 @@ struct mem_ctl_info {
*/
u32 ce_noinfo_count, ue_noinfo_count;
u32 ue_mc, ce_mc;
- u32 *ce_per_layer[EDAC_MAX_LAYERS], *ue_per_layer[EDAC_MAX_LAYERS];
struct completion complete;
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* Re: [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two
2020-01-23 9:02 [PATCH v3 00/10] EDAC: Rework core and ghes drivers, part two Robert Richter
` (9 preceding siblings ...)
2020-01-23 9:03 ` [PATCH v3 10/10] EDAC/mc: Remove per layer counters Robert Richter
@ 2020-02-17 15:49 ` Borislav Petkov
10 siblings, 0 replies; 17+ messages in thread
From: Borislav Petkov @ 2020-02-17 15:49 UTC (permalink / raw)
To: Robert Richter
Cc: Mauro Carvalho Chehab, Tony Luck, James Morse, Aristeu Rozanski,
linux-edac, linux-kernel
On Thu, Jan 23, 2020 at 09:02:47AM +0000, Robert Richter wrote:
> This patch set is part two of a rework of the ghes_edac and edac_mc
> driver. It addresses issues found during code review and while working
> with the code. Part one has been included to v5.5, see:
>
> https://lore.kernel.org/patchwork/cover/1093488/
>
> The changes of this series include:
>
> * add helper functions and factor out code (#1, #2, #5)
>
> * improve function interfaces and data structures to decrease
> complexity such as number of function arguments, unused data, etc.
> (#3, #4, #7, #8, #9, #10),
>
> * minor functional fixes (#6)
>
> * improve code readability (#9)
>
> V3:
> * rebased onto edac-for-next + "EDAC/mc: Fix use-after-free and
> memleaks during device removal", no code changes:
> 7e5d6cf35329 ("EDAC/amd64: Do not warn when removing instances")
> https://lore.kernel.org/patchwork/patch/1169444/
> * added Aristeu's ACKs
>
> V2:
> * fixed documentation issue in #3 found by "kbuild test robot
> <lkp@intel.com>"
>
> Changes compared to part one:
> * rebased onto 5781823fd0d3 ("EDAC/altera: Use the Altera System
> Manager driver")
> * reworded patch subjects
> * reordered patches
> * collected Mauro's Reviewed-by-tags (note: I kept them though there
> has been small conflicts but dropped it when reworked)
> * dropped: "EDAC/mc: Rework edac_raw_mc_handle_error() to use struct
> dimm_info"
> * split "EDAC/mc: Remove per layer counters" into smaller changes
> * added:
> "EDAC/mc: Report "unknown memory" on too many DIMM labels found"
> "EDAC/mc: Remove enable_per_layer_report function arguments"
> "EDAC/mc: Pass the error descriptor to error reporting functions"
> "EDAC/mc: Remove detail[] string and cleanup error string
> generation"
> * moved to the end:
> "EDAC/mc: Remove per layer counters"
>
>
> Robert Richter (10):
> EDAC/mc: Split edac_mc_alloc() into smaller functions
> EDAC/mc: Reorder functions edac_mc_alloc*()
> EDAC: Store error type in struct edac_raw_error_desc
> EDAC/mc: Determine mci pointer from the error descriptor
> EDAC/mc: Create new function edac_inc_csrow()
> EDAC/mc: Report "unknown memory" on too many DIMM labels found
> EDAC/mc: Remove enable_per_layer_report function arguments
> EDAC/mc: Pass the error descriptor to error reporting functions
> EDAC/mc: Remove detail[] string and cleanup error string generation
> EDAC/mc: Remove per layer counters
>
> drivers/edac/edac_mc.c | 502 ++++++++++++++++-------------------
> drivers/edac/edac_mc.h | 6 +-
> drivers/edac/edac_mc_sysfs.c | 20 +-
> drivers/edac/ghes_edac.c | 16 +-
> include/linux/edac.h | 9 +-
> 5 files changed, 252 insertions(+), 301 deletions(-)
Ok, all queued and will appear in linux-next soon. Let's see what falls
out.
Thx for the nice cleanup!
--
Regards/Gruss,
Boris.
https://people.kernel.org/tglx/notes-about-netiquette
^ permalink raw reply [flat|nested] 17+ messages in thread