Re: [PATCH V3 14/18] coresight: tmc: keep track of memory width

From: Suzuki K Poulose
Date: Mon Apr 25 2016 - 11:10:04 EST


On 25/04/16 15:55, Mathieu Poirier wrote:
On 25 April 2016 at 08:41, Suzuki K Poulose <Suzuki.Poulose@xxxxxxx> wrote:
On 22/04/16 18:14, Mathieu Poirier wrote:

Accessing the HW configuration register each time the memory
width is needed simply doesn't make sense. It is much more
efficient to read the value once and keep a reference for
later use.

Signed-off-by: Mathieu Poirier <mathieu.poirier@xxxxxxxxxx>
---
drivers/hwtracing/coresight/coresight-tmc-etf.c | 14 +---------
drivers/hwtracing/coresight/coresight-tmc.c | 34
+++++++++++++++++++++++++
drivers/hwtracing/coresight/coresight-tmc.h | 10 +++++---
3 files changed, 41 insertions(+), 17 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-tmc-etf.c
b/drivers/hwtracing/coresight/coresight-tmc-etf.c
index cc88c15ba45c..981c5ca75e36 100644
--- a/drivers/hwtracing/coresight/coresight-tmc-etf.c
+++ b/drivers/hwtracing/coresight/coresight-tmc-etf.c
@@ -41,25 +41,13 @@ void tmc_etb_enable_hw(struct tmc_drvdata *drvdata)

static void tmc_etb_dump_hw(struct tmc_drvdata *drvdata)
{
- enum tmc_mem_intf_width memwidth;
- u8 memwords;
char *bufp;
u32 read_data;
int i;

- memwidth = BMVAL(readl_relaxed(drvdata->base + CORESIGHT_DEVID),
8, 10);
- if (memwidth == TMC_MEM_INTF_WIDTH_32BITS)
- memwords = 1;
- else if (memwidth == TMC_MEM_INTF_WIDTH_64BITS)
- memwords = 2;
- else if (memwidth == TMC_MEM_INTF_WIDTH_128BITS)
- memwords = 4;
- else
- memwords = 8;
-
bufp = drvdata->buf;
while (1) {
- for (i = 0; i < memwords; i++) {
+ for (i = 0; i < drvdata->memwidth; i++) {
read_data = readl_relaxed(drvdata->base +
TMC_RRD);
if (read_data == 0xFFFFFFFF)
return;
diff --git a/drivers/hwtracing/coresight/coresight-tmc.c
b/drivers/hwtracing/coresight/coresight-tmc.c
index 55806352b1f1..cb030a09659d 100644
--- a/drivers/hwtracing/coresight/coresight-tmc.c
+++ b/drivers/hwtracing/coresight/coresight-tmc.c
@@ -193,6 +193,39 @@ static const struct file_operations tmc_fops = {
.llseek = no_llseek,
};

+static enum tmc_mem_intf_width tmc_get_memwidth(u32 devid)
+{
+ enum tmc_mem_intf_width memwidth;
+
+ /*
+ * Excerpt from the TRM:
+ *
+ * DEVID::MEMWIDTH[10:8]
+ * 0x2 Memory interface databus is 32 bits wide.
+ * 0x3 Memory interface databus is 64 bits wide.
+ * 0x4 Memory interface databus is 128 bits wide.
+ * 0x5 Memory interface databus is 256 bits wide.
+ */
+ switch (BMVAL(devid, 8, 10)) {
+ case 0x2:
+ memwidth = TMC_MEM_INTF_WIDTH_32BITS;
+ break;
+ case 0x3:
+ memwidth = TMC_MEM_INTF_WIDTH_64BITS;
+ break;
+ case 0x4:
+ memwidth = TMC_MEM_INTF_WIDTH_128BITS;
+ break;
+ case 0x5:
+ memwidth = TMC_MEM_INTF_WIDTH_256BITS;
+ break;
+ default:
+ memwidth = 0;
+ }
+
+ return memwidth;
+}
+
#define coresight_tmc_simple_func(name, offset) \
coresight_simple_func(struct tmc_drvdata, name, offset)

@@ -306,6 +339,7 @@ static int tmc_probe(struct amba_device *adev, const
struct amba_id *id)

devid = readl_relaxed(drvdata->base + CORESIGHT_DEVID);
drvdata->config_type = BMVAL(devid, 6, 7);
+ drvdata->memwidth = tmc_get_memwidth(devid);

if (drvdata->config_type == TMC_CONFIG_TYPE_ETR) {
if (np)
diff --git a/drivers/hwtracing/coresight/coresight-tmc.h
b/drivers/hwtracing/coresight/coresight-tmc.h
index 9b4c215d2b6b..12a097f3d06c 100644
--- a/drivers/hwtracing/coresight/coresight-tmc.h
+++ b/drivers/hwtracing/coresight/coresight-tmc.h
@@ -81,10 +81,10 @@ enum tmc_mode {
};

enum tmc_mem_intf_width {
- TMC_MEM_INTF_WIDTH_32BITS = 0x2,
- TMC_MEM_INTF_WIDTH_64BITS = 0x3,
- TMC_MEM_INTF_WIDTH_128BITS = 0x4,
- TMC_MEM_INTF_WIDTH_256BITS = 0x5,
+ TMC_MEM_INTF_WIDTH_32BITS = 1,
+ TMC_MEM_INTF_WIDTH_64BITS = 2,
+ TMC_MEM_INTF_WIDTH_128BITS = 4,
+ TMC_MEM_INTF_WIDTH_256BITS = 8,
};


I think this would cause confusion for the reader. It would be good to
leave the definitions above as before and tmc_get_memwidth() doing:

i.e,
case TMC_MEM_INTF_WIDTH_32BITS:
memwidth = 1;
break;

But we still store the memwidth in bytes.

If we proceed this way we have to do a case statement on hard coded
values (or introduce a new enumeration) in tmc_update_etf_buffer(). In

But then we are doing that already with this patch in tmc_get_memwidth already.
My point is, we expect named symbols for values defined in the standards, so that
you don't make a mistake when dealing with them. It is really difficult to track
down a bug if we do make a mistake.

my opinion the current solution scales better.

I would prefer doing a hardcoded check in tmc_update_etf_buffer() like :

+ /*
+ * The value written to RRP must be byte-address aligned to
+ * the width of the trace memory databus _and_ to a frame
+ * boundary (16 byte), whichever is the biggest. For example,
+ * for 32-bit, 64-bit and 128-bit wide trace memory, the four
+ * LSBs must be 0s. For 256-bit wide trace memory, the five
+ * LSBs must be 0s.
+ */
+ if (drvdata->memwidth == 8)
+ mask = GENMASK(31, 6);
+ else
+ mask = GENMASK(31, 5);


Suzuki