From: Suzuki.Poulose@arm•com (Suzuki K Poulose)
To: linux-arm-kernel@lists•infradead.org
Subject: [PATCH V3 14/18] coresight: tmc: keep track of memory width
Date: Mon, 25 Apr 2016 15:41:13 +0100 [thread overview]
Message-ID: <571E2C89.7010003@arm.com> (raw)
In-Reply-To: <1461345255-11758-15-git-send-email-mathieu.poirier@linaro.org>
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@linaro•org>
> ---
> 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.
Suzuki
next prev parent reply other threads:[~2016-04-25 14:41 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-04-22 17:13 [PATCH V3 00/18] coresight: tmc: make driver usable by Perf Mathieu Poirier
2016-04-22 17:13 ` [PATCH V3 01/18] coresight: tmc: modifying naming convention Mathieu Poirier
2016-04-22 17:13 ` [PATCH V3 02/18] coresight: tmc: waiting for TMCReady bit before programming Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 03/18] coresight: tmc: re-implementing tmc_read_prepare/unprepare() functions Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 04/18] coresight: tmc: clearly define number of transfers per burst Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 05/18] coresight: tmc: introducing new header file Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 06/18] coresight: tmc: cleaning up " Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 07/18] coresight: tmc: splitting driver in ETB/ETF and ETR components Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 08/18] coresight: tmc: making prepare/unprepare functions generic Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 09/18] coresight: tmc: allocating memory when needed Mathieu Poirier
2016-04-25 10:20 ` Suzuki K Poulose
2016-04-25 14:24 ` Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 10/18] coresight: tmc: getting the right read_count on tmc_open() Mathieu Poirier
2016-04-25 10:47 ` Suzuki K Poulose
2016-04-25 14:25 ` Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 11/18] coresight: tmc: adding mode of operation for link/sinks Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 12/18] coresight: tmc: dump system memory content only when needed Mathieu Poirier
2016-04-25 11:16 ` Suzuki K Poulose
2016-04-25 14:38 ` Mathieu Poirier
2016-04-25 14:49 ` Suzuki K Poulose
2016-04-22 17:14 ` [PATCH V3 13/18] coresight: tmc: make sysFS and Perf mode mutually exclusive Mathieu Poirier
2016-04-25 14:32 ` Suzuki K Poulose
2016-04-25 14:48 ` Mathieu Poirier
2016-04-25 14:52 ` Suzuki K Poulose
2016-04-25 15:05 ` Mathieu Poirier
2016-04-25 15:11 ` Suzuki K Poulose
2016-04-25 15:18 ` Mathieu Poirier
2016-04-26 9:23 ` Suzuki K Poulose
2016-04-22 17:14 ` [PATCH V3 14/18] coresight: tmc: keep track of memory width Mathieu Poirier
2016-04-25 14:41 ` Suzuki K Poulose [this message]
2016-04-25 14:55 ` Mathieu Poirier
2016-04-25 15:09 ` Suzuki K Poulose
2016-04-25 15:25 ` Mathieu Poirier
2016-04-25 15:28 ` Suzuki K Poulose
2016-04-22 17:14 ` [PATCH V3 15/18] coresight: moving struct cs_buffers to header file Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 16/18] coresight: tmc: implementing TMC-ETF AUX space API Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 17/18] coresight: tmc: implementing TMC-ETR " Mathieu Poirier
2016-04-22 17:14 ` [PATCH V3 18/18] coresight: configuring ETF in FIFO mode when acting as link Mathieu Poirier
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=571E2C89.7010003@arm.com \
--to=suzuki.poulose@arm$(echo .)com \
--cc=linux-arm-kernel@lists$(echo .)infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox