On 25/11/2022 14:26, hejunhao wrote:
>
> On 2022/11/24 21:45, Suzuki K Poulose wrote:
>> On 24/11/2022 13:33, hejunhao wrote:
>>>
>>> On 2022/11/23 22:03, Suzuki K Poulose wrote:
>>>> On 23/11/2022 12:38, Junhao He wrote:
>>>>> From: Qi Liu <[email protected]>
>>>>>
>>>>> Add driver for UltraSoc SMB(System Memory Buffer) device.
>>>>> SMB provides a way to buffer messages from ETM, and store
>>>>> these "CPU instructions trace" in system memory.
>>>>> The SMB device is identifier as ACPI HID "HISI03A1". Device
>>>>> system memory address resources are allocated using the _CRS
>>>>> method and buffer modes is the circular buffer mode.
>>>>>
>>>>> SMB is developed by UltraSoc technology, which is acquired by
>>>>> Siemens, and we still use "UltraSoc" to name driver.
>>>>>
>>>>> Signed-off-by: Qi Liu <[email protected]>
>>>>> Signed-off-by: Junhao He <[email protected]>
>>>>> Tested-by: JunHao He <[email protected]>
>>>>> Reviewed-by: Jonathan Cameron <[email protected]>
>>>>> ---
>>>>> drivers/hwtracing/coresight/Kconfig | 12 +
>>>>> drivers/hwtracing/coresight/Makefile | 1 +
>>>>> drivers/hwtracing/coresight/ultrasoc-smb.c | 658
>>>>> +++++++++++++++++++++
>>>>> drivers/hwtracing/coresight/ultrasoc-smb.h | 129 ++++
>>>>> 4 files changed, 800 insertions(+)
>>>>> create mode 100644 drivers/hwtracing/coresight/ultrasoc-smb.c
>>>>> create mode 100644 drivers/hwtracing/coresight/ultrasoc-smb.h
>>>>>
>>>>
>>>>> +static void smb_sync_perf_buffer(struct smb_drv_data *drvdata,
>>>>> + struct cs_buffers *buf,
>>>>> + unsigned long head,
>>>>> + unsigned long data_size)
>>>>> +{
>>>>> + struct smb_data_buffer *sdb = &drvdata->sdb;
>>>>> + char **dst_pages = (char **)buf->data_pages;
>>>>> + unsigned long to_copy;
>>>>> + long pg_idx, pg_offset;
>>>>> +
>>>>> + pg_idx = head >> PAGE_SHIFT;
>>>>> + pg_offset = head & (PAGE_SIZE - 1);
>>>>> +
>>>>> + while (data_size) {
>>>>> + unsigned long pg_space = PAGE_SIZE - pg_offset;
>>>>> +
>>>>> + /* Copy parts of trace data when read pointer wrap around */
>>>>> + if (sdb->rd_offset + pg_space > sdb->buf_size)
>>>>> + to_copy = sdb->buf_size - sdb->rd_offset;
>>>>> + else
>>>>> + to_copy = min(data_size, pg_space);
>>>>> +
>>>>> + memcpy(dst_pages[pg_idx] + pg_offset,
>>>>> + sdb->buf_base + sdb->rd_offset, to_copy);
>>>>> +
>>>>> + pg_offset += to_copy;
>>>>> + if (pg_offset >= PAGE_SIZE) {
>>>>> + pg_offset = 0;
>>>>> + pg_idx++;
>>>>> + pg_idx %= buf->nr_pages;
>>>>> + }
>>>>> + data_size -= to_copy;
>>>>> + sdb->rd_offset += to_copy;
>>>>> + sdb->rd_offset %= sdb->buf_size;
>>>>> + }
>>>>> +
>>>>> + sdb->data_size = 0;
>>>>
>>>>
>>>> --8>-- cut here --<8--
>>>>
>>>>> + writel(sdb->start_addr + sdb->rd_offset,
>>>>> + drvdata->base + SMB_LB_RD_ADDR_REG);
>>>>> +
>>>>> + /*
>>>>> + * Data remained in link cannot be purged when SMB is full, so
>>>>> + * synchronize the read pointer to write pointer, to make sure
>>>>> + * these remained data won't influence next trace.
>>>>> + */
>>>>> + if (sdb->full) {
>>>>> + smb_purge_data(drvdata);
>>>>> + writel(readl(drvdata->base + SMB_LB_WR_ADDR_REG),
>>>>> + drvdata->base + SMB_LB_RD_ADDR_REG);
>>>>> + }
>>>>
>>>> --<8-- end here --8>--
>>>>
>>>> As pointed out in the last review, we must do this step
>>>> everytime for perf mode irrespective of whether the buffer
>>>> was "FULL" or not.
>>>>
>>>> i.e, the above block should simply be:
>>>>
>>>> if (sdb->full)
>>>> smb_purge_data(drvdata);
>>>>
>>>> /*
>>>> * The uncollected Data must be discarded for perf,
>>>> * as it cannot be clubbed with next schedule. We
>>>> * any way TRUNCATE the buffer in this case.
>>>> */
>>>> writel(readl(drvdata->base + SMB_LB_WR_ADDR_REG),
>>>> drvdata->base + SMB_LB_RD_ADDR_REG);
>>>>
>>>> Suzuki
>>>
>>> Hi Suzuki,
>>>
>>> We need to update SMB_LB_RD_ADDR_REG register first, then
>>> check the "full" flag, whether the register needs to be
>>> updated again.
>>
>> Why ? sdb->full is not updated after the write to RD_ADDR_REG.
>>
> Hi Suzuki,
>
> Maybe using the code below is more appropriate.
> i.e,
>
> writel(sdb->start_addr + sdb->rd_offset,
> drvdata->base + SMB_LB_RD_ADDR_REG);
>
> /*
> * The uncollected Data must be discarded for perf,
> * as it cannot be clubbed with next schedule. We
> * any way TRUNCATE the buffer in this case.
> */
> smb_update_data_size(drvdata);
> if (sdb->data_size)
> writel(readl(drvdata->base + SMB_LB_WR_ADDR_REG),
> drvdata->base + SMB_LB_RD_ADDR_REG);
That should work. However, my question is, we must set the
RD_ADDR_REG to the WR_ADDR_REG in all cases. So, why not :
/*
* We must flush and discard any data left in the
* buffer for a perf session, to avoid mixing
* the data from sessions.
* We TRUNCATE the buffer anyways to indicate
* that the buffer was lost.
*/
smb_purge_data();
writel(readl(drvdata->base + SMB_LB_WR_ADDR_REG),
drvdata->base + SMB_LB_RD_ADDR_REG);
and avoid the extra write ?
>>>
>>> If we don`t update the value of SMB_LB_RD_ADDR_REG register
>>> or reset buffer state, the buffer state will still be "full".
>>> The buffer has not free area,so the data will still remain
>>> in link.
>>
>> My issue here is with potentially "leaving the trace from a previous
>> session for the next session". i.e., at the end of a run, we must always
>> make sure that the buffer is left empty (unlike the sysfs mode).
>>
>> e.g.,
>>
>> perf_session_x: RUN0: RD=0x0, WR=0x5000, HANDLE_SIZE=0x3000, full=false.
>> At the end of the above routine we will have :
>> RD=0x3000, WR=0x5000,
>>
>> and if a different perf session comes in, say perf_session_y, it will
>> consume trace written by "x" at 0x3000-0x50000, right ?
>>
>> This is fine in the sysfs mode as we expect the entire sysfs mode
>> to be owned by a single session and is left to the user to split it.
>> But for perf mode we cannot do that and thus must make sure we don't
>> leak trace from one session to antoher.
>>
>> Suzuki
>>
>
> In this cace: RUN0: RD=0x0, WR=0x5000, HANDLE_SIZE=0x3000, full=false.
> We will update the "rd_offset" in smb_update_buffer() function.
> like this:
>
> ...
> if (data_size > handle->size) {
> sdb->rd_offset += data_size - handle->size;
> sdb->rd_offset %= sdb->buf_size;
> ...
> }
> ...
>
>
> So the rd_offset will advance to 0x2000 first,
> then we dump the latest trace data (0x2000 - 0x5000) from "buf_base +
> rd_offset".
Right, that makes sense.
Thanks
Suzuki