From: Mathieu Poirier Date: Wed, 19 Jun 2019 17:29:08 +0000 (-0600) Subject: coresight: tmc-etf: Fix snapshot mode update function X-Git-Tag: v5.3-rc1~127^2~88 X-Git-Url: https://asedeno.scripts.mit.edu/gitweb/?a=commitdiff_plain;h=99f81eb9c51d499f65bac25597e9def22f6a32dc;p=linux.git coresight: tmc-etf: Fix snapshot mode update function When working in snapshot mode function perf_aux_output_begin() does not set the handle->size because the size is expected to be deduced by the placement of the "head" and "old" pointers in user space. As such there is no point in trying to adjust the amount of data to copy to the ring buffer. Signed-off-by: Mathieu Poirier Reviewed-by: Leo Yan Reviewed-by: Suzuki K Poulose Tested-by: Leo Yan Signed-off-by: Greg Kroah-Hartman --- diff --git a/drivers/hwtracing/coresight/coresight-tmc-etf.c b/drivers/hwtracing/coresight/coresight-tmc-etf.c index d026bd04a6af..31d41e2ad955 100644 --- a/drivers/hwtracing/coresight/coresight-tmc-etf.c +++ b/drivers/hwtracing/coresight/coresight-tmc-etf.c @@ -477,9 +477,11 @@ static unsigned long tmc_update_etf_buffer(struct coresight_device *csdev, /* * The TMC RAM buffer may be bigger than the space available in the * perf ring buffer (handle->size). If so advance the RRP so that we - * get the latest trace data. + * get the latest trace data. In snapshot mode none of that matters + * since we are expected to clobber stale data in favour of the latest + * traces. */ - if (to_read > handle->size) { + if (!buf->snapshot && to_read > handle->size) { u32 mask = 0; /*