Lines Matching full:reports
75 * the constraints on HW configuration require reports to be filtered before it
114 * The OA unit writes reports of counters into a circular buffer, without
134 * feature and forwarded OA reports to userspace via perf's 'raw' sample
138 * them. For Mesa it's also convenient to be forwarded raw, periodic reports
139 * for combining with the side-band raw reports it captures using
169 * attach GPU timestamps to our OA samples. We were shoehorning OA reports
238 * Although this can be observed explicitly while copying reports to userspace
239 * by checking for a zeroed report-id field in tail reports, we want to account
243 * We workaround this issue in oa_buffer_check_unlocked() by reading the reports
259 * non-periodic reports (such as on context switch) or the OA unit may be
266 * reports to the circular OA buffer...
275 * of the 64bit timestamp bits to trigger reports from) but there's currently
278 * Since the timestamps included in OA reports are only 32bits it seems
286 /* On Gen8+ automatically triggered OA reports include a 'reason' field... */
314 * code assumes all reports have a power-of-two size and ~(size - 1) can
588 * progress by increments of 64 bytes and that reports can be up
720 * gen8_append_oa_reports - Copies all buffered OA reports into
733 * Note: reports are consumed from the head, and appended to the
827 * when GPU goes idle. In all other reports, it is set to 1.
831 * reports it is set to 0.
834 * reports. MMIO triggered reports have the markers in the
838 * buffer. Since MMIO triggered reports exist only on XEHP,
854 * Automatic reports now include a context ID so reports can be
863 * needs be forwarded bookend context-switch reports so that it
867 * automatically triggers reports when switching to a new
870 * reading ahead while parsing reports to try and minimize
871 * forwarding redundant context switch reports (i.e. between
904 * to detect unlanded reports.
948 * gen8_oa_read - copy status records then buffered OA reports
956 * initiate appending any buffered OA reports.
991 * automatically triggered reports in this condition and so we
992 * have to assume that old reports are now being trampled
1038 * gen7_append_oa_reports - Copies all buffered OA reports into
1051 * Note: reports are consumed from the head, and appended to the
1142 * reports.
1168 * gen7_oa_read - copy status records then buffered OA reports
1176 * condition) and then initiate appending any buffered OA reports.
1214 * reports it may also affect our sanity checks for invalid
1215 * reports when copying to userspace that assume new reports
1587 * On Haswell we don't do any post processing of the reports
1606 * ID in the OA reports, so squash those lower bits.
1786 * the assumption that new reports are being written to zeroed
1823 * reports we will forward to userspace while filtering for a single
1839 * the assumption that new reports are being written to zeroed
1876 * reports we will forward to userspace while filtering for a single
1892 * the assumption that new reports are being written to zeroed
2857 * We disable slice/unslice clock ratio change reports on SKL since
2858 * they are too noisy. The HW generates a lot of redundant reports
2864 * feature it's worth noting that clock ratio reports have to be
2866 * correctly block these reports.
2876 * be read back from automatically triggered reports, as part of the
2930 /* Disable clk ratio reports, like previous Gens. */
2934 * If the user didn't require OA reports, instruct
2935 * the hardware not to emit ctx switch reports.
3033 * Reset buf pointers so we don't forward reports from before now.
3037 * in error paths, and we have checks for invalid reports based
3059 * Reset buf pointers so we don't forward reports from before now.
3063 * in error paths, and we have checks for invalid reports based
3072 * field of reports
3085 * If we don't want OA reports from the OA buffer, then we don't even
3176 * Stops the OA unit from periodically writing counter reports into the
3338 * counter reports and marshal to the appropriate client
3542 * any reports really belong to the current context
3925 * filter reports based on the included context ID we can't block
5255 * _NOTES about spurious, invalid OA reports which we don't
5265 /* Since we use a DRM_NOTE for spurious reports it would be
5357 * accessible from a delta of MI_RPC reports without looking at the