From: "Jouni Högander" <jouni.hogander@intel.com>
To: igt-dev@lists.freedesktop.org
Cc: ramanaidu.naladala@intel.com,
"Jouni Högander" <jouni.hogander@intel.com>
Subject: [PATCH i-g-t v2 1/5] lib/igt_psr: Add mechanism to check sink status as well
Date: Wed, 11 Sep 2024 08:49:54 +0300 [thread overview]
Message-ID: <20240911054958.1023925-2-jouni.hogander@intel.com> (raw)
In-Reply-To: <20240911054958.1023925-1-jouni.hogander@intel.com>
We have seen passing PSR testcases even though panel is not even aware of
PSR being used. This can happen because we currently not checking sink PSR
statuses at all. Also sink might have detected errors but our testcase
currently don't care about that.
Help the gap described above by adding new interface to check sink error
statuses. Also add sink status check to be part of psr_is_active check.
Signed-off-by: Jouni Högander <jouni.hogander@intel.com>
Reviewed-by: Naladala Ramanaidu <ramanaidu.naladala@intel.com>
---
lib/igt_psr.c | 52 ++++++++++++++++++++++++++++++++++++++++++++++++++-
lib/igt_psr.h | 1 +
2 files changed, 52 insertions(+), 1 deletion(-)
diff --git a/lib/igt_psr.c b/lib/igt_psr.c
index e3e7577eb..47517b5dc 100644
--- a/lib/igt_psr.c
+++ b/lib/igt_psr.c
@@ -52,6 +52,19 @@ bool selective_fetch_check(int debugfs_fd, igt_output_t *output)
return strstr(buf, "PSR2 selective fetch: enabled");
}
+static bool psr_active_sink_check(int debugfs_fd, igt_output_t *output)
+{
+ char debugfs_file[128] = {0};
+ char buf[PSR_STATUS_MAX_LEN];
+ int ret;
+
+ sprintf(debugfs_file, "%s/i915_psr_sink_status", output->name);
+ ret = igt_debugfs_simple_read(debugfs_fd, debugfs_file, buf,
+ sizeof(buf));
+ igt_assert_f(ret >= 1, "Failed to read sink status\n");
+
+ return strstr(buf, "0x2 [active, display from RFB]");
+}
/*
* Checks if Early Transport is enabled in PSR status by reading the debugfs.
@@ -72,6 +85,7 @@ static bool psr_active_check(int debugfs_fd, enum psr_mode mode, igt_output_t *o
char buf[PSR_STATUS_MAX_LEN];
drmModeConnector *c;
const char *state;
+ bool active;
int ret;
if (mode == PR_MODE || mode == PR_MODE_SEL_FETCH) {
@@ -100,7 +114,11 @@ static bool psr_active_check(int debugfs_fd, enum psr_mode mode, igt_output_t *o
igt_skip_on(strstr(buf, "PSR sink not reliable: yes"));
- return strstr(buf, state);
+ active = strstr(buf, state);
+ if (active && output)
+ active = psr_active_sink_check(debugfs_fd, output);
+
+ return active;
}
/*
@@ -297,6 +315,38 @@ bool psr_sink_support(int device, int debugfs_fd, enum psr_mode mode, igt_output
}
}
+/**
+ * psr_sink_error_check
+ * Check and assert on PSR errors detected by panel
+ *
+ * Returns:
+ * None
+ */
+void psr_sink_error_check(int debugfs_fd, enum psr_mode mode, igt_output_t *output)
+{
+ char *line;
+ char debugfs_file[128] = {0};
+ char buf[PSR_STATUS_MAX_LEN];
+ int ret;
+
+ sprintf(debugfs_file, "%s/i915_psr_sink_status", output->name);
+ ret = igt_debugfs_simple_read(debugfs_fd, debugfs_file, buf,
+ sizeof(buf));
+ igt_assert_f(ret >= 1, "Failed to read sink status\n");
+
+ line = strstr(buf, "error status: 0x0");
+
+ /*
+ * On certain PSR1 panels we are seeing "PSR VSC SDP
+ * uncorrectable error" bit set even it is applicable for PSR1
+ * only
+ */
+ if (!line && mode == PSR_MODE_1)
+ line = strstr(buf, "Sink PSR error status: 0x4");
+
+ igt_assert_f(line, "Sink detected PSR error(s):\n%s\n", buf);
+}
+
#define PSR2_SU_BLOCK_STR_LOOKUP "PSR2 SU blocks:\n0\t"
/* Return the the last or last but one su blocks */
diff --git a/lib/igt_psr.h b/lib/igt_psr.h
index a7ebd0739..7639f8d46 100644
--- a/lib/igt_psr.h
+++ b/lib/igt_psr.h
@@ -56,6 +56,7 @@ bool psr_long_wait_update(int debugfs_fd, enum psr_mode mode, igt_output_t *outp
bool psr_enable(int device, int debugfs_fd, enum psr_mode, igt_output_t *output);
bool psr_disable(int device, int debugfs_fd, igt_output_t *output);
bool psr_sink_support(int device, int debugfs_fd, enum psr_mode mode, igt_output_t *output);
+void psr_sink_error_check(int debugfs_fd, enum psr_mode mode, igt_output_t *output);
bool psr2_wait_su(int debugfs_fd, uint16_t *num_su_blocks);
void psr_print_debugfs(int debugfs_fd);
enum psr_mode psr_get_mode(int debugfs_fd, igt_output_t *output);
--
2.34.1
next prev parent reply other threads:[~2024-09-11 5:50 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-09-11 5:49 [PATCH i-g-t v2 0/5] PSR testing improvement by checking sink status Jouni Högander
2024-09-11 5:49 ` Jouni Högander [this message]
2024-09-11 5:49 ` [PATCH i-g-t v2 2/5] tests/intel/kms_psr*: Add psr_sink_error_check to PSR tests Jouni Högander
2024-09-11 10:53 ` Thasleem, Mohammed
2024-09-11 5:49 ` [PATCH i-g-t v2 3/5] tests/intel/kms_fbt: " Jouni Högander
2024-09-11 11:34 ` Thasleem, Mohammed
2024-09-11 5:49 ` [PATCH i-g-t v2 4/5] test/intel/kms_pm_dc: " Jouni Högander
2024-09-11 9:54 ` Thasleem, Mohammed
2024-09-11 5:49 ` [PATCH i-g-t v2 5/5] tests/intel/kms_dirtyfb: Check features after rendering Jouni Högander
2024-09-12 3:34 ` ✓ CI.xeBAT: success for PSR testing improvement by checking sink status Patchwork
2024-09-12 3:46 ` ✓ Fi.CI.BAT: " Patchwork
2024-09-12 8:12 ` ✓ CI.xeFULL: " Patchwork
2024-09-12 22:50 ` ✗ Fi.CI.IGT: failure " Patchwork
2024-09-17 11:38 ` Hogander, Jouni
2024-09-17 11:59 ` [PATCH i-g-t v2 0/5] " Hogander, Jouni
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=20240911054958.1023925-2-jouni.hogander@intel.com \
--to=jouni.hogander@intel.com \
--cc=igt-dev@lists.freedesktop.org \
--cc=ramanaidu.naladala@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
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