linuxppc-dev.lists.ozlabs.org archive mirror
 help / color / mirror / Atom feed
From: Michael Ellerman <mpe@ellerman.id.au>
To: <linuxppc-dev@ozlabs.org>
Cc: sam.bobroff@au1.ibm.com
Subject: [PATCH 08/10] selftests/powerpc: Add cycles test with MMCR2 handling
Date: Wed, 23 Jul 2014 17:31:38 +1000	[thread overview]
Message-ID: <1406100700-9702-8-git-send-email-mpe@ellerman.id.au> (raw)
In-Reply-To: <1406100700-9702-1-git-send-email-mpe@ellerman.id.au>

Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
---
 tools/testing/selftests/powerpc/pmu/ebb/Makefile   |  3 +-
 .../powerpc/pmu/ebb/cycles_with_mmcr2_test.c       | 91 ++++++++++++++++++++++
 2 files changed, 93 insertions(+), 1 deletion(-)
 create mode 100644 tools/testing/selftests/powerpc/pmu/ebb/cycles_with_mmcr2_test.c

diff --git a/tools/testing/selftests/powerpc/pmu/ebb/Makefile b/tools/testing/selftests/powerpc/pmu/ebb/Makefile
index 251447eafe49..3dc4332698cb 100644
--- a/tools/testing/selftests/powerpc/pmu/ebb/Makefile
+++ b/tools/testing/selftests/powerpc/pmu/ebb/Makefile
@@ -13,7 +13,8 @@ PROGS := reg_access_test event_attributes_test cycles_test	\
 	 close_clears_pmcc_test instruction_count_test		\
 	 fork_cleanup_test ebb_on_child_test			\
 	 ebb_on_willing_child_test back_to_back_ebbs_test	\
-	 lost_exception_test no_handler_test
+	 lost_exception_test no_handler_test			\
+	 cycles_with_mmcr2_test
 
 all: $(PROGS)
 
diff --git a/tools/testing/selftests/powerpc/pmu/ebb/cycles_with_mmcr2_test.c b/tools/testing/selftests/powerpc/pmu/ebb/cycles_with_mmcr2_test.c
new file mode 100644
index 000000000000..d43029b0800c
--- /dev/null
+++ b/tools/testing/selftests/powerpc/pmu/ebb/cycles_with_mmcr2_test.c
@@ -0,0 +1,91 @@
+/*
+ * Copyright 2014, Michael Ellerman, IBM Corp.
+ * Licensed under GPLv2.
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdbool.h>
+
+#include "ebb.h"
+
+
+/*
+ * Test of counting cycles while manipulating the user accessible bits in MMCR2.
+ */
+
+/* We use two values because the first freezes PMC1 and so we would get no EBBs */
+#define MMCR2_EXPECTED_1 0x4020100804020000UL /* (FC1P|FC2P|FC3P|FC4P|FC5P|FC6P) */
+#define MMCR2_EXPECTED_2 0x0020100804020000UL /* (     FC2P|FC3P|FC4P|FC5P|FC6P) */
+
+
+int cycles_with_mmcr2(void)
+{
+	struct event event;
+	uint64_t val, expected[2], actual;
+	int i;
+	bool bad_mmcr2;
+
+	event_init_named(&event, 0x1001e, "cycles");
+	event_leader_ebb_init(&event);
+
+	event.attr.exclude_kernel = 1;
+	event.attr.exclude_hv = 1;
+	event.attr.exclude_idle = 1;
+
+	FAIL_IF(event_open(&event));
+
+	ebb_enable_pmc_counting(1);
+	setup_ebb_handler(standard_ebb_callee);
+	ebb_global_enable();
+
+	FAIL_IF(ebb_event_enable(&event));
+
+	mtspr(SPRN_PMC1, pmc_sample_period(sample_period));
+
+	/* XXX Set of MMCR2 must be after enable */
+	expected[0] = MMCR2_EXPECTED_1;
+	expected[1] = MMCR2_EXPECTED_2;
+	i = 0;
+	bad_mmcr2 = false;
+
+	/* Make sure we loop until we take at least one EBB */
+	while ((ebb_state.stats.ebb_count < 20 && !bad_mmcr2) ||
+		ebb_state.stats.ebb_count < 1)
+	{
+		mtspr(SPRN_MMCR2, expected[i % 2]);
+
+		FAIL_IF(core_busy_loop());
+
+		val = mfspr(SPRN_MMCR2);
+		if (val != expected[i % 2]) {
+			bad_mmcr2 = true;
+			actual = val;
+		}
+
+		i++;
+	}
+
+	ebb_global_disable();
+	ebb_freeze_pmcs();
+
+	count_pmc(1, sample_period);
+
+	dump_ebb_state();
+
+	event_close(&event);
+
+	FAIL_IF(ebb_state.stats.ebb_count == 0);
+
+	if (bad_mmcr2)
+		printf("Bad MMCR2 value seen is 0x%lx\n", actual);
+
+	FAIL_IF(bad_mmcr2);
+
+	return 0;
+}
+
+int main(void)
+{
+	return test_harness(cycles_with_mmcr2, "cycles_with_mmcr2");
+}
-- 
1.9.1

  parent reply	other threads:[~2014-07-23  7:31 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-07-23  7:31 [PATCH 01/10] selftests/powerpc: Don't ignore errors from sub Makefiles Michael Ellerman
2014-07-23  7:31 ` [PATCH 02/10] selftests/powerpc: Fix parse_proc_maps() Michael Ellerman
2014-07-23  7:31 ` [PATCH 03/10] selftests/powerpc: Move core_busy_loop() into asm Michael Ellerman
2014-07-23  7:31 ` [PATCH 04/10] selftests/powerpc: Add test of L3 bank handling Michael Ellerman
2014-07-23  7:31 ` [PATCH 05/10] selftests/powerpc: Count instructions under scheduler pressure Michael Ellerman
2014-07-23  7:31 ` [PATCH 06/10] selftests/powerpc: Count more instructions & use decimal Michael Ellerman
2014-07-23  7:31 ` [PATCH 07/10] selftests/powerpc: Dump MMCR2 as part of the EBB HW state Michael Ellerman
2014-07-23  7:31 ` Michael Ellerman [this message]
2014-07-23  7:31 ` [PATCH 09/10] selftests/powerpc: Add a routine for retrieving an AUXV entry Michael Ellerman
2014-07-23  7:31 ` [PATCH 10/10] selftests/powerpc: Add test of per-event excludes Michael Ellerman

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=1406100700-9702-8-git-send-email-mpe@ellerman.id.au \
    --to=mpe@ellerman.id.au \
    --cc=linuxppc-dev@ozlabs.org \
    --cc=sam.bobroff@au1.ibm.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;
as well as URLs for NNTP newsgroup(s).