* [kvm-unit-tests PATCH 01/14] lib: asm-generic: add missing casts
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 02/14] lib: share arm-selftest utility functions Andrew Jones
` (14 subsequent siblings)
15 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
lib/asm-generic/io.h | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/lib/asm-generic/io.h b/lib/asm-generic/io.h
index a9939d3a5921f..abd245676e97a 100644
--- a/lib/asm-generic/io.h
+++ b/lib/asm-generic/io.h
@@ -104,27 +104,27 @@ static inline u64 __bswap64(u64 x)
#endif
#define le16_to_cpu(x) \
- ({ u16 __r = __cpu_is_be() ? __bswap16(x) : (x); __r; })
+ ({ u16 __r = __cpu_is_be() ? __bswap16((u16)x) : ((u16)x); __r; })
#define cpu_to_le16 le16_to_cpu
#define le32_to_cpu(x) \
- ({ u32 __r = __cpu_is_be() ? __bswap32(x) : (x); __r; })
+ ({ u32 __r = __cpu_is_be() ? __bswap32((u32)x) : ((u32)x); __r; })
#define cpu_to_le32 le32_to_cpu
#define le64_to_cpu(x) \
- ({ u64 __r = __cpu_is_be() ? __bswap64(x) : (x); __r; })
+ ({ u64 __r = __cpu_is_be() ? __bswap64((u64)x) : ((u64)x); __r; })
#define cpu_to_le64 le64_to_cpu
#define be16_to_cpu(x) \
- ({ u16 __r = !__cpu_is_be() ? __bswap16(x) : (x); __r; })
+ ({ u16 __r = !__cpu_is_be() ? __bswap16((u16)x) : ((u16)x); __r; })
#define cpu_to_be16 be16_to_cpu
#define be32_to_cpu(x) \
- ({ u32 __r = !__cpu_is_be() ? __bswap32(x) : (x); __r; })
+ ({ u32 __r = !__cpu_is_be() ? __bswap32((u32)x) : ((u32)x); __r; })
#define cpu_to_be32 be32_to_cpu
#define be64_to_cpu(x) \
- ({ u64 __r = !__cpu_is_be() ? __bswap64(x) : (x); __r; })
+ ({ u64 __r = !__cpu_is_be() ? __bswap64((u64)x) : ((u64)x); __r; })
#define cpu_to_be64 be64_to_cpu
#ifndef rmb
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* [kvm-unit-tests PATCH 02/14] lib: share arm-selftest utility functions
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 01/14] lib: asm-generic: add missing casts Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 03/14] config: no need to mix arch makefiles Andrew Jones
` (13 subsequent siblings)
15 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
arm-selftest has a couple utility functions that could be useful
to other unit tests, even other architectures. So move them out.
split_var moves to lib/util, where we can add other random
utilities over time. assert_args inspires report_abort, which
allows us to report a message, using the current prefix, that
we're aborting (outputs ABORT vs. PASS/FAIL). This is useful
for cases when unit tests can't complete due to missing
dependencies of some sort, such as missing/invalid inputs from
the user.
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
arm/selftest.c | 33 ++++++++-------------------------
config/config-arm-common.mak | 1 +
lib/libcflat.h | 11 ++++++-----
lib/report.c | 16 ++++++++++++++++
lib/util.c | 20 ++++++++++++++++++++
lib/util.h | 23 +++++++++++++++++++++++
6 files changed, 74 insertions(+), 30 deletions(-)
create mode 100644 lib/util.c
create mode 100644 lib/util.h
diff --git a/arm/selftest.c b/arm/selftest.c
index fc9ec609d875e..74abd048add6c 100644
--- a/arm/selftest.c
+++ b/arm/selftest.c
@@ -6,6 +6,7 @@
* This work is licensed under the terms of the GNU LGPL, version 2.
*/
#include <libcflat.h>
+#include <util.h>
#include <alloc.h>
#include <devicetree.h>
#include <asm/setup.h>
@@ -18,28 +19,6 @@
#include <asm/cpumask.h>
#include <asm/barrier.h>
-static void assert_args(int num_args, int needed_args)
-{
- if (num_args < needed_args) {
- printf("selftest: not enough arguments\n");
- abort();
- }
-}
-
-static char *split_var(char *s, long *val)
-{
- char *p;
-
- p = strchr(s, '=');
- if (!p)
- return NULL;
-
- *val = atol(p+1);
- *p = '\0';
-
- return s;
-}
-
static void check_setup(int argc, char **argv)
{
int nr_tests = 0, i;
@@ -48,7 +27,7 @@ static void check_setup(int argc, char **argv)
for (i = 0; i < argc; ++i) {
- var = split_var(argv[i], &val);
+ var = parse_keyval(argv[i], &val);
if (!var)
continue;
@@ -72,7 +51,8 @@ static void check_setup(int argc, char **argv)
report_prefix_pop();
}
- assert_args(nr_tests, 2);
+ if (nr_tests < 2)
+ report_abort("missing input");
}
static struct pt_regs expected_regs;
@@ -344,7 +324,10 @@ static void cpu_report(void)
int main(int argc, char **argv)
{
report_prefix_push("selftest");
- assert_args(argc, 1);
+
+ if (argc < 1)
+ report_abort("no test specified");
+
report_prefix_push(argv[0]);
if (strcmp(argv[0], "setup") == 0) {
diff --git a/config/config-arm-common.mak b/config/config-arm-common.mak
index f35b8b9231c03..4cff846930796 100644
--- a/config/config-arm-common.mak
+++ b/config/config-arm-common.mak
@@ -32,6 +32,7 @@ CFLAGS += -I lib -I lib/libfdt
asm-offsets = lib/$(ARCH)/asm-offsets.h
include config/asm-offsets.mak
+cflatobjs += lib/util.o
cflatobjs += lib/alloc.o
cflatobjs += lib/devicetree.o
cflatobjs += lib/virtio.o
diff --git a/lib/libcflat.h b/lib/libcflat.h
index 9747ccdbc9f1d..8411f6c5d92e3 100644
--- a/lib/libcflat.h
+++ b/lib/libcflat.h
@@ -57,11 +57,12 @@ extern int snprintf(char *buf, int size, const char *fmt, ...);
extern int vsnprintf(char *buf, int size, const char *fmt, va_list va);
extern long atol(const char *ptr);
-void report_prefix_push(const char *prefix);
-void report_prefix_pop(void);
-void report(const char *msg_fmt, bool pass, ...);
-void report_xfail(const char *msg_fmt, bool xfail, bool pass, ...);
-int report_summary(void);
+extern void report_prefix_push(const char *prefix);
+extern void report_prefix_pop(void);
+extern void report(const char *msg_fmt, bool pass, ...);
+extern void report_xfail(const char *msg_fmt, bool xfail, bool pass, ...);
+extern void report_abort(const char *msg_fmt, ...);
+extern int report_summary(void);
#define ARRAY_SIZE(_a) (sizeof(_a)/sizeof((_a)[0]))
diff --git a/lib/report.c b/lib/report.c
index 35e664108a921..62916c4ac3c8a 100644
--- a/lib/report.c
+++ b/lib/report.c
@@ -96,3 +96,19 @@ int report_summary(void)
spin_unlock(&lock);
}
+
+void report_abort(const char *msg_fmt, ...)
+{
+ va_list va;
+ char buf[2000];
+
+ puts("ABORT: ");
+ puts(prefixes);
+ va_start(va, msg_fmt);
+ vsnprintf(buf, sizeof(buf), msg_fmt, va);
+ va_end(va);
+ puts(buf);
+ puts("\n");
+ report_summary();
+ abort();
+}
diff --git a/lib/util.c b/lib/util.c
new file mode 100644
index 0000000000000..9dcec30ac8b5f
--- /dev/null
+++ b/lib/util.c
@@ -0,0 +1,20 @@
+/*
+ * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2.
+ */
+#include <libcflat.h>
+
+char *parse_keyval(char *s, long *val)
+{
+ char *p;
+
+ p = strchr(s, '=');
+ if (!p)
+ return NULL;
+
+ *val = atol(p+1);
+ *p = '\0';
+
+ return s;
+}
diff --git a/lib/util.h b/lib/util.h
new file mode 100644
index 0000000000000..9a12bda4f61df
--- /dev/null
+++ b/lib/util.h
@@ -0,0 +1,23 @@
+#ifndef _UTIL_H_
+#define _UTIL_H_
+/*
+ * Collection of utility functions to share between unit tests.
+ *
+ * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2.
+ */
+
+/*
+ * parse_keyval extracts the integer from a string formatted as
+ * string=integer. This is useful for passing expected values in
+ * to the unit test on the command line, i.e. it helps parse QEMU
+ * command lines that include something like -append var1=1 var2=2
+ * @s is the input string, likely a command line parameter, and
+ * @val is a pointer to where the integer will be stored.
+ *
+ * Returns a pointer to the key string of the parsed key-value pair.
+ */
+extern char *parse_keyval(char *s, long *val);
+
+#endif
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* [kvm-unit-tests PATCH 03/14] config: no need to mix arch makefiles
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 01/14] lib: asm-generic: add missing casts Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 02/14] lib: share arm-selftest utility functions Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 04/14] powerpc/ppc64: start skeleton framework Andrew Jones
` (12 subsequent siblings)
15 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
Move the config/config-*.mak files to their own directories and
rename to Makefile.* README is also updated to reflect the change,
as well as to remove an optimistic reference to a non-existent
docs directory.
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
Makefile | 2 +-
README | 3 +--
arm/Makefile | 1 +
config/config-arm.mak => arm/Makefile.arm | 2 +-
config/config-arm64.mak => arm/Makefile.arm64 | 2 +-
config/config-arm-common.mak => arm/Makefile.common | 2 +-
{config => scripts}/asm-offsets.mak | 0
x86/Makefile | 1 +
config/config-x86-common.mak => x86/Makefile.common | 0
config/config-i386.mak => x86/Makefile.i386 | 2 +-
config/config-x86_64.mak => x86/Makefile.x86_64 | 2 +-
11 files changed, 9 insertions(+), 8 deletions(-)
create mode 100644 arm/Makefile
rename config/config-arm.mak => arm/Makefile.arm (90%)
rename config/config-arm64.mak => arm/Makefile.arm64 (89%)
rename config/config-arm-common.mak => arm/Makefile.common (98%)
rename {config => scripts}/asm-offsets.mak (100%)
create mode 100644 x86/Makefile
rename config/config-x86-common.mak => x86/Makefile.common (100%)
rename config/config-i386.mak => x86/Makefile.i386 (91%)
rename config/config-x86_64.mak => x86/Makefile.x86_64 (92%)
diff --git a/Makefile b/Makefile
index 0d5933474cd8c..2dcc323592299 100644
--- a/Makefile
+++ b/Makefile
@@ -30,7 +30,7 @@ LIBFDT_include = $(addprefix $(LIBFDT_srcdir)/,$(LIBFDT_INCLUDES))
LIBFDT_version = $(addprefix $(LIBFDT_srcdir)/,$(LIBFDT_VERSION))
#include architecure specific make rules
-include config/config-$(ARCH).mak
+include $(TEST_DIR)/Makefile
# cc-option
# Usage: OP_CFLAGS+=$(call cc-option, -falign-functions=0, -malign-functions=0)
diff --git a/README b/README
index 0884aafa21d18..4cb30892699f4 100644
--- a/README
+++ b/README
@@ -28,8 +28,7 @@ each as a standalone test.
Directory structure:
.: configure script, top-level Makefile, and run_tests.sh
-./config: collection of architecture dependent makefiles
-./docs: documentation files
+./scripts: helper scripts for building and running tests
./lib: general architecture neutral services for the tests
./lib/<ARCH>: architecture dependent services for the tests
./<ARCH>: the sources of the tests and the created objects/images
diff --git a/arm/Makefile b/arm/Makefile
new file mode 100644
index 0000000000000..369a38b2d1703
--- /dev/null
+++ b/arm/Makefile
@@ -0,0 +1 @@
+include $(TEST_DIR)/Makefile.$(ARCH)
diff --git a/config/config-arm.mak b/arm/Makefile.arm
similarity index 90%
rename from config/config-arm.mak
rename to arm/Makefile.arm
index ae6c2e7134883..946422872532d 100644
--- a/config/config-arm.mak
+++ b/arm/Makefile.arm
@@ -18,6 +18,6 @@ cflatobjs += lib/arm/processor.o
# arm specific tests
tests =
-include config/config-arm-common.mak
+include $(TEST_DIR)/Makefile.common
arch_clean: arm_clean
diff --git a/config/config-arm64.mak b/arm/Makefile.arm64
similarity index 89%
rename from config/config-arm64.mak
rename to arm/Makefile.arm64
index d61b703c8140e..0b0761c729c7c 100644
--- a/config/config-arm64.mak
+++ b/arm/Makefile.arm64
@@ -14,7 +14,7 @@ cflatobjs += lib/arm64/spinlock.o
# arm64 specific tests
tests =
-include config/config-arm-common.mak
+include $(TEST_DIR)/Makefile.common
arch_clean: arm_clean
$(RM) lib/arm64/.*.d
diff --git a/config/config-arm-common.mak b/arm/Makefile.common
similarity index 98%
rename from config/config-arm-common.mak
rename to arm/Makefile.common
index 4cff846930796..f50d9f9d4e0f6 100644
--- a/config/config-arm-common.mak
+++ b/arm/Makefile.common
@@ -30,7 +30,7 @@ CFLAGS += -O2
CFLAGS += -I lib -I lib/libfdt
asm-offsets = lib/$(ARCH)/asm-offsets.h
-include config/asm-offsets.mak
+include scripts/asm-offsets.mak
cflatobjs += lib/util.o
cflatobjs += lib/alloc.o
diff --git a/config/asm-offsets.mak b/scripts/asm-offsets.mak
similarity index 100%
rename from config/asm-offsets.mak
rename to scripts/asm-offsets.mak
diff --git a/x86/Makefile b/x86/Makefile
new file mode 100644
index 0000000000000..369a38b2d1703
--- /dev/null
+++ b/x86/Makefile
@@ -0,0 +1 @@
+include $(TEST_DIR)/Makefile.$(ARCH)
diff --git a/config/config-x86-common.mak b/x86/Makefile.common
similarity index 100%
rename from config/config-x86-common.mak
rename to x86/Makefile.common
diff --git a/config/config-i386.mak b/x86/Makefile.i386
similarity index 91%
rename from config/config-i386.mak
rename to x86/Makefile.i386
index 691381ca69c85..ec189ce735bad 100644
--- a/config/config-i386.mak
+++ b/x86/Makefile.i386
@@ -6,7 +6,7 @@ CFLAGS += -I $(KERNELDIR)/include
tests = $(TEST_DIR)/taskswitch.flat $(TEST_DIR)/taskswitch2.flat \
$(TEST_DIR)/cmpxchg8b.flat
-include config/config-x86-common.mak
+include $(TEST_DIR)/Makefile.common
$(TEST_DIR)/cmpxchg8b.elf: $(cstart.o) $(TEST_DIR)/cmpxchg8b.o
$(TEST_DIR)/taskswitch.elf: $(cstart.o) $(TEST_DIR)/taskswitch.o
diff --git a/config/config-x86_64.mak b/x86/Makefile.x86_64
similarity index 92%
rename from config/config-x86_64.mak
rename to x86/Makefile.x86_64
index 7d4eb3494647e..e7799fe62b032 100644
--- a/config/config-x86_64.mak
+++ b/x86/Makefile.x86_64
@@ -12,4 +12,4 @@ tests += $(TEST_DIR)/svm.flat
tests += $(TEST_DIR)/vmx.flat
tests += $(TEST_DIR)/tscdeadline_latency.flat
-include config/config-x86-common.mak
+include $(TEST_DIR)/Makefile.common
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* [kvm-unit-tests PATCH 04/14] powerpc/ppc64: start skeleton framework
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (2 preceding siblings ...)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 03/14] config: no need to mix arch makefiles Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 05/14] powerpc/pp64: ppc-ify makefiles and linker script Andrew Jones
` (11 subsequent siblings)
15 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
Use the arm/arm64 framework as a template to start a skeleton
for powerpc/ppc64. Copy the arm makefiles and linker script
verbatim. We'll modify them for powerpc with the next patch,
making it clear what needs to be changed.
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
configure | 2 ++
lib/powerpc/.gitignore | 1 +
lib/powerpc/io.c | 20 ++++++++++++
lib/powerpc/mmu.c | 0
lib/powerpc/setup.c | 0
lib/powerpc/smp.c | 0
lib/ppc64/.gitignore | 1 +
lib/ppc64/asm-offsets.c | 12 +++++++
lib/ppc64/asm/asm-offsets.h | 1 +
lib/ppc64/asm/io.h | 5 +++
lib/ppc64/asm/page.h | 1 +
lib/ppc64/asm/spinlock.h | 11 +++++++
lib/ppc64/processor.c | 0
lib/ppc64/spinlock.c | 11 +++++++
powerpc/Makefile | 1 +
powerpc/Makefile.common | 80 +++++++++++++++++++++++++++++++++++++++++++++
powerpc/Makefile.ppc64 | 20 ++++++++++++
powerpc/cstart64.S | 19 +++++++++++
powerpc/flat.lds | 27 +++++++++++++++
powerpc/selftest.c | 7 ++++
20 files changed, 219 insertions(+)
create mode 100644 lib/powerpc/.gitignore
create mode 100644 lib/powerpc/io.c
create mode 100644 lib/powerpc/mmu.c
create mode 100644 lib/powerpc/setup.c
create mode 100644 lib/powerpc/smp.c
create mode 100644 lib/ppc64/.gitignore
create mode 100644 lib/ppc64/asm-offsets.c
create mode 100644 lib/ppc64/asm/asm-offsets.h
create mode 100644 lib/ppc64/asm/io.h
create mode 100644 lib/ppc64/asm/page.h
create mode 100644 lib/ppc64/asm/spinlock.h
create mode 100644 lib/ppc64/processor.c
create mode 100644 lib/ppc64/spinlock.c
create mode 100644 powerpc/Makefile
create mode 100644 powerpc/Makefile.common
create mode 100644 powerpc/Makefile.ppc64
create mode 100644 powerpc/cstart64.S
create mode 100644 powerpc/flat.lds
create mode 100644 powerpc/selftest.c
diff --git a/configure b/configure
index 078b70ce096a6..b367224093369 100755
--- a/configure
+++ b/configure
@@ -80,6 +80,8 @@ if [ "$arch" = "i386" ] || [ "$arch" = "x86_64" ]; then
testdir=x86
elif [ "$arch" = "arm" ] || [ "$arch" = "arm64" ]; then
testdir=arm
+elif [ "$arch" = "ppc64" ]; then
+ testdir=powerpc
else
testdir=$arch
fi
diff --git a/lib/powerpc/.gitignore b/lib/powerpc/.gitignore
new file mode 100644
index 0000000000000..84872bf197c67
--- /dev/null
+++ b/lib/powerpc/.gitignore
@@ -0,0 +1 @@
+asm-offsets.[hs]
diff --git a/lib/powerpc/io.c b/lib/powerpc/io.c
new file mode 100644
index 0000000000000..cf3b6347e1e46
--- /dev/null
+++ b/lib/powerpc/io.c
@@ -0,0 +1,20 @@
+/*
+ * Each architecture must implement puts() and exit().
+ *
+ * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2.
+ */
+#include <libcflat.h>
+
+void io_init(void)
+{
+}
+
+void puts(const char *s __unused)
+{
+}
+
+void exit(int code __unused)
+{
+}
diff --git a/lib/powerpc/mmu.c b/lib/powerpc/mmu.c
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/lib/powerpc/setup.c b/lib/powerpc/setup.c
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/lib/powerpc/smp.c b/lib/powerpc/smp.c
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/lib/ppc64/.gitignore b/lib/ppc64/.gitignore
new file mode 100644
index 0000000000000..84872bf197c67
--- /dev/null
+++ b/lib/ppc64/.gitignore
@@ -0,0 +1 @@
+asm-offsets.[hs]
diff --git a/lib/ppc64/asm-offsets.c b/lib/ppc64/asm-offsets.c
new file mode 100644
index 0000000000000..895296e19b060
--- /dev/null
+++ b/lib/ppc64/asm-offsets.c
@@ -0,0 +1,12 @@
+/*
+ * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2.
+ */
+#include <libcflat.h>
+#include <kbuild.h>
+
+int main(void)
+{
+ return 0;
+}
diff --git a/lib/ppc64/asm/asm-offsets.h b/lib/ppc64/asm/asm-offsets.h
new file mode 100644
index 0000000000000..d370ee36a182b
--- /dev/null
+++ b/lib/ppc64/asm/asm-offsets.h
@@ -0,0 +1 @@
+#include <generated/asm-offsets.h>
diff --git a/lib/ppc64/asm/io.h b/lib/ppc64/asm/io.h
new file mode 100644
index 0000000000000..c0801d4762ba6
--- /dev/null
+++ b/lib/ppc64/asm/io.h
@@ -0,0 +1,5 @@
+#ifndef _ASMPPC64_IO_H_
+#define _ASMPPC64_IO_H_
+#define __cpu_is_be() (1)
+#include <asm-generic/io.h>
+#endif
diff --git a/lib/ppc64/asm/page.h b/lib/ppc64/asm/page.h
new file mode 100644
index 0000000000000..1a8b62711f288
--- /dev/null
+++ b/lib/ppc64/asm/page.h
@@ -0,0 +1 @@
+#include <asm-generic/page.h>
diff --git a/lib/ppc64/asm/spinlock.h b/lib/ppc64/asm/spinlock.h
new file mode 100644
index 0000000000000..002cdb194b1db
--- /dev/null
+++ b/lib/ppc64/asm/spinlock.h
@@ -0,0 +1,11 @@
+#ifndef _ASMPPC64_SPINLOCK_H_
+#define _ASMPPC64_SPINLOCK_H_
+
+struct spinlock {
+ int v;
+};
+
+extern void spin_lock(struct spinlock *lock);
+extern void spin_unlock(struct spinlock *lock);
+
+#endif /* _ASMPPC64_SPINLOCK_H_ */
diff --git a/lib/ppc64/processor.c b/lib/ppc64/processor.c
new file mode 100644
index 0000000000000..e69de29bb2d1d
diff --git a/lib/ppc64/spinlock.c b/lib/ppc64/spinlock.c
new file mode 100644
index 0000000000000..522f8d047c970
--- /dev/null
+++ b/lib/ppc64/spinlock.c
@@ -0,0 +1,11 @@
+#include <asm/spinlock.h>
+
+void spin_lock(struct spinlock *lock)
+{
+ lock->v = 1;
+}
+
+void spin_unlock(struct spinlock *lock)
+{
+ lock->v = 0;
+}
diff --git a/powerpc/Makefile b/powerpc/Makefile
new file mode 100644
index 0000000000000..369a38b2d1703
--- /dev/null
+++ b/powerpc/Makefile
@@ -0,0 +1 @@
+include $(TEST_DIR)/Makefile.$(ARCH)
diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
new file mode 100644
index 0000000000000..f50d9f9d4e0f6
--- /dev/null
+++ b/powerpc/Makefile.common
@@ -0,0 +1,80 @@
+#
+# arm common makefile
+#
+# Authors: Andrew Jones <drjones@redhat.com>
+#
+
+ifeq ($(LOADADDR),)
+ # qemu mach-virt default load address
+ LOADADDR = 0x40000000
+endif
+
+tests-common = \
+ $(TEST_DIR)/selftest.flat \
+ $(TEST_DIR)/spinlock-test.flat
+
+ifneq ($(TEST),)
+ tests = $(TEST_DIR)/$(TEST).flat
+ tests-common =
+endif
+
+all: test_cases
+
+##################################################################
+phys_base = $(LOADADDR)
+
+CFLAGS += -std=gnu99
+CFLAGS += -ffreestanding
+CFLAGS += -Wextra
+CFLAGS += -O2
+CFLAGS += -I lib -I lib/libfdt
+
+asm-offsets = lib/$(ARCH)/asm-offsets.h
+include scripts/asm-offsets.mak
+
+cflatobjs += lib/util.o
+cflatobjs += lib/alloc.o
+cflatobjs += lib/devicetree.o
+cflatobjs += lib/virtio.o
+cflatobjs += lib/virtio-mmio.o
+cflatobjs += lib/chr-testdev.o
+cflatobjs += lib/prng.o
+cflatobjs += lib/arm/io.o
+cflatobjs += lib/arm/setup.o
+cflatobjs += lib/arm/mmu.o
+cflatobjs += lib/arm/bitops.o
+cflatobjs += lib/arm/psci.o
+cflatobjs += lib/arm/smp.o
+
+libeabi = lib/arm/libeabi.a
+eabiobjs = lib/arm/eabi_compat.o
+
+libgcc := $(shell $(CC) $(machine) --print-libgcc-file-name)
+start_addr := $(shell printf "%x\n" $$(( $(phys_base) + $(kernel_offset) )))
+
+FLATLIBS = $(libcflat) $(LIBFDT_archive) $(libgcc) $(libeabi)
+%.elf: LDFLAGS = $(CFLAGS) -nostdlib
+%.elf: %.o $(FLATLIBS) arm/flat.lds
+ $(CC) $(LDFLAGS) -o $@ \
+ -Wl,-T,arm/flat.lds,--build-id=none,-Ttext=$(start_addr) \
+ $(filter %.o, $^) $(FLATLIBS)
+
+%.flat: %.elf
+ $(OBJCOPY) -O binary $^ $@
+
+$(libeabi): $(eabiobjs)
+ $(AR) rcs $@ $^
+
+arm_clean: libfdt_clean asm_offsets_clean
+ $(RM) $(TEST_DIR)/*.{o,flat,elf} $(libeabi) $(eabiobjs) \
+ $(TEST_DIR)/.*.d lib/arm/.*.d
+
+##################################################################
+
+generated_files = $(asm-offsets)
+
+test_cases: $(generated_files) $(tests-common) $(tests)
+
+$(TEST_DIR)/$(TEST).elf: $(cstart.o) $(TEST_DIR)/$(TEST).o
+$(TEST_DIR)/selftest.elf: $(cstart.o) $(TEST_DIR)/selftest.o
+$(TEST_DIR)/spinlock-test.elf: $(cstart.o) $(TEST_DIR)/spinlock-test.o
diff --git a/powerpc/Makefile.ppc64 b/powerpc/Makefile.ppc64
new file mode 100644
index 0000000000000..0b0761c729c7c
--- /dev/null
+++ b/powerpc/Makefile.ppc64
@@ -0,0 +1,20 @@
+#
+# arm64 makefile
+#
+# Authors: Andrew Jones <drjones@redhat.com>
+#
+bits = 64
+ldarch = elf64-littleaarch64
+kernel_offset = 0x80000
+
+cstart.o = $(TEST_DIR)/cstart64.o
+cflatobjs += lib/arm64/processor.o
+cflatobjs += lib/arm64/spinlock.o
+
+# arm64 specific tests
+tests =
+
+include $(TEST_DIR)/Makefile.common
+
+arch_clean: arm_clean
+ $(RM) lib/arm64/.*.d
diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
new file mode 100644
index 0000000000000..d9b77f44f9e0e
--- /dev/null
+++ b/powerpc/cstart64.S
@@ -0,0 +1,19 @@
+/*
+ * Entry point and assembler functions for ppc64 tests.
+ *
+ * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2.
+ */
+
+.section .init
+
+.globl start
+start:
+ b halt
+
+.text
+
+.globl halt
+halt:
+1: b 1b
diff --git a/powerpc/flat.lds b/powerpc/flat.lds
new file mode 100644
index 0000000000000..efdf5d7109ffb
--- /dev/null
+++ b/powerpc/flat.lds
@@ -0,0 +1,27 @@
+
+SECTIONS
+{
+ .text : { *(.init) *(.text) *(.text.*) }
+ . = ALIGN(64K);
+ etext = .;
+ .data : {
+ *(.data)
+ }
+ . = ALIGN(16);
+ .rodata : { *(.rodata) }
+ . = ALIGN(16);
+ .bss : { *(.bss) }
+ . = ALIGN(64K);
+ edata = .;
+ . += 64K;
+ . = ALIGN(64K);
+ /*
+ * stack depth is 16K for arm and PAGE_SIZE for arm64, see THREAD_SIZE
+ * sp must be 16 byte aligned for arm64, and 8 byte aligned for arm
+ * sp must always be strictly less than the true stacktop
+ */
+ stackptr = . - 16;
+ stacktop = .;
+}
+
+ENTRY(start)
diff --git a/powerpc/selftest.c b/powerpc/selftest.c
new file mode 100644
index 0000000000000..2f2a5215dd55c
--- /dev/null
+++ b/powerpc/selftest.c
@@ -0,0 +1,7 @@
+#include <libcflat.h>
+
+int main(void)
+{
+ printf("hello world\n");
+ return 0;
+}
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* [kvm-unit-tests PATCH 05/14] powerpc/pp64: ppc-ify makefiles and linker script
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (3 preceding siblings ...)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 04/14] powerpc/ppc64: start skeleton framework Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 06/14] powerpc/ppc64: add boot rom source Andrew Jones
` (10 subsequent siblings)
15 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
powerpc/Makefile.common | 44 +++++++++++++-------------------------------
powerpc/Makefile.ppc64 | 16 ++++++++--------
powerpc/flat.lds | 14 ++++++++++----
3 files changed, 31 insertions(+), 43 deletions(-)
diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
index f50d9f9d4e0f6..5cd3ea8085038 100644
--- a/powerpc/Makefile.common
+++ b/powerpc/Makefile.common
@@ -1,17 +1,15 @@
#
-# arm common makefile
+# powerpc common makefile
#
# Authors: Andrew Jones <drjones@redhat.com>
#
ifeq ($(LOADADDR),)
- # qemu mach-virt default load address
- LOADADDR = 0x40000000
+ LOADADDR = 0x0
endif
tests-common = \
- $(TEST_DIR)/selftest.flat \
- $(TEST_DIR)/spinlock-test.flat
+ $(TEST_DIR)/selftest.elf
ifneq ($(TEST),)
tests = $(TEST_DIR)/$(TEST).flat
@@ -35,39 +33,24 @@ include scripts/asm-offsets.mak
cflatobjs += lib/util.o
cflatobjs += lib/alloc.o
cflatobjs += lib/devicetree.o
-cflatobjs += lib/virtio.o
-cflatobjs += lib/virtio-mmio.o
-cflatobjs += lib/chr-testdev.o
-cflatobjs += lib/prng.o
-cflatobjs += lib/arm/io.o
-cflatobjs += lib/arm/setup.o
-cflatobjs += lib/arm/mmu.o
-cflatobjs += lib/arm/bitops.o
-cflatobjs += lib/arm/psci.o
-cflatobjs += lib/arm/smp.o
-
-libeabi = lib/arm/libeabi.a
-eabiobjs = lib/arm/eabi_compat.o
+cflatobjs += lib/powerpc/io.o
+cflatobjs += lib/powerpc/setup.o
+cflatobjs += lib/powerpc/mmu.o
+cflatobjs += lib/powerpc/smp.o
libgcc := $(shell $(CC) $(machine) --print-libgcc-file-name)
start_addr := $(shell printf "%x\n" $$(( $(phys_base) + $(kernel_offset) )))
-FLATLIBS = $(libcflat) $(LIBFDT_archive) $(libgcc) $(libeabi)
+FLATLIBS = $(libcflat) $(LIBFDT_archive) $(libgcc)
%.elf: LDFLAGS = $(CFLAGS) -nostdlib
-%.elf: %.o $(FLATLIBS) arm/flat.lds
+%.elf: %.o $(FLATLIBS) powerpc/flat.lds
$(CC) $(LDFLAGS) -o $@ \
- -Wl,-T,arm/flat.lds,--build-id=none,-Ttext=$(start_addr) \
+ -Wl,-T,powerpc/flat.lds,--build-id=none,-Ttext=$(start_addr) \
$(filter %.o, $^) $(FLATLIBS)
-%.flat: %.elf
- $(OBJCOPY) -O binary $^ $@
-
-$(libeabi): $(eabiobjs)
- $(AR) rcs $@ $^
-
-arm_clean: libfdt_clean asm_offsets_clean
- $(RM) $(TEST_DIR)/*.{o,flat,elf} $(libeabi) $(eabiobjs) \
- $(TEST_DIR)/.*.d lib/arm/.*.d
+powerpc_clean: libfdt_clean asm_offsets_clean
+ $(RM) $(TEST_DIR)/*.{o,elf} \
+ $(TEST_DIR)/.*.d lib/powerpc/.*.d
##################################################################
@@ -77,4 +60,3 @@ test_cases: $(generated_files) $(tests-common) $(tests)
$(TEST_DIR)/$(TEST).elf: $(cstart.o) $(TEST_DIR)/$(TEST).o
$(TEST_DIR)/selftest.elf: $(cstart.o) $(TEST_DIR)/selftest.o
-$(TEST_DIR)/spinlock-test.elf: $(cstart.o) $(TEST_DIR)/spinlock-test.o
diff --git a/powerpc/Makefile.ppc64 b/powerpc/Makefile.ppc64
index 0b0761c729c7c..7c61933dfa8ba 100644
--- a/powerpc/Makefile.ppc64
+++ b/powerpc/Makefile.ppc64
@@ -1,20 +1,20 @@
#
-# arm64 makefile
+# ppc64 makefile
#
# Authors: Andrew Jones <drjones@redhat.com>
#
bits = 64
-ldarch = elf64-littleaarch64
-kernel_offset = 0x80000
+ldarch = elf64-powerpc #elf64-powerpcle (eventually)
+kernel_offset = 0x0
cstart.o = $(TEST_DIR)/cstart64.o
-cflatobjs += lib/arm64/processor.o
-cflatobjs += lib/arm64/spinlock.o
+cflatobjs += lib/ppc64/processor.o
+cflatobjs += lib/ppc64/spinlock.o
-# arm64 specific tests
+# ppc64 specific tests
tests =
include $(TEST_DIR)/Makefile.common
-arch_clean: arm_clean
- $(RM) lib/arm64/.*.d
+arch_clean: powerpc_clean
+ $(RM) lib/ppc64/.*.d
diff --git a/powerpc/flat.lds b/powerpc/flat.lds
index efdf5d7109ffb..bd075efb2c51b 100644
--- a/powerpc/flat.lds
+++ b/powerpc/flat.lds
@@ -4,6 +4,8 @@ SECTIONS
.text : { *(.init) *(.text) *(.text.*) }
. = ALIGN(64K);
etext = .;
+ .opd : { *(.opd) }
+ . = ALIGN(16);
.data : {
*(.data)
}
@@ -11,16 +13,20 @@ SECTIONS
.rodata : { *(.rodata) }
. = ALIGN(16);
.bss : { *(.bss) }
+ . = ALIGN(16);
+ /*
+ * tocptr is tocbase + 32K, allowing toc offsets to be +-32K
+ */
+ tocptr = . + 32K;
+ .got : { *(.toc) *(.got) }
. = ALIGN(64K);
edata = .;
. += 64K;
. = ALIGN(64K);
/*
- * stack depth is 16K for arm and PAGE_SIZE for arm64, see THREAD_SIZE
- * sp must be 16 byte aligned for arm64, and 8 byte aligned for arm
- * sp must always be strictly less than the true stacktop
+ * stackptr set with initial stack frame (64 bytes) preallocated
*/
- stackptr = . - 16;
+ stackptr = . - 64;
stacktop = .;
}
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* [kvm-unit-tests PATCH 06/14] powerpc/ppc64: add boot rom source
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (4 preceding siblings ...)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 05/14] powerpc/pp64: ppc-ify makefiles and linker script Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 07/14] powerpc/ppc64: add bootloader to bounce into memory Andrew Jones
` (9 subsequent siblings)
15 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
Make a one line replacement for SLOF. This bootloader just
jumps to 0x400000, because we know the kernel will be there.
This commit allows rebuilding of the bootloader, as long
as the SLOF build_romfs tool is present, but since it would be
a pain to require that tool for ever build, we'll just commit
the binary rom image too with the next commit.
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
powerpc/rom/.gitignore | 1 +
powerpc/rom/Makefile | 36 ++++++++++++++++++++++++++++++++++++
| 1 +
powerpc/rom/rom.ffs | 4 ++++
powerpc/rom/stage1.S | 5 +++++
5 files changed, 47 insertions(+)
create mode 100644 powerpc/rom/.gitignore
create mode 100644 powerpc/rom/Makefile
create mode 100644 powerpc/rom/header.img
create mode 100644 powerpc/rom/rom.ffs
create mode 100644 powerpc/rom/stage1.S
diff --git a/powerpc/rom/.gitignore b/powerpc/rom/.gitignore
new file mode 100644
index 0000000000000..a8a0dcec44720
--- /dev/null
+++ b/powerpc/rom/.gitignore
@@ -0,0 +1 @@
+*.bin
diff --git a/powerpc/rom/Makefile b/powerpc/rom/Makefile
new file mode 100644
index 0000000000000..03b8bf7e0b5d2
--- /dev/null
+++ b/powerpc/rom/Makefile
@@ -0,0 +1,36 @@
+#
+# powerpc boot_rom makefile
+#
+# Authors: Andrew Jones <drjones@redhat.com>
+#
+
+# Use SLOF's build_romfs tool (git://git.qemu-project.org/SLOF.git)
+BUILD_ROMFS ?= build_romfs
+
+ifeq ($(wildcard ../../config.mak),)
+$(error run ./configure first. See ./configure -h)
+endif
+
+include ../../config.mak
+
+.PHONY: clean distclean
+
+all: boot_rom.bin
+
+boot_rom.bin: stage1.bin
+ $(BUILD_ROMFS) rom.ffs boot_rom.bin
+
+stage1.bin: stage1.elf
+ $(OBJCOPY) -O binary $^ $@
+
+stage1.elf: stage1.o
+ $(LD) -nostdlib -Ttext=0x100 --entry=start --build-id=none -o $@ $<
+
+stage1.o: stage1.S
+ $(CC) -c -o $@ $<
+
+clean:
+ $(RM) stage1.{o,elf,bin}
+
+distclean: clean
+ $(RM) boot_rom.bin
--git a/powerpc/rom/header.img b/powerpc/rom/header.img
new file mode 100644
index 0000000000000..794129c95eec5
--- /dev/null
+++ b/powerpc/rom/header.img
@@ -0,0 +1 @@
+Key.Polynome....XXXXXXXX..Mask..XXXXXXXX.Polynome.Length....XXXX.Header.and.File.lengthXXXXXXXX... und weiter im Text!
diff --git a/powerpc/rom/rom.ffs b/powerpc/rom/rom.ffs
new file mode 100644
index 0000000000000..b7851c00c462b
--- /dev/null
+++ b/powerpc/rom/rom.ffs
@@ -0,0 +1,4 @@
+# FFile-Name Real Filename Flags ROM-Offset i/a
+#--------------|-------------------------------|-----------------------|--------------
+header header.img 0 0
+stage1 stage1.bin 1 0x100
diff --git a/powerpc/rom/stage1.S b/powerpc/rom/stage1.S
new file mode 100644
index 0000000000000..ae2c08ddce3c1
--- /dev/null
+++ b/powerpc/rom/stage1.S
@@ -0,0 +1,5 @@
+#define SPAPR_KERNEL_LOAD_ADDR 0x400000
+.text
+.globl start
+start:
+ b SPAPR_KERNEL_LOAD_ADDR - 0x100
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* [kvm-unit-tests PATCH 07/14] powerpc/ppc64: add bootloader to bounce into memory
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (5 preceding siblings ...)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 06/14] powerpc/ppc64: add boot rom source Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 08/14] powerpc/ppc64: add HV putchar Andrew Jones
` (8 subsequent siblings)
15 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
Committing the binary, since there's little point in rebuilding,
and it requires SLOF's build_romfs tool, which is not part of
this source repository.
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
powerpc/boot_rom.bin | Bin 0 -> 280 bytes
1 file changed, 0 insertions(+), 0 deletions(-)
create mode 100644 powerpc/boot_rom.bin
diff --git a/powerpc/boot_rom.bin b/powerpc/boot_rom.bin
new file mode 100644
index 0000000000000000000000000000000000000000..606e1204047afb21e1a90469d7e708496e596e7e
GIT binary patch
literal 280
zcmZQz0D~J4ilGBa!zhi6)Wnq3A_j)s#PrN$LnC8|1fxVsQC5ClYB5Y5Lt$!esR5ja
zjiDgQE-vZ}lQD!bfON3u)_MP-07;mdOctnNj8IymxFj(>73c^L`~P4eWcwelS1bkq
DUsxnq
literal 0
HcmV?d00001
--
2.4.3
^ permalink raw reply [flat|nested] 36+ messages in thread* [kvm-unit-tests PATCH 08/14] powerpc/ppc64: add HV putchar
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (6 preceding siblings ...)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 07/14] powerpc/ppc64: add bootloader to bounce into memory Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-04 3:50 ` David Gibson
2015-08-03 14:41 ` [kvm-unit-tests PATCH 09/14] powerpc/ppc64: adapt arm's setup Andrew Jones
` (7 subsequent siblings)
15 siblings, 1 reply; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
Add the hvcall for putchar and use it in puts. That, along with a
couple more lines in start to prepare for C code, and a branch to
main(), gets us "hello world". Run with
qemu-system-ppc64 -M pseries \
-bios powerpc/boot_rom.bin \
-display none -serial stdio \
-kernel powerpc/selftest.elf
(We're still not relocating yet, that comes in a later patch. Thus,
testing hello-world at this point requires a hacked QEMU and linking
the unit test at QEMU's kernel load address.)
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
lib/powerpc/io.c | 15 +++++++++++++--
powerpc/Makefile.common | 1 +
powerpc/cstart64.S | 27 +++++++++++++++++++++++++++
3 files changed, 41 insertions(+), 2 deletions(-)
diff --git a/lib/powerpc/io.c b/lib/powerpc/io.c
index cf3b6347e1e46..a7eaafeca9205 100644
--- a/lib/powerpc/io.c
+++ b/lib/powerpc/io.c
@@ -6,15 +6,26 @@
* This work is licensed under the terms of the GNU LGPL, version 2.
*/
#include <libcflat.h>
+#include <asm/spinlock.h>
+
+extern void halt(int code);
+extern void putchar(int c);
+
+static struct spinlock uart_lock;
void io_init(void)
{
}
-void puts(const char *s __unused)
+void puts(const char *s)
{
+ spin_lock(&uart_lock);
+ while (*s)
+ putchar(*s++);
+ spin_unlock(&uart_lock);
}
-void exit(int code __unused)
+void exit(int code)
{
+ halt(code);
}
diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
index 5cd3ea8085038..d6356540918a5 100644
--- a/powerpc/Makefile.common
+++ b/powerpc/Makefile.common
@@ -26,6 +26,7 @@ CFLAGS += -ffreestanding
CFLAGS += -Wextra
CFLAGS += -O2
CFLAGS += -I lib -I lib/libfdt
+CFLAGS += -Wa,-mregnames
asm-offsets = lib/$(ARCH)/asm-offsets.h
include scripts/asm-offsets.mak
diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
index d9b77f44f9e0e..d7c51cd352ee4 100644
--- a/powerpc/cstart64.S
+++ b/powerpc/cstart64.S
@@ -6,14 +6,41 @@
* This work is licensed under the terms of the GNU LGPL, version 2.
*/
+#define HVSC .long 0x44000022
+#define H_PUT_TERM_CHAR 0x58
+
+#define LOAD_REG_IMMEDIATE(reg,expr) \
+ lis reg,(expr)@highest; \
+ ori reg,reg,(expr)@higher; \
+ rldicr reg,reg,32,31; \
+ oris reg,reg,(expr)@h; \
+ ori reg,reg,(expr)@l;
+
+#define LOAD_REG_ADDR(reg,name) \
+ ld reg,name@got(r2)
+
.section .init
.globl start
start:
+ LOAD_REG_IMMEDIATE(r1, stackptr)
+ LOAD_REG_IMMEDIATE(r2, tocptr)
+ bl .main
+ bl .exit
b halt
.text
+.align 3
.globl halt
halt:
1: b 1b
+
+.globl putchar
+putchar:
+ sldi r6, r3, 56
+ li r3, H_PUT_TERM_CHAR
+ li r4, 0 /* vty-reg 0 means to use the default vty */
+ li r5, 1 /* sending just 1 byte */
+ HVSC
+ blr
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* Re: [kvm-unit-tests PATCH 08/14] powerpc/ppc64: add HV putchar
2015-08-03 14:41 ` [kvm-unit-tests PATCH 08/14] powerpc/ppc64: add HV putchar Andrew Jones
@ 2015-08-04 3:50 ` David Gibson
2015-08-04 7:33 ` Andrew Jones
0 siblings, 1 reply; 36+ messages in thread
From: David Gibson @ 2015-08-04 3:50 UTC (permalink / raw)
To: Andrew Jones; +Cc: kvm, kvm-ppc, dgibson, agraf, thuth, lvivier, pbonzini
[-- Attachment #1: Type: text/plain, Size: 3671 bytes --]
On Mon, Aug 03, 2015 at 04:41:25PM +0200, Andrew Jones wrote:
> Add the hvcall for putchar and use it in puts. That, along with a
> couple more lines in start to prepare for C code, and a branch to
> main(), gets us "hello world". Run with
>
> qemu-system-ppc64 -M pseries \
> -bios powerpc/boot_rom.bin \
> -display none -serial stdio \
> -kernel powerpc/selftest.elf
>
> (We're still not relocating yet, that comes in a later patch. Thus,
> testing hello-world at this point requires a hacked QEMU and linking
> the unit test at QEMU's kernel load address.)
>
> Signed-off-by: Andrew Jones <drjones@redhat.com>
> ---
> lib/powerpc/io.c | 15 +++++++++++++--
> powerpc/Makefile.common | 1 +
> powerpc/cstart64.S | 27 +++++++++++++++++++++++++++
> 3 files changed, 41 insertions(+), 2 deletions(-)
>
> diff --git a/lib/powerpc/io.c b/lib/powerpc/io.c
> index cf3b6347e1e46..a7eaafeca9205 100644
> --- a/lib/powerpc/io.c
> +++ b/lib/powerpc/io.c
> @@ -6,15 +6,26 @@
> * This work is licensed under the terms of the GNU LGPL, version 2.
> */
> #include <libcflat.h>
> +#include <asm/spinlock.h>
> +
> +extern void halt(int code);
> +extern void putchar(int c);
> +
> +static struct spinlock uart_lock;
>
> void io_init(void)
> {
> }
>
> -void puts(const char *s __unused)
> +void puts(const char *s)
> {
> + spin_lock(&uart_lock);
> + while (*s)
> + putchar(*s++);
> + spin_unlock(&uart_lock);
> }
>
> -void exit(int code __unused)
> +void exit(int code)
> {
> + halt(code);
> }
> diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
> index 5cd3ea8085038..d6356540918a5 100644
> --- a/powerpc/Makefile.common
> +++ b/powerpc/Makefile.common
> @@ -26,6 +26,7 @@ CFLAGS += -ffreestanding
> CFLAGS += -Wextra
> CFLAGS += -O2
> CFLAGS += -I lib -I lib/libfdt
> +CFLAGS += -Wa,-mregnames
>
> asm-offsets = lib/$(ARCH)/asm-offsets.h
> include scripts/asm-offsets.mak
> diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
> index d9b77f44f9e0e..d7c51cd352ee4 100644
> --- a/powerpc/cstart64.S
> +++ b/powerpc/cstart64.S
> @@ -6,14 +6,41 @@
> * This work is licensed under the terms of the GNU LGPL, version 2.
> */
>
> +#define HVSC .long 0x44000022
Unless your assembler is really old, you could probably use "sc 1"
inline, instead of this macro. Not that it really matters.
> +#define H_PUT_TERM_CHAR 0x58
> +
> +#define LOAD_REG_IMMEDIATE(reg,expr) \
> + lis reg,(expr)@highest; \
> + ori reg,reg,(expr)@higher; \
> + rldicr reg,reg,32,31; \
> + oris reg,reg,(expr)@h; \
> + ori reg,reg,(expr)@l;
> +
> +#define LOAD_REG_ADDR(reg,name) \
> + ld reg,name@got(r2)
> +
> .section .init
>
> .globl start
> start:
> + LOAD_REG_IMMEDIATE(r1, stackptr)
> + LOAD_REG_IMMEDIATE(r2, tocptr)
> + bl .main
> + bl .exit
Is this built for ppc64 or ppc64le? IIUC under the new ABI version
usually used on ppc64le, function descriptors are no longer used. And
even under the old ABI, I think the assembler now has the smarts to
avoid explicitly referencing the .-symbols.
> b halt
>
> .text
> +.align 3
>
> .globl halt
> halt:
> 1: b 1b
> +
> +.globl putchar
> +putchar:
> + sldi r6, r3, 56
> + li r3, H_PUT_TERM_CHAR
> + li r4, 0 /* vty-reg 0 means to use the default vty */
> + li r5, 1 /* sending just 1 byte */
> + HVSC
> + blr
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
[-- Attachment #2: Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply [flat|nested] 36+ messages in thread* Re: [kvm-unit-tests PATCH 08/14] powerpc/ppc64: add HV putchar
2015-08-04 3:50 ` David Gibson
@ 2015-08-04 7:33 ` Andrew Jones
0 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-04 7:33 UTC (permalink / raw)
To: David Gibson; +Cc: kvm, kvm-ppc, dgibson, agraf, thuth, lvivier, pbonzini
On Tue, Aug 04, 2015 at 01:50:39PM +1000, David Gibson wrote:
> On Mon, Aug 03, 2015 at 04:41:25PM +0200, Andrew Jones wrote:
> > Add the hvcall for putchar and use it in puts. That, along with a
> > couple more lines in start to prepare for C code, and a branch to
> > main(), gets us "hello world". Run with
> >
> > qemu-system-ppc64 -M pseries \
> > -bios powerpc/boot_rom.bin \
> > -display none -serial stdio \
> > -kernel powerpc/selftest.elf
> >
> > (We're still not relocating yet, that comes in a later patch. Thus,
> > testing hello-world at this point requires a hacked QEMU and linking
> > the unit test at QEMU's kernel load address.)
> >
> > Signed-off-by: Andrew Jones <drjones@redhat.com>
> > ---
> > lib/powerpc/io.c | 15 +++++++++++++--
> > powerpc/Makefile.common | 1 +
> > powerpc/cstart64.S | 27 +++++++++++++++++++++++++++
> > 3 files changed, 41 insertions(+), 2 deletions(-)
> >
> > diff --git a/lib/powerpc/io.c b/lib/powerpc/io.c
> > index cf3b6347e1e46..a7eaafeca9205 100644
> > --- a/lib/powerpc/io.c
> > +++ b/lib/powerpc/io.c
> > @@ -6,15 +6,26 @@
> > * This work is licensed under the terms of the GNU LGPL, version 2.
> > */
> > #include <libcflat.h>
> > +#include <asm/spinlock.h>
> > +
> > +extern void halt(int code);
> > +extern void putchar(int c);
> > +
> > +static struct spinlock uart_lock;
> >
> > void io_init(void)
> > {
> > }
> >
> > -void puts(const char *s __unused)
> > +void puts(const char *s)
> > {
> > + spin_lock(&uart_lock);
> > + while (*s)
> > + putchar(*s++);
> > + spin_unlock(&uart_lock);
> > }
> >
> > -void exit(int code __unused)
> > +void exit(int code)
> > {
> > + halt(code);
> > }
> > diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
> > index 5cd3ea8085038..d6356540918a5 100644
> > --- a/powerpc/Makefile.common
> > +++ b/powerpc/Makefile.common
> > @@ -26,6 +26,7 @@ CFLAGS += -ffreestanding
> > CFLAGS += -Wextra
> > CFLAGS += -O2
> > CFLAGS += -I lib -I lib/libfdt
> > +CFLAGS += -Wa,-mregnames
> >
> > asm-offsets = lib/$(ARCH)/asm-offsets.h
> > include scripts/asm-offsets.mak
> > diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
> > index d9b77f44f9e0e..d7c51cd352ee4 100644
> > --- a/powerpc/cstart64.S
> > +++ b/powerpc/cstart64.S
> > @@ -6,14 +6,41 @@
> > * This work is licensed under the terms of the GNU LGPL, version 2.
> > */
> >
> > +#define HVSC .long 0x44000022
>
> Unless your assembler is really old, you could probably use "sc 1"
> inline, instead of this macro. Not that it really matters.
It makes sense to switch to sc for a new project. I'll do that.
>
> > +#define H_PUT_TERM_CHAR 0x58
> > +
> > +#define LOAD_REG_IMMEDIATE(reg,expr) \
> > + lis reg,(expr)@highest; \
> > + ori reg,reg,(expr)@higher; \
> > + rldicr reg,reg,32,31; \
> > + oris reg,reg,(expr)@h; \
> > + ori reg,reg,(expr)@l;
> > +
> > +#define LOAD_REG_ADDR(reg,name) \
> > + ld reg,name@got(r2)
> > +
> > .section .init
> >
> > .globl start
> > start:
> > + LOAD_REG_IMMEDIATE(r1, stackptr)
> > + LOAD_REG_IMMEDIATE(r2, tocptr)
> > + bl .main
> > + bl .exit
>
> Is this built for ppc64 or ppc64le? IIUC under the new ABI version
> usually used on ppc64le, function descriptors are no longer used. And
> even under the old ABI, I think the assembler now has the smarts to
> avoid explicitly referencing the .-symbols.
For now only building for ppc64, but eventually both. I'll remove the
dots.
Thanks,
drew
>
> > b halt
> >
> > .text
> > +.align 3
> >
> > .globl halt
> > halt:
> > 1: b 1b
> > +
> > +.globl putchar
> > +putchar:
> > + sldi r6, r3, 56
> > + li r3, H_PUT_TERM_CHAR
> > + li r4, 0 /* vty-reg 0 means to use the default vty */
> > + li r5, 1 /* sending just 1 byte */
> > + HVSC
> > + blr
>
> --
> David Gibson | I'll have my music baroque, and my code
> david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
> | _way_ _around_!
> http://www.ozlabs.org/~dgibson
^ permalink raw reply [flat|nested] 36+ messages in thread
* [kvm-unit-tests PATCH 09/14] powerpc/ppc64: adapt arm's setup
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (7 preceding siblings ...)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 08/14] powerpc/ppc64: add HV putchar Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 10/14] powerpc/ppc64: relocate linker VMAs Andrew Jones
` (6 subsequent siblings)
15 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
Copy arm's setup code (also DT based) over to powerpc, adapting
it a bit. Also bring over arm's setup selftest, giving powerpc
its first test.
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
lib/powerpc/asm/setup.h | 27 ++++++++++++++++
lib/powerpc/setup.c | 82 +++++++++++++++++++++++++++++++++++++++++++++++++
lib/ppc64/asm/setup.h | 1 +
powerpc/cstart64.S | 9 ++++++
powerpc/selftest.c | 64 ++++++++++++++++++++++++++++++++++++--
5 files changed, 180 insertions(+), 3 deletions(-)
create mode 100644 lib/powerpc/asm/setup.h
create mode 100644 lib/ppc64/asm/setup.h
diff --git a/lib/powerpc/asm/setup.h b/lib/powerpc/asm/setup.h
new file mode 100644
index 0000000000000..745226a8cf067
--- /dev/null
+++ b/lib/powerpc/asm/setup.h
@@ -0,0 +1,27 @@
+#ifndef _ASMPOWERPC_SETUP_H_
+#define _ASMPOWERPC_SETUP_H_
+/*
+ * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2.
+ */
+#include <libcflat.h>
+
+#define NR_CPUS 8 /* arbitrarily set for now */
+extern u32 cpus[NR_CPUS];
+extern int nr_cpus;
+
+extern phys_addr_t __physical_start, __physical_end;
+
+#define PHYSICAL_START (__physical_start)
+#define PHYSICAL_END (__physical_end)
+
+#ifdef __powerpc64__
+#define L1_CACHE_SHIFT 7
+#else
+#define L1_CACHE_SHIFT 5
+#endif
+#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT)
+#define SMP_CACHE_BYTES L1_CACHE_BYTES
+
+#endif /* _ASMPOWERPC_SETUP_H_ */
diff --git a/lib/powerpc/setup.c b/lib/powerpc/setup.c
index e69de29bb2d1d..1794d1f39066a 100644
--- a/lib/powerpc/setup.c
+++ b/lib/powerpc/setup.c
@@ -0,0 +1,82 @@
+/*
+ * Initialize machine setup information and I/O.
+ *
+ * After running setup() unit tests may query how many cpus they have
+ * (nr_cpus), how much memory they have (PHYSICAL_END - PHYSICAL_START),
+ * may use dynamic memory allocation (malloc, etc.), printf, and exit.
+ * Finally, argc and argv are also ready to be passed to main().
+ *
+ * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2.
+ */
+#include <libcflat.h>
+#include <libfdt/libfdt.h>
+#include <devicetree.h>
+#include <alloc.h>
+#include <asm/setup.h>
+#include <asm/page.h>
+
+extern unsigned long stacktop;
+extern void io_init(void);
+extern void setup_args(const char *args);
+
+u32 cpus[NR_CPUS] = { [0 ... NR_CPUS-1] = (~0U) };
+int nr_cpus;
+
+phys_addr_t __physical_start, __physical_end;
+
+static void cpu_set(int fdtnode __unused, u32 regval, void *info __unused)
+{
+ int cpu = nr_cpus++;
+ assert(cpu < NR_CPUS);
+ cpus[cpu] = regval;
+}
+
+static void cpu_init(void)
+{
+ nr_cpus = 0;
+ assert(dt_for_each_cpu_node(cpu_set, NULL) == 0);
+}
+
+static void mem_init(phys_addr_t freemem_start)
+{
+ /* we only expect one membank to be defined in the DT */
+ struct dt_pbus_reg regs[1];
+ phys_addr_t mem_start, mem_end;
+
+ assert(dt_get_memory_params(regs, 1));
+
+ mem_start = regs[0].addr;
+ mem_end = mem_start + regs[0].size;
+
+// assert(!(mem_start & ~PHYS_MASK) && !((mem_end-1) & ~PHYS_MASK));
+ assert(freemem_start >= mem_start && freemem_start < mem_end);
+
+ __physical_start = mem_start; /* PHYSICAL_START */
+ __physical_end = mem_end; /* PHYSICAL_END */
+
+ phys_alloc_init(freemem_start, mem_end - freemem_start);
+ phys_alloc_set_minimum_alignment(SMP_CACHE_BYTES);
+}
+
+void setup(const void *fdt)
+{
+ const char *bootargs;
+ u32 fdt_size;
+
+ /*
+ * Move the fdt to just above the stack. The free memory
+ * then starts just after the fdt.
+ */
+ fdt_size = fdt_totalsize(fdt);
+ assert(fdt_move(fdt, &stacktop, fdt_size) == 0);
+ assert(dt_init(&stacktop) == 0);
+
+ mem_init(PAGE_ALIGN((unsigned long)&stacktop + fdt_size));
+ io_init();
+ cpu_init();
+
+ assert(dt_get_bootargs(&bootargs) == 0);
+ setup_args(bootargs);
+}
diff --git a/lib/ppc64/asm/setup.h b/lib/ppc64/asm/setup.h
new file mode 100644
index 0000000000000..20192985928a4
--- /dev/null
+++ b/lib/ppc64/asm/setup.h
@@ -0,0 +1 @@
+#include "../../powerpc/asm/setup.h"
diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
index d7c51cd352ee4..141d4563563d5 100644
--- a/powerpc/cstart64.S
+++ b/powerpc/cstart64.S
@@ -21,10 +21,19 @@
.section .init
+/*
+ * start is the entry point. r3 points to the DTB
+ */
.globl start
start:
LOAD_REG_IMMEDIATE(r1, stackptr)
LOAD_REG_IMMEDIATE(r2, tocptr)
+ bl .setup
+
+ /* run the test */
+ LOAD_REG_IMMEDIATE(r5, __argc)
+ LOAD_REG_IMMEDIATE(r4, __argv)
+ lwz r3, 0(r5)
bl .main
bl .exit
b halt
diff --git a/powerpc/selftest.c b/powerpc/selftest.c
index 2f2a5215dd55c..b1d5e73029f46 100644
--- a/powerpc/selftest.c
+++ b/powerpc/selftest.c
@@ -1,7 +1,65 @@
+/*
+ * Test the framework itself. These tests confirm that setup works.
+ *
+ * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2.
+ */
#include <libcflat.h>
+#include <util.h>
+#include <alloc.h> /* phys_addr_t */
+#include <asm/setup.h>
-int main(void)
+static void check_setup(int argc, char **argv)
{
- printf("hello world\n");
- return 0;
+ int nr_tests = 0, i;
+ char *var;
+ long val;
+
+ for (i = 0; i < argc; ++i) {
+
+ var = parse_keyval(argv[i], &val);
+ if (!var)
+ continue;
+
+ report_prefix_push(var);
+
+ if (strcmp(var, "mem") == 0) {
+
+ phys_addr_t memsize = PHYSICAL_END - PHYSICAL_START;
+ phys_addr_t expected = ((phys_addr_t)val)*1024*1024;
+
+ report("size = %d MB", memsize == expected,
+ memsize/1024/1024);
+ ++nr_tests;
+
+ } else if (strcmp(var, "smp") == 0) {
+
+ report("nr_cpus = %d", nr_cpus == (int)val, nr_cpus);
+ ++nr_tests;
+ }
+
+ report_prefix_pop();
+ }
+
+ if (nr_tests < 2)
+ report_abort("missing input");
+}
+
+int main(int argc, char **argv)
+{
+ report_prefix_push("selftest");
+
+ if (argc < 1)
+ report_abort("no test specified");
+
+ report_prefix_push(argv[0]);
+
+ if (strcmp(argv[0], "setup") == 0) {
+
+ check_setup(argc-1, &argv[1]);
+
+ }
+
+ return report_summary();
}
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* [kvm-unit-tests PATCH 10/14] powerpc/ppc64: relocate linker VMAs
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (8 preceding siblings ...)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 09/14] powerpc/ppc64: adapt arm's setup Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-04 3:53 ` David Gibson
2015-08-03 14:41 ` [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off Andrew Jones
` (5 subsequent siblings)
15 siblings, 1 reply; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
QEMU loads the unit test, but due to the way it translates the
unit test's linker VMA to the LMA, we can't just link such that
VMA == LMA. Thus, we link with VMA == 0x0, and then deal with
relocation.
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
configure | 2 ++
powerpc/Makefile.common | 13 +++++++++---
powerpc/Makefile.ppc64 | 1 +
powerpc/cstart64.S | 40 +++++++++++++++++++++++++++++++----
powerpc/flat.lds | 13 +++++++++++-
powerpc/reloc64.c | 55 +++++++++++++++++++++++++++++++++++++++++++++++++
6 files changed, 116 insertions(+), 8 deletions(-)
create mode 100644 powerpc/reloc64.c
diff --git a/configure b/configure
index b367224093369..b2ad199da7873 100755
--- a/configure
+++ b/configure
@@ -5,6 +5,7 @@ kerneldir=/lib/modules/$(uname -r)/build
cc=gcc
ld=ld
objcopy=objcopy
+objdump=objdump
ar=ar
arch=`uname -m | sed -e s/i.86/i386/ | sed -e 's/arm.*/arm/'`
host=$arch
@@ -132,6 +133,7 @@ PROCESSOR=$processor
CC=$cross_prefix$cc
LD=$cross_prefix$ld
OBJCOPY=$cross_prefix$objcopy
+OBJDUMP=$cross_prefix$objdump
AR=$cross_prefix$ar
API=$api
TEST_DIR=$testdir
diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
index d6356540918a5..b130342dee60e 100644
--- a/powerpc/Makefile.common
+++ b/powerpc/Makefile.common
@@ -27,6 +27,7 @@ CFLAGS += -Wextra
CFLAGS += -O2
CFLAGS += -I lib -I lib/libfdt
CFLAGS += -Wa,-mregnames
+CFLAGS += -fpie
asm-offsets = lib/$(ARCH)/asm-offsets.h
include scripts/asm-offsets.mak
@@ -43,11 +44,17 @@ libgcc := $(shell $(CC) $(machine) --print-libgcc-file-name)
start_addr := $(shell printf "%x\n" $$(( $(phys_base) + $(kernel_offset) )))
FLATLIBS = $(libcflat) $(LIBFDT_archive) $(libgcc)
-%.elf: LDFLAGS = $(CFLAGS) -nostdlib
+%.elf: LDFLAGS = $(CFLAGS) -nostdlib -pie
%.elf: %.o $(FLATLIBS) powerpc/flat.lds
$(CC) $(LDFLAGS) -o $@ \
-Wl,-T,powerpc/flat.lds,--build-id=none,-Ttext=$(start_addr) \
$(filter %.o, $^) $(FLATLIBS)
+ @echo -n Checking $@ for unsupported reloc types...
+ @if $(OBJDUMP) -R $@ | grep R_ | grep -v R_PPC64_RELATIVE; then \
+ false; \
+ else \
+ echo " looks good."; \
+ fi
powerpc_clean: libfdt_clean asm_offsets_clean
$(RM) $(TEST_DIR)/*.{o,elf} \
@@ -59,5 +66,5 @@ generated_files = $(asm-offsets)
test_cases: $(generated_files) $(tests-common) $(tests)
-$(TEST_DIR)/$(TEST).elf: $(cstart.o) $(TEST_DIR)/$(TEST).o
-$(TEST_DIR)/selftest.elf: $(cstart.o) $(TEST_DIR)/selftest.o
+$(TEST_DIR)/$(TEST).elf: $(cstart.o) $(reloc.o) $(TEST_DIR)/$(TEST).o
+$(TEST_DIR)/selftest.elf: $(cstart.o) $(reloc.o) $(TEST_DIR)/selftest.o
diff --git a/powerpc/Makefile.ppc64 b/powerpc/Makefile.ppc64
index 7c61933dfa8ba..7274e0d98b5a5 100644
--- a/powerpc/Makefile.ppc64
+++ b/powerpc/Makefile.ppc64
@@ -8,6 +8,7 @@ ldarch = elf64-powerpc #elf64-powerpcle (eventually)
kernel_offset = 0x0
cstart.o = $(TEST_DIR)/cstart64.o
+reloc.o = $(TEST_DIR)/reloc64.o
cflatobjs += lib/ppc64/processor.o
cflatobjs += lib/ppc64/spinlock.o
diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
index 141d4563563d5..8edaaa6e251fc 100644
--- a/powerpc/cstart64.S
+++ b/powerpc/cstart64.S
@@ -26,18 +26,50 @@
*/
.globl start
start:
- LOAD_REG_IMMEDIATE(r1, stackptr)
- LOAD_REG_IMMEDIATE(r2, tocptr)
+ /*
+ * We were loaded at QEMU's kernel load address, but we're not
+ * allowed to link there due to how QEMU deals with linker VMAs,
+ * so we just linked at zero. This means the first thing to do is
+ * to find our stack and toc, and then do a relocate.
+ */
+ bl . + 4
+0: mflr r31
+ subi r31, r31, 0b - start /* QEMU's kernel load address */
+ ld r1, (p_stack - start)(r31)
+ ld r2, (p_toc - start)(r31)
+ add r1, r1, r31
+ add r2, r2, r31
+
+ /* save DTB pointer */
+ std r3, 56(r1)
+
+ /*
+ * Call relocate. relocate is C code, but careful to not use
+ * any global references, as they may use absolute addresses,
+ * which are, obviously, not yet relocated.
+ */
+ mr r3, r31
+ ld r4, (p_dyn - start)(r31)
+ add r4, r4, r31
+ bl .relocate
+
+ /* complete setup */
+ ld r3, 56(r1)
bl .setup
/* run the test */
- LOAD_REG_IMMEDIATE(r5, __argc)
- LOAD_REG_IMMEDIATE(r4, __argv)
+ LOAD_REG_ADDR(r5, __argc)
+ LOAD_REG_ADDR(r4, __argv)
lwz r3, 0(r5)
bl .main
bl .exit
b halt
+.align 3
+p_stack: .llong stackptr
+p_toc: .llong tocptr
+p_dyn: .llong dynamic_start
+
.text
.align 3
diff --git a/powerpc/flat.lds b/powerpc/flat.lds
index bd075efb2c51b..8a573d27346de 100644
--- a/powerpc/flat.lds
+++ b/powerpc/flat.lds
@@ -6,11 +6,22 @@ SECTIONS
etext = .;
.opd : { *(.opd) }
. = ALIGN(16);
+ .dynamic : {
+ dynamic_start = .;
+ *(.dynamic)
+ }
+ .dynsym : {
+ dynsym_start = .;
+ *(.dynsym)
+ }
+ .rela.dyn : { *(.rela*) }
+ . = ALIGN(16);
.data : {
*(.data)
+ *(.data.rel*)
}
. = ALIGN(16);
- .rodata : { *(.rodata) }
+ .rodata : { *(.rodata) *(.rodata.*) }
. = ALIGN(16);
.bss : { *(.bss) }
. = ALIGN(16);
diff --git a/powerpc/reloc64.c b/powerpc/reloc64.c
new file mode 100644
index 0000000000000..2804823bdfee3
--- /dev/null
+++ b/powerpc/reloc64.c
@@ -0,0 +1,55 @@
+/*
+ * relocate R_PPC_RELATIVE RELA entries. Normally this is done in
+ * assembly code to avoid the risk of using absolute addresses before
+ * they're relocated. We use C, but cautiously (no global references).
+ *
+ * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2.
+ */
+#define DT_NULL 0
+#define DT_RELA 7
+#define DT_RELACOUNT 0x6ffffff9
+#define R_PPC_RELATIVE 22
+
+struct elf64_dyn {
+ signed long long tag;
+ unsigned long long val;
+};
+
+#define RELA_GET_TYPE(rela_ptr) ((rela_ptr)->info & 0xffffffff)
+struct elf64_rela {
+ unsigned long long offset;
+ unsigned long long info;
+ signed long long addend;
+};
+
+void relocate(unsigned long load_addr, struct elf64_dyn *dyn_table)
+{
+ unsigned long long rela_addr = 0, rela_count = 0, *addr;
+ struct elf64_dyn *d = dyn_table;
+ struct elf64_rela *r;
+
+ while (d && d->tag != DT_NULL) {
+ if (d->tag == DT_RELA)
+ rela_addr = d->val;
+ else if (d->tag == DT_RELACOUNT)
+ rela_count = d->val;
+ if (rela_addr && rela_count)
+ break;
+ ++d;
+ }
+
+ if (!rela_addr || !rela_count)
+ return;
+
+ r = (void *)(rela_addr + load_addr);
+
+ while (rela_count--) {
+ if (RELA_GET_TYPE(r) == R_PPC_RELATIVE) {
+ addr = (void *)(r->offset + load_addr);
+ *addr = r->addend + load_addr;
+ }
+ ++r;
+ }
+}
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* Re: [kvm-unit-tests PATCH 10/14] powerpc/ppc64: relocate linker VMAs
2015-08-03 14:41 ` [kvm-unit-tests PATCH 10/14] powerpc/ppc64: relocate linker VMAs Andrew Jones
@ 2015-08-04 3:53 ` David Gibson
2015-08-04 7:35 ` Andrew Jones
0 siblings, 1 reply; 36+ messages in thread
From: David Gibson @ 2015-08-04 3:53 UTC (permalink / raw)
To: Andrew Jones; +Cc: kvm, kvm-ppc, dgibson, agraf, thuth, lvivier, pbonzini
[-- Attachment #1: Type: text/plain, Size: 7603 bytes --]
On Mon, Aug 03, 2015 at 04:41:27PM +0200, Andrew Jones wrote:
> QEMU loads the unit test, but due to the way it translates the
> unit test's linker VMA to the LMA, we can't just link such that
> VMA == LMA. Thus, we link with VMA == 0x0, and then deal with
> relocation.
>
> Signed-off-by: Andrew Jones <drjones@redhat.com>
> ---
> configure | 2 ++
> powerpc/Makefile.common | 13 +++++++++---
> powerpc/Makefile.ppc64 | 1 +
> powerpc/cstart64.S | 40 +++++++++++++++++++++++++++++++----
> powerpc/flat.lds | 13 +++++++++++-
> powerpc/reloc64.c | 55 +++++++++++++++++++++++++++++++++++++++++++++++++
> 6 files changed, 116 insertions(+), 8 deletions(-)
> create mode 100644 powerpc/reloc64.c
>
> diff --git a/configure b/configure
> index b367224093369..b2ad199da7873 100755
> --- a/configure
> +++ b/configure
> @@ -5,6 +5,7 @@ kerneldir=/lib/modules/$(uname -r)/build
> cc=gcc
> ld=ld
> objcopy=objcopy
> +objdump=objdump
> ar=ar
> arch=`uname -m | sed -e s/i.86/i386/ | sed -e 's/arm.*/arm/'`
> host=$arch
> @@ -132,6 +133,7 @@ PROCESSOR=$processor
> CC=$cross_prefix$cc
> LD=$cross_prefix$ld
> OBJCOPY=$cross_prefix$objcopy
> +OBJDUMP=$cross_prefix$objdump
> AR=$cross_prefix$ar
> API=$api
> TEST_DIR=$testdir
> diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
> index d6356540918a5..b130342dee60e 100644
> --- a/powerpc/Makefile.common
> +++ b/powerpc/Makefile.common
> @@ -27,6 +27,7 @@ CFLAGS += -Wextra
> CFLAGS += -O2
> CFLAGS += -I lib -I lib/libfdt
> CFLAGS += -Wa,-mregnames
> +CFLAGS += -fpie
>
> asm-offsets = lib/$(ARCH)/asm-offsets.h
> include scripts/asm-offsets.mak
> @@ -43,11 +44,17 @@ libgcc := $(shell $(CC) $(machine) --print-libgcc-file-name)
> start_addr := $(shell printf "%x\n" $$(( $(phys_base) + $(kernel_offset) )))
>
> FLATLIBS = $(libcflat) $(LIBFDT_archive) $(libgcc)
> -%.elf: LDFLAGS = $(CFLAGS) -nostdlib
> +%.elf: LDFLAGS = $(CFLAGS) -nostdlib -pie
> %.elf: %.o $(FLATLIBS) powerpc/flat.lds
> $(CC) $(LDFLAGS) -o $@ \
> -Wl,-T,powerpc/flat.lds,--build-id=none,-Ttext=$(start_addr) \
> $(filter %.o, $^) $(FLATLIBS)
> + @echo -n Checking $@ for unsupported reloc types...
> + @if $(OBJDUMP) -R $@ | grep R_ | grep -v R_PPC64_RELATIVE; then \
> + false; \
> + else \
> + echo " looks good."; \
> + fi
>
> powerpc_clean: libfdt_clean asm_offsets_clean
> $(RM) $(TEST_DIR)/*.{o,elf} \
> @@ -59,5 +66,5 @@ generated_files = $(asm-offsets)
>
> test_cases: $(generated_files) $(tests-common) $(tests)
>
> -$(TEST_DIR)/$(TEST).elf: $(cstart.o) $(TEST_DIR)/$(TEST).o
> -$(TEST_DIR)/selftest.elf: $(cstart.o) $(TEST_DIR)/selftest.o
> +$(TEST_DIR)/$(TEST).elf: $(cstart.o) $(reloc.o) $(TEST_DIR)/$(TEST).o
> +$(TEST_DIR)/selftest.elf: $(cstart.o) $(reloc.o) $(TEST_DIR)/selftest.o
> diff --git a/powerpc/Makefile.ppc64 b/powerpc/Makefile.ppc64
> index 7c61933dfa8ba..7274e0d98b5a5 100644
> --- a/powerpc/Makefile.ppc64
> +++ b/powerpc/Makefile.ppc64
> @@ -8,6 +8,7 @@ ldarch = elf64-powerpc #elf64-powerpcle (eventually)
> kernel_offset = 0x0
>
> cstart.o = $(TEST_DIR)/cstart64.o
> +reloc.o = $(TEST_DIR)/reloc64.o
> cflatobjs += lib/ppc64/processor.o
> cflatobjs += lib/ppc64/spinlock.o
>
> diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
> index 141d4563563d5..8edaaa6e251fc 100644
> --- a/powerpc/cstart64.S
> +++ b/powerpc/cstart64.S
> @@ -26,18 +26,50 @@
> */
> .globl start
> start:
> - LOAD_REG_IMMEDIATE(r1, stackptr)
> - LOAD_REG_IMMEDIATE(r2, tocptr)
> + /*
> + * We were loaded at QEMU's kernel load address, but we're not
> + * allowed to link there due to how QEMU deals with linker VMAs,
> + * so we just linked at zero. This means the first thing to do is
> + * to find our stack and toc, and then do a relocate.
> + */
> + bl . + 4
"bl 0f" might make the connection to the following instructions
clearer.
> +0: mflr r31
> + subi r31, r31, 0b - start /* QEMU's kernel load address */
> + ld r1, (p_stack - start)(r31)
> + ld r2, (p_toc - start)(r31)
> + add r1, r1, r31
> + add r2, r2, r31
> +
> + /* save DTB pointer */
> + std r3, 56(r1)
> +
> + /*
> + * Call relocate. relocate is C code, but careful to not use
> + * any global references, as they may use absolute addresses,
> + * which are, obviously, not yet relocated.
> + */
> + mr r3, r31
> + ld r4, (p_dyn - start)(r31)
> + add r4, r4, r31
> + bl .relocate
> +
> + /* complete setup */
> + ld r3, 56(r1)
> bl .setup
>
> /* run the test */
> - LOAD_REG_IMMEDIATE(r5, __argc)
> - LOAD_REG_IMMEDIATE(r4, __argv)
> + LOAD_REG_ADDR(r5, __argc)
> + LOAD_REG_ADDR(r4, __argv)
> lwz r3, 0(r5)
> bl .main
> bl .exit
> b halt
>
> +.align 3
> +p_stack: .llong stackptr
> +p_toc: .llong tocptr
> +p_dyn: .llong dynamic_start
> +
> .text
> .align 3
>
> diff --git a/powerpc/flat.lds b/powerpc/flat.lds
> index bd075efb2c51b..8a573d27346de 100644
> --- a/powerpc/flat.lds
> +++ b/powerpc/flat.lds
> @@ -6,11 +6,22 @@ SECTIONS
> etext = .;
> .opd : { *(.opd) }
> . = ALIGN(16);
> + .dynamic : {
> + dynamic_start = .;
> + *(.dynamic)
> + }
> + .dynsym : {
> + dynsym_start = .;
> + *(.dynsym)
> + }
> + .rela.dyn : { *(.rela*) }
> + . = ALIGN(16);
> .data : {
> *(.data)
> + *(.data.rel*)
> }
> . = ALIGN(16);
> - .rodata : { *(.rodata) }
> + .rodata : { *(.rodata) *(.rodata.*) }
> . = ALIGN(16);
> .bss : { *(.bss) }
> . = ALIGN(16);
> diff --git a/powerpc/reloc64.c b/powerpc/reloc64.c
> new file mode 100644
> index 0000000000000..2804823bdfee3
> --- /dev/null
> +++ b/powerpc/reloc64.c
> @@ -0,0 +1,55 @@
> +/*
> + * relocate R_PPC_RELATIVE RELA entries. Normally this is done in
> + * assembly code to avoid the risk of using absolute addresses before
> + * they're relocated. We use C, but cautiously (no global references).
> + *
> + * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
> + *
> + * This work is licensed under the terms of the GNU LGPL, version 2.
> + */
> +#define DT_NULL 0
> +#define DT_RELA 7
> +#define DT_RELACOUNT 0x6ffffff9
> +#define R_PPC_RELATIVE 22
> +
> +struct elf64_dyn {
> + signed long long tag;
> + unsigned long long val;
> +};
> +
> +#define RELA_GET_TYPE(rela_ptr) ((rela_ptr)->info & 0xffffffff)
> +struct elf64_rela {
> + unsigned long long offset;
> + unsigned long long info;
> + signed long long addend;
> +};
> +
> +void relocate(unsigned long load_addr, struct elf64_dyn *dyn_table)
> +{
> + unsigned long long rela_addr = 0, rela_count = 0, *addr;
> + struct elf64_dyn *d = dyn_table;
> + struct elf64_rela *r;
> +
> + while (d && d->tag != DT_NULL) {
> + if (d->tag == DT_RELA)
> + rela_addr = d->val;
> + else if (d->tag == DT_RELACOUNT)
> + rela_count = d->val;
> + if (rela_addr && rela_count)
> + break;
> + ++d;
> + }
> +
> + if (!rela_addr || !rela_count)
> + return;
> +
> + r = (void *)(rela_addr + load_addr);
> +
> + while (rela_count--) {
> + if (RELA_GET_TYPE(r) == R_PPC_RELATIVE) {
> + addr = (void *)(r->offset + load_addr);
> + *addr = r->addend + load_addr;
> + }
> + ++r;
> + }
> +}
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
[-- Attachment #2: Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply [flat|nested] 36+ messages in thread* Re: [kvm-unit-tests PATCH 10/14] powerpc/ppc64: relocate linker VMAs
2015-08-04 3:53 ` David Gibson
@ 2015-08-04 7:35 ` Andrew Jones
0 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-04 7:35 UTC (permalink / raw)
To: David Gibson; +Cc: kvm, kvm-ppc, dgibson, agraf, thuth, lvivier, pbonzini
On Tue, Aug 04, 2015 at 01:53:24PM +1000, David Gibson wrote:
> On Mon, Aug 03, 2015 at 04:41:27PM +0200, Andrew Jones wrote:
> > QEMU loads the unit test, but due to the way it translates the
> > unit test's linker VMA to the LMA, we can't just link such that
> > VMA == LMA. Thus, we link with VMA == 0x0, and then deal with
> > relocation.
> >
> > Signed-off-by: Andrew Jones <drjones@redhat.com>
> > ---
> > configure | 2 ++
> > powerpc/Makefile.common | 13 +++++++++---
> > powerpc/Makefile.ppc64 | 1 +
> > powerpc/cstart64.S | 40 +++++++++++++++++++++++++++++++----
> > powerpc/flat.lds | 13 +++++++++++-
> > powerpc/reloc64.c | 55 +++++++++++++++++++++++++++++++++++++++++++++++++
> > 6 files changed, 116 insertions(+), 8 deletions(-)
> > create mode 100644 powerpc/reloc64.c
> >
> > diff --git a/configure b/configure
> > index b367224093369..b2ad199da7873 100755
> > --- a/configure
> > +++ b/configure
> > @@ -5,6 +5,7 @@ kerneldir=/lib/modules/$(uname -r)/build
> > cc=gcc
> > ld=ld
> > objcopy=objcopy
> > +objdump=objdump
> > ar=ar
> > arch=`uname -m | sed -e s/i.86/i386/ | sed -e 's/arm.*/arm/'`
> > host=$arch
> > @@ -132,6 +133,7 @@ PROCESSOR=$processor
> > CC=$cross_prefix$cc
> > LD=$cross_prefix$ld
> > OBJCOPY=$cross_prefix$objcopy
> > +OBJDUMP=$cross_prefix$objdump
> > AR=$cross_prefix$ar
> > API=$api
> > TEST_DIR=$testdir
> > diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
> > index d6356540918a5..b130342dee60e 100644
> > --- a/powerpc/Makefile.common
> > +++ b/powerpc/Makefile.common
> > @@ -27,6 +27,7 @@ CFLAGS += -Wextra
> > CFLAGS += -O2
> > CFLAGS += -I lib -I lib/libfdt
> > CFLAGS += -Wa,-mregnames
> > +CFLAGS += -fpie
> >
> > asm-offsets = lib/$(ARCH)/asm-offsets.h
> > include scripts/asm-offsets.mak
> > @@ -43,11 +44,17 @@ libgcc := $(shell $(CC) $(machine) --print-libgcc-file-name)
> > start_addr := $(shell printf "%x\n" $$(( $(phys_base) + $(kernel_offset) )))
> >
> > FLATLIBS = $(libcflat) $(LIBFDT_archive) $(libgcc)
> > -%.elf: LDFLAGS = $(CFLAGS) -nostdlib
> > +%.elf: LDFLAGS = $(CFLAGS) -nostdlib -pie
> > %.elf: %.o $(FLATLIBS) powerpc/flat.lds
> > $(CC) $(LDFLAGS) -o $@ \
> > -Wl,-T,powerpc/flat.lds,--build-id=none,-Ttext=$(start_addr) \
> > $(filter %.o, $^) $(FLATLIBS)
> > + @echo -n Checking $@ for unsupported reloc types...
> > + @if $(OBJDUMP) -R $@ | grep R_ | grep -v R_PPC64_RELATIVE; then \
> > + false; \
> > + else \
> > + echo " looks good."; \
> > + fi
> >
> > powerpc_clean: libfdt_clean asm_offsets_clean
> > $(RM) $(TEST_DIR)/*.{o,elf} \
> > @@ -59,5 +66,5 @@ generated_files = $(asm-offsets)
> >
> > test_cases: $(generated_files) $(tests-common) $(tests)
> >
> > -$(TEST_DIR)/$(TEST).elf: $(cstart.o) $(TEST_DIR)/$(TEST).o
> > -$(TEST_DIR)/selftest.elf: $(cstart.o) $(TEST_DIR)/selftest.o
> > +$(TEST_DIR)/$(TEST).elf: $(cstart.o) $(reloc.o) $(TEST_DIR)/$(TEST).o
> > +$(TEST_DIR)/selftest.elf: $(cstart.o) $(reloc.o) $(TEST_DIR)/selftest.o
> > diff --git a/powerpc/Makefile.ppc64 b/powerpc/Makefile.ppc64
> > index 7c61933dfa8ba..7274e0d98b5a5 100644
> > --- a/powerpc/Makefile.ppc64
> > +++ b/powerpc/Makefile.ppc64
> > @@ -8,6 +8,7 @@ ldarch = elf64-powerpc #elf64-powerpcle (eventually)
> > kernel_offset = 0x0
> >
> > cstart.o = $(TEST_DIR)/cstart64.o
> > +reloc.o = $(TEST_DIR)/reloc64.o
> > cflatobjs += lib/ppc64/processor.o
> > cflatobjs += lib/ppc64/spinlock.o
> >
> > diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
> > index 141d4563563d5..8edaaa6e251fc 100644
> > --- a/powerpc/cstart64.S
> > +++ b/powerpc/cstart64.S
> > @@ -26,18 +26,50 @@
> > */
> > .globl start
> > start:
> > - LOAD_REG_IMMEDIATE(r1, stackptr)
> > - LOAD_REG_IMMEDIATE(r2, tocptr)
> > + /*
> > + * We were loaded at QEMU's kernel load address, but we're not
> > + * allowed to link there due to how QEMU deals with linker VMAs,
> > + * so we just linked at zero. This means the first thing to do is
> > + * to find our stack and toc, and then do a relocate.
> > + */
> > + bl . + 4
>
> "bl 0f" might make the connection to the following instructions
> clearer.
Thanks, I'll change it.
drew
>
> > +0: mflr r31
> > + subi r31, r31, 0b - start /* QEMU's kernel load address */
> > + ld r1, (p_stack - start)(r31)
> > + ld r2, (p_toc - start)(r31)
> > + add r1, r1, r31
> > + add r2, r2, r31
> > +
> > + /* save DTB pointer */
> > + std r3, 56(r1)
> > +
> > + /*
> > + * Call relocate. relocate is C code, but careful to not use
> > + * any global references, as they may use absolute addresses,
> > + * which are, obviously, not yet relocated.
> > + */
> > + mr r3, r31
> > + ld r4, (p_dyn - start)(r31)
> > + add r4, r4, r31
> > + bl .relocate
> > +
> > + /* complete setup */
> > + ld r3, 56(r1)
> > bl .setup
> >
> > /* run the test */
> > - LOAD_REG_IMMEDIATE(r5, __argc)
> > - LOAD_REG_IMMEDIATE(r4, __argv)
> > + LOAD_REG_ADDR(r5, __argc)
> > + LOAD_REG_ADDR(r4, __argv)
> > lwz r3, 0(r5)
> > bl .main
> > bl .exit
> > b halt
> >
> > +.align 3
> > +p_stack: .llong stackptr
> > +p_toc: .llong tocptr
> > +p_dyn: .llong dynamic_start
> > +
> > .text
> > .align 3
> >
> > diff --git a/powerpc/flat.lds b/powerpc/flat.lds
> > index bd075efb2c51b..8a573d27346de 100644
> > --- a/powerpc/flat.lds
> > +++ b/powerpc/flat.lds
> > @@ -6,11 +6,22 @@ SECTIONS
> > etext = .;
> > .opd : { *(.opd) }
> > . = ALIGN(16);
> > + .dynamic : {
> > + dynamic_start = .;
> > + *(.dynamic)
> > + }
> > + .dynsym : {
> > + dynsym_start = .;
> > + *(.dynsym)
> > + }
> > + .rela.dyn : { *(.rela*) }
> > + . = ALIGN(16);
> > .data : {
> > *(.data)
> > + *(.data.rel*)
> > }
> > . = ALIGN(16);
> > - .rodata : { *(.rodata) }
> > + .rodata : { *(.rodata) *(.rodata.*) }
> > . = ALIGN(16);
> > .bss : { *(.bss) }
> > . = ALIGN(16);
> > diff --git a/powerpc/reloc64.c b/powerpc/reloc64.c
> > new file mode 100644
> > index 0000000000000..2804823bdfee3
> > --- /dev/null
> > +++ b/powerpc/reloc64.c
> > @@ -0,0 +1,55 @@
> > +/*
> > + * relocate R_PPC_RELATIVE RELA entries. Normally this is done in
> > + * assembly code to avoid the risk of using absolute addresses before
> > + * they're relocated. We use C, but cautiously (no global references).
> > + *
> > + * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
> > + *
> > + * This work is licensed under the terms of the GNU LGPL, version 2.
> > + */
> > +#define DT_NULL 0
> > +#define DT_RELA 7
> > +#define DT_RELACOUNT 0x6ffffff9
> > +#define R_PPC_RELATIVE 22
> > +
> > +struct elf64_dyn {
> > + signed long long tag;
> > + unsigned long long val;
> > +};
> > +
> > +#define RELA_GET_TYPE(rela_ptr) ((rela_ptr)->info & 0xffffffff)
> > +struct elf64_rela {
> > + unsigned long long offset;
> > + unsigned long long info;
> > + signed long long addend;
> > +};
> > +
> > +void relocate(unsigned long load_addr, struct elf64_dyn *dyn_table)
> > +{
> > + unsigned long long rela_addr = 0, rela_count = 0, *addr;
> > + struct elf64_dyn *d = dyn_table;
> > + struct elf64_rela *r;
> > +
> > + while (d && d->tag != DT_NULL) {
> > + if (d->tag == DT_RELA)
> > + rela_addr = d->val;
> > + else if (d->tag == DT_RELACOUNT)
> > + rela_count = d->val;
> > + if (rela_addr && rela_count)
> > + break;
> > + ++d;
> > + }
> > +
> > + if (!rela_addr || !rela_count)
> > + return;
> > +
> > + r = (void *)(rela_addr + load_addr);
> > +
> > + while (rela_count--) {
> > + if (RELA_GET_TYPE(r) == R_PPC_RELATIVE) {
> > + addr = (void *)(r->offset + load_addr);
> > + *addr = r->addend + load_addr;
> > + }
> > + ++r;
> > + }
> > +}
>
> --
> David Gibson | I'll have my music baroque, and my code
> david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
> | _way_ _around_!
> http://www.ozlabs.org/~dgibson
^ permalink raw reply [flat|nested] 36+ messages in thread
* [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (9 preceding siblings ...)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 10/14] powerpc/ppc64: relocate linker VMAs Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-03 17:08 ` Paolo Bonzini
` (2 more replies)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 12/14] scripts: add exit code snooper Andrew Jones
` (4 subsequent siblings)
15 siblings, 3 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
Add enough RTAS support to support power-off, and apply it to
exit().
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
lib/powerpc/asm/rtas.h | 27 ++++++++++++++++
lib/powerpc/io.c | 3 ++
lib/powerpc/rtas.c | 84 +++++++++++++++++++++++++++++++++++++++++++++++++
lib/ppc64/asm/rtas.h | 1 +
powerpc/Makefile.common | 1 +
powerpc/cstart64.S | 9 ++++++
6 files changed, 125 insertions(+)
create mode 100644 lib/powerpc/asm/rtas.h
create mode 100644 lib/powerpc/rtas.c
create mode 100644 lib/ppc64/asm/rtas.h
diff --git a/lib/powerpc/asm/rtas.h b/lib/powerpc/asm/rtas.h
new file mode 100644
index 0000000000000..53ce126c69a42
--- /dev/null
+++ b/lib/powerpc/asm/rtas.h
@@ -0,0 +1,27 @@
+#ifndef _ASMPOWERPC_RTAS_H_
+#define _ASMPOWERPC_RTAS_H_
+/*
+ * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2.
+ */
+#include <libcflat.h>
+
+#define RTAS_UNKNOWN_SERVICE (-1)
+
+struct rtas_args {
+ u32 token;
+ u32 nargs;
+ u32 nret;
+ u32 args[16];
+ u32 *rets;
+};
+
+extern void rtas_init(void);
+extern void enter_rtas(unsigned long args);
+extern int rtas_token(const char *service);
+extern int rtas_call(int token, int nargs, int nret, int *outputs, ...);
+
+extern void rtas_power_off(void);
+
+#endif /* _ASMPOWERPC_RTAS_H_ */
diff --git a/lib/powerpc/io.c b/lib/powerpc/io.c
index a7eaafeca9205..25cdd0ad961ee 100644
--- a/lib/powerpc/io.c
+++ b/lib/powerpc/io.c
@@ -7,6 +7,7 @@
*/
#include <libcflat.h>
#include <asm/spinlock.h>
+#include <asm/rtas.h>
extern void halt(int code);
extern void putchar(int c);
@@ -15,6 +16,7 @@ static struct spinlock uart_lock;
void io_init(void)
{
+ rtas_init();
}
void puts(const char *s)
@@ -27,5 +29,6 @@ void puts(const char *s)
void exit(int code)
{
+ rtas_power_off();
halt(code);
}
diff --git a/lib/powerpc/rtas.c b/lib/powerpc/rtas.c
new file mode 100644
index 0000000000000..39f7d4428ee77
--- /dev/null
+++ b/lib/powerpc/rtas.c
@@ -0,0 +1,84 @@
+/*
+ * powerpc RTAS
+ *
+ * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2.
+ */
+#include <libcflat.h>
+#include <libfdt/libfdt.h>
+#include <devicetree.h>
+#include <asm/spinlock.h>
+#include <asm/io.h>
+#include <asm/rtas.h>
+
+static struct spinlock lock;
+struct rtas_args rtas_args;
+static int rtas_node;
+
+void rtas_init(void)
+{
+ if (!dt_available()) {
+ printf("%s: No device tree!\n", __func__);
+ abort();
+ }
+
+ rtas_node = fdt_path_offset(dt_fdt(), "/rtas");
+ if (rtas_node < 0) {
+ printf("%s: /rtas node: %s\n", __func__,
+ fdt_strerror(rtas_node));
+ abort();
+ }
+}
+
+int rtas_token(const char *service)
+{
+ const struct fdt_property *prop;
+ u32 *token;
+
+ prop = fdt_get_property(dt_fdt(), rtas_node, service, NULL);
+ if (prop) {
+ token = (u32 *)prop->data;
+ return fdt32_to_cpu(*token);
+ }
+ return RTAS_UNKNOWN_SERVICE;
+}
+
+int rtas_call(int token, int nargs, int nret, int *outputs, ...)
+{
+ va_list list;
+ int ret, i;
+
+ spin_lock(&lock);
+
+ rtas_args.token = cpu_to_be32(token);
+ rtas_args.nargs = cpu_to_be32(nargs);
+ rtas_args.nret = cpu_to_be32(nret);
+ rtas_args.rets = &rtas_args.args[nargs];
+
+ va_start(list, outputs);
+ for (i = 0; i < nargs; ++i)
+ rtas_args.args[i] = cpu_to_be32(va_arg(list, u32));
+ va_end(list);
+
+ for (i = 0; i < nret; ++i)
+ rtas_args.rets[i] = 0;
+
+ enter_rtas(__pa(&rtas_args));
+
+ if (nret > 1 && outputs != NULL)
+ for (i = 0; i < nret - 1; ++i)
+ outputs[i] = be32_to_cpu(rtas_args.rets[i + 1]);
+
+ ret = nret > 0 ? be32_to_cpu(rtas_args.rets[0]) : 0;
+
+ spin_unlock(&lock);
+ return ret;
+}
+
+void rtas_power_off(void)
+{
+ printf("RTAS power-off returned %d\n",
+ rtas_call(rtas_token("power-off"), 2, 1, NULL, -1, -1));
+ assert(0);
+}
diff --git a/lib/ppc64/asm/rtas.h b/lib/ppc64/asm/rtas.h
new file mode 100644
index 0000000000000..fe77f635cd860
--- /dev/null
+++ b/lib/ppc64/asm/rtas.h
@@ -0,0 +1 @@
+#include "../../powerpc/asm/rtas.h"
diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
index b130342dee60e..f4eff7f8eccb4 100644
--- a/powerpc/Makefile.common
+++ b/powerpc/Makefile.common
@@ -39,6 +39,7 @@ cflatobjs += lib/powerpc/io.o
cflatobjs += lib/powerpc/setup.o
cflatobjs += lib/powerpc/mmu.o
cflatobjs += lib/powerpc/smp.o
+cflatobjs += lib/powerpc/rtas.o
libgcc := $(shell $(CC) $(machine) --print-libgcc-file-name)
start_addr := $(shell printf "%x\n" $$(( $(phys_base) + $(kernel_offset) )))
diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
index 8edaaa6e251fc..1c48083efa33c 100644
--- a/powerpc/cstart64.S
+++ b/powerpc/cstart64.S
@@ -8,6 +8,7 @@
#define HVSC .long 0x44000022
#define H_PUT_TERM_CHAR 0x58
+#define KVMPPC_H_RTAS 0xf000
#define LOAD_REG_IMMEDIATE(reg,expr) \
lis reg,(expr)@highest; \
@@ -85,3 +86,11 @@ putchar:
li r5, 1 /* sending just 1 byte */
HVSC
blr
+
+.globl enter_rtas
+enter_rtas:
+ mr r4, r3
+ lis r3, KVMPPC_H_RTAS@h
+ ori r3, r3, KVMPPC_H_RTAS@l
+ HVSC
+ blr
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-03 14:41 ` [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off Andrew Jones
@ 2015-08-03 17:08 ` Paolo Bonzini
2015-08-03 18:02 ` Andrew Jones
2015-08-04 4:09 ` David Gibson
2015-08-04 4:03 ` David Gibson
2015-08-04 4:11 ` David Gibson
2 siblings, 2 replies; 36+ messages in thread
From: Paolo Bonzini @ 2015-08-03 17:08 UTC (permalink / raw)
To: Andrew Jones, kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier
On 03/08/2015 16:41, Andrew Jones wrote:
> Add enough RTAS support to support power-off, and apply it to
> exit().
>
> Signed-off-by: Andrew Jones <drjones@redhat.com>
Why not use virtio-mmio + testdev on ppc as well? Similar to how we're
not using PSCI on ARM or ACPI on x86.
Paolo
^ permalink raw reply [flat|nested] 36+ messages in thread
* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-03 17:08 ` Paolo Bonzini
@ 2015-08-03 18:02 ` Andrew Jones
2015-08-03 22:27 ` Alexander Graf
2015-08-04 4:09 ` David Gibson
1 sibling, 1 reply; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 18:02 UTC (permalink / raw)
To: Paolo Bonzini; +Cc: kvm, kvm-ppc, dgibson, david, agraf, thuth, lvivier
On Mon, Aug 03, 2015 at 07:08:17PM +0200, Paolo Bonzini wrote:
>
>
> On 03/08/2015 16:41, Andrew Jones wrote:
> > Add enough RTAS support to support power-off, and apply it to
> > exit().
> >
> > Signed-off-by: Andrew Jones <drjones@redhat.com>
>
> Why not use virtio-mmio + testdev on ppc as well? Similar to how we're
> not using PSCI on ARM or ACPI on x86.
I have some longer term plans to add minimal virtio-pci support to
kvm-unit-tests, and then we could plug virtio-serial+chr-testdev into
that. I didn't think I could use virtio-mmio directly with spapr, but
maybe I can? Actually, I sort of like this approach more in some
respects though, as it doesn't require a special testdev or virtio
support, keeping the unit test extra minimal. In fact, I was even
thinking about posting patches (which I've already written) that
allow chr-testdev to be optional for ARM too, now that it could use
the exitcode snooper.
Thanks,
drew
>
> Paolo
> --
> To unsubscribe from this list: send the line "unsubscribe kvm" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 36+ messages in thread
* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-03 18:02 ` Andrew Jones
@ 2015-08-03 22:27 ` Alexander Graf
0 siblings, 0 replies; 36+ messages in thread
From: Alexander Graf @ 2015-08-03 22:27 UTC (permalink / raw)
To: Andrew Jones, Paolo Bonzini; +Cc: kvm, kvm-ppc, dgibson, david, thuth, lvivier
On 03.08.15 19:02, Andrew Jones wrote:
> On Mon, Aug 03, 2015 at 07:08:17PM +0200, Paolo Bonzini wrote:
>>
>>
>> On 03/08/2015 16:41, Andrew Jones wrote:
>>> Add enough RTAS support to support power-off, and apply it to
>>> exit().
>>>
>>> Signed-off-by: Andrew Jones <drjones@redhat.com>
>>
>> Why not use virtio-mmio + testdev on ppc as well? Similar to how we're
>> not using PSCI on ARM or ACPI on x86.
>
> I have some longer term plans to add minimal virtio-pci support to
> kvm-unit-tests, and then we could plug virtio-serial+chr-testdev into
> that. I didn't think I could use virtio-mmio directly with spapr, but
> maybe I can? Actually, I sort of like this approach more in some
You would need to add support for the dynamic sysbus device allocation
in the spapr machine, but then I don't see why it wouldn't work.
PCI however is the more natural choice on sPAPR if you want to do virtio.
That said, if all you need is a chr transport, IIRC there should be a
way to get you additional channels on the existing "serial port" - which
really is just a simply hypercall interface. But David is the best
person to guide you to the best path forward here.
Alex
> respects though, as it doesn't require a special testdev or virtio
> support, keeping the unit test extra minimal. In fact, I was even
> thinking about posting patches (which I've already written) that
> allow chr-testdev to be optional for ARM too, now that it could use
> the exitcode snooper.
>
> Thanks,
> drew
>
>>
>> Paolo
>> --
>> To unsubscribe from this list: send the line "unsubscribe kvm" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 36+ messages in thread
* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-03 17:08 ` Paolo Bonzini
2015-08-03 18:02 ` Andrew Jones
@ 2015-08-04 4:09 ` David Gibson
2015-08-04 7:47 ` Andrew Jones
1 sibling, 1 reply; 36+ messages in thread
From: David Gibson @ 2015-08-04 4:09 UTC (permalink / raw)
To: Paolo Bonzini; +Cc: Andrew Jones, kvm, kvm-ppc, dgibson, agraf, thuth, lvivier
[-- Attachment #1: Type: text/plain, Size: 1111 bytes --]
On Mon, Aug 03, 2015 at 07:08:17PM +0200, Paolo Bonzini wrote:
>
>
> On 03/08/2015 16:41, Andrew Jones wrote:
> > Add enough RTAS support to support power-off, and apply it to
> > exit().
> >
> > Signed-off-by: Andrew Jones <drjones@redhat.com>
>
> Why not use virtio-mmio + testdev on ppc as well? Similar to how we're
> not using PSCI on ARM or ACPI on x86.
Strange as it seems, MMIO is actually a PITA for a simple pseries
guest like this. Basically, you have to enable the MMU - which
requires a whole bunch of setup - in order to perform cache-inhibited
loads and stores, which is what you need for IO.
There are hypercalls to sidestep this (H_LOGICAL_CI_LOAD and
H_LOGICAL_CI_STORE), but having a hypercall and KVM exit for every IO
access may be hideously slow.
In early development we did have a hypercall mediated virtio model,
but it was abandoned once we got PCI working.
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
[-- Attachment #2: Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply [flat|nested] 36+ messages in thread
* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-04 4:09 ` David Gibson
@ 2015-08-04 7:47 ` Andrew Jones
2015-08-04 13:15 ` Paolo Bonzini
2015-08-05 0:06 ` David Gibson
0 siblings, 2 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-04 7:47 UTC (permalink / raw)
To: David Gibson; +Cc: Paolo Bonzini, kvm, kvm-ppc, dgibson, agraf, thuth, lvivier
On Tue, Aug 04, 2015 at 02:09:52PM +1000, David Gibson wrote:
> On Mon, Aug 03, 2015 at 07:08:17PM +0200, Paolo Bonzini wrote:
> >
> >
> > On 03/08/2015 16:41, Andrew Jones wrote:
> > > Add enough RTAS support to support power-off, and apply it to
> > > exit().
> > >
> > > Signed-off-by: Andrew Jones <drjones@redhat.com>
> >
> > Why not use virtio-mmio + testdev on ppc as well? Similar to how we're
> > not using PSCI on ARM or ACPI on x86.
>
> Strange as it seems, MMIO is actually a PITA for a simple pseries
> guest like this. Basically, you have to enable the MMU - which
> requires a whole bunch of setup - in order to perform cache-inhibited
> loads and stores, which is what you need for IO.
>
> There are hypercalls to sidestep this (H_LOGICAL_CI_LOAD and
> H_LOGICAL_CI_STORE), but having a hypercall and KVM exit for every IO
> access may be hideously slow.
>
> In early development we did have a hypercall mediated virtio model,
> but it was abandoned once we got PCI working.
So I think by yours and Alex's responses, if we want testdev support
then we should target using pci to expose it. I'm ok with that, but
prefer not to be distracted with it while getting ppc kickstarted.
So, question for Paolo, are you OK with the exitcode snooper "cheat"?
Thanks,
drew
>
> --
> David Gibson | I'll have my music baroque, and my code
> david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
> | _way_ _around_!
> http://www.ozlabs.org/~dgibson
^ permalink raw reply [flat|nested] 36+ messages in thread
* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-04 7:47 ` Andrew Jones
@ 2015-08-04 13:15 ` Paolo Bonzini
2015-08-04 13:21 ` Andrew Jones
2015-08-05 0:06 ` David Gibson
1 sibling, 1 reply; 36+ messages in thread
From: Paolo Bonzini @ 2015-08-04 13:15 UTC (permalink / raw)
To: Andrew Jones, David Gibson; +Cc: kvm, kvm-ppc, dgibson, agraf, thuth, lvivier
On 04/08/2015 09:47, Andrew Jones wrote:
>> > In early development we did have a hypercall mediated virtio model,
>> > but it was abandoned once we got PCI working.
> So I think by yours and Alex's responses, if we want testdev support
> then we should target using pci to expose it. I'm ok with that, but
> prefer not to be distracted with it while getting ppc kickstarted.
> So, question for Paolo, are you OK with the exitcode snooper "cheat"?
If we want that we should use it for x86 and ARM as well. However,
attaching the char testdev to a spapr-vty (virtio-mmio made little
sense) should be easy.
Paolo
^ permalink raw reply [flat|nested] 36+ messages in thread
* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-04 13:15 ` Paolo Bonzini
@ 2015-08-04 13:21 ` Andrew Jones
0 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-04 13:21 UTC (permalink / raw)
To: Paolo Bonzini; +Cc: David Gibson, kvm, kvm-ppc, dgibson, agraf, thuth, lvivier
On Tue, Aug 04, 2015 at 03:15:25PM +0200, Paolo Bonzini wrote:
>
>
> On 04/08/2015 09:47, Andrew Jones wrote:
> >> > In early development we did have a hypercall mediated virtio model,
> >> > but it was abandoned once we got PCI working.
> > So I think by yours and Alex's responses, if we want testdev support
> > then we should target using pci to expose it. I'm ok with that, but
> > prefer not to be distracted with it while getting ppc kickstarted.
> > So, question for Paolo, are you OK with the exitcode snooper "cheat"?
>
> If we want that we should use it for x86 and ARM as well. However,
> attaching the char testdev to a spapr-vty (virtio-mmio made little
> sense) should be easy.
OK, I'll look at doing the vty approach first.
Thanks,
drew
^ permalink raw reply [flat|nested] 36+ messages in thread
* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-04 7:47 ` Andrew Jones
2015-08-04 13:15 ` Paolo Bonzini
@ 2015-08-05 0:06 ` David Gibson
1 sibling, 0 replies; 36+ messages in thread
From: David Gibson @ 2015-08-05 0:06 UTC (permalink / raw)
To: Andrew Jones; +Cc: Paolo Bonzini, kvm, kvm-ppc, dgibson, agraf, thuth, lvivier
[-- Attachment #1: Type: text/plain, Size: 1815 bytes --]
On Tue, Aug 04, 2015 at 09:47:59AM +0200, Andrew Jones wrote:
> On Tue, Aug 04, 2015 at 02:09:52PM +1000, David Gibson wrote:
> > On Mon, Aug 03, 2015 at 07:08:17PM +0200, Paolo Bonzini wrote:
> > >
> > >
> > > On 03/08/2015 16:41, Andrew Jones wrote:
> > > > Add enough RTAS support to support power-off, and apply it to
> > > > exit().
> > > >
> > > > Signed-off-by: Andrew Jones <drjones@redhat.com>
> > >
> > > Why not use virtio-mmio + testdev on ppc as well? Similar to how we're
> > > not using PSCI on ARM or ACPI on x86.
> >
> > Strange as it seems, MMIO is actually a PITA for a simple pseries
> > guest like this. Basically, you have to enable the MMU - which
> > requires a whole bunch of setup - in order to perform cache-inhibited
> > loads and stores, which is what you need for IO.
> >
> > There are hypercalls to sidestep this (H_LOGICAL_CI_LOAD and
> > H_LOGICAL_CI_STORE), but having a hypercall and KVM exit for every IO
> > access may be hideously slow.
> >
> > In early development we did have a hypercall mediated virtio model,
> > but it was abandoned once we got PCI working.
>
> So I think by yours and Alex's responses, if we want testdev support
> then we should target using pci to expose it.
Um.. maybe. I'm not really familiar with these testdevs, so I can't
answer directly.
> I'm ok with that, but
> prefer not to be distracted with it while getting ppc kickstarted.
> So, question for Paolo, are you OK with the exitcode snooper "cheat"?
If you wanted to add a special hypercall channel for use by the tests
I'd be ok with that too.
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
[-- Attachment #2: Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply [flat|nested] 36+ messages in thread
* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-03 14:41 ` [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off Andrew Jones
2015-08-03 17:08 ` Paolo Bonzini
@ 2015-08-04 4:03 ` David Gibson
2015-08-04 7:51 ` Andrew Jones
2015-08-04 4:11 ` David Gibson
2 siblings, 1 reply; 36+ messages in thread
From: David Gibson @ 2015-08-04 4:03 UTC (permalink / raw)
To: Andrew Jones; +Cc: kvm, kvm-ppc, dgibson, agraf, thuth, lvivier, pbonzini
[-- Attachment #1: Type: text/plain, Size: 6668 bytes --]
On Mon, Aug 03, 2015 at 04:41:28PM +0200, Andrew Jones wrote:
> Add enough RTAS support to support power-off, and apply it to
> exit().
>
> Signed-off-by: Andrew Jones <drjones@redhat.com>
> ---
> lib/powerpc/asm/rtas.h | 27 ++++++++++++++++
> lib/powerpc/io.c | 3 ++
> lib/powerpc/rtas.c | 84 +++++++++++++++++++++++++++++++++++++++++++++++++
> lib/ppc64/asm/rtas.h | 1 +
> powerpc/Makefile.common | 1 +
> powerpc/cstart64.S | 9 ++++++
> 6 files changed, 125 insertions(+)
> create mode 100644 lib/powerpc/asm/rtas.h
> create mode 100644 lib/powerpc/rtas.c
> create mode 100644 lib/ppc64/asm/rtas.h
>
> diff --git a/lib/powerpc/asm/rtas.h b/lib/powerpc/asm/rtas.h
> new file mode 100644
> index 0000000000000..53ce126c69a42
> --- /dev/null
> +++ b/lib/powerpc/asm/rtas.h
> @@ -0,0 +1,27 @@
> +#ifndef _ASMPOWERPC_RTAS_H_
> +#define _ASMPOWERPC_RTAS_H_
> +/*
> + * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
> + *
> + * This work is licensed under the terms of the GNU LGPL, version 2.
> + */
> +#include <libcflat.h>
> +
> +#define RTAS_UNKNOWN_SERVICE (-1)
> +
> +struct rtas_args {
> + u32 token;
> + u32 nargs;
> + u32 nret;
> + u32 args[16];
> + u32 *rets;
> +};
> +
> +extern void rtas_init(void);
> +extern void enter_rtas(unsigned long args);
> +extern int rtas_token(const char *service);
> +extern int rtas_call(int token, int nargs, int nret, int *outputs, ...);
> +
> +extern void rtas_power_off(void);
> +
> +#endif /* _ASMPOWERPC_RTAS_H_ */
> diff --git a/lib/powerpc/io.c b/lib/powerpc/io.c
> index a7eaafeca9205..25cdd0ad961ee 100644
> --- a/lib/powerpc/io.c
> +++ b/lib/powerpc/io.c
> @@ -7,6 +7,7 @@
> */
> #include <libcflat.h>
> #include <asm/spinlock.h>
> +#include <asm/rtas.h>
>
> extern void halt(int code);
> extern void putchar(int c);
> @@ -15,6 +16,7 @@ static struct spinlock uart_lock;
>
> void io_init(void)
> {
> + rtas_init();
> }
>
> void puts(const char *s)
> @@ -27,5 +29,6 @@ void puts(const char *s)
>
> void exit(int code)
> {
> + rtas_power_off();
> halt(code);
> }
> diff --git a/lib/powerpc/rtas.c b/lib/powerpc/rtas.c
> new file mode 100644
> index 0000000000000..39f7d4428ee77
> --- /dev/null
> +++ b/lib/powerpc/rtas.c
> @@ -0,0 +1,84 @@
> +/*
> + * powerpc RTAS
> + *
> + * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
> + *
> + * This work is licensed under the terms of the GNU LGPL, version 2.
> + */
> +#include <libcflat.h>
> +#include <libfdt/libfdt.h>
> +#include <devicetree.h>
> +#include <asm/spinlock.h>
> +#include <asm/io.h>
> +#include <asm/rtas.h>
> +
> +static struct spinlock lock;
> +struct rtas_args rtas_args;
> +static int rtas_node;
> +
> +void rtas_init(void)
> +{
> + if (!dt_available()) {
> + printf("%s: No device tree!\n", __func__);
> + abort();
> + }
> +
> + rtas_node = fdt_path_offset(dt_fdt(), "/rtas");
> + if (rtas_node < 0) {
> + printf("%s: /rtas node: %s\n", __func__,
> + fdt_strerror(rtas_node));
> + abort();
> + }
> +}
> +
> +int rtas_token(const char *service)
> +{
> + const struct fdt_property *prop;
> + u32 *token;
> +
> + prop = fdt_get_property(dt_fdt(), rtas_node, service, NULL);
> + if (prop) {
> + token = (u32 *)prop->data;
> + return fdt32_to_cpu(*token);
> + }
> + return RTAS_UNKNOWN_SERVICE;
> +}
> +
> +int rtas_call(int token, int nargs, int nret, int *outputs, ...)
> +{
> + va_list list;
> + int ret, i;
> +
> + spin_lock(&lock);
> +
> + rtas_args.token = cpu_to_be32(token);
> + rtas_args.nargs = cpu_to_be32(nargs);
> + rtas_args.nret = cpu_to_be32(nret);
> + rtas_args.rets = &rtas_args.args[nargs];
> +
> + va_start(list, outputs);
> + for (i = 0; i < nargs; ++i)
> + rtas_args.args[i] = cpu_to_be32(va_arg(list, u32));
> + va_end(list);
> +
> + for (i = 0; i < nret; ++i)
> + rtas_args.rets[i] = 0;
> +
> + enter_rtas(__pa(&rtas_args));
> +
> + if (nret > 1 && outputs != NULL)
> + for (i = 0; i < nret - 1; ++i)
> + outputs[i] = be32_to_cpu(rtas_args.rets[i + 1]);
> +
> + ret = nret > 0 ? be32_to_cpu(rtas_args.rets[0]) : 0;
> +
> + spin_unlock(&lock);
> + return ret;
> +}
> +
> +void rtas_power_off(void)
> +{
> + printf("RTAS power-off returned %d\n",
> + rtas_call(rtas_token("power-off"), 2, 1, NULL, -1, -1));
> + assert(0);
> +}
> diff --git a/lib/ppc64/asm/rtas.h b/lib/ppc64/asm/rtas.h
> new file mode 100644
> index 0000000000000..fe77f635cd860
> --- /dev/null
> +++ b/lib/ppc64/asm/rtas.h
> @@ -0,0 +1 @@
> +#include "../../powerpc/asm/rtas.h"
> diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
> index b130342dee60e..f4eff7f8eccb4 100644
> --- a/powerpc/Makefile.common
> +++ b/powerpc/Makefile.common
> @@ -39,6 +39,7 @@ cflatobjs += lib/powerpc/io.o
> cflatobjs += lib/powerpc/setup.o
> cflatobjs += lib/powerpc/mmu.o
> cflatobjs += lib/powerpc/smp.o
> +cflatobjs += lib/powerpc/rtas.o
>
> libgcc := $(shell $(CC) $(machine) --print-libgcc-file-name)
> start_addr := $(shell printf "%x\n" $$(( $(phys_base) + $(kernel_offset) )))
> diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
> index 8edaaa6e251fc..1c48083efa33c 100644
> --- a/powerpc/cstart64.S
> +++ b/powerpc/cstart64.S
> @@ -8,6 +8,7 @@
>
> #define HVSC .long 0x44000022
> #define H_PUT_TERM_CHAR 0x58
> +#define KVMPPC_H_RTAS 0xf000
>
> #define LOAD_REG_IMMEDIATE(reg,expr) \
> lis reg,(expr)@highest; \
> @@ -85,3 +86,11 @@ putchar:
> li r5, 1 /* sending just 1 byte */
> HVSC
> blr
> +
> +.globl enter_rtas
> +enter_rtas:
> + mr r4, r3
> + lis r3, KVMPPC_H_RTAS@h
> + ori r3, r3, KVMPPC_H_RTAS@l
> + HVSC
> + blr
So, strictly speaking this isn't how a client program is supposed to
access RTAS. Instead, you're supposed to call into SLOF to
instantiate RTAS, then jump into the blob it gives you. Or, if you're
replacing rather than using SLOF, then you're supposed to get the
existing RTAS blob address from the device tree and jump into that (or
copy it somewhere else yourself).
In practice that RTAS blob will have the same 5 instructions as your
enter_rtas, so it's probably ok to do this.
What concerns me slightly, though, is that if we ever messed up the
RTAS blob handling in qemu or SLOF, then bypassing it in the kvm tests
like this means they wouldn't catch that breakage.
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
[-- Attachment #2: Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply [flat|nested] 36+ messages in thread* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-04 4:03 ` David Gibson
@ 2015-08-04 7:51 ` Andrew Jones
0 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-04 7:51 UTC (permalink / raw)
To: David Gibson; +Cc: kvm, kvm-ppc, dgibson, agraf, thuth, lvivier, pbonzini
On Tue, Aug 04, 2015 at 02:03:24PM +1000, David Gibson wrote:
> On Mon, Aug 03, 2015 at 04:41:28PM +0200, Andrew Jones wrote:
> > Add enough RTAS support to support power-off, and apply it to
> > exit().
> >
> > Signed-off-by: Andrew Jones <drjones@redhat.com>
> > ---
> > lib/powerpc/asm/rtas.h | 27 ++++++++++++++++
> > lib/powerpc/io.c | 3 ++
> > lib/powerpc/rtas.c | 84 +++++++++++++++++++++++++++++++++++++++++++++++++
> > lib/ppc64/asm/rtas.h | 1 +
> > powerpc/Makefile.common | 1 +
> > powerpc/cstart64.S | 9 ++++++
> > 6 files changed, 125 insertions(+)
> > create mode 100644 lib/powerpc/asm/rtas.h
> > create mode 100644 lib/powerpc/rtas.c
> > create mode 100644 lib/ppc64/asm/rtas.h
> >
> > diff --git a/lib/powerpc/asm/rtas.h b/lib/powerpc/asm/rtas.h
> > new file mode 100644
> > index 0000000000000..53ce126c69a42
> > --- /dev/null
> > +++ b/lib/powerpc/asm/rtas.h
> > @@ -0,0 +1,27 @@
> > +#ifndef _ASMPOWERPC_RTAS_H_
> > +#define _ASMPOWERPC_RTAS_H_
> > +/*
> > + * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
> > + *
> > + * This work is licensed under the terms of the GNU LGPL, version 2.
> > + */
> > +#include <libcflat.h>
> > +
> > +#define RTAS_UNKNOWN_SERVICE (-1)
> > +
> > +struct rtas_args {
> > + u32 token;
> > + u32 nargs;
> > + u32 nret;
> > + u32 args[16];
> > + u32 *rets;
> > +};
> > +
> > +extern void rtas_init(void);
> > +extern void enter_rtas(unsigned long args);
> > +extern int rtas_token(const char *service);
> > +extern int rtas_call(int token, int nargs, int nret, int *outputs, ...);
> > +
> > +extern void rtas_power_off(void);
> > +
> > +#endif /* _ASMPOWERPC_RTAS_H_ */
> > diff --git a/lib/powerpc/io.c b/lib/powerpc/io.c
> > index a7eaafeca9205..25cdd0ad961ee 100644
> > --- a/lib/powerpc/io.c
> > +++ b/lib/powerpc/io.c
> > @@ -7,6 +7,7 @@
> > */
> > #include <libcflat.h>
> > #include <asm/spinlock.h>
> > +#include <asm/rtas.h>
> >
> > extern void halt(int code);
> > extern void putchar(int c);
> > @@ -15,6 +16,7 @@ static struct spinlock uart_lock;
> >
> > void io_init(void)
> > {
> > + rtas_init();
> > }
> >
> > void puts(const char *s)
> > @@ -27,5 +29,6 @@ void puts(const char *s)
> >
> > void exit(int code)
> > {
> > + rtas_power_off();
> > halt(code);
> > }
> > diff --git a/lib/powerpc/rtas.c b/lib/powerpc/rtas.c
> > new file mode 100644
> > index 0000000000000..39f7d4428ee77
> > --- /dev/null
> > +++ b/lib/powerpc/rtas.c
> > @@ -0,0 +1,84 @@
> > +/*
> > + * powerpc RTAS
> > + *
> > + * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
> > + *
> > + * This work is licensed under the terms of the GNU LGPL, version 2.
> > + */
> > +#include <libcflat.h>
> > +#include <libfdt/libfdt.h>
> > +#include <devicetree.h>
> > +#include <asm/spinlock.h>
> > +#include <asm/io.h>
> > +#include <asm/rtas.h>
> > +
> > +static struct spinlock lock;
> > +struct rtas_args rtas_args;
> > +static int rtas_node;
> > +
> > +void rtas_init(void)
> > +{
> > + if (!dt_available()) {
> > + printf("%s: No device tree!\n", __func__);
> > + abort();
> > + }
> > +
> > + rtas_node = fdt_path_offset(dt_fdt(), "/rtas");
> > + if (rtas_node < 0) {
> > + printf("%s: /rtas node: %s\n", __func__,
> > + fdt_strerror(rtas_node));
> > + abort();
> > + }
> > +}
> > +
> > +int rtas_token(const char *service)
> > +{
> > + const struct fdt_property *prop;
> > + u32 *token;
> > +
> > + prop = fdt_get_property(dt_fdt(), rtas_node, service, NULL);
> > + if (prop) {
> > + token = (u32 *)prop->data;
> > + return fdt32_to_cpu(*token);
> > + }
> > + return RTAS_UNKNOWN_SERVICE;
> > +}
> > +
> > +int rtas_call(int token, int nargs, int nret, int *outputs, ...)
> > +{
> > + va_list list;
> > + int ret, i;
> > +
> > + spin_lock(&lock);
> > +
> > + rtas_args.token = cpu_to_be32(token);
> > + rtas_args.nargs = cpu_to_be32(nargs);
> > + rtas_args.nret = cpu_to_be32(nret);
> > + rtas_args.rets = &rtas_args.args[nargs];
> > +
> > + va_start(list, outputs);
> > + for (i = 0; i < nargs; ++i)
> > + rtas_args.args[i] = cpu_to_be32(va_arg(list, u32));
> > + va_end(list);
> > +
> > + for (i = 0; i < nret; ++i)
> > + rtas_args.rets[i] = 0;
> > +
> > + enter_rtas(__pa(&rtas_args));
> > +
> > + if (nret > 1 && outputs != NULL)
> > + for (i = 0; i < nret - 1; ++i)
> > + outputs[i] = be32_to_cpu(rtas_args.rets[i + 1]);
> > +
> > + ret = nret > 0 ? be32_to_cpu(rtas_args.rets[0]) : 0;
> > +
> > + spin_unlock(&lock);
> > + return ret;
> > +}
> > +
> > +void rtas_power_off(void)
> > +{
> > + printf("RTAS power-off returned %d\n",
> > + rtas_call(rtas_token("power-off"), 2, 1, NULL, -1, -1));
> > + assert(0);
> > +}
> > diff --git a/lib/ppc64/asm/rtas.h b/lib/ppc64/asm/rtas.h
> > new file mode 100644
> > index 0000000000000..fe77f635cd860
> > --- /dev/null
> > +++ b/lib/ppc64/asm/rtas.h
> > @@ -0,0 +1 @@
> > +#include "../../powerpc/asm/rtas.h"
> > diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
> > index b130342dee60e..f4eff7f8eccb4 100644
> > --- a/powerpc/Makefile.common
> > +++ b/powerpc/Makefile.common
> > @@ -39,6 +39,7 @@ cflatobjs += lib/powerpc/io.o
> > cflatobjs += lib/powerpc/setup.o
> > cflatobjs += lib/powerpc/mmu.o
> > cflatobjs += lib/powerpc/smp.o
> > +cflatobjs += lib/powerpc/rtas.o
> >
> > libgcc := $(shell $(CC) $(machine) --print-libgcc-file-name)
> > start_addr := $(shell printf "%x\n" $$(( $(phys_base) + $(kernel_offset) )))
> > diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
> > index 8edaaa6e251fc..1c48083efa33c 100644
> > --- a/powerpc/cstart64.S
> > +++ b/powerpc/cstart64.S
> > @@ -8,6 +8,7 @@
> >
> > #define HVSC .long 0x44000022
> > #define H_PUT_TERM_CHAR 0x58
> > +#define KVMPPC_H_RTAS 0xf000
> >
> > #define LOAD_REG_IMMEDIATE(reg,expr) \
> > lis reg,(expr)@highest; \
> > @@ -85,3 +86,11 @@ putchar:
> > li r5, 1 /* sending just 1 byte */
> > HVSC
> > blr
> > +
> > +.globl enter_rtas
> > +enter_rtas:
> > + mr r4, r3
> > + lis r3, KVMPPC_H_RTAS@h
> > + ori r3, r3, KVMPPC_H_RTAS@l
> > + HVSC
> > + blr
>
> So, strictly speaking this isn't how a client program is supposed to
> access RTAS. Instead, you're supposed to call into SLOF to
> instantiate RTAS, then jump into the blob it gives you. Or, if you're
> replacing rather than using SLOF, then you're supposed to get the
> existing RTAS blob address from the device tree and jump into that (or
> copy it somewhere else yourself).
>
> In practice that RTAS blob will have the same 5 instructions as your
> enter_rtas, so it's probably ok to do this.
>
> What concerns me slightly, though, is that if we ever messed up the
> RTAS blob handling in qemu or SLOF, then bypassing it in the kvm tests
> like this means they wouldn't catch that breakage.
I'll look into doing this properly. It doesn't sound like it'd be much
effort.
Thanks,
drew
>
> --
> David Gibson | I'll have my music baroque, and my code
> david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
> | _way_ _around_!
> http://www.ozlabs.org/~dgibson
^ permalink raw reply [flat|nested] 36+ messages in thread
* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-03 14:41 ` [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off Andrew Jones
2015-08-03 17:08 ` Paolo Bonzini
2015-08-04 4:03 ` David Gibson
@ 2015-08-04 4:11 ` David Gibson
2015-08-04 7:54 ` Andrew Jones
2 siblings, 1 reply; 36+ messages in thread
From: David Gibson @ 2015-08-04 4:11 UTC (permalink / raw)
To: Andrew Jones; +Cc: kvm, kvm-ppc, dgibson, agraf, thuth, lvivier, pbonzini
[-- Attachment #1: Type: text/plain, Size: 6327 bytes --]
On Mon, Aug 03, 2015 at 04:41:28PM +0200, Andrew Jones wrote:
> Add enough RTAS support to support power-off, and apply it to
> exit().
>
> Signed-off-by: Andrew Jones <drjones@redhat.com>
> ---
> lib/powerpc/asm/rtas.h | 27 ++++++++++++++++
> lib/powerpc/io.c | 3 ++
> lib/powerpc/rtas.c | 84 +++++++++++++++++++++++++++++++++++++++++++++++++
> lib/ppc64/asm/rtas.h | 1 +
> powerpc/Makefile.common | 1 +
> powerpc/cstart64.S | 9 ++++++
> 6 files changed, 125 insertions(+)
> create mode 100644 lib/powerpc/asm/rtas.h
> create mode 100644 lib/powerpc/rtas.c
> create mode 100644 lib/ppc64/asm/rtas.h
>
> diff --git a/lib/powerpc/asm/rtas.h b/lib/powerpc/asm/rtas.h
> new file mode 100644
> index 0000000000000..53ce126c69a42
> --- /dev/null
> +++ b/lib/powerpc/asm/rtas.h
> @@ -0,0 +1,27 @@
> +#ifndef _ASMPOWERPC_RTAS_H_
> +#define _ASMPOWERPC_RTAS_H_
> +/*
> + * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
> + *
> + * This work is licensed under the terms of the GNU LGPL, version 2.
> + */
> +#include <libcflat.h>
> +
> +#define RTAS_UNKNOWN_SERVICE (-1)
> +
> +struct rtas_args {
> + u32 token;
> + u32 nargs;
> + u32 nret;
> + u32 args[16];
> + u32 *rets;
> +};
> +
> +extern void rtas_init(void);
> +extern void enter_rtas(unsigned long args);
> +extern int rtas_token(const char *service);
> +extern int rtas_call(int token, int nargs, int nret, int *outputs, ...);
> +
> +extern void rtas_power_off(void);
> +
> +#endif /* _ASMPOWERPC_RTAS_H_ */
> diff --git a/lib/powerpc/io.c b/lib/powerpc/io.c
> index a7eaafeca9205..25cdd0ad961ee 100644
> --- a/lib/powerpc/io.c
> +++ b/lib/powerpc/io.c
> @@ -7,6 +7,7 @@
> */
> #include <libcflat.h>
> #include <asm/spinlock.h>
> +#include <asm/rtas.h>
>
> extern void halt(int code);
> extern void putchar(int c);
> @@ -15,6 +16,7 @@ static struct spinlock uart_lock;
>
> void io_init(void)
> {
> + rtas_init();
> }
>
> void puts(const char *s)
> @@ -27,5 +29,6 @@ void puts(const char *s)
>
> void exit(int code)
> {
> + rtas_power_off();
> halt(code);
> }
> diff --git a/lib/powerpc/rtas.c b/lib/powerpc/rtas.c
> new file mode 100644
> index 0000000000000..39f7d4428ee77
> --- /dev/null
> +++ b/lib/powerpc/rtas.c
> @@ -0,0 +1,84 @@
> +/*
> + * powerpc RTAS
> + *
> + * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
> + *
> + * This work is licensed under the terms of the GNU LGPL, version 2.
> + */
> +#include <libcflat.h>
> +#include <libfdt/libfdt.h>
> +#include <devicetree.h>
> +#include <asm/spinlock.h>
> +#include <asm/io.h>
> +#include <asm/rtas.h>
> +
> +static struct spinlock lock;
> +struct rtas_args rtas_args;
> +static int rtas_node;
> +
> +void rtas_init(void)
> +{
> + if (!dt_available()) {
> + printf("%s: No device tree!\n", __func__);
> + abort();
> + }
> +
> + rtas_node = fdt_path_offset(dt_fdt(), "/rtas");
> + if (rtas_node < 0) {
> + printf("%s: /rtas node: %s\n", __func__,
> + fdt_strerror(rtas_node));
> + abort();
> + }
> +}
> +
> +int rtas_token(const char *service)
> +{
> + const struct fdt_property *prop;
> + u32 *token;
> +
> + prop = fdt_get_property(dt_fdt(), rtas_node, service, NULL);
Oh.. one other thing here.
This is only safe if you never alter the device tree blob you're given
(or at least, not after rtas_init()). That may well be the case, but
I don't know your code well enough to be sure.
Otherwise, the rtas node could get moved by other dt changes, meaning
the offset stored in rtas_node is no longer valid.
> + if (prop) {
> + token = (u32 *)prop->data;
> + return fdt32_to_cpu(*token);
> + }
> + return RTAS_UNKNOWN_SERVICE;
> +}
> +
> +int rtas_call(int token, int nargs, int nret, int *outputs, ...)
> +{
> + va_list list;
> + int ret, i;
> +
> + spin_lock(&lock);
> +
> + rtas_args.token = cpu_to_be32(token);
> + rtas_args.nargs = cpu_to_be32(nargs);
> + rtas_args.nret = cpu_to_be32(nret);
> + rtas_args.rets = &rtas_args.args[nargs];
> +
> + va_start(list, outputs);
> + for (i = 0; i < nargs; ++i)
> + rtas_args.args[i] = cpu_to_be32(va_arg(list, u32));
> + va_end(list);
> +
> + for (i = 0; i < nret; ++i)
> + rtas_args.rets[i] = 0;
> +
> + enter_rtas(__pa(&rtas_args));
> +
> + if (nret > 1 && outputs != NULL)
> + for (i = 0; i < nret - 1; ++i)
> + outputs[i] = be32_to_cpu(rtas_args.rets[i + 1]);
> +
> + ret = nret > 0 ? be32_to_cpu(rtas_args.rets[0]) : 0;
> +
> + spin_unlock(&lock);
> + return ret;
> +}
> +
> +void rtas_power_off(void)
> +{
> + printf("RTAS power-off returned %d\n",
> + rtas_call(rtas_token("power-off"), 2, 1, NULL, -1, -1));
> + assert(0);
> +}
> diff --git a/lib/ppc64/asm/rtas.h b/lib/ppc64/asm/rtas.h
> new file mode 100644
> index 0000000000000..fe77f635cd860
> --- /dev/null
> +++ b/lib/ppc64/asm/rtas.h
> @@ -0,0 +1 @@
> +#include "../../powerpc/asm/rtas.h"
> diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
> index b130342dee60e..f4eff7f8eccb4 100644
> --- a/powerpc/Makefile.common
> +++ b/powerpc/Makefile.common
> @@ -39,6 +39,7 @@ cflatobjs += lib/powerpc/io.o
> cflatobjs += lib/powerpc/setup.o
> cflatobjs += lib/powerpc/mmu.o
> cflatobjs += lib/powerpc/smp.o
> +cflatobjs += lib/powerpc/rtas.o
>
> libgcc := $(shell $(CC) $(machine) --print-libgcc-file-name)
> start_addr := $(shell printf "%x\n" $$(( $(phys_base) + $(kernel_offset) )))
> diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
> index 8edaaa6e251fc..1c48083efa33c 100644
> --- a/powerpc/cstart64.S
> +++ b/powerpc/cstart64.S
> @@ -8,6 +8,7 @@
>
> #define HVSC .long 0x44000022
> #define H_PUT_TERM_CHAR 0x58
> +#define KVMPPC_H_RTAS 0xf000
>
> #define LOAD_REG_IMMEDIATE(reg,expr) \
> lis reg,(expr)@highest; \
> @@ -85,3 +86,11 @@ putchar:
> li r5, 1 /* sending just 1 byte */
> HVSC
> blr
> +
> +.globl enter_rtas
> +enter_rtas:
> + mr r4, r3
> + lis r3, KVMPPC_H_RTAS@h
> + ori r3, r3, KVMPPC_H_RTAS@l
> + HVSC
> + blr
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
[-- Attachment #2: Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply [flat|nested] 36+ messages in thread* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-04 4:11 ` David Gibson
@ 2015-08-04 7:54 ` Andrew Jones
2015-08-05 0:12 ` David Gibson
0 siblings, 1 reply; 36+ messages in thread
From: Andrew Jones @ 2015-08-04 7:54 UTC (permalink / raw)
To: David Gibson; +Cc: kvm, kvm-ppc, dgibson, agraf, thuth, lvivier, pbonzini
On Tue, Aug 04, 2015 at 02:11:30PM +1000, David Gibson wrote:
> On Mon, Aug 03, 2015 at 04:41:28PM +0200, Andrew Jones wrote:
> > Add enough RTAS support to support power-off, and apply it to
> > exit().
> >
> > Signed-off-by: Andrew Jones <drjones@redhat.com>
> > ---
> > lib/powerpc/asm/rtas.h | 27 ++++++++++++++++
> > lib/powerpc/io.c | 3 ++
> > lib/powerpc/rtas.c | 84 +++++++++++++++++++++++++++++++++++++++++++++++++
> > lib/ppc64/asm/rtas.h | 1 +
> > powerpc/Makefile.common | 1 +
> > powerpc/cstart64.S | 9 ++++++
> > 6 files changed, 125 insertions(+)
> > create mode 100644 lib/powerpc/asm/rtas.h
> > create mode 100644 lib/powerpc/rtas.c
> > create mode 100644 lib/ppc64/asm/rtas.h
> >
> > diff --git a/lib/powerpc/asm/rtas.h b/lib/powerpc/asm/rtas.h
> > new file mode 100644
> > index 0000000000000..53ce126c69a42
> > --- /dev/null
> > +++ b/lib/powerpc/asm/rtas.h
> > @@ -0,0 +1,27 @@
> > +#ifndef _ASMPOWERPC_RTAS_H_
> > +#define _ASMPOWERPC_RTAS_H_
> > +/*
> > + * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
> > + *
> > + * This work is licensed under the terms of the GNU LGPL, version 2.
> > + */
> > +#include <libcflat.h>
> > +
> > +#define RTAS_UNKNOWN_SERVICE (-1)
> > +
> > +struct rtas_args {
> > + u32 token;
> > + u32 nargs;
> > + u32 nret;
> > + u32 args[16];
> > + u32 *rets;
> > +};
> > +
> > +extern void rtas_init(void);
> > +extern void enter_rtas(unsigned long args);
> > +extern int rtas_token(const char *service);
> > +extern int rtas_call(int token, int nargs, int nret, int *outputs, ...);
> > +
> > +extern void rtas_power_off(void);
> > +
> > +#endif /* _ASMPOWERPC_RTAS_H_ */
> > diff --git a/lib/powerpc/io.c b/lib/powerpc/io.c
> > index a7eaafeca9205..25cdd0ad961ee 100644
> > --- a/lib/powerpc/io.c
> > +++ b/lib/powerpc/io.c
> > @@ -7,6 +7,7 @@
> > */
> > #include <libcflat.h>
> > #include <asm/spinlock.h>
> > +#include <asm/rtas.h>
> >
> > extern void halt(int code);
> > extern void putchar(int c);
> > @@ -15,6 +16,7 @@ static struct spinlock uart_lock;
> >
> > void io_init(void)
> > {
> > + rtas_init();
> > }
> >
> > void puts(const char *s)
> > @@ -27,5 +29,6 @@ void puts(const char *s)
> >
> > void exit(int code)
> > {
> > + rtas_power_off();
> > halt(code);
> > }
> > diff --git a/lib/powerpc/rtas.c b/lib/powerpc/rtas.c
> > new file mode 100644
> > index 0000000000000..39f7d4428ee77
> > --- /dev/null
> > +++ b/lib/powerpc/rtas.c
> > @@ -0,0 +1,84 @@
> > +/*
> > + * powerpc RTAS
> > + *
> > + * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
> > + *
> > + * This work is licensed under the terms of the GNU LGPL, version 2.
> > + */
> > +#include <libcflat.h>
> > +#include <libfdt/libfdt.h>
> > +#include <devicetree.h>
> > +#include <asm/spinlock.h>
> > +#include <asm/io.h>
> > +#include <asm/rtas.h>
> > +
> > +static struct spinlock lock;
> > +struct rtas_args rtas_args;
> > +static int rtas_node;
> > +
> > +void rtas_init(void)
> > +{
> > + if (!dt_available()) {
> > + printf("%s: No device tree!\n", __func__);
> > + abort();
> > + }
> > +
> > + rtas_node = fdt_path_offset(dt_fdt(), "/rtas");
> > + if (rtas_node < 0) {
> > + printf("%s: /rtas node: %s\n", __func__,
> > + fdt_strerror(rtas_node));
> > + abort();
> > + }
> > +}
> > +
> > +int rtas_token(const char *service)
> > +{
> > + const struct fdt_property *prop;
> > + u32 *token;
> > +
> > + prop = fdt_get_property(dt_fdt(), rtas_node, service, NULL);
>
> Oh.. one other thing here.
>
> This is only safe if you never alter the device tree blob you're given
> (or at least, not after rtas_init()). That may well be the case, but
> I don't know your code well enough to be sure.
>
> Otherwise, the rtas node could get moved by other dt changes, meaning
> the offset stored in rtas_node is no longer valid.
I hadn't planned on modifying the DT, but if you can conceive of a test
they may want to, then we should do this right. I guess doing it right
just means to hunt down rtas_node each time, that's easy.
Thanks,
drew
>
> > + if (prop) {
> > + token = (u32 *)prop->data;
> > + return fdt32_to_cpu(*token);
> > + }
> > + return RTAS_UNKNOWN_SERVICE;
> > +}
> > +
> > +int rtas_call(int token, int nargs, int nret, int *outputs, ...)
> > +{
> > + va_list list;
> > + int ret, i;
> > +
> > + spin_lock(&lock);
> > +
> > + rtas_args.token = cpu_to_be32(token);
> > + rtas_args.nargs = cpu_to_be32(nargs);
> > + rtas_args.nret = cpu_to_be32(nret);
> > + rtas_args.rets = &rtas_args.args[nargs];
> > +
> > + va_start(list, outputs);
> > + for (i = 0; i < nargs; ++i)
> > + rtas_args.args[i] = cpu_to_be32(va_arg(list, u32));
> > + va_end(list);
> > +
> > + for (i = 0; i < nret; ++i)
> > + rtas_args.rets[i] = 0;
> > +
> > + enter_rtas(__pa(&rtas_args));
> > +
> > + if (nret > 1 && outputs != NULL)
> > + for (i = 0; i < nret - 1; ++i)
> > + outputs[i] = be32_to_cpu(rtas_args.rets[i + 1]);
> > +
> > + ret = nret > 0 ? be32_to_cpu(rtas_args.rets[0]) : 0;
> > +
> > + spin_unlock(&lock);
> > + return ret;
> > +}
> > +
> > +void rtas_power_off(void)
> > +{
> > + printf("RTAS power-off returned %d\n",
> > + rtas_call(rtas_token("power-off"), 2, 1, NULL, -1, -1));
> > + assert(0);
> > +}
> > diff --git a/lib/ppc64/asm/rtas.h b/lib/ppc64/asm/rtas.h
> > new file mode 100644
> > index 0000000000000..fe77f635cd860
> > --- /dev/null
> > +++ b/lib/ppc64/asm/rtas.h
> > @@ -0,0 +1 @@
> > +#include "../../powerpc/asm/rtas.h"
> > diff --git a/powerpc/Makefile.common b/powerpc/Makefile.common
> > index b130342dee60e..f4eff7f8eccb4 100644
> > --- a/powerpc/Makefile.common
> > +++ b/powerpc/Makefile.common
> > @@ -39,6 +39,7 @@ cflatobjs += lib/powerpc/io.o
> > cflatobjs += lib/powerpc/setup.o
> > cflatobjs += lib/powerpc/mmu.o
> > cflatobjs += lib/powerpc/smp.o
> > +cflatobjs += lib/powerpc/rtas.o
> >
> > libgcc := $(shell $(CC) $(machine) --print-libgcc-file-name)
> > start_addr := $(shell printf "%x\n" $$(( $(phys_base) + $(kernel_offset) )))
> > diff --git a/powerpc/cstart64.S b/powerpc/cstart64.S
> > index 8edaaa6e251fc..1c48083efa33c 100644
> > --- a/powerpc/cstart64.S
> > +++ b/powerpc/cstart64.S
> > @@ -8,6 +8,7 @@
> >
> > #define HVSC .long 0x44000022
> > #define H_PUT_TERM_CHAR 0x58
> > +#define KVMPPC_H_RTAS 0xf000
> >
> > #define LOAD_REG_IMMEDIATE(reg,expr) \
> > lis reg,(expr)@highest; \
> > @@ -85,3 +86,11 @@ putchar:
> > li r5, 1 /* sending just 1 byte */
> > HVSC
> > blr
> > +
> > +.globl enter_rtas
> > +enter_rtas:
> > + mr r4, r3
> > + lis r3, KVMPPC_H_RTAS@h
> > + ori r3, r3, KVMPPC_H_RTAS@l
> > + HVSC
> > + blr
>
> --
> David Gibson | I'll have my music baroque, and my code
> david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
> | _way_ _around_!
> http://www.ozlabs.org/~dgibson
^ permalink raw reply [flat|nested] 36+ messages in thread* Re: [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off
2015-08-04 7:54 ` Andrew Jones
@ 2015-08-05 0:12 ` David Gibson
0 siblings, 0 replies; 36+ messages in thread
From: David Gibson @ 2015-08-05 0:12 UTC (permalink / raw)
To: Andrew Jones; +Cc: kvm, kvm-ppc, dgibson, agraf, thuth, lvivier, pbonzini
[-- Attachment #1: Type: text/plain, Size: 5278 bytes --]
On Tue, Aug 04, 2015 at 09:54:44AM +0200, Andrew Jones wrote:
> On Tue, Aug 04, 2015 at 02:11:30PM +1000, David Gibson wrote:
> > On Mon, Aug 03, 2015 at 04:41:28PM +0200, Andrew Jones wrote:
> > > Add enough RTAS support to support power-off, and apply it to
> > > exit().
> > >
> > > Signed-off-by: Andrew Jones <drjones@redhat.com>
> > > ---
> > > lib/powerpc/asm/rtas.h | 27 ++++++++++++++++
> > > lib/powerpc/io.c | 3 ++
> > > lib/powerpc/rtas.c | 84 +++++++++++++++++++++++++++++++++++++++++++++++++
> > > lib/ppc64/asm/rtas.h | 1 +
> > > powerpc/Makefile.common | 1 +
> > > powerpc/cstart64.S | 9 ++++++
> > > 6 files changed, 125 insertions(+)
> > > create mode 100644 lib/powerpc/asm/rtas.h
> > > create mode 100644 lib/powerpc/rtas.c
> > > create mode 100644 lib/ppc64/asm/rtas.h
> > >
> > > diff --git a/lib/powerpc/asm/rtas.h b/lib/powerpc/asm/rtas.h
> > > new file mode 100644
> > > index 0000000000000..53ce126c69a42
> > > --- /dev/null
> > > +++ b/lib/powerpc/asm/rtas.h
> > > @@ -0,0 +1,27 @@
> > > +#ifndef _ASMPOWERPC_RTAS_H_
> > > +#define _ASMPOWERPC_RTAS_H_
> > > +/*
> > > + * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
> > > + *
> > > + * This work is licensed under the terms of the GNU LGPL, version 2.
> > > + */
> > > +#include <libcflat.h>
> > > +
> > > +#define RTAS_UNKNOWN_SERVICE (-1)
> > > +
> > > +struct rtas_args {
> > > + u32 token;
> > > + u32 nargs;
> > > + u32 nret;
> > > + u32 args[16];
> > > + u32 *rets;
> > > +};
> > > +
> > > +extern void rtas_init(void);
> > > +extern void enter_rtas(unsigned long args);
> > > +extern int rtas_token(const char *service);
> > > +extern int rtas_call(int token, int nargs, int nret, int *outputs, ...);
> > > +
> > > +extern void rtas_power_off(void);
> > > +
> > > +#endif /* _ASMPOWERPC_RTAS_H_ */
> > > diff --git a/lib/powerpc/io.c b/lib/powerpc/io.c
> > > index a7eaafeca9205..25cdd0ad961ee 100644
> > > --- a/lib/powerpc/io.c
> > > +++ b/lib/powerpc/io.c
> > > @@ -7,6 +7,7 @@
> > > */
> > > #include <libcflat.h>
> > > #include <asm/spinlock.h>
> > > +#include <asm/rtas.h>
> > >
> > > extern void halt(int code);
> > > extern void putchar(int c);
> > > @@ -15,6 +16,7 @@ static struct spinlock uart_lock;
> > >
> > > void io_init(void)
> > > {
> > > + rtas_init();
> > > }
> > >
> > > void puts(const char *s)
> > > @@ -27,5 +29,6 @@ void puts(const char *s)
> > >
> > > void exit(int code)
> > > {
> > > + rtas_power_off();
> > > halt(code);
> > > }
> > > diff --git a/lib/powerpc/rtas.c b/lib/powerpc/rtas.c
> > > new file mode 100644
> > > index 0000000000000..39f7d4428ee77
> > > --- /dev/null
> > > +++ b/lib/powerpc/rtas.c
> > > @@ -0,0 +1,84 @@
> > > +/*
> > > + * powerpc RTAS
> > > + *
> > > + * Copyright (C) 2015, Red Hat Inc, Andrew Jones <drjones@redhat.com>
> > > + *
> > > + * This work is licensed under the terms of the GNU LGPL, version 2.
> > > + */
> > > +#include <libcflat.h>
> > > +#include <libfdt/libfdt.h>
> > > +#include <devicetree.h>
> > > +#include <asm/spinlock.h>
> > > +#include <asm/io.h>
> > > +#include <asm/rtas.h>
> > > +
> > > +static struct spinlock lock;
> > > +struct rtas_args rtas_args;
> > > +static int rtas_node;
> > > +
> > > +void rtas_init(void)
> > > +{
> > > + if (!dt_available()) {
> > > + printf("%s: No device tree!\n", __func__);
> > > + abort();
> > > + }
> > > +
> > > + rtas_node = fdt_path_offset(dt_fdt(), "/rtas");
> > > + if (rtas_node < 0) {
> > > + printf("%s: /rtas node: %s\n", __func__,
> > > + fdt_strerror(rtas_node));
> > > + abort();
> > > + }
> > > +}
> > > +
> > > +int rtas_token(const char *service)
> > > +{
> > > + const struct fdt_property *prop;
> > > + u32 *token;
> > > +
> > > + prop = fdt_get_property(dt_fdt(), rtas_node, service, NULL);
> >
> > Oh.. one other thing here.
> >
> > This is only safe if you never alter the device tree blob you're given
> > (or at least, not after rtas_init()). That may well be the case, but
> > I don't know your code well enough to be sure.
> >
> > Otherwise, the rtas node could get moved by other dt changes, meaning
> > the offset stored in rtas_node is no longer valid.
>
> I hadn't planned on modifying the DT, but if you can conceive of a test
> they may want to, then we should do this right.
That's probably ok then, as long as it's made clear that you can't
mess with the DT. Or if tests might want to modify it, you could give
them their own copy of it.
>I guess doing it right
> just means to hunt down rtas_node each time, that's easy.
That's right.
Not having persistent handles is a bit of a pain, but it's the
necessary tradeoff to work with the tree in flattened form without
having some complicated pile of "context" state to go with it.
(Attempting to remind people that these aren't persistent handles is,
incidentally, why so many of the libfdt functions explicitly mention
"offset" in their names).
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
[-- Attachment #2: Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply [flat|nested] 36+ messages in thread
* [kvm-unit-tests PATCH 12/14] scripts: add exit code snooper
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (10 preceding siblings ...)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 11/14] powerpc/ppc64: add rtas_power_off Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 13/14] powerpc/ppc64: add run script and unittests.cfg Andrew Jones
` (3 subsequent siblings)
15 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
Add a function that reads test output streams and just forwards it
on to stdout, except when a line is 'EXIT: STATUS=%d', at which
point it grabs the status to use as the exit code. Tests that
always exit with zero (due to how they inform QEMU to quit), can
now return other status codes. Run with
. scripts/functions.bash
$test_cmdline | snoop_exitcode
The unit test must call print_exit_line(code) before invoking
its method to quit QEMU. For arm and powerpc, we just add the
print_exit_line calls to their exit() functions right now.
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
configure | 5 ++++-
lib/arm/io.c | 1 +
lib/libcflat.h | 3 +++
lib/powerpc/io.c | 1 +
scripts/functions.bash | 11 +++++++++++
5 files changed, 20 insertions(+), 1 deletion(-)
diff --git a/configure b/configure
index b2ad199da7873..e1f70bd79d4f7 100755
--- a/configure
+++ b/configure
@@ -10,6 +10,7 @@ ar=ar
arch=`uname -m | sed -e s/i.86/i386/ | sed -e 's/arm.*/arm/'`
host=$arch
cross_prefix=
+snooper=
usage() {
cat <<-EOF
@@ -81,8 +82,9 @@ if [ "$arch" = "i386" ] || [ "$arch" = "x86_64" ]; then
testdir=x86
elif [ "$arch" = "arm" ] || [ "$arch" = "arm64" ]; then
testdir=arm
-elif [ "$arch" = "ppc64" ]; then
+elif [ "$arch" = "powerpc" ] || [ "$arch" = "ppc64" ]; then
testdir=powerpc
+ snooper=snoop_exitcode
else
testdir=$arch
fi
@@ -137,4 +139,5 @@ OBJDUMP=$cross_prefix$objdump
AR=$cross_prefix$ar
API=$api
TEST_DIR=$testdir
+SNOOPER="$snooper"
EOF
diff --git a/lib/arm/io.c b/lib/arm/io.c
index a08d394e4aa1c..487071d2edcd3 100644
--- a/lib/arm/io.c
+++ b/lib/arm/io.c
@@ -78,6 +78,7 @@ void puts(const char *s)
void exit(int code)
{
+ print_exit_line(code);
chr_testdev_exit(code);
halt(code);
}
diff --git a/lib/libcflat.h b/lib/libcflat.h
index 8411f6c5d92e3..c3d90a0af831a 100644
--- a/lib/libcflat.h
+++ b/lib/libcflat.h
@@ -77,4 +77,7 @@ do { \
abort(); \
} while (0)
+#define print_exit_line(code) \
+ printf("\nEXIT: STATUS=%d\n", ((code) << 1) | 1)
+
#endif
diff --git a/lib/powerpc/io.c b/lib/powerpc/io.c
index 25cdd0ad961ee..b68d33b9ef740 100644
--- a/lib/powerpc/io.c
+++ b/lib/powerpc/io.c
@@ -29,6 +29,7 @@ void puts(const char *s)
void exit(int code)
{
+ print_exit_line(code);
rtas_power_off();
halt(code);
}
diff --git a/scripts/functions.bash b/scripts/functions.bash
index 7ed5a517250bc..7b4a5e4e63c9b 100644
--- a/scripts/functions.bash
+++ b/scripts/functions.bash
@@ -40,3 +40,14 @@ function for_each_unittest()
"$cmd" "$testname" "$groups" "$smp" "$kernel" "$opts" "$arch" "$check"
exec {fd}<&-
}
+
+function snoop_exitcode()
+{
+ while read line; do
+ if [[ $line =~ ^EXIT:\ STATUS=(.*)$ ]]; then
+ code=${BASH_REMATCH[1]}
+ fi
+ echo $line
+ done
+ exit $code
+}
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* [kvm-unit-tests PATCH 13/14] powerpc/ppc64: add run script and unittests.cfg
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (11 preceding siblings ...)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 12/14] scripts: add exit code snooper Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-03 14:41 ` [kvm-unit-tests PATCH 14/14] mkstandalone: add support for powerpc Andrew Jones
` (2 subsequent siblings)
15 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
Now that we have all the pieces, pull them together into the
run script. To run a test do ./powerpc-run powerpc/test.elf
To run all tests in unittests.cfg do ./run_tests.sh
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
powerpc/run | 57 +++++++++++++++++++++++++++++++++++++++++++++++++++
powerpc/unittests.cfg | 19 +++++++++++++++++
2 files changed, 76 insertions(+)
create mode 100755 powerpc/run
create mode 100644 powerpc/unittests.cfg
diff --git a/powerpc/run b/powerpc/run
new file mode 100755
index 0000000000000..5bc826765eff8
--- /dev/null
+++ b/powerpc/run
@@ -0,0 +1,57 @@
+#!/bin/bash
+
+if [ ! -f config.mak ]; then
+ echo run ./configure first. See ./configure -h
+ exit 2
+fi
+source config.mak
+source scripts/functions.bash
+
+processor="$PROCESSOR"
+
+# Default to using KVM if available
+if [ -c /dev/kvm ]; then
+ if [ "$HOST" = "ppc64" ] && [ "$ARCH" = "ppc64" ]; then
+ usingkvm=yes
+ fi
+fi
+
+qemu="${QEMU:-qemu-system-$ARCH_NAME}"
+qpath=$(which $qemu 2>/dev/null)
+
+if [ -z "$qpath" ]; then
+ echo $qemu not found.
+ exit 2
+fi
+
+if ! $qemu -machine '?' 2>&1 | grep 'pseries' > /dev/null; then
+ echo "$qpath doesn't support pSeries ('-machine pseries'). Exiting."
+ exit 2
+fi
+
+boot_rom='powerpc/boot_rom.bin'
+if [ -f powerpc/rom/boot_rom.bin ]; then
+ boot_rom='powerpc/rom/boot_rom.bin'
+fi
+
+M='-machine pseries'
+if [ "$DRYRUN" != "yes" ]; then
+ if [ "$usingkvm" = "yes" ]; then
+ M+=',accel=kvm'
+ else
+ echo "Running with TCG"
+ M+=',accel=tcg'
+ fi
+else
+ M+=',accel=kvm:tcg'
+fi
+
+command="$qemu $M -bios $boot_rom -display none -serial stdio -kernel"
+echo $command "$@"
+
+if [ "$DRYRUN" != "yes" ]; then
+ $command "$@" | $SNOOPER
+ ret=$?
+ echo Return value from qemu: $ret
+ exit $ret
+fi
diff --git a/powerpc/unittests.cfg b/powerpc/unittests.cfg
new file mode 100644
index 0000000000000..5da2c6242a1f9
--- /dev/null
+++ b/powerpc/unittests.cfg
@@ -0,0 +1,19 @@
+# Define your new unittest following the convention:
+# [unittest_name]
+# file = foo.elf # Name of the "kernel" to be used
+# smp = 2 # Number of processors the VM will use during this test
+# # Use $MAX_SMP to use the maximum the host supports.
+# extra_params = -append <params...> # Additional parameters used
+# arch = powerpc/ppc64 # Only if test case is specific to one
+# groups = group1 group2 # Used to identify test cases with run_tests -g ...
+
+#
+# Test that the configured number of processors (smp = <num>), and
+# that the configured amount of memory (-m <MB>) are correctly setup
+# by the framework.
+#
+[selftest-setup]
+file = selftest.elf
+smp = 2
+extra_params = -m 256 -append 'setup smp=2 mem=256'
+groups = selftest
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* [kvm-unit-tests PATCH 14/14] mkstandalone: add support for powerpc
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (12 preceding siblings ...)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 13/14] powerpc/ppc64: add run script and unittests.cfg Andrew Jones
@ 2015-08-03 14:41 ` Andrew Jones
2015-08-03 17:06 ` [kvm-unit-tests PATCH 00/14] ppc64: initial drop Paolo Bonzini
2015-11-03 7:08 ` Thomas Huth
15 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-08-03 14:41 UTC (permalink / raw)
To: kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier, pbonzini
PowerPC needs firmware and an exit code snooper.
Signed-off-by: Andrew Jones <drjones@redhat.com>
---
configure | 3 +++
powerpc/run | 2 +-
scripts/mkstandalone.sh | 50 +++++++++++++++++++++++++++++++++++++++++++++----
3 files changed, 50 insertions(+), 5 deletions(-)
diff --git a/configure b/configure
index e1f70bd79d4f7..7b332269e0d2e 100755
--- a/configure
+++ b/configure
@@ -11,6 +11,7 @@ arch=`uname -m | sed -e s/i.86/i386/ | sed -e 's/arm.*/arm/'`
host=$arch
cross_prefix=
snooper=
+firmware=
usage() {
cat <<-EOF
@@ -85,6 +86,7 @@ elif [ "$arch" = "arm" ] || [ "$arch" = "arm64" ]; then
elif [ "$arch" = "powerpc" ] || [ "$arch" = "ppc64" ]; then
testdir=powerpc
snooper=snoop_exitcode
+ firmware=$testdir/boot_rom.bin
else
testdir=$arch
fi
@@ -140,4 +142,5 @@ AR=$cross_prefix$ar
API=$api
TEST_DIR=$testdir
SNOOPER="$snooper"
+FIRMWARE=$firmware
EOF
diff --git a/powerpc/run b/powerpc/run
index 5bc826765eff8..1d5b1c9135299 100755
--- a/powerpc/run
+++ b/powerpc/run
@@ -29,7 +29,7 @@ if ! $qemu -machine '?' 2>&1 | grep 'pseries' > /dev/null; then
exit 2
fi
-boot_rom='powerpc/boot_rom.bin'
+boot_rom=$FIRMWARE
if [ -f powerpc/rom/boot_rom.bin ]; then
boot_rom='powerpc/rom/boot_rom.bin'
fi
diff --git a/scripts/mkstandalone.sh b/scripts/mkstandalone.sh
index 4a6e4a3ada1a3..dd10ece6940de 100755
--- a/scripts/mkstandalone.sh
+++ b/scripts/mkstandalone.sh
@@ -17,6 +17,10 @@ elif [ -n "$one_kernel" ] && [ -z "$one_testname" ]; then
one_testname="${one_kernel_base%.*}"
fi
+if [ "$SNOOPER" ]; then
+ snooper="| $SNOOPER"
+fi
+
unittests=$TEST_DIR/unittests.cfg
mkdir -p tests
@@ -51,6 +55,22 @@ function mkstandalone()
#!/bin/sh
EOF
+if [ "$SNOOPER" = "snoop_exitcode" ]; then
+ cat <<EOF >> $standalone
+snoop_exitcode()
+{
+ while read line; do
+ if echo \$line | grep 'EXIT: STATUS=' > /dev/null; then
+ code=\`echo \$line | cut -d= -f2\`
+ fi
+ echo \$line
+ done
+ exit \$code
+}
+
+EOF
+fi
+
if [ "$arch" ]; then
cat <<EOF >> $standalone
ARCH=\`uname -m | sed -e s/i.86/i386/ | sed -e 's/arm.*/arm/'\`
@@ -82,7 +102,21 @@ exit 1
EOF
else
cat <<EOF >> $standalone
-trap 'rm -f \$bin; exit 1' HUP INT TERM
+trap 'rm -f \$fwbin \$bin; exit 1' HUP INT TERM
+
+EOF
+if [ -f "$FIRMWARE" ]; then
+ cat <<EOF >> $standalone
+fwbin=\`mktemp\`
+base64 -d << 'BIN_EOF' | zcat > \$fwbin &&
+EOF
+gzip - < $FIRMWARE | base64 >> $standalone
+ cat <<EOF >> $standalone
+BIN_EOF
+
+EOF
+fi
+ cat <<EOF >> $standalone
bin=\`mktemp\`
base64 -d << 'BIN_EOF' | zcat > \$bin &&
EOF
@@ -101,9 +135,17 @@ while \$qemu \$cmdline -smp \$MAX_SMP 2>&1 | grep 'exceeds max cpus' > /dev/null
MAX_SMP=\`expr \$MAX_SMP - 1\`
done
-cmdline="\`echo '$cmdline' | sed s%$kernel%\$bin%\`"
+cmdline='$cmdline'
+EOF
+if [ "$FIRMWARE" ]; then
+ cat <<EOF >> $standalone
+cmdline="\`echo \$cmdline | sed s%$FIRMWARE%\$fwbin%\`"
+EOF
+fi
+ cat <<EOF >> $standalone
+cmdline="\`echo \$cmdline | sed s%$kernel%\$bin%\`"
echo \$qemu $cmdline -smp $smp $opts
-\$qemu \$cmdline -smp $smp $opts
+\$qemu \$cmdline -smp $smp $opts $snooper
ret=\$?
echo Return value from qemu: \$ret
if [ \$ret -le 1 ]; then
@@ -111,7 +153,7 @@ if [ \$ret -le 1 ]; then
else
echo FAIL $testname 1>&2
fi
-rm -f \$bin
+rm -f \$fwbin \$bin
exit 0
EOF
fi
--
2.4.3
^ permalink raw reply related [flat|nested] 36+ messages in thread* Re: [kvm-unit-tests PATCH 00/14] ppc64: initial drop
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (13 preceding siblings ...)
2015-08-03 14:41 ` [kvm-unit-tests PATCH 14/14] mkstandalone: add support for powerpc Andrew Jones
@ 2015-08-03 17:06 ` Paolo Bonzini
2015-11-03 7:08 ` Thomas Huth
15 siblings, 0 replies; 36+ messages in thread
From: Paolo Bonzini @ 2015-08-03 17:06 UTC (permalink / raw)
To: Andrew Jones, kvm, kvm-ppc; +Cc: dgibson, david, agraf, thuth, lvivier
On 03/08/2015 16:41, Andrew Jones wrote:
> This series is the first series of a series of series that will
> bring support to kvm-unit-tests for ppc64, and eventually ppc64le.
> (Yes, the word 'series' is four times in that sentence. And now I've
> typed 'series' five times! Wait, now six times! OK, moving on...)
>
> This series brings basic setup; starts a test's C entry point, main(),
> and printf, exit, and malloc work. Three more series following this
> one will bring; vector support, mmu support, and smp support, at which
> point I believe the framework can just evolve with the creation of unit
> tests.
>
> Executive summary of the patches is as follows
> 01: A needed fix to asm-generic/io.h. This could go in without the
> rest of the series, but it's only needed for BE.
> 02: Share arm/selftest.c utility functions (for powerpc/selftest.c)
> 03: Split makefiles up and rename them. This is almost an RFC,
> since it generates some churn, but I think it has value now
> that we're adding another architecture.
> 04-05: Here's the framework start; empty skeleton plus some tweaks.
> 06-07: Adds a bootloader. As we target the SPAPR machine type that
> requires one.
> 08: printf support
> 09: More of setup's framework, and do more setup.
> 10: Relocate support (since we can't link LMA == VMA)
> 11: exit() support
> 12-14: Unit test running scripts
>
> These patches are available here
> https://github.com/rhdrjones/kvm-unit-tests/commits/ppc64/initial-drop
>
> and they're based on this branch
> https://github.com/rhdrjones/kvm-unit-tests/commits/staging
>
> which is based on (upstream's next branch)
> http://git.kernel.org/cgit/virt/kvm/kvm-unit-tests.git/log/?h=next
Now pushed to master. Thanks for the ppc64 tests, I'll review them as
well as Alex's mttcg suite.
Paolo
> Thanks,
> drew
>
> Andrew Jones (14):
> lib: asm-generic: add missing casts
> lib: share arm-selftest utility functions
> config: no need to mix arch makefiles
> powerpc/ppc64: start skeleton framework
> powerpc/pp64: ppc-ify makefiles and linker script
> powerpc/ppc64: add boot rom source
> powerpc/ppc64: add bootloader to bounce into memory
> powerpc/ppc64: add HV putchar
> powerpc/ppc64: adapt arm's setup
> powerpc/ppc64: relocate linker VMAs
> powerpc/ppc64: add rtas_power_off
> scripts: add exit code snooper
> powerpc/ppc64: add run script and unittests.cfg
> mkstandalone: add support for powerpc
>
> Makefile | 2 +-
> README | 3 +-
> arm/Makefile | 1 +
> config/config-arm.mak => arm/Makefile.arm | 2 +-
> config/config-arm64.mak => arm/Makefile.arm64 | 2 +-
> .../config-arm-common.mak => arm/Makefile.common | 3 +-
> arm/selftest.c | 33 ++-----
> configure | 10 +++
> lib/arm/io.c | 1 +
> lib/asm-generic/io.h | 12 +--
> lib/libcflat.h | 14 +--
> lib/powerpc/.gitignore | 1 +
> lib/powerpc/asm/rtas.h | 27 ++++++
> lib/powerpc/asm/setup.h | 27 ++++++
> lib/powerpc/io.c | 35 ++++++++
> lib/powerpc/mmu.c | 0
> lib/powerpc/rtas.c | 84 ++++++++++++++++++
> lib/powerpc/setup.c | 82 ++++++++++++++++++
> lib/powerpc/smp.c | 0
> lib/ppc64/.gitignore | 1 +
> lib/ppc64/asm-offsets.c | 12 +++
> lib/ppc64/asm/asm-offsets.h | 1 +
> lib/ppc64/asm/io.h | 5 ++
> lib/ppc64/asm/page.h | 1 +
> lib/ppc64/asm/rtas.h | 1 +
> lib/ppc64/asm/setup.h | 1 +
> lib/ppc64/asm/spinlock.h | 11 +++
> lib/ppc64/processor.c | 0
> lib/ppc64/spinlock.c | 11 +++
> lib/report.c | 16 ++++
> lib/util.c | 20 +++++
> lib/util.h | 23 +++++
> powerpc/Makefile | 1 +
> powerpc/Makefile.common | 71 +++++++++++++++
> powerpc/Makefile.ppc64 | 21 +++++
> powerpc/boot_rom.bin | Bin 0 -> 280 bytes
> powerpc/cstart64.S | 96 +++++++++++++++++++++
> powerpc/flat.lds | 44 ++++++++++
> powerpc/reloc64.c | 55 ++++++++++++
> powerpc/rom/.gitignore | 1 +
> powerpc/rom/Makefile | 36 ++++++++
> powerpc/rom/header.img | 1 +
> powerpc/rom/rom.ffs | 4 +
> powerpc/rom/stage1.S | 5 ++
> powerpc/run | 57 ++++++++++++
> powerpc/selftest.c | 65 ++++++++++++++
> powerpc/unittests.cfg | 19 ++++
> {config => scripts}/asm-offsets.mak | 0
> scripts/functions.bash | 11 +++
> scripts/mkstandalone.sh | 50 ++++++++++-
> x86/Makefile | 1 +
> .../config-x86-common.mak => x86/Makefile.common | 0
> config/config-i386.mak => x86/Makefile.i386 | 2 +-
> config/config-x86_64.mak => x86/Makefile.x86_64 | 2 +-
> 54 files changed, 936 insertions(+), 48 deletions(-)
> create mode 100644 arm/Makefile
> rename config/config-arm.mak => arm/Makefile.arm (90%)
> rename config/config-arm64.mak => arm/Makefile.arm64 (89%)
> rename config/config-arm-common.mak => arm/Makefile.common (97%)
> create mode 100644 lib/powerpc/.gitignore
> create mode 100644 lib/powerpc/asm/rtas.h
> create mode 100644 lib/powerpc/asm/setup.h
> create mode 100644 lib/powerpc/io.c
> create mode 100644 lib/powerpc/mmu.c
> create mode 100644 lib/powerpc/rtas.c
> create mode 100644 lib/powerpc/setup.c
> create mode 100644 lib/powerpc/smp.c
> create mode 100644 lib/ppc64/.gitignore
> create mode 100644 lib/ppc64/asm-offsets.c
> create mode 100644 lib/ppc64/asm/asm-offsets.h
> create mode 100644 lib/ppc64/asm/io.h
> create mode 100644 lib/ppc64/asm/page.h
> create mode 100644 lib/ppc64/asm/rtas.h
> create mode 100644 lib/ppc64/asm/setup.h
> create mode 100644 lib/ppc64/asm/spinlock.h
> create mode 100644 lib/ppc64/processor.c
> create mode 100644 lib/ppc64/spinlock.c
> create mode 100644 lib/util.c
> create mode 100644 lib/util.h
> create mode 100644 powerpc/Makefile
> create mode 100644 powerpc/Makefile.common
> create mode 100644 powerpc/Makefile.ppc64
> create mode 100644 powerpc/boot_rom.bin
> create mode 100644 powerpc/cstart64.S
> create mode 100644 powerpc/flat.lds
> create mode 100644 powerpc/reloc64.c
> create mode 100644 powerpc/rom/.gitignore
> create mode 100644 powerpc/rom/Makefile
> create mode 100644 powerpc/rom/header.img
> create mode 100644 powerpc/rom/rom.ffs
> create mode 100644 powerpc/rom/stage1.S
> create mode 100755 powerpc/run
> create mode 100644 powerpc/selftest.c
> create mode 100644 powerpc/unittests.cfg
> rename {config => scripts}/asm-offsets.mak (100%)
> create mode 100644 x86/Makefile
> rename config/config-x86-common.mak => x86/Makefile.common (100%)
> rename config/config-i386.mak => x86/Makefile.i386 (91%)
> rename config/config-x86_64.mak => x86/Makefile.x86_64 (92%)
>
^ permalink raw reply [flat|nested] 36+ messages in thread* Re: [kvm-unit-tests PATCH 00/14] ppc64: initial drop
2015-08-03 14:41 [kvm-unit-tests PATCH 00/14] ppc64: initial drop Andrew Jones
` (14 preceding siblings ...)
2015-08-03 17:06 ` [kvm-unit-tests PATCH 00/14] ppc64: initial drop Paolo Bonzini
@ 2015-11-03 7:08 ` Thomas Huth
2015-11-03 9:40 ` Paolo Bonzini
15 siblings, 1 reply; 36+ messages in thread
From: Thomas Huth @ 2015-11-03 7:08 UTC (permalink / raw)
To: Andrew Jones, kvm, kvm-ppc; +Cc: dgibson, david, agraf, lvivier, pbonzini
On 03/08/15 16:41, Andrew Jones wrote:
> This series is the first series of a series of series that will
> bring support to kvm-unit-tests for ppc64, and eventually ppc64le.
Hi Andrew,
may I ask about the current state of ppc64 support in the
kvm-unit-tests? Is there a newer version available than the one you
posted three months ago?
Thanks,
Thomas
^ permalink raw reply [flat|nested] 36+ messages in thread* Re: [kvm-unit-tests PATCH 00/14] ppc64: initial drop
2015-11-03 7:08 ` Thomas Huth
@ 2015-11-03 9:40 ` Paolo Bonzini
2015-11-03 14:56 ` Andrew Jones
0 siblings, 1 reply; 36+ messages in thread
From: Paolo Bonzini @ 2015-11-03 9:40 UTC (permalink / raw)
To: Thomas Huth, Andrew Jones, kvm, kvm-ppc; +Cc: dgibson, david, agraf, lvivier
On 03/11/2015 08:08, Thomas Huth wrote:
> On 03/08/15 16:41, Andrew Jones wrote:
>> > This series is the first series of a series of series that will
>> > bring support to kvm-unit-tests for ppc64, and eventually ppc64le.
> Hi Andrew,
>
> may I ask about the current state of ppc64 support in the
> kvm-unit-tests? Is there a newer version available than the one you
> posted three months ago?
I've been a slob with all the kvm-unit-tests patches. Andrew, can you
send a single submission of all the patches, so that I can review them
and apply them?
Paolo
^ permalink raw reply [flat|nested] 36+ messages in thread
* Re: [kvm-unit-tests PATCH 00/14] ppc64: initial drop
2015-11-03 9:40 ` Paolo Bonzini
@ 2015-11-03 14:56 ` Andrew Jones
0 siblings, 0 replies; 36+ messages in thread
From: Andrew Jones @ 2015-11-03 14:56 UTC (permalink / raw)
To: Paolo Bonzini; +Cc: Thomas Huth, kvm, kvm-ppc, dgibson, david, agraf, lvivier
On Tue, Nov 03, 2015 at 10:40:18AM +0100, Paolo Bonzini wrote:
>
>
> On 03/11/2015 08:08, Thomas Huth wrote:
> > On 03/08/15 16:41, Andrew Jones wrote:
> >> > This series is the first series of a series of series that will
> >> > bring support to kvm-unit-tests for ppc64, and eventually ppc64le.
> > Hi Andrew,
> >
> > may I ask about the current state of ppc64 support in the
> > kvm-unit-tests? Is there a newer version available than the one you
> > posted three months ago?
>
Hi Thomas,
I haven't gotten around to preparing the v2 yet :-( I do have it on
my TODO list, and I'm looking forward to working on it. Now that I know
you're looking for it, I'll try to bump it up in priority. Thanks for
the interest!
> I've been a slob with all the kvm-unit-tests patches. Andrew, can you
> send a single submission of all the patches, so that I can review them
> and apply them?
Hi Paolo,
I've got several patches on my staging branch that I believe are ready.
I plan to send those as a big "pull" series for your review soon.
Thanks,
drew
>
> Paolo
> --
> To unsubscribe from this list: send the line "unsubscribe kvm" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 36+ messages in thread