[PATCH RFC 1/5] samples/bpf: Use getppid instead of getpgrp for array map stress

[PATCH RFC 1/5] samples/bpf: Use getppid instead of getpgrp for array map stress

[Joel Fernandes]

When cross-compiling the bpf sample map_perf_test for aarch64, I find that
__NR_getpgrp is undefined. This causes build errors. This syscall is deprecated
and requires defining __ARCH_WANT_SYSCALL_DEPRECATED. To avoid having to define
that, just use a different syscall (getppid) for the array map stress test.

Signed-off-by: Joel Fernandes <joelaf@google.com>
---
 samples/bpf/map_perf_test_kern.c | 2 +-
 samples/bpf/map_perf_test_user.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/samples/bpf/map_perf_test_kern.c b/samples/bpf/map_perf_test_kern.c
index 245165817fbe..038ffec295cf 100644
--- a/samples/bpf/map_perf_test_kern.c
+++ b/samples/bpf/map_perf_test_kern.c
@@ -232,7 +232,7 @@ int stress_hash_map_lookup(struct pt_regs *ctx)
 	return 0;
 }
 
-SEC("kprobe/sys_getpgrp")
+SEC("kprobe/sys_getppid")
 int stress_array_map_lookup(struct pt_regs *ctx)
 {
 	u32 key = 1, i;
diff --git a/samples/bpf/map_perf_test_user.c b/samples/bpf/map_perf_test_user.c
index 1a8894b5ac51..1e9e68942197 100644
--- a/samples/bpf/map_perf_test_user.c
+++ b/samples/bpf/map_perf_test_user.c
@@ -232,7 +232,7 @@ static void test_array_lookup(int cpu)
 
 	start_time = time_get_ns();
 	for (i = 0; i < max_cnt; i++)
-		syscall(__NR_getpgrp, 0);
+		syscall(__NR_getppid, 0);
 	printf("%d:array_lookup %lld lookups per sec\n",
 	       cpu, max_cnt * 1000000000ll * 64 / (time_get_ns() - start_time));
 }
-- 
2.14.0.rc1.383.gd1ce394fe2-goog

[PATCH RFC 2/5] samples/bpf: Enable cross compiler support

[Joel Fernandes]

When cross compiling, bpf samples use HOSTCC, however what we really want is to
use the cross compiler to build for the cross target since that is what will
help run the BPF target code.  Detect this and also set -static as LDFLAGS
since often times we don't have control over what C library the cross target is
running and its not smart to rely on it.

Signed-off-by: Joel Fernandes <joelaf@google.com>
---
 samples/bpf/Makefile | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/samples/bpf/Makefile b/samples/bpf/Makefile
index 6c7468eb3684..e5642c8c144d 100644
--- a/samples/bpf/Makefile
+++ b/samples/bpf/Makefile
@@ -152,6 +152,12 @@ HOSTLOADLIBES_test_map_in_map += -lelf
 LLC ?= llc
 CLANG ?= clang
 
+# Detect that we're cross compiling and use the right compilers and flags
+ifdef CROSS_COMPILE
+HOSTCC = $(CROSS_COMPILE)gcc
+HOSTLDFLAGS += -static
+endif
+
 # Trick to allow make to be run from this directory
 all:
 	$(MAKE) -C ../../ $(CURDIR)/
-- 
2.14.0.rc1.383.gd1ce394fe2-goog

[PATCH RFC 3/5] samples/bpf: Fix inline asm issues building samples on arm64

[Joel Fernandes]

inline assembly has haunted building samples on arm64 for quite sometime.
This patch uses the pre-processor to noop all occurences of inline asm when
compiling the BPF sample for the BPF target.

This patch reintroduces inclusion of asm/sysregs.h which needs to be included
to avoid compiler errors now, see [1]. Previously a hack prevented this
inclusion [2] (to avoid the exact problem this patch fixes - skipping inline
assembler) but the hack causes other errors now and no longer works.

Using the preprocessor to noop the inline asm occurences, we also avoid
any future unstable hackery needed (such as those that skip asm headers)
and provides information that asm headers may have which could have been
used but which the inline asm issues prevented. This is the least messy
of all hacks in my opinion.

[1] https://lkml.org/lkml/2017/8/5/143
[2] https://lists.linaro.org/pipermail/linaro-kernel/2015-November/024036.html

Signed-off-by: Joel Fernandes <joelaf@google.com>
---
 samples/bpf/Makefile           | 40 +++++++++++++++++++++++++++++++++-------
 samples/bpf/arm64_asmstubs.h   |  3 +++
 samples/bpf/bpf_helpers.h      | 12 ++++++------
 samples/bpf/generic_asmstubs.h |  4 ++++
 4 files changed, 46 insertions(+), 13 deletions(-)
 create mode 100644 samples/bpf/arm64_asmstubs.h
 create mode 100644 samples/bpf/generic_asmstubs.h

diff --git a/samples/bpf/Makefile b/samples/bpf/Makefile
index e5642c8c144d..7591cdd7fe69 100644
--- a/samples/bpf/Makefile
+++ b/samples/bpf/Makefile
@@ -151,6 +151,8 @@ HOSTLOADLIBES_test_map_in_map += -lelf
 #  make samples/bpf/ LLC=~/git/llvm/build/bin/llc CLANG=~/git/llvm/build/bin/clang
 LLC ?= llc
 CLANG ?= clang
+PERL ?= perl
+RM ?= rm
 
 # Detect that we're cross compiling and use the right compilers and flags
 ifdef CROSS_COMPILE
@@ -186,14 +188,38 @@ verify_target_bpf: verify_cmds
 
 $(src)/*.c: verify_target_bpf
 
-# asm/sysreg.h - inline assembly used by it is incompatible with llvm.
-# But, there is no easy way to fix it, so just exclude it since it is
-# useless for BPF samples.
-$(obj)/%.o: $(src)/%.c
-	$(CLANG) $(NOSTDINC_FLAGS) $(LINUXINCLUDE) $(EXTRA_CFLAGS) \
-		-D__KERNEL__ -D__ASM_SYSREG_H -Wno-unused-value -Wno-pointer-sign \
+curdir := $(shell dirname $(realpath $(lastword $(MAKEFILE_LIST))))
+ifeq ($(wildcard  $(curdir)/${ARCH}_asmstubs.h),)
+    ARCH_ASM_STUBS :=
+else
+    ARCH_ASM_STUBS := -include $(src)/${ARCH}_asmstubs.h
+endif
+
+ASM_STUBS := ${ARCH_ASM_STUBS} -include $(src)/generic_asmstubs.h
+
+CLANG_ARGS = $(NOSTDINC_FLAGS) $(LINUXINCLUDE) $(EXTRA_CFLAGS) \
+		-D__KERNEL__ -Wno-unused-value -Wno-pointer-sign \
+		$(ASM_STUBS) \
 		-Wno-compare-distinct-pointer-types \
 		-Wno-gnu-variable-sized-type-not-at-end \
 		-Wno-address-of-packed-member -Wno-tautological-compare \
 		-Wno-unknown-warning-option \
-		-O2 -emit-llvm -c $< -o -| $(LLC) -march=bpf -filetype=obj -o $@
+		-O2 -emit-llvm
+
+$(obj)/%.o: $(src)/%.c
+	# Steps to compile BPF sample while getting rid of inline asm
+	# This has the advantage of not having to skip important asm headers
+	# Step 1. Use clang preprocessor to stub out asm() calls
+	# Step 2. Replace all "asm volatile" with single keyword "asmvolatile"
+	# Step 3. Use clang preprocessor to noop all asm volatile() calls
+	#         and restore asm_bpf to asm for BPF's asm directives
+	# Step 4. Compile and link
+
+	$(CLANG) -E $(CLANG_ARGS) -c $< -o - | \
+	$(PERL) -pe "s/[_\s]*asm[_\s]*volatile[_\s]*/asmvolatile/g" | \
+	$(CLANG) -E $(ASM_STUBS) - -o - | \
+	$(CLANG) -E -Dasm_bpf=asm - -o $@.tmp.c
+
+	$(CLANG) $(CLANG_ARGS) -c $@.tmp.c \
+		-o - | $(LLC) -march=bpf -filetype=obj -o $@
+	$(RM) $@.tmp.c
diff --git a/samples/bpf/arm64_asmstubs.h b/samples/bpf/arm64_asmstubs.h
new file mode 100644
index 000000000000..23d47dbe61b1
--- /dev/null
+++ b/samples/bpf/arm64_asmstubs.h
@@ -0,0 +1,3 @@
+/* Special handing for current_stack_pointer */
+#define __ASM_STACK_POINTER_H
+#define current_stack_pointer 0
diff --git a/samples/bpf/bpf_helpers.h b/samples/bpf/bpf_helpers.h
index 9a9c95f2c9fb..67c9c4438e4b 100644
--- a/samples/bpf/bpf_helpers.h
+++ b/samples/bpf/bpf_helpers.h
@@ -64,12 +64,12 @@ static int (*bpf_xdp_adjust_head)(void *ctx, int offset) =
  * emit BPF_LD_ABS and BPF_LD_IND instructions
  */
 struct sk_buff;
-unsigned long long load_byte(void *skb,
-			     unsigned long long off) asm("llvm.bpf.load.byte");
-unsigned long long load_half(void *skb,
-			     unsigned long long off) asm("llvm.bpf.load.half");
-unsigned long long load_word(void *skb,
-			     unsigned long long off) asm("llvm.bpf.load.word");
+unsigned long long load_byte(void *skb, unsigned long long off)
+			     asm_bpf("llvm.bpf.load.byte");
+unsigned long long load_half(void *skb, unsigned long long off)
+			     asm_bpf("llvm.bpf.load.half");
+unsigned long long load_word(void *skb, unsigned long long off)
+			     asm_bpf("llvm.bpf.load.word");
 
 /* a helper structure used by eBPF C program
  * to describe map attributes to elf_bpf loader
diff --git a/samples/bpf/generic_asmstubs.h b/samples/bpf/generic_asmstubs.h
new file mode 100644
index 000000000000..1b9e9f5094d8
--- /dev/null
+++ b/samples/bpf/generic_asmstubs.h
@@ -0,0 +1,4 @@
+#define bpf_noop_stub
+#define asm(...) bpf_noop_stub
+#define __asm__(...) bpf_noop_stub
+#define asmvolatile(...) bpf_noop_stub
-- 
2.14.0.rc1.383.gd1ce394fe2-goog

[PATCH RFC 4/5] samples/bpf: Fix pt_regs issues when cross-compiling

[Joel Fernandes]

BPF samples fail to build when cross-compiling for ARM64 because of incorrect
pt_regs param selection. This is because clang defines __x86_64__ and
bpf_headers thinks we're building for x86. Since clang is building for the BPF
target, it shouldn't make assumptions about what target the BPF program is
going to run on. To fix this, lets pass ARCH so the header knows which target
the BPF program is being compiled for and can use the correct pt_regs code.

Signed-off-by: Joel Fernandes <joelaf@google.com>
---
 samples/bpf/Makefile      |  2 +-
 samples/bpf/bpf_helpers.h | 49 +++++++++++++++++++++++++++++++++++++++++------
 2 files changed, 44 insertions(+), 7 deletions(-)

diff --git a/samples/bpf/Makefile b/samples/bpf/Makefile
index 7591cdd7fe69..8cbcaffe4001 100644
--- a/samples/bpf/Makefile
+++ b/samples/bpf/Makefile
@@ -199,7 +199,7 @@ ASM_STUBS := ${ARCH_ASM_STUBS} -include $(src)/generic_asmstubs.h
 
 CLANG_ARGS = $(NOSTDINC_FLAGS) $(LINUXINCLUDE) $(EXTRA_CFLAGS) \
 		-D__KERNEL__ -Wno-unused-value -Wno-pointer-sign \
-		$(ASM_STUBS) \
+		-D__TARGET_ARCH_$(ARCH) $(ASM_STUBS) \
 		-Wno-compare-distinct-pointer-types \
 		-Wno-gnu-variable-sized-type-not-at-end \
 		-Wno-address-of-packed-member -Wno-tautological-compare \
diff --git a/samples/bpf/bpf_helpers.h b/samples/bpf/bpf_helpers.h
index 67c9c4438e4b..199d2e32703a 100644
--- a/samples/bpf/bpf_helpers.h
+++ b/samples/bpf/bpf_helpers.h
@@ -96,7 +96,42 @@ static int (*bpf_skb_under_cgroup)(void *ctx, void *map, int index) =
 static int (*bpf_skb_change_head)(void *, int len, int flags) =
 	(void *) BPF_FUNC_skb_change_head;
 
+/* Scan the ARCH passed in from ARCH env variable (see Makefile) */
+#if defined(__TARGET_ARCH_x86)
+	#define bpf_target_x86
+	#define bpf_target_defined
+#elif defined(__TARGET_ARCH_s930x)
+	#define bpf_target_s930x
+	#define bpf_target_defined
+#elif defined(__TARGET_ARCH_arm64)
+	#define bpf_target_arm64
+	#define bpf_target_defined
+#elif defined(__TARGET_ARCH_powerpc)
+	#define bpf_target_powerpc
+	#define bpf_target_defined
+#elif defined(__TARGET_ARCH_sparc)
+	#define bpf_target_sparc
+	#define bpf_target_defined
+#else
+	#undef bpf_target_defined
+#endif
+
+/* Fall back to what the compiler says */
+#ifndef bpf_target_defined
 #if defined(__x86_64__)
+	#define bpf_target_x86
+#elif defined(__s390x__)
+	#define bpf_target_s930x
+#elif defined(__aarch64__)
+	#define bpf_target_arm64
+#elif defined(__powerpc__)
+	#define bpf_target_powerpc
+#elif defined(__sparc__)
+	#define bpf_target_sparc
+#endif
+#endif
+
+#if defined(bpf_target_x86)
 
 #define PT_REGS_PARM1(x) ((x)->di)
 #define PT_REGS_PARM2(x) ((x)->si)
@@ -109,7 +144,7 @@ static int (*bpf_skb_change_head)(void *, int len, int flags) =
 #define PT_REGS_SP(x) ((x)->sp)
 #define PT_REGS_IP(x) ((x)->ip)
 
-#elif defined(__s390x__)
+#elif defined(bpf_target_s390x)
 
 #define PT_REGS_PARM1(x) ((x)->gprs[2])
 #define PT_REGS_PARM2(x) ((x)->gprs[3])
@@ -122,7 +157,7 @@ static int (*bpf_skb_change_head)(void *, int len, int flags) =
 #define PT_REGS_SP(x) ((x)->gprs[15])
 #define PT_REGS_IP(x) ((x)->psw.addr)
 
-#elif defined(__aarch64__)
+#elif defined(bpf_target_arm64)
 
 #define PT_REGS_PARM1(x) ((x)->regs[0])
 #define PT_REGS_PARM2(x) ((x)->regs[1])
@@ -135,7 +170,7 @@ static int (*bpf_skb_change_head)(void *, int len, int flags) =
 #define PT_REGS_SP(x) ((x)->sp)
 #define PT_REGS_IP(x) ((x)->pc)
 
-#elif defined(__powerpc__)
+#elif defined(bpf_target_powerpc)
 
 #define PT_REGS_PARM1(x) ((x)->gpr[3])
 #define PT_REGS_PARM2(x) ((x)->gpr[4])
@@ -146,7 +181,7 @@ static int (*bpf_skb_change_head)(void *, int len, int flags) =
 #define PT_REGS_SP(x) ((x)->sp)
 #define PT_REGS_IP(x) ((x)->nip)
 
-#elif defined(__sparc__)
+#elif defined(bpf_target_sparc)
 
 #define PT_REGS_PARM1(x) ((x)->u_regs[UREG_I0])
 #define PT_REGS_PARM2(x) ((x)->u_regs[UREG_I1])
@@ -156,6 +191,8 @@ static int (*bpf_skb_change_head)(void *, int len, int flags) =
 #define PT_REGS_RET(x) ((x)->u_regs[UREG_I7])
 #define PT_REGS_RC(x) ((x)->u_regs[UREG_I0])
 #define PT_REGS_SP(x) ((x)->u_regs[UREG_FP])
+
+/* Should this also be a bpf_target check for the sparc case? */
 #if defined(__arch64__)
 #define PT_REGS_IP(x) ((x)->tpc)
 #else
@@ -164,10 +201,10 @@ static int (*bpf_skb_change_head)(void *, int len, int flags) =
 
 #endif
 
-#ifdef __powerpc__
+#ifdef bpf_target_powerpc
 #define BPF_KPROBE_READ_RET_IP(ip, ctx)		({ (ip) = (ctx)->link; })
 #define BPF_KRETPROBE_READ_RET_IP		BPF_KPROBE_READ_RET_IP
-#elif defined(__sparc__)
+#elif bpf_target_sparc
 #define BPF_KPROBE_READ_RET_IP(ip, ctx)		({ (ip) = PT_REGS_RET(ctx); })
 #define BPF_KRETPROBE_READ_RET_IP		BPF_KPROBE_READ_RET_IP
 #else
-- 
2.14.0.rc1.383.gd1ce394fe2-goog

[PATCH RFC 5/5] samples/bpf: Add documentation on cross compilation

[Joel Fernandes]

Signed-off-by: Joel Fernandes <joelaf@google.com>
---
 samples/bpf/README.rst | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/samples/bpf/README.rst b/samples/bpf/README.rst
index 79f9a58f1872..2b906127ef54 100644
--- a/samples/bpf/README.rst
+++ b/samples/bpf/README.rst
@@ -64,3 +64,13 @@ It is also possible to point make to the newly compiled 'llc' or
 'clang' command via redefining LLC or CLANG on the make command line::
 
  make samples/bpf/ LLC=~/git/llvm/build/bin/llc CLANG=~/git/llvm/build/bin/clang
+
+Cross compiling samples
+-----------------------
+Inorder to cross-compile, say for arm64 targets, export CROSS_COMPILE and ARCH
+environment variables before calling make. This will direct make to build
+samples for the cross target.
+
+export ARCH=arm64
+export CROSS_COMPILE="aarch64-linux-gnu-"
+make samples/bpf/ LLC=~/git/llvm/build/bin/llc CLANG=~/git/llvm/build/bin/clang
-- 
2.14.0.rc1.383.gd1ce394fe2-goog