date:20150407

[dpdk-dev] [RFC PATCHv3 0/7] add OSv support

2015-04-07 Thread Takuya ASADA

This is 3rd version of OSv support patchset.

Changelist:
 - sync with latest master
 - updated Capstan build script for OSv v0.19
 - rte_eal_version.map fix
 - CONFIG_RTE_BUILD_SHARED_LIB=y and CONFIG_RTE_BUILD_COMBINE_LIBS=y by default
 - Check OSv NIC driver is loaded for each device, attach PMD only if OSv 
driver is not loaded
 - Use --maxnic on cmdline to skip loading OSv NIC driver
 - Don't enable MSI-X on virtio, since it's not enable when OSv NIC driver is 
not loaded

Takuya ASADA (7):
  mk: support compiling C++ code
  eal: Add extern C on eal_hugepages.h
  eal: Add extern C on eal_thread.h
  eal: Add extern C on eal_private.h
  add OSv support
  app/test: support OSv
  doc: Add Getting Started Guide for OSv

 app/test/test_eal_flags.c  |  34 +--
 app/test/test_timer_perf.c |   2 +-
 config/{common_linuxapp => common_osvapp}  |  29 +-
 ...xapp-gcc => defconfig_x86_64-native-osvapp-gcc} |   2 +-
 doc/guides/index.rst   |   1 +
 doc/guides/osv_gsg/build_dpdk.rst  | 282 +++
 doc/guides/osv_gsg/build_sample_apps.rst   | 123 
 doc/guides/{freebsd_gsg => osv_gsg}/index.rst  |   3 +-
 doc/guides/{freebsd_gsg => osv_gsg}/intro.rst  |  15 +-
 lib/librte_eal/Makefile|   2 +
 lib/librte_eal/common/eal_hugepages.h  |   8 +
 lib/librte_eal/common/eal_private.h|   8 +
 lib/librte_eal/common/eal_thread.h |   8 +
 Makefile => lib/librte_eal/osvapp/Makefile |   5 +-
 lib/librte_eal/osvapp/capstan/Capstanfile  |   6 +
 lib/librte_eal/osvapp/capstan/GET  |  22 ++
 lib/librte_eal/osvapp/eal/Makefile | 115 
 lib/librte_eal/{linuxapp => osvapp}/eal/eal.c  | 123 +---
 .../{linuxapp => osvapp}/eal/eal_alarm.c   |   0
 .../{linuxapp => osvapp}/eal/eal_debug.c   |   0
 lib/librte_eal/osvapp/eal/eal_hugepage_info.cc |  63 +
 .../{bsdapp => osvapp}/eal/eal_interrupts.c|   0
 .../eal/eal_lcore.c => osvapp/eal/eal_lcore.cc}|  53 ++--
 lib/librte_eal/{bsdapp => osvapp}/eal/eal_log.c|   0
 lib/librte_eal/osvapp/eal/eal_memory.cc| 148 ++
 lib/librte_eal/osvapp/eal/eal_pci.cc   | 312 +
 .../{linuxapp => osvapp}/eal/eal_thread.c  |   0
 lib/librte_eal/osvapp/eal/eal_timer.c  | 130 +
 .../eal/include/exec-env/rte_interrupts.h  |   0
 .../{bsdapp => osvapp}/eal/rte_eal_version.map |   6 -
 mk/exec-env/{linuxapp => osvapp}/rte.app.mk|   0
 mk/exec-env/{linuxapp => osvapp}/rte.vars.mk   |   6 +-
 mk/internal/rte.compile-pre.mk |  41 ++-
 mk/target/generic/rte.vars.mk  |   4 +
 mk/toolchain/gcc/rte.vars.mk   |   5 +-
 35 files changed, 1345 insertions(+), 211 deletions(-)
 copy config/{common_linuxapp => common_osvapp} (95%)
 copy config/{defconfig_x86_64-native-linuxapp-gcc => 
defconfig_x86_64-native-osvapp-gcc} (98%)
 create mode 100644 doc/guides/osv_gsg/build_dpdk.rst
 create mode 100644 doc/guides/osv_gsg/build_sample_apps.rst
 copy doc/guides/{freebsd_gsg => osv_gsg}/index.rst (96%)
 copy doc/guides/{freebsd_gsg => osv_gsg}/intro.rst (84%)
 copy Makefile => lib/librte_eal/osvapp/Makefile (93%)
 create mode 100644 lib/librte_eal/osvapp/capstan/Capstanfile
 create mode 100755 lib/librte_eal/osvapp/capstan/GET
 create mode 100644 lib/librte_eal/osvapp/eal/Makefile
 copy lib/librte_eal/{linuxapp => osvapp}/eal/eal.c (87%)
 copy lib/librte_eal/{linuxapp => osvapp}/eal/eal_alarm.c (100%)
 copy lib/librte_eal/{linuxapp => osvapp}/eal/eal_debug.c (100%)
 create mode 100644 lib/librte_eal/osvapp/eal/eal_hugepage_info.cc
 copy lib/librte_eal/{bsdapp => osvapp}/eal/eal_interrupts.c (100%)
 copy lib/librte_eal/{bsdapp/eal/eal_lcore.c => osvapp/eal/eal_lcore.cc} (80%)
 copy lib/librte_eal/{bsdapp => osvapp}/eal/eal_log.c (100%)
 create mode 100644 lib/librte_eal/osvapp/eal/eal_memory.cc
 create mode 100644 lib/librte_eal/osvapp/eal/eal_pci.cc
 copy lib/librte_eal/{linuxapp => osvapp}/eal/eal_thread.c (100%)
 create mode 100644 lib/librte_eal/osvapp/eal/eal_timer.c
 copy lib/librte_eal/{bsdapp => osvapp}/eal/include/exec-env/rte_interrupts.h 
(100%)
 copy lib/librte_eal/{bsdapp => osvapp}/eal/rte_eal_version.map (92%)
 copy mk/exec-env/{linuxapp => osvapp}/rte.app.mk (100%)
 copy mk/exec-env/{linuxapp => osvapp}/rte.vars.mk (95%)

-- 
2.1.0

[dpdk-dev] [RFC PATCHv3 1/7] mk: support compiling C++ code

2015-04-07 Thread Takuya ASADA

Since OSv is written in C++, we need to write OSv EAL in C++.
To do so, we need to compile .cc files by $(CXX).

This patch does not contain diff for clang and icc, but OSv EAL does not 
supported these toolchain, this is enough for now.

Signed-off-by: Takuya ASADA 
---
 mk/internal/rte.compile-pre.mk | 41 ++---
 mk/target/generic/rte.vars.mk  |  4 
 mk/toolchain/gcc/rte.vars.mk   |  5 -
 3 files changed, 46 insertions(+), 4 deletions(-)

diff --git a/mk/internal/rte.compile-pre.mk b/mk/internal/rte.compile-pre.mk
index b9bff4a..142f996 100644
--- a/mk/internal/rte.compile-pre.mk
+++ b/mk/internal/rte.compile-pre.mk
@@ -37,7 +37,7 @@ SRCS-all := $(SRCS-y) $(SRCS-n) $(SRCS-)

 # convert source to obj file
 src2obj = $(strip $(patsubst %.c,%.o,\
-   $(patsubst %.S,%_s.o,$(1
+   $(patsubst %.S,%_s.o,$(patsubst %.cc,%.o,$(1)

 # add a dot in front of the file name
 dotfile = $(strip $(foreach f,$(1),\
@@ -46,12 +46,12 @@ dotfile = $(strip $(foreach f,$(1),\
 # convert source/obj files into dot-dep filename (does not
 # include .S files)
 src2dep = $(strip $(call dotfile,$(patsubst %.c,%.o.d, \
-   $(patsubst %.S,,$(1)
+   $(patsubst %.S,,$(patsubst %.cc,%.o.d,$(1))
 obj2dep = $(strip $(call dotfile,$(patsubst %.o,%.o.d,$(1

 # convert source/obj files into dot-cmd filename
 src2cmd = $(strip $(call dotfile,$(patsubst %.c,%.o.cmd, \
-   $(patsubst %.S,%_s.o.cmd,$(1)
+   $(patsubst %.S,%_s.o.cmd,$(patsubst %.cc,%.o.cmd,$(1))
 obj2cmd = $(strip $(call dotfile,$(patsubst %.o,%.o.cmd,$(1

 OBJS-y := $(call src2obj,$(SRCS-y))
@@ -78,11 +78,19 @@ C_TO_O = $(HOSTCC) -Wp,-MD,$(call obj2dep,$(@)).tmp 
$(HOST_CFLAGS) \
$(CFLAGS_$(@)) $(HOST_EXTRA_CFLAGS) -o $@ -c $<
 C_TO_O_STR = $(subst ','\'',$(C_TO_O)) #'# fix syntax highlight
 C_TO_O_DISP = $(if $(V),"$(C_TO_O_STR)","  HOSTCC $(@)")
+CXX_TO_O = $(HOSTCXX) -Wp,-MD,$(call obj2dep,$(@)).tmp $(HOST_CXXFLAGS) \
+   $(CXXFLAGS_$(@)) $(HOST_EXTRA_CXXFLAGS) -o $@ -c $<
+CXX_TO_O_STR = $(subst ','\'',$(CXX_TO_O)) #'# fix syntax highlight
+CXX_TO_O_DISP = $(if $(V),"$(CXX_TO_O_STR)","  HOSTCXX $(@)")
 else
 C_TO_O = $(CC) -Wp,-MD,$(call obj2dep,$(@)).tmp $(CFLAGS) \
$(CFLAGS_$(@)) $(EXTRA_CFLAGS) -o $@ -c $<
 C_TO_O_STR = $(subst ','\'',$(C_TO_O)) #'# fix syntax highlight
 C_TO_O_DISP = $(if $(V),"$(C_TO_O_STR)","  CC $(@)")
+CXX_TO_O = $(CXX) -Wp,-MD,$(call obj2dep,$(@)).tmp $(CXXFLAGS) \
+   $(CXXFLAGS_$(@)) $(EXTRA_CXXFLAGS) -o $@ -c $<
+CXX_TO_O_STR = $(subst ','\'',$(CXX_TO_O)) #'# fix syntax highlight
+CXX_TO_O_DISP = $(if $(V),"$(CXX_TO_O_STR)","  CXX $(@)")
 endif
 C_TO_O_CMD = 'cmd_$@ = $(C_TO_O_STR)'
 C_TO_O_DO = @set -e; \
@@ -91,6 +99,13 @@ C_TO_O_DO = @set -e; \
echo $(C_TO_O_CMD) > $(call obj2cmd,$(@)) && \
sed 's,'$@':,dep_'$@' =,' $(call obj2dep,$(@)).tmp > $(call 
obj2dep,$(@)) && \
rm -f $(call obj2dep,$(@)).tmp
+CXX_TO_O_CMD = 'cmd_$@ = $(CXX_TO_O_STR)'
+CXX_TO_O_DO = @set -e; \
+   echo $(CXX_TO_O_DISP); \
+   $(CXX_TO_O) && \
+   echo $(CXX_TO_O_CMD) > $(call obj2cmd,$(@)) && \
+   sed 's,'$@':,dep_'$@' =,' $(call obj2dep,$(@)).tmp > $(call 
obj2dep,$(@)) && \
+   rm -f $(call obj2dep,$(@)).tmp

 # return an empty string if string are equal
 compare = $(strip $(subst $(1),,$(2)) $(subst $(2),,$(1)))
@@ -136,6 +151,26 @@ boolean = $(if $1,1,0)
$(depfile_missing),\
$(depfile_newer)),\
$(C_TO_O_DO))
+#
+# Compile .cc file if needed
+# Note: dep_$$@ is from the .d file and DEP_$$@ can be specified by
+# user (by default it is empty)
+#
+.SECONDEXPANSION:
+%.o: %.cc $$(wildcard $$(dep_$$@)) $$(DEP_$$(@)) FORCE
+   @[ -d $(dir $@) ] || mkdir -p $(dir $@)
+   $(if $(D),\
+   @echo -n "$< -> $@ " ; \
+   echo -n "file_missing=$(call boolean,$(file_missing)) " ; \
+   echo -n "cmdline_changed=$(call boolean,$(call 
cmdline_changed,$(CXX_TO_O))) " ; \
+   echo -n "depfile_missing=$(call boolean,$(depfile_missing)) " ; 
\
+   echo "depfile_newer=$(call boolean,$(depfile_newer))")
+   $(if $(or \
+   $(file_missing),\
+   $(call cmdline_changed,$(CXX_TO_O)),\
+   $(depfile_missing),\
+   $(depfile_newer)),\
+   $(CXX_TO_O_DO))

 # command to assemble a .S file to generate an object
 ifeq ($(USE_HOST),1)
diff --git a/mk/target/generic/rte.vars.mk b/mk/target/generic/rte.vars.mk
index 53650c3..47d845b 100644
--- a/mk/target/generic/rte.vars.mk
+++ b/mk/target/generic/rte.vars.mk
@@ -146,7 +146,11 @@ endif
 LDFLAGS += -L$(RTE_SDK_BIN)/lib
 endif

+# copy CFLAGS to CXXFLAGS
+CXXFLAGS := $(CFLAGS)
+
 export CFLAGS
 export LDFLAGS
+export CXXFLAGS

 endif
diff --git a/mk/toolchain/gcc/rte.vars.mk b/mk/toolchain/gcc/rte.vars.mk
index 88f235c..4bdf2eb 100644
--- a/mk/toolchain/gcc/rte.vars.mk
+++ b/mk/

[dpdk-dev] [RFC PATCHv3 2/7] eal: Add extern C on eal_hugepages.h

2015-04-07 Thread Takuya ASADA

This is required to link with OSv EAL.

Signed-off-by: Takuya ASADA 
---
 lib/librte_eal/common/eal_hugepages.h | 8 
 1 file changed, 8 insertions(+)

diff --git a/lib/librte_eal/common/eal_hugepages.h 
b/lib/librte_eal/common/eal_hugepages.h
index 38edac0..b722aee 100644
--- a/lib/librte_eal/common/eal_hugepages.h
+++ b/lib/librte_eal/common/eal_hugepages.h
@@ -34,6 +34,10 @@
 #ifndef EAL_HUGEPAGES_H
 #define EAL_HUGEPAGES_H

+#ifdef __cplusplus
+extern "C" {
+#endif
+
 #include 
 #include 
 #include 
@@ -64,4 +68,8 @@ struct hugepage_file {
  */
 int eal_hugepage_info_init(void);

+#ifdef __cplusplus
+}
+#endif
+
 #endif /* EAL_HUGEPAGES_H */
-- 
2.1.0

[dpdk-dev] [RFC PATCHv3 3/7] eal: Add extern C on eal_thread.h

2015-04-07 Thread Takuya ASADA

This is required to link with OSv EAL.

Signed-off-by: Takuya ASADA 
---
 lib/librte_eal/common/eal_thread.h | 8 
 1 file changed, 8 insertions(+)

diff --git a/lib/librte_eal/common/eal_thread.h 
b/lib/librte_eal/common/eal_thread.h
index e4e76b9..794137f 100644
--- a/lib/librte_eal/common/eal_thread.h
+++ b/lib/librte_eal/common/eal_thread.h
@@ -34,6 +34,10 @@
 #ifndef EAL_THREAD_H
 #define EAL_THREAD_H

+#ifdef __cplusplus
+extern "C" {
+#endif
+
 #include 

 /**
@@ -97,4 +101,8 @@ int eal_cpuset_socket_id(rte_cpuset_t *cpusetp);
 int
 eal_thread_dump_affinity(char *str, unsigned size);

+#ifdef __cplusplus
+}
+#endif
+
 #endif /* EAL_THREAD_H */
-- 
2.1.0

[dpdk-dev] [RFC PATCHv3 4/7] eal: Add extern C on eal_private.h

2015-04-07 Thread Takuya ASADA

This is required to link with OSv EAL.

Signed-off-by: Takuya ASADA 
---
 lib/librte_eal/common/eal_private.h | 8 
 1 file changed, 8 insertions(+)

diff --git a/lib/librte_eal/common/eal_private.h 
b/lib/librte_eal/common/eal_private.h
index 4acf5a0..80b3d44 100644
--- a/lib/librte_eal/common/eal_private.h
+++ b/lib/librte_eal/common/eal_private.h
@@ -34,6 +34,10 @@
 #ifndef _EAL_PRIVATE_H_
 #define _EAL_PRIVATE_H_

+#ifdef __cplusplus
+extern "C" {
+#endif
+
 #include 

 /**
@@ -232,4 +236,8 @@ int rte_eal_dev_init(void);
  */
 int rte_eal_check_module(const char *module_name);

+#ifdef __cplusplus
+}
+#endif
+
 #endif /* _EAL_PRIVATE_H_ */
-- 
2.1.0

[dpdk-dev] [RFC PATCHv3 5/7] add OSv support

2015-04-07 Thread Takuya ASADA

Adding OSv support.
Based on Linux/FreeBSD EAL, but calling OSv kernel APIs to access devices, 
allocate contiguous memory, etc.

Signed-off-by: Takuya ASADA 
---
 config/{common_linuxapp => common_osvapp}  |  29 +-
 ...xapp-gcc => defconfig_x86_64-native-osvapp-gcc} |   2 +-
 lib/librte_eal/Makefile|   2 +
 Makefile => lib/librte_eal/osvapp/Makefile |   5 +-
 lib/librte_eal/osvapp/capstan/Capstanfile  |   6 +
 lib/librte_eal/osvapp/capstan/GET  |  22 ++
 lib/librte_eal/osvapp/eal/Makefile | 115 
 lib/librte_eal/{linuxapp => osvapp}/eal/eal.c  | 123 +---
 .../{linuxapp => osvapp}/eal/eal_alarm.c   |   0
 .../{linuxapp => osvapp}/eal/eal_debug.c   |   0
 lib/librte_eal/osvapp/eal/eal_hugepage_info.cc |  63 +
 .../{bsdapp => osvapp}/eal/eal_interrupts.c|   0
 .../eal/eal_lcore.c => osvapp/eal/eal_lcore.cc}|  53 ++--
 lib/librte_eal/{bsdapp => osvapp}/eal/eal_log.c|   0
 lib/librte_eal/osvapp/eal/eal_memory.cc| 148 ++
 lib/librte_eal/osvapp/eal/eal_pci.cc   | 312 +
 .../{linuxapp => osvapp}/eal/eal_thread.c  |   0
 lib/librte_eal/osvapp/eal/eal_timer.c  | 130 +
 .../eal/include/exec-env/rte_interrupts.h  |   0
 .../{bsdapp => osvapp}/eal/rte_eal_version.map |   6 -
 mk/exec-env/{linuxapp => osvapp}/rte.app.mk|   0
 mk/exec-env/{linuxapp => osvapp}/rte.vars.mk   |   6 +-
 22 files changed, 847 insertions(+), 175 deletions(-)
 copy config/{common_linuxapp => common_osvapp} (95%)
 copy config/{defconfig_x86_64-native-linuxapp-gcc => 
defconfig_x86_64-native-osvapp-gcc} (98%)
 copy Makefile => lib/librte_eal/osvapp/Makefile (93%)
 create mode 100644 lib/librte_eal/osvapp/capstan/Capstanfile
 create mode 100755 lib/librte_eal/osvapp/capstan/GET
 create mode 100644 lib/librte_eal/osvapp/eal/Makefile
 copy lib/librte_eal/{linuxapp => osvapp}/eal/eal.c (87%)
 copy lib/librte_eal/{linuxapp => osvapp}/eal/eal_alarm.c (100%)
 copy lib/librte_eal/{linuxapp => osvapp}/eal/eal_debug.c (100%)
 create mode 100644 lib/librte_eal/osvapp/eal/eal_hugepage_info.cc
 copy lib/librte_eal/{bsdapp => osvapp}/eal/eal_interrupts.c (100%)
 copy lib/librte_eal/{bsdapp/eal/eal_lcore.c => osvapp/eal/eal_lcore.cc} (80%)
 copy lib/librte_eal/{bsdapp => osvapp}/eal/eal_log.c (100%)
 create mode 100644 lib/librte_eal/osvapp/eal/eal_memory.cc
 create mode 100644 lib/librte_eal/osvapp/eal/eal_pci.cc
 copy lib/librte_eal/{linuxapp => osvapp}/eal/eal_thread.c (100%)
 create mode 100644 lib/librte_eal/osvapp/eal/eal_timer.c
 copy lib/librte_eal/{bsdapp => osvapp}/eal/include/exec-env/rte_interrupts.h 
(100%)
 copy lib/librte_eal/{bsdapp => osvapp}/eal/rte_eal_version.map (92%)
 copy mk/exec-env/{linuxapp => osvapp}/rte.app.mk (100%)
 copy mk/exec-env/{linuxapp => osvapp}/rte.vars.mk (95%)

diff --git a/config/common_linuxapp b/config/common_osvapp
similarity index 95%
copy from config/common_linuxapp
copy to config/common_osvapp
index 0b25f34..e26762c 100644
--- a/config/common_linuxapp
+++ b/config/common_osvapp
@@ -35,8 +35,8 @@
 #
 # CONFIG_RTE_EXEC_ENV can be linuxapp, bsdapp
 #
-CONFIG_RTE_EXEC_ENV="linuxapp"
-CONFIG_RTE_EXEC_ENV_LINUXAPP=y
+CONFIG_RTE_EXEC_ENV="osvapp"
+CONFIG_RTE_EXEC_ENV_OSVAPP=y

 ##
 ## machine can define specific variables or action for a specific board
@@ -76,12 +76,12 @@ CONFIG_RTE_FORCE_INTRINSICS=n
 #
 # Compile to share library
 #
-CONFIG_RTE_BUILD_SHARED_LIB=n
+CONFIG_RTE_BUILD_SHARED_LIB=y

 #
 # Combine to one single library
 #
-CONFIG_RTE_BUILD_COMBINE_LIBS=n
+CONFIG_RTE_BUILD_COMBINE_LIBS=y
 CONFIG_RTE_LIBNAME="intel_dpdk"

 #
@@ -89,7 +89,7 @@ CONFIG_RTE_LIBNAME="intel_dpdk"
 #
 CONFIG_RTE_LIBRTE_EAL=y
 CONFIG_RTE_MAX_LCORE=128
-CONFIG_RTE_MAX_NUMA_NODES=8
+CONFIG_RTE_MAX_NUMA_NODES=1
 CONFIG_RTE_MAX_MEMSEG=256
 CONFIG_RTE_MAX_MEMZONE=2560
 CONFIG_RTE_MAX_TAILQ=32
@@ -98,8 +98,8 @@ CONFIG_RTE_LOG_HISTORY=256
 CONFIG_RTE_LIBEAL_USE_HPET=n
 CONFIG_RTE_EAL_ALLOW_INV_SOCKET_ID=n
 CONFIG_RTE_EAL_ALWAYS_PANIC_ON_ERROR=n
-CONFIG_RTE_EAL_IGB_UIO=y
-CONFIG_RTE_EAL_VFIO=y
+CONFIG_RTE_EAL_IGB_UIO=n
+CONFIG_RTE_EAL_VFIO=n

 #
 # Special configurations in PCI Config Space for high performance
@@ -111,7 +111,13 @@ CONFIG_RTE_PCI_MAX_READ_REQUEST_SIZE=0
 #
 # Compile Environment Abstraction Layer for linux
 #
-CONFIG_RTE_LIBRTE_EAL_LINUXAPP=y
+CONFIG_RTE_LIBRTE_EAL_OSVAPP=y
+
+#
+# Compile OSv specific parameters
+#
+CONFIG_RTE_CONTIGUOUS_CHUNK_SIZE=32M
+CONFIG_RTE_DEFAULT_NUM_CHUNKS=4

 #
 # Compile Environment Abstraction Layer to support hotplug
@@ -235,7 +241,7 @@ CONFIG_RTE_LIBRTE_VMXNET3_DEBUG_DRIVER=n
 #
 # Compile example software rings based PMD
 #
-CONFIG_RTE_LIBRTE_PMD_RING=y
+CONFIG_RTE_LIBRTE_PMD_RING=n
 CONFIG_RTE_PMD_RING_MAX_RX_RINGS=16
 CONFIG_RTE_PMD_RING_MAX_TX_RINGS=16

@@ -247,7 +253,7 @@ CONFIG_RTE_LIBRTE_PMD_PCAP=n
 #
 # Compile link bonding PMD library
 #
-CONFIG

[dpdk-dev] [RFC PATCHv3 6/7] app/test: support OSv

2015-04-07 Thread Takuya ASADA

Add support OSv EAL.

Signed-off-by: Takuya ASADA 
---
 app/test/test_eal_flags.c  | 34 +-
 app/test/test_timer_perf.c |  2 +-
 2 files changed, 18 insertions(+), 18 deletions(-)

diff --git a/app/test/test_eal_flags.c b/app/test/test_eal_flags.c
index 0352f87..40a5c7e 100644
--- a/app/test/test_eal_flags.c
+++ b/app/test/test_eal_flags.c
@@ -287,7 +287,7 @@ static int
 test_whitelist_flag(void)
 {
unsigned i;
-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
/* BSD target doesn't support prefixes at this point */
const char * prefix = "";
 #else
@@ -353,7 +353,7 @@ test_whitelist_flag(void)
 static int
 test_invalid_b_flag(void)
 {
-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
/* BSD target doesn't support prefixes at this point */
const char * prefix = "";
 #else
@@ -400,7 +400,7 @@ test_invalid_b_flag(void)
 static int
 test_invalid_vdev_flag(void)
 {
-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
/* BSD target doesn't support prefixes at this point, and we also need 
to
 * run another primary process here */
const char * prefix = no_shconf;
@@ -454,7 +454,7 @@ test_invalid_vdev_flag(void)
 static int
 test_invalid_r_flag(void)
 {
-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
/* BSD target doesn't support prefixes at this point */
const char * prefix = "";
 #else
@@ -498,7 +498,7 @@ test_invalid_r_flag(void)
 static int
 test_missing_c_flag(void)
 {
-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
/* BSD target doesn't support prefixes at this point */
const char * prefix = "";
 #else
@@ -629,7 +629,7 @@ test_missing_c_flag(void)
 static int
 test_master_lcore_flag(void)
 {
-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
/* BSD target doesn't support prefixes at this point */
const char *prefix = "";
 #else
@@ -677,7 +677,7 @@ test_master_lcore_flag(void)
 static int
 test_missing_n_flag(void)
 {
-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
/* BSD target doesn't support prefixes at this point */
const char * prefix = "";
 #else
@@ -722,7 +722,7 @@ test_no_hpet_flag(void)
 {
char prefix[PATH_MAX], tmp[PATH_MAX];

-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
return 0;
 #endif
if (get_current_prefix(tmp, sizeof(tmp)) == NULL) {
@@ -754,7 +754,7 @@ test_no_hpet_flag(void)
 static int
 test_no_huge_flag(void)
 {
-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
/* BSD target doesn't support prefixes at this point, and we also need 
to
 * run another primary process here */
const char * prefix = no_shconf;
@@ -782,7 +782,7 @@ test_no_huge_flag(void)
printf("Error - process run ok with --no-huge and -m flags\n");
return -1;
}
-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
/* BSD target does not support NUMA, hence no --socket-mem tests */
return 0;
 #endif
@@ -870,7 +870,7 @@ static int
 test_misc_flags(void)
 {
char hugepath[PATH_MAX] = {0};
-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
/* BSD target doesn't support prefixes at this point */
const char * prefix = "";
const char * nosh_prefix = "";
@@ -942,7 +942,7 @@ test_misc_flags(void)
const char *argv6[] = {prgname, "-c", "1", "-n", "2", "-m", 
DEFAULT_MEM_SIZE,
no_shconf, nosh_prefix };

-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
return 0;
 #endif
/* With --huge-dir */
@@ -1007,7 +1007,7 @@ test_misc_flags(void)
printf("Error - process did not run ok with --no-shconf 
flag\n");
return -1;
}
-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
return 0;
 #endif
if (launch_proc(argv7) != 0) {
@@ -1068,7 +1068,7 @@ test_file_prefix(void)
 * 7. check that only memtest2 hugefiles are present in the hugedir
 */

-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
return 0;
 #endif

@@ -1175,7 +1175,7 @@ test_file_prefix(void)
 static int
 test_memory_flags(void)
 {
-#ifdef RTE_EXEC_ENV_BSDAPP
+#if defined(RTE_EXEC_ENV_BSDAPP) || defined(RTE_EXEC_ENV_OSVAPP)
/* BSD target doesn't support prefixes at this point */
const char * prefix = "";
 #else
@@ -1228,7 +1228,7 @@ test_memory_flags(void)

[dpdk-dev] [RFC PATCHv3 7/7] doc: Add Getting Started Guide for OSv

2015-04-07 Thread Takuya ASADA

Described how to build DPDK for OSv, using "Capstan".

Signed-off-by: Takuya ASADA 
---
 doc/guides/index.rst  |   1 +
 doc/guides/osv_gsg/build_dpdk.rst | 282 ++
 doc/guides/osv_gsg/build_sample_apps.rst  | 123 +++
 doc/guides/{freebsd_gsg => osv_gsg}/index.rst |   3 +-
 doc/guides/{freebsd_gsg => osv_gsg}/intro.rst |  15 +-
 5 files changed, 410 insertions(+), 14 deletions(-)
 create mode 100644 doc/guides/osv_gsg/build_dpdk.rst
 create mode 100644 doc/guides/osv_gsg/build_sample_apps.rst
 copy doc/guides/{freebsd_gsg => osv_gsg}/index.rst (96%)
 copy doc/guides/{freebsd_gsg => osv_gsg}/intro.rst (84%)

diff --git a/doc/guides/index.rst b/doc/guides/index.rst
index 44e8432..3b45ca3 100644
--- a/doc/guides/index.rst
+++ b/doc/guides/index.rst
@@ -40,6 +40,7 @@ Contents:
linux_gsg/index
freebsd_gsg/index
xen/index
+   osv_gsg/index
prog_guide/index
nics/index
sample_app_ug/index
diff --git a/doc/guides/osv_gsg/build_dpdk.rst 
b/doc/guides/osv_gsg/build_dpdk.rst
new file mode 100644
index 000..4eadb68
--- /dev/null
+++ b/doc/guides/osv_gsg/build_dpdk.rst
@@ -0,0 +1,282 @@
+..  BSD LICENSE
+Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions
+are met:
+
+* Redistributions of source code must retain the above copyright
+notice, this list of conditions and the following disclaimer.
+* Redistributions in binary form must reproduce the above copyright
+notice, this list of conditions and the following disclaimer in
+the documentation and/or other materials provided with the
+distribution.
+* Neither the name of Intel Corporation nor the names of its
+contributors may be used to endorse or promote products derived
+from this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+.. _building_from_source:
+
+Compiling the DPDK Target from Source
+=
+
+System Requirements
+---
+
+To building DPDK for OSv, you will need to use Linux/x86_64 with g++ 4.8 or 
later.
+
+Install the DPDK and Browse Sources
+---
+
+First, uncompress the archive and move to the DPDK source directory:
+
+.. code-block:: console
+
+[user at host ~]$ unzip DPDK-zip
+[user at host ~]$ cd DPDK-
+[user at host DPDK]$ ls
+app/ config/ examples/ lib/ LICENSE.GPL LICENSE.LGPL Makefile mk/ scripts/ 
tools/
+
+The DPDK is composed of several directories:
+
+*   lib: Source code of DPDK libraries
+
+*   app: Source code of DPDK applications (automatic tests)
+
+*   examples: Source code of DPDK applications
+
+*   config, tools, scripts, mk: Framework-related makefiles, scripts and 
configuration
+
+Install Capstan
+
+
+Before start building VM image, you need to install Capstan*:
+
+`http://osv.io/capstan/`
+
+.. code-block:: console
+
+[user at host ~]$ curl 
https://raw.githubusercontent.com/cloudius-systems/capstan/master/scripts/download
 | bash
+
+Build DPDK for OSv VM image
+
+
+Build VM image using Capstan:
+
+.. code-block:: console
+
+[user at host ~]$ cd DPDK-/lib/librte_eal/osvapp/capstan/
+[user at host capstan]$ capstan build osv-dpdk
+Building osv-dpdk...
+Downloading cloudius/osv-base/index.yaml...
+145 B / 145 B [==] 
100.00 %
+Downloading cloudius/osv-base/osv-base.qemu.gz...
+20.09 MB / 20.09 MB [] 
100.00 %
+Uploading files...
+10 / 10 [] 
100.00 %
+
+Run DPDK for OSv VM image
+
+
+Run VM image using Capstan:
+
+.. code-block:: console
+
+[user at host ~]$ capstan run osv-dpdk
+Created instance: osv-dpdk
+OSv v0.19
+eth0: 192.168.122.15
+EAL: Detected lcore 0 as core 0 on

[dpdk-dev] [RFC PATCHv2 5/8] add OSv support

2015-04-07 Thread Takuya ASADA

On Tue, Mar 24, 2015 at 10:52 PM, Neil Horman  wrote:

> On Sat, Mar 21, 2015 at 12:23:02PM +0900, Takuya ASADA wrote:
> > Adding OSv support.
> > Based on Linux/FreeBSD EAL, but calling OSv kernel APIs to access
> devices, allocate contiguous memory, etc.
> >
> > Signed-off-by: Takuya ASADA 
> > ---
> > diff --git a/lib/librte_eal/bsdapp/eal/include/exec-env/rte_interrupts.h
> b/lib/librte_eal/osvapp/eal/include/exec-env/rte_interrupts.h
> > similarity index 100%
> > copy from lib/librte_eal/bsdapp/eal/include/exec-env/rte_interrupts.h
> > copy to lib/librte_eal/osvapp/eal/include/exec-env/rte_interrupts.h
> > diff --git a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> b/lib/librte_eal/osvapp/eal/rte_eal_version.map
> > similarity index 92%
> > copy from lib/librte_eal/bsdapp/eal/rte_eal_version.map
> > copy to lib/librte_eal/osvapp/eal/rte_eal_version.map
> > index 67b6a6c..ebb584a 100644
> > --- a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> > +++ b/lib/librte_eal/osvapp/eal/rte_eal_version.map
> > @@ -31,24 +31,26 @@ DPDK_2.0 {
> >   rte_eal_get_physmem_layout;
> >   rte_eal_get_physmem_size;
> >   rte_eal_has_hugepages;
> > - rte_eal_hpet_init;
> >   rte_eal_init;
> >   rte_eal_iopl_init;
> >   rte_eal_lcore_role;
> >   rte_eal_mp_remote_launch;
> >   rte_eal_mp_wait_lcore;
> > + rte_eal_parse_devargs_str;
> > + rte_eal_pci_close_one;
> >   rte_eal_pci_dump;
> >   rte_eal_pci_probe;
> > + rte_eal_pci_probe_one;
> >   rte_eal_pci_register;
> >   rte_eal_pci_unregister;
> >   rte_eal_process_type;
> >   rte_eal_remote_launch;
> >   rte_eal_tailq_lookup;
> >   rte_eal_tailq_register;
> > + rte_eal_vdev_init;
> > + rte_eal_vdev_uninit;
> >   rte_eal_wait_lcore;
> >   rte_exit;
> > - rte_get_hpet_cycles;
> > - rte_get_hpet_hz;
> >   rte_get_tsc_hz;
> >   rte_hexdump;
> >   rte_intr_callback_register;
> > @@ -86,9 +88,6 @@ DPDK_2.0 {
> >   rte_thread_get_affinity;
> >   rte_thread_set_affinity;
> >   rte_vlog;
> > - rte_xen_dom0_memory_attach;
> > - rte_xen_dom0_memory_init;
> > - test_mp_secondary;
> >
> I understand removing symbols that you don't support, but why are you
> adding the
> vdev, pci and parsing symbols?  I don't see you using them anywhere in your
> code...
>

This was because I copied rte_eal_version.map from linuxapp, but git
format-patch generated diff from bsdapp.
But osvapp don't have these symbols, so I dropped these lines.
Please check v3 patchset which I just posted.


> Neil
>
>

[dpdk-dev] CROSS_COMPILATION with custom toolchain

2015-04-07 Thread Shyam Sundar Govindaraj

Hi

  I am trying to build DPDK-2.0 with custom toolchain (binutils 2.24 , gcc 
4.8.2 , glibc 2.19, linux-3.13-header). With CROSS variable properly set , even 
though DPDK is picking the right gcc from toolchain ,it ends up with below 
error since the libraries are taken from the build system and not from 
toolchain.

/lib64/libc.so.6: version `GLIBC_2.14' not found (required by 
toolchain/x86_64-unknown-linux-gnu/bin/x86_64-unknown-linux-gnu-gcc)

Note: Build system has glibc 2.12

Can you please answer below questions and suggest me if there is any other 
possible way to proceed from here


1)  Is there any variable like (CROSS) available to point to the right 
library?OR

2)  Is it possible to make use of sysroot available in the toolchain. If 
yes, how?

Thanks
Shyam

[dpdk-dev] [RFC] Cuckoo hash for DPDK 2.1

2015-04-07 Thread yangguangje...@hotmail.com


hi Pablo,
rte_hash uses Jenkins hash (http://burtleburtle.net/bob/hash/ ) in older 
dpdk veriosn,which is originated lookup2.c in 1996.Bob Jenkins updates his hash 
function named lookup3.c in 2006. The hash function is more faster than 
lookup2.c. 
   why not continue to adopt the new hash function lookup3.c ?






At 2015-04-04 04:28:06, "De Lara Guarch, Pablo"  wrote:
>Hi all,
>
>This RFC is to describe a proposed replacement for the existing rte_hash 
>implementation,
>using the cuckoo hash scheme (see 
>http://www.cs.cmu.edu/~dongz/papers/cuckooswitch.pdf),
>which should provide better performance, in terms of lookup time, as well as a 
>higher load factor.
>
>This new implementation also shall offer several improvements compared to the 
>existing one, such as:
>
>-Data return: existing implementation returns an index to the bucket 
>where the key is stored,
>
>whereas the new implementation shall return 8-byte integers or pointers to 
>external data.
>
>
>
>-Increased key length: key length shall be increased more than the 
>existing 64 bytes,
>
>without having a big penalty on performance
>
>
>
>-Increased burst size: current implementation only allows 16 lookups 
>at the same time,
>
>whereas the new implementation shall allow more than that (probably 64)
>
>
>
>-Opening addressing: current implementation does not allow new keys to 
>be added
>
>if its target bucket is full, whereas with Cuckoo hash, it offers an 
>alternative location to store the key.
>
>I am currently working on the implementation, considering several options:
>
>
>-Using a single table to store all the signatures, regardless they 
>have used their primary or secondary hash function.
>
>
>
>-Using two tables to store the signatures, one for primary hashes and 
>another for the secondary hashes.
>
>
>I need to do some testing on both implementations to know which one is more 
>suitable for DPDK.
>
>Any comments/ideas welcome.
>
>Thanks
>Pablo
>



yangguangjerry at hotmail.com

[dpdk-dev] Running DPDK Binaries on a different Target

2015-04-07 Thread Venkat Thummala

Hi,

I have built a DPDK application [based on version 2.0] and run on the
native machine successfully.

I have tried running the binary on a different machine, but it resulted in
a CRASH with the following back trace.

Please find the CPU info of the machines from the attachment.

cpuinfo-1 - Native Machine
cpuinfo-2 - Non Native Machine

Could someone please help me in understanding the issue here and making it
work?

Regards
Venkat

Program terminated with signal SIGILL, Illegal instruction.
#0  0x004209f2 in rte_cpu_get_flag_enabled (feature=2147483656,
feature at entry=RTE_CPUFLAG_EM64T)
at
/home/vthummala/src/vwlc/dpdk/dpdk-2.0.0/x86_64-native-linuxapp-gcc/include/rte_cpuflags.h:303
303return (regs[feat->reg] >> feat->bit) & 1;
(gdb) bt
#0  0x004209f2 in rte_cpu_get_flag_enabled (feature=2147483656,
feature at entry=RTE_CPUFLAG_EM64T)
at
/home/vthummala/src/vwlc/dpdk/dpdk-2.0.0/x86_64-native-linuxapp-gcc/include/rte_cpuflags.h:303
#1  0x00420a1b in rte_hash_crc_set_alg (alg=6 '\006')
at
/home/vthummala/src/vwlc/dpdk/dpdk-2.0.0/x86_64-native-linuxapp-gcc/include/rte_hash_crc.h:429
#2  rte_hash_crc_init_alg () at
/home/vthummala/src/vwlc/dpdk/dpdk-2.0.0/x86_64-native-linuxapp-gcc/include/rte_hash_crc.h:445
#3  0x0054968d in __libc_csu_init ()
#4  0x7fc3ca474e55 in group_nodes_into_DFAstates
(dests_ch=0x940f507ab10ff0c8, dests_node=0x4a8b44de74c084c0,
state=0x4420528b48a8ebc9,
dfa=) at regexec.c:3614
#5  build_trtable (dfa=0x840fc139c0014468, state=0x4420528b48a8ebc9) at
regexec.c:3354
#6  0x41d589495541f689 in ?? ()
#7  0x00251630258d4c54 in ?? ()
#8  0x002517302d8d4855 in ?? ()
#9  0xc148db31e5294c53 in ?? ()
#10 0x65e808ec834803fd in ?? ()
#11 0x1e74ed8548ffed48 in ?? ()
#12 0x00841f0f in ?? ()

[dpdk-dev] Running DPDK Binaries on a different Target

2015-04-07 Thread Venkat Thummala

Attaching the CPU Info.

On 7 April 2015 at 17:28, Venkat Thummala 
wrote:

> Hi,
>
> I have built a DPDK application [based on version 2.0] and run on the
> native machine successfully.
>
> I have tried running the binary on a different machine, but it resulted in
> a CRASH with the following back trace.
>
> Please find the CPU info of the machines from the attachment.
>
> cpuinfo-1 - Native Machine
> cpuinfo-2 - Non Native Machine
>
> Could someone please help me in understanding the issue here and making it
> work?
>
> Regards
> Venkat
>
> Program terminated with signal SIGILL, Illegal instruction.
> #0  0x004209f2 in rte_cpu_get_flag_enabled (feature=2147483656,
> feature at entry=RTE_CPUFLAG_EM64T)
> at
> /home/vthummala/src/vwlc/dpdk/dpdk-2.0.0/x86_64-native-linuxapp-gcc/include/rte_cpuflags.h:303
> 303return (regs[feat->reg] >> feat->bit) & 1;
> (gdb) bt
> #0  0x004209f2 in rte_cpu_get_flag_enabled (feature=2147483656,
> feature at entry=RTE_CPUFLAG_EM64T)
> at
> /home/vthummala/src/vwlc/dpdk/dpdk-2.0.0/x86_64-native-linuxapp-gcc/include/rte_cpuflags.h:303
> #1  0x00420a1b in rte_hash_crc_set_alg (alg=6 '\006')
> at
> /home/vthummala/src/vwlc/dpdk/dpdk-2.0.0/x86_64-native-linuxapp-gcc/include/rte_hash_crc.h:429
> #2  rte_hash_crc_init_alg () at
> /home/vthummala/src/vwlc/dpdk/dpdk-2.0.0/x86_64-native-linuxapp-gcc/include/rte_hash_crc.h:445
> #3  0x0054968d in __libc_csu_init ()
> #4  0x7fc3ca474e55 in group_nodes_into_DFAstates
> (dests_ch=0x940f507ab10ff0c8, dests_node=0x4a8b44de74c084c0,
> state=0x4420528b48a8ebc9,
> dfa=) at regexec.c:3614
> #5  build_trtable (dfa=0x840fc139c0014468, state=0x4420528b48a8ebc9) at
> regexec.c:3354
> #6  0x41d589495541f689 in ?? ()
> #7  0x00251630258d4c54 in ?? ()
> #8  0x002517302d8d4855 in ?? ()
> #9  0xc148db31e5294c53 in ?? ()
> #10 0x65e808ec834803fd in ?? ()
> #11 0x1e74ed8548ffed48 in ?? ()
> #12 0x00841f0f in ?? ()
>

[dpdk-dev] [PATCH v3 1/5] mbuf: fix clone support when application uses private mbuf data

2015-04-07 Thread Ananyev, Konstantin

Hi Olivier,

> -Original Message-
> From: Olivier MATZ [mailto:olivier.matz at 6wind.com]
> Sent: Monday, April 06, 2015 10:50 PM
> To: Ananyev, Konstantin; dev at dpdk.org
> Cc: zoltan.kiss at linaro.org; Richardson, Bruce
> Subject: Re: [PATCH v3 1/5] mbuf: fix clone support when application uses 
> private mbuf data
> 
> Hi Konstantin,
> 
> Thanks for your comments.
> 
> On 04/02/2015 07:21 PM, Ananyev, Konstantin wrote:
> > Hi Olivier,
> >
> >> -Original Message-
> >> From: Olivier Matz [mailto:olivier.matz at 6wind.com]
> >> Sent: Tuesday, March 31, 2015 8:23 PM
> >> To: dev at dpdk.org
> >> Cc: Ananyev, Konstantin; zoltan.kiss at linaro.org; Richardson, Bruce; 
> >> Olivier Matz
> >> Subject: [PATCH v3 1/5] mbuf: fix clone support when application uses 
> >> private mbuf data
> >>
> >> From: Olivier Matz 
> >>
> >> Add a new private_size field in mbuf structure that should
> >> be initialized at mbuf pool creation. This field contains the
> >> size of the application private data in mbufs.
> >>
> >> Introduce new static inline functions rte_mbuf_from_indirect()
> >> and rte_mbuf_to_baddr() to replace the existing macros, which
> >> take the private size in account when attaching and detaching
> >> mbufs.
> >>
> >> Signed-off-by: Olivier Matz 
> >> ---
> >>  app/test-pmd/testpmd.c |  1 +
> >>  examples/vhost/main.c  |  4 +--
> >>  lib/librte_mbuf/rte_mbuf.c |  1 +
> >>  lib/librte_mbuf/rte_mbuf.h | 77 
> >> +++---
> >>  4 files changed, 63 insertions(+), 20 deletions(-)
> >>
> >> diff --git a/app/test-pmd/testpmd.c b/app/test-pmd/testpmd.c
> >> index 3057791..c5a195a 100644
> >> --- a/app/test-pmd/testpmd.c
> >> +++ b/app/test-pmd/testpmd.c
> >> @@ -425,6 +425,7 @@ testpmd_mbuf_ctor(struct rte_mempool *mp,
> >>mb->tx_offload   = 0;
> >>mb->vlan_tci = 0;
> >>mb->hash.rss = 0;
> >> +  mb->priv_size= 0;
> >>  }
> >>
> >>  static void
> >> diff --git a/examples/vhost/main.c b/examples/vhost/main.c
> >> index c3fcb80..e44e82f 100644
> >> --- a/examples/vhost/main.c
> >> +++ b/examples/vhost/main.c
> >> @@ -139,7 +139,7 @@
> >>  /* Number of descriptors per cacheline. */
> >>  #define DESC_PER_CACHELINE (RTE_CACHE_LINE_SIZE / sizeof(struct 
> >> vring_desc))
> >>
> >> -#define MBUF_EXT_MEM(mb)   (RTE_MBUF_FROM_BADDR((mb)->buf_addr) != (mb))
> >> +#define MBUF_EXT_MEM(mb)   (rte_mbuf_from_indirect(mb) != (mb))
> >>
> >>  /* mask of enabled ports */
> >>  static uint32_t enabled_port_mask = 0;
> >> @@ -1550,7 +1550,7 @@ attach_rxmbuf_zcp(struct virtio_net *dev)
> >>  static inline void pktmbuf_detach_zcp(struct rte_mbuf *m)
> >>  {
> >>const struct rte_mempool *mp = m->pool;
> >> -  void *buf = RTE_MBUF_TO_BADDR(m);
> >> +  void *buf = rte_mbuf_to_baddr(m);
> >>uint32_t buf_ofs;
> >>uint32_t buf_len = mp->elt_size - sizeof(*m);
> >>m->buf_physaddr = rte_mempool_virt2phy(mp, m) + sizeof(*m);
> >> diff --git a/lib/librte_mbuf/rte_mbuf.c b/lib/librte_mbuf/rte_mbuf.c
> >> index 526b18d..e095999 100644
> >> --- a/lib/librte_mbuf/rte_mbuf.c
> >> +++ b/lib/librte_mbuf/rte_mbuf.c
> >> @@ -125,6 +125,7 @@ rte_pktmbuf_init(struct rte_mempool *mp,
> >>m->pool = mp;
> >>m->nb_segs = 1;
> >>m->port = 0xff;
> >> +  m->priv_size = 0;
> >
> > Why it is 0?
> > Shouldn't it be the same calulations as in detach() below:
> > m->priv_size = /*get private size from mempool private*/;
> > m->buf_addr = (char *)m + sizeof(struct rte_mbuf) + m->priv_size;
> > m->buf_len = mp->elt_size - sizeof(struct rte_mbuf) - m->priv_size;
> > ?
> 
> It's 0 because we also have in the function (not visible in the
> patch):
> 
>   m->buf_addr = (char *)m + sizeof(struct rte_mbuf);

Yep, that's why as I wrote above, I think we need to setup here all 3 fields:
priv_size, buf_addr, buf_len exactly in the same way as in detach().  

> 
> It means that an application that wants to use a private area has
> to provide another init function derived from this default function.

After your changes, attach/free and other functions from public mbuf API
rely on priv_size being set properly.
So I suppose 'official' pktmbuf_init() should also set it in a proper manner. 

> This was already the case before the patch series.

Before this patch series, we don't have priv_size, so we have nothing to setup.

> 
> As we discussed in previous mail, I plan to propose a rework of
> mbuf pool initialization in another series, and my initial idea was to
> change this at the same time. But on the other hand it does not hurt
> to do this change now. I'll include it in next version.

Ok.

> 
> 
> > BTW, don't see changes in rte_pktmbuf_pool_init() to setup
> > mbp_priv->mbuf_data_room_size properly.
> > Without that changes, how can people start using that feature?
> > It seems that the only way now - setup priv_size and buf_len for each mbuf 
> > manually.
> 
> It's the same reason than above. To use a private are, the user has
> to provide its own function that

[dpdk-dev] [PATCH 4/5] bond mode 4: allow external state machine

2015-04-07 Thread Pawel Wodkowski

On 2015-04-06 19:01, Eric Kinzie wrote:

Interesting patch. I will closer look at this tomorrow.

For now I have first comments:

> +static void bond_mode_8023ad_ext_periodic_cb(void *arg);
> +
>   #ifdef RTE_LIBRTE_BOND_DEBUG_8023AD
>   #define MODE4_DEBUG(fmt, ...) RTE_LOG(DEBUG, PMD, "%6u [Port %u: %s] " fmt, 
> \
>   bond_dbg_get_time_diff_ms(), slave_id, \
> @@ -1014,6 +1016,8 @@ bond_mode_8023ad_conf_get(struct rte_eth_dev *dev,
>   conf->tx_period_ms = mode4->tx_period_timeout / ms_ticks;
>   conf->update_timeout_ms = mode4->update_timeout_us / 1000;
>   conf->rx_marker_period_ms = mode4->rx_marker_timeout / ms_ticks;
> + conf->slowrx_cb = mode4->slowrx_cb;
> + conf->external_sm = mode4->external_sm;

mode4->external_sm flag realy needed? Why do not use mode4->slowrx_cb as 
external state machine indicator?

>   }
>
>   void
> @@ -1035,6 +1039,8 @@ bond_mode_8023ad_setup(struct rte_eth_dev *dev,
>   conf->tx_period_ms = BOND_8023AD_TX_MACHINE_PERIOD_MS;
>   conf->rx_marker_period_ms = BOND_8023AD_RX_MARKER_PERIOD_MS;
>   conf->update_timeout_ms = BOND_MODE_8023AX_UPDATE_TIMEOUT_MS;
> + conf->slowrx_cb = NULL;
> + conf->external_sm = 0;
>   }
>
>   mode4->fast_periodic_timeout = conf->fast_periodic_ms * ms_ticks;
> @@ -1045,6 +1051,8 @@ bond_mode_8023ad_setup(struct rte_eth_dev *dev,
>   mode4->tx_period_timeout = conf->tx_period_ms * ms_ticks;
>   mode4->rx_marker_timeout = conf->rx_marker_period_ms * ms_ticks;
>   mode4->update_timeout_us = conf->update_timeout_ms * 1000;
> + mode4->slowrx_cb = conf->slowrx_cb;
> + mode4->external_sm = conf->external_sm;
>   }
>
>   int
> @@ -1062,6 +1070,13 @@ bond_mode_8023ad_enable(struct rte_eth_dev *bond_dev)
>   int
>   bond_mode_8023ad_start(struct rte_eth_dev *bond_dev)
>   {
> + struct bond_dev_private *internals = bond_dev->data->dev_private;
> + struct mode8023ad_private *mode4 = &internals->mode4;
> +
> + if (mode4->external_sm)
> + return rte_eal_alarm_set(BOND_MODE_8023AX_UPDATE_TIMEOUT_MS * 
> 1000,
> + &bond_mode_8023ad_ext_periodic_cb, bond_dev);
> +
>   return rte_eal_alarm_set(BOND_MODE_8023AX_UPDATE_TIMEOUT_MS * 1000,
>   &bond_mode_8023ad_periodic_cb, bond_dev);
>   }
> @@ -1069,6 +1084,13 @@ bond_mode_8023ad_start(struct rte_eth_dev *bond_dev)
>   void
>   bond_mode_8023ad_stop(struct rte_eth_dev *bond_dev)
>   {
> + struct bond_dev_private *internals = bond_dev->data->dev_private;
> + struct mode8023ad_private *mode4 = &internals->mode4;
> +
> + if (mode4->external_sm) {

This is bad idea. If bond_mode_8023ad_setup will be called you might 
have two handlers running for while. You should stop mode 4 by invoking 
bond_mode_8023ad_stop() before you set mode4->external_sm and then, if 
mode 4 was running, start it again.

Also, maybe a renaming "external_sm" to "state_machine_cb", set it to 
against default one and using it without "if()" will simplify code. It 
is no crucial but will eliminate couple of if's. In 
rte_eth_bond_8023ad_ext_slowtx() you can compare it against default one.

> + rte_eal_alarm_cancel(&bond_mode_8023ad_ext_periodic_cb, 
> bond_dev);
> + return;
> + }
>   rte_eal_alarm_cancel(&bond_mode_8023ad_periodic_cb, bond_dev);
>   }
>
> @@ -1215,3 +1237,156 @@ rte_eth_bond_8023ad_slave_info(uint8_t port_id, 
> uint8_t slave_id,
>   info->agg_port_id = port->aggregator_port_id;
>   return 0;
>   }

-- 
Pawel

[dpdk-dev] [snabb-devel] Re: memory barriers in virtq.lua?

2015-04-07 Thread Luke Gorrie

Hi Michael,

I'm writing to follow up the previous discussion about memory barriers in
virtio-net device implementations, and Cc'ing the DPDK list because I
believe this is relevant to them too.

First, thanks again for getting in touch and reviewing our code.

I have now found a missed case where we *do* require a hardware memory
barrier on x86 in our vhost/virtio-net device. That is when checking the
interrupt suppression flag after updating used->idx. This is needed because
x86 can reorder the write to used->idx after the read from avail->flags,
and that causes the guest to see a stale value of used->idx after it
toggles interrupt suppression.

If I may spell out my mental model, for the sake of being corrected and/or
as an example of how third party developers are reading and interpreting
the Virtio-net spec:

Relating this to Virtio 1.0, the most relevant section is 3.2.1 (Supplying
Buffers to the Device) which calls for two "suitable memory barriers". The
spec talks about these from the driver perspective, but they are both
relevant to the device side too.

The first barrier (write to descriptor table before write to used->idx) is
implicit on x86 because writes by the same core are not reordered. This
means that no explicit hardware barrier is needed. (A compiler barrier may
be needed, however.)

The second memory barrier (write to used->idx before reading avail->flags)
is not implicit on x86 because stores are reordered after loads. So an
explicit hardware memory barrier is needed.

I hope that is a correct assessment of the situation. (Forgive my
x86centricity, I am sure that seems very foreign to kernel hackers.)

If this assessment is correct then the DPDK developers might also want to
review librte_vhost/vhost_rxtx.c and consider adding a hardware memory
barrier between writing used->idx and reading avail->flags.

Cheers,
-Luke

P.S. I notice that the Linux virtio-net driver does not seem to tolerate
spurious interrupts, even though the Virtio 1.0 spec requires this
("must"). On 3.13.11-ckt15 I see them trigger an "irq nobody cared" kernel
log message and then the irq is disabled. If that sounds suspicious I can
supply more information.

[dpdk-dev] Running DPDK Binaries on a different Target

2015-04-07 Thread Neil Horman

On Tue, Apr 07, 2015 at 05:30:15PM +0530, Venkat Thummala wrote:
> Attaching the CPU Info.
> 
> On 7 April 2015 at 17:28, Venkat Thummala 
> wrote:
> 
> > Hi,
> >
> > I have built a DPDK application [based on version 2.0] and run on the
> > native machine successfully.
> >
> > I have tried running the binary on a different machine, but it resulted in
> > a CRASH with the following back trace.
> >
> > Please find the CPU info of the machines from the attachment.
> >
> > cpuinfo-1 - Native Machine
> > cpuinfo-2 - Non Native Machine
> >
> > Could someone please help me in understanding the issue here and making it
> > work?
> >
> > Regards
> > Venkat
> >
> > Program terminated with signal SIGILL, Illegal instruction.
> > #0  0x004209f2 in rte_cpu_get_flag_enabled (feature=2147483656,
> > feature at entry=RTE_CPUFLAG_EM64T)
> > at
> > /home/vthummala/src/vwlc/dpdk/dpdk-2.0.0/x86_64-native-linuxapp-gcc/include/rte_cpuflags.h:303
> > 303return (regs[feat->reg] >> feat->bit) & 1;

Looks like this somehow compiled into an instruction that the native system
understands but the system you're running on doesn't.  disassemble the code here
to see what instruction that is to confirm, then you either need to:

1) Change the machine you're compiling for to be more specific to your target
system

or

2) Understand that the hardware you're targeting doesn't support dpdk

Neil

> > (gdb) bt
> > #0  0x004209f2 in rte_cpu_get_flag_enabled (feature=2147483656,
> > feature at entry=RTE_CPUFLAG_EM64T)
> > at
> > /home/vthummala/src/vwlc/dpdk/dpdk-2.0.0/x86_64-native-linuxapp-gcc/include/rte_cpuflags.h:303
> > #1  0x00420a1b in rte_hash_crc_set_alg (alg=6 '\006')
> > at
> > /home/vthummala/src/vwlc/dpdk/dpdk-2.0.0/x86_64-native-linuxapp-gcc/include/rte_hash_crc.h:429
> > #2  rte_hash_crc_init_alg () at
> > /home/vthummala/src/vwlc/dpdk/dpdk-2.0.0/x86_64-native-linuxapp-gcc/include/rte_hash_crc.h:445
> > #3  0x0054968d in __libc_csu_init ()
> > #4  0x7fc3ca474e55 in group_nodes_into_DFAstates
> > (dests_ch=0x940f507ab10ff0c8, dests_node=0x4a8b44de74c084c0,
> > state=0x4420528b48a8ebc9,
> > dfa=) at regexec.c:3614
> > #5  build_trtable (dfa=0x840fc139c0014468, state=0x4420528b48a8ebc9) at
> > regexec.c:3354
> > #6  0x41d589495541f689 in ?? ()
> > #7  0x00251630258d4c54 in ?? ()
> > #8  0x002517302d8d4855 in ?? ()
> > #9  0xc148db31e5294c53 in ?? ()
> > #10 0x65e808ec834803fd in ?? ()
> > #11 0x1e74ed8548ffed48 in ?? ()
> > #12 0x00841f0f in ?? ()
> >
>

[dpdk-dev] [PATCH 4/5] bond mode 4: allow external state machine

2015-04-07 Thread Pawel Wodkowski

On 2015-04-07 16:18, Pawel Wodkowski wrote:

>
> Also, maybe a renaming "external_sm" to "state_machine_cb", set it to
> against default one and using it without "if()" will simplify code. It
> is no crucial but will eliminate couple of if's. In
> rte_eth_bond_8023ad_ext_slowtx() you can compare it against default one.

Oh, I read what I wrote :)
Please ignore that.

-- 
Pawel

[dpdk-dev] [snabb-devel] Re: memory barriers in virtq.lua?

2015-04-07 Thread Michael S. Tsirkin

On Tue, Apr 07, 2015 at 04:22:42PM +0200, Luke Gorrie wrote:
> Hi Michael,
> 
> I'm writing to follow up the previous discussion about memory barriers in
> virtio-net device implementations, and Cc'ing the DPDK list because I believe
> this is relevant to them too.
> 
> First, thanks again for getting in touch and reviewing our code.
> 
> I have now found a missed case where we *do* require a hardware memory barrier
> on x86 in our vhost/virtio-net device. That is when checking the interrupt
> suppression flag after updating used->idx. This is needed because x86 can
> reorder the write to used->idx after the read from avail->flags, and that
> causes the guest to see a stale value of used->idx after it toggles interrupt
> suppression.
> 
> If I may spell out my mental model, for the sake of being corrected and/or as
> an example of how third party developers are reading and interpreting the
> Virtio-net spec:
> 
> Relating this to Virtio 1.0, the most relevant section is 3.2.1 (Supplying
> Buffers to the Device) which calls for two "suitable memory barriers". The 
> spec
> talks about these from the driver perspective, but they are both relevant to
> the device side too.
> 
> The first barrier (write to descriptor table before write to used->idx) is
> implicit on x86 because writes by the same core are not reordered. This means
> that no explicit hardware barrier is needed. (A compiler barrier may be 
> needed,
> however.)
> 
> The second memory barrier (write to used->idx before reading avail->flags) is
> not implicit on x86 because stores are reordered after loads. So an explicit
> hardware memory barrier is needed.
> 
> I hope that is a correct assessment of the situation. (Forgive my
> x86centricity, I am sure that seems very foreign to kernel hackers.)
> 
> If this assessment is correct then the DPDK developers might also want to
> review librte_vhost/vhost_rxtx.c and consider adding a hardware memory barrier
> between writing used->idx and reading avail->flags.
> 
> Cheers,
> -Luke

I agree, this looks like a bug in dpdk.

> P.S. I notice that the Linux virtio-net driver does not seem to tolerate
> spurious interrupts, even though the Virtio 1.0 spec requires this ("must"). 
> On
> 3.13.11-ckt15 I see them trigger an "irq nobody cared" kernel log message and
> then the irq is disabled. If that sounds suspicious I can supply more
> information.
> 
>

More information might be useful, yes.

Just guessing from the available info:

I think you refer to this:
The driver MUST handle spurious interrupts from the device.

The intent is to be able to handle some spurious interrupts once in a
while.  AFAIK linux triggers the message if it gets a huge number of
spurious interrupts for an extended period of time.
For example, this will trigger if the device does not clear interrupt
line after interrupt register read.

[dpdk-dev] Crash related to virtio NICs in DPDK 2.0.0 on Freebsd 10.1 VM

2015-04-07 Thread Raz Amir

I am moving from dpdk 1.7.1 to 2.0.0 and I am experiencing a crash in any
dpdk application, when rte_eal_init initializes the virtio NICs.

I have Ubuntu 14.04.02, running a Freebsd 10.1 VM with 3 virtio NICs over
qemu with kvm.

I run testpmd (for example) inside the Freebsd VM and it crashes (more info
below).

On the same setup, testpmd from dpdk 1.7.1 works without an issue.

Can you tell if you know that a Freebsd VM with virtio NICs is working for
dpdk 2.0.0?



Additional information from gdb:



user@:/dpdk/dpdk-2.0.0 # gdb testpmd

GNU gdb (GDB) 7.8.1 [GDB v7.8.1 for FreeBSD]

Copyright (C) 2014 Free Software Foundation, Inc.

License GPLv3+: GNU GPL version 3 or later


This is free software: you are free to change and redistribute it.

There is NO WARRANTY, to the extent permitted by law.  Type "show copying"

and "show warranty" for details.

This GDB was configured as "x86_64-portbld-freebsd10.1".

Type "show configuration" for configuration details.

For bug reporting instructions, please see:

.

Find the GDB manual and other documentation resources online at:

.

For help, type "help".

Type "apropos word" to search for commands related to "word"...

Reading symbols from testpmd...done.





(gdb) r -c 3 -n 2

Starting program: /root/testpmd -c 3 -n 2

EAL: Sysctl reports 2 cpus

EAL: Detected lcore 0

EAL: Detected lcore 1

EAL: Support maximum 128 logical core(s) by configuration.

EAL: Detected 2 lcore(s)

EAL: Contigmem driver has 2 buffers, each of size 64MB

EAL: Setting up physically contiguous memory...

EAL: Mapped memory segment 1 @ 0x80280: physaddr:0x1000, len
67108864

EAL: Mapped memory segment 2 @ 0x80680: physaddr:0x1400, len
67108864

EAL: TSC is not safe to use in SMP mode

EAL: TSC is not invariant

EAL: TSC frequency is ~2933513 KHz

EAL: PCI scan found 8 devices

EAL: Master lcore 0 is ready (tid=0x802406400;cpuset=[0])

PMD: ENICPMD trace: rte_enic_pmd_init

EAL: lcore 1 is ready (tid=0x802406800;cpuset=[1])

EAL: PCI device :00:03.0 on NUMA socket 0

EAL:   probe driver: 1af4:1000 rte_virtio_pmd

[New Thread 802406400 (LWP 100111)]



Program received signal SIGBUS, Bus error.

[Switching to Thread 802406400 (LWP 100111)]

0x00492f2c in outb (port=49170, data=0 '\000') at
/usr/include/machine/cpufunc.h:244

244 /usr/include/machine/cpufunc.h: No such file or directory.





(gdb) bt

#0  0x00492f2c in outb (port=49170, data=0 '\000')

at /usr/include/machine/cpufunc.h:244

#1  0x00492f7a in outb_p (data=0 '\000', port=49170)

at /dpdk/dpdk-2.0.0/lib/librte_pmd_virtio/virtio_pci.h:211

#2  0x0049328d in vtpci_set_status (hw=0x80331f380, status=0 '\000')

at /dpdk/dpdk-2.0.0/lib/librte_pmd_virtio/virtio_pci.c:130

#3  0x004931fe in vtpci_reset (hw=0x80331f380)

at /dpdk/dpdk-2.0.0/lib/librte_pmd_virtio/virtio_pci.c:108

#4  0x004a175e in eth_virtio_dev_init (eth_dev=0x831b80
)

at /dpdk/dpdk-2.0.0/lib/librte_pmd_virtio/virtio_ethdev.c:1150

#5  0x00462c09 in rte_eth_dev_init (pci_drv=0x79d1a0
,

pci_dev=0x802417560) at
/dpdk/dpdk-2.0.0/lib/librte_ether/rte_ethdev.c:326

#6  0x0046f03f in rte_eal_pci_probe_one_driver (dr=0x79d1a0
,

dev=0x802417560) at
/dpdk/dpdk-2.0.0/lib/librte_eal/bsdapp/eal/eal_pci.c:487

#7  0x00475b06 in pci_probe_all_drivers (dev=0x802417560)

at /dpdk/dpdk-2.0.0/lib/librte_eal/common/eal_common_pci.c:116

#8  0x00475bb9 in rte_eal_pci_probe ()

at /dpdk/dpdk-2.0.0/lib/librte_eal/common/eal_common_pci.c:246

#9  0x0046cd63 in rte_eal_init (argc=5, argv=0x7fffeaf0)

at /dpdk/dpdk-2.0.0/lib/librte_eal/bsdapp/eal/eal.c:554

#10 0x00404544 in main ()





(gdb) disassemble inb

Dump of assembler code for function inb:

   0x00492ed6 <+0>: push   %rbp

   0x00492ed7 <+1>: mov%rsp,%rbp

   0x00492eda <+4>: mov%edi,-0x14(%rbp)

   0x00492edd <+7>: mov-0x14(%rbp),%eax

   0x00492ee0 <+10>:mov%eax,%edx

   0x00492ee2 <+12>:in (%dx),%al

   0x00492ee3 <+13>:mov%al,-0x1(%rbp)

   0x00492ee6 <+16>:movzbl -0x1(%rbp),%eax

   0x00492eea <+20>:pop%rbp

   0x00492eeb <+21>:retq

End of assembler dump.





(gdb) p $dx

$1 = -16366





(gdb) info registers

rax0x0  0

rbx0x80280  34401681408

rcx0x0  0

rdx0xc012   49170

rsi0x0  0

rdi0xc012   49170

rbp0x7fffe708   0x7fffe708

rsp0x7fffe708   0x7fffe708

r8 0x80082b63c  34368304700

r9 0x80082b61c  34368304668

r100x1  1

r110x80331f300  34413343488

r120x7fffeaf0   140737488349

[dpdk-dev] [PATCH v3 1/5] mbuf: fix clone support when application uses private mbuf data

2015-04-07 Thread Olivier MATZ

Hi Konstantin,

On 04/07/2015 02:40 PM, Ananyev, Konstantin wrote:
> Hi Olivier,
>
>> -Original Message-
>> From: Olivier MATZ [mailto:olivier.matz at 6wind.com]
>> Sent: Monday, April 06, 2015 10:50 PM
>> To: Ananyev, Konstantin; dev at dpdk.org
>> Cc: zoltan.kiss at linaro.org; Richardson, Bruce
>> Subject: Re: [PATCH v3 1/5] mbuf: fix clone support when application uses 
>> private mbuf data
>>
>> Hi Konstantin,
>>
>> Thanks for your comments.
>>
>> On 04/02/2015 07:21 PM, Ananyev, Konstantin wrote:
>>> Hi Olivier,
>>>
 -Original Message-
 From: Olivier Matz [mailto:olivier.matz at 6wind.com]
 Sent: Tuesday, March 31, 2015 8:23 PM
 To: dev at dpdk.org
 Cc: Ananyev, Konstantin; zoltan.kiss at linaro.org; Richardson, Bruce; 
 Olivier Matz
 Subject: [PATCH v3 1/5] mbuf: fix clone support when application uses 
 private mbuf data

 From: Olivier Matz 

 Add a new private_size field in mbuf structure that should
 be initialized at mbuf pool creation. This field contains the
 size of the application private data in mbufs.

 Introduce new static inline functions rte_mbuf_from_indirect()
 and rte_mbuf_to_baddr() to replace the existing macros, which
 take the private size in account when attaching and detaching
 mbufs.

 Signed-off-by: Olivier Matz 
 ---
   app/test-pmd/testpmd.c |  1 +
   examples/vhost/main.c  |  4 +--
   lib/librte_mbuf/rte_mbuf.c |  1 +
   lib/librte_mbuf/rte_mbuf.h | 77 
 +++---
   4 files changed, 63 insertions(+), 20 deletions(-)

 diff --git a/app/test-pmd/testpmd.c b/app/test-pmd/testpmd.c
 index 3057791..c5a195a 100644
 --- a/app/test-pmd/testpmd.c
 +++ b/app/test-pmd/testpmd.c
 @@ -425,6 +425,7 @@ testpmd_mbuf_ctor(struct rte_mempool *mp,
mb->tx_offload   = 0;
mb->vlan_tci = 0;
mb->hash.rss = 0;
 +  mb->priv_size= 0;
   }

   static void
 diff --git a/examples/vhost/main.c b/examples/vhost/main.c
 index c3fcb80..e44e82f 100644
 --- a/examples/vhost/main.c
 +++ b/examples/vhost/main.c
 @@ -139,7 +139,7 @@
   /* Number of descriptors per cacheline. */
   #define DESC_PER_CACHELINE (RTE_CACHE_LINE_SIZE / sizeof(struct 
 vring_desc))

 -#define MBUF_EXT_MEM(mb)   (RTE_MBUF_FROM_BADDR((mb)->buf_addr) != (mb))
 +#define MBUF_EXT_MEM(mb)   (rte_mbuf_from_indirect(mb) != (mb))

   /* mask of enabled ports */
   static uint32_t enabled_port_mask = 0;
 @@ -1550,7 +1550,7 @@ attach_rxmbuf_zcp(struct virtio_net *dev)
   static inline void pktmbuf_detach_zcp(struct rte_mbuf *m)
   {
const struct rte_mempool *mp = m->pool;
 -  void *buf = RTE_MBUF_TO_BADDR(m);
 +  void *buf = rte_mbuf_to_baddr(m);
uint32_t buf_ofs;
uint32_t buf_len = mp->elt_size - sizeof(*m);
m->buf_physaddr = rte_mempool_virt2phy(mp, m) + sizeof(*m);
 diff --git a/lib/librte_mbuf/rte_mbuf.c b/lib/librte_mbuf/rte_mbuf.c
 index 526b18d..e095999 100644
 --- a/lib/librte_mbuf/rte_mbuf.c
 +++ b/lib/librte_mbuf/rte_mbuf.c
 @@ -125,6 +125,7 @@ rte_pktmbuf_init(struct rte_mempool *mp,
m->pool = mp;
m->nb_segs = 1;
m->port = 0xff;
 +  m->priv_size = 0;
>>>
>>> Why it is 0?
>>> Shouldn't it be the same calulations as in detach() below:
>>> m->priv_size = /*get private size from mempool private*/;
>>> m->buf_addr = (char *)m + sizeof(struct rte_mbuf) + m->priv_size;
>>> m->buf_len = mp->elt_size - sizeof(struct rte_mbuf) - m->priv_size;
>>> ?
>>
>> It's 0 because we also have in the function (not visible in the
>> patch):
>>
>>m->buf_addr = (char *)m + sizeof(struct rte_mbuf);
>
> Yep, that's why as I wrote above, I think we need to setup here all 3 fields:
> priv_size, buf_addr, buf_len exactly in the same way as in detach().
>
>>
>> It means that an application that wants to use a private area has
>> to provide another init function derived from this default function.
>
> After your changes, attach/free and other functions from public mbuf API
> rely on priv_size being set properly.
> So I suppose 'official' pktmbuf_init() should also set it in a proper manner.
>
>> This was already the case before the patch series.
>
> Before this patch series, we don't have priv_size, so we have nothing to 
> setup.
>
>>
>> As we discussed in previous mail, I plan to propose a rework of
>> mbuf pool initialization in another series, and my initial idea was to
>> change this at the same time. But on the other hand it does not hurt
>> to do this change now. I'll include it in next version.
>
> Ok.

Just to be sure we're on the same line:

- before the patch series

   - private area was working before that patch series if clones were not
 used. To use a private are, the user had to provide another
 function derived from pktmbuf_init() to chan

[dpdk-dev] [PATCH] doc: fix code-block syntax

2015-04-07 Thread Thomas Monjalon

Some blocks are not visible with some Sphinx versions because
they are using the wrong keyword for code.

Tested with Sphinx v1.1.3.

Fixes: 1733be6d3147 ("doc: new eal multi-pthread feature")
Fixes: ccefe752cab0 ("doc: add jobstats sample guide")

Signed-off-by: Thomas Monjalon 
---
 doc/guides/prog_guide/env_abstraction_layer.rst   | 2 +-
 doc/guides/sample_app_ug/l2_forward_job_stats.rst | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/doc/guides/prog_guide/env_abstraction_layer.rst 
b/doc/guides/prog_guide/env_abstraction_layer.rst
index 1b531e2..06289ed 100644
--- a/doc/guides/prog_guide/env_abstraction_layer.rst
+++ b/doc/guides/prog_guide/env_abstraction_layer.rst
@@ -349,7 +349,7 @@ cgroup control
 The following is a simple example of cgroup control usage, there are two 
pthreads(t0 and t1) doing packet I/O on the same core ($CPU).
 We expect only 50% of CPU spend on packet IO.

-  .. code::
+  .. code-block:: console

 mkdir /sys/fs/cgroup/cpu/pkt_io
 mkdir /sys/fs/cgroup/cpuset/pkt_io
diff --git a/doc/guides/sample_app_ug/l2_forward_job_stats.rst 
b/doc/guides/sample_app_ug/l2_forward_job_stats.rst
index eafb8df..54d25cb 100644
--- a/doc/guides/sample_app_ug/l2_forward_job_stats.rst
+++ b/doc/guides/sample_app_ug/l2_forward_job_stats.rst
@@ -494,7 +494,7 @@ The rte_eth_rx_burst() function writes the mbuf pointers in 
a local table and re

 After first read second try is issued.

-.. code::
+.. code-block:: c

 if (total_nb_rx == MAX_PKT_BURST) {
 const uint16_t nb_rx = rte_eth_rx_burst((uint8_t) portid, 0, 
pkts_burst,
@@ -510,7 +510,7 @@ After first read second try is issued.

 This second read is important to give job stats library a feedback how many 
packets was processed.

-.. code::
+.. code-block:: c

 /* Adjust period time in which we are running here. */
 if (rte_jobstats_finish(job, total_nb_rx) != 0) {
-- 
2.2.2

[dpdk-dev] [PATCH] eth_dev: make ether dev_ops const

2015-04-07 Thread Don Provan

-Original Message-
>From: Stephen Hemminger [mailto:stephen at networkplumber.org] 
>Sent: Monday, April 06, 2015 11:05 AM
>To: dev at dpdk.org
>Subject: [dpdk-dev] [PATCH] eth_dev: make ether dev_ops const
>
>Ethernet device function tables should be immutable for correctness and 
>security. Special case for the test code driver.
...
>diff --git a/app/test/virtual_pmd.c b/app/test/virtual_pmd.c index 
>f163562..f579558 100644
>--- a/app/test/virtual_pmd.c
>+++ b/app/test/virtual_pmd.c
...
>+/* This driver uses private mutable eth_dev_ops for each
>+ * instance so it is safe to override const here.
>+ */
>+#pragma GCC diagnostic push
>+#pragma GCC diagnostic ignored "-Wcast-qual"
> void
> virtual_ethdev_start_fn_set_success(uint8_t port_id, uint8_t success)  {
>   struct rte_eth_dev *vrtl_eth_dev = &rte_eth_devices[port_id];
>+  struct eth_dev_ops *dev_ops
>+  = (struct eth_dev_ops *) vrtl_eth_dev->dev_ops;
 ...

If this is really safe, then you should be able to accomplish it
without disabling a bunch of protection. I suggest adding a
pointer that isn't const to the private data block and adjusting
the allocated dispatch table through that instead of through
the pointer to the immutable dispatch table you've established
in struct rte_eth_dev. That reinforces the fact that modifying
the dispatch table is a private matter within the driver while
showing structurally exactly why it's safe to change it.

And it's not nearly so ugly.

-don provan
dprovan at bivio.net

[dpdk-dev] [PATCH v3 1/5] mbuf: fix clone support when application uses private mbuf data

2015-04-07 Thread Ananyev, Konstantin

Hi Olivier,

> -Original Message-
> From: Olivier MATZ [mailto:olivier.matz at 6wind.com]
> Sent: Tuesday, April 07, 2015 4:46 PM
> To: Ananyev, Konstantin; dev at dpdk.org
> Cc: zoltan.kiss at linaro.org; Richardson, Bruce
> Subject: Re: [PATCH v3 1/5] mbuf: fix clone support when application uses 
> private mbuf data
> 
> Hi Konstantin,
> 
> On 04/07/2015 02:40 PM, Ananyev, Konstantin wrote:
> > Hi Olivier,
> >
> >> -Original Message-
> >> From: Olivier MATZ [mailto:olivier.matz at 6wind.com]
> >> Sent: Monday, April 06, 2015 10:50 PM
> >> To: Ananyev, Konstantin; dev at dpdk.org
> >> Cc: zoltan.kiss at linaro.org; Richardson, Bruce
> >> Subject: Re: [PATCH v3 1/5] mbuf: fix clone support when application uses 
> >> private mbuf data
> >>
> >> Hi Konstantin,
> >>
> >> Thanks for your comments.
> >>
> >> On 04/02/2015 07:21 PM, Ananyev, Konstantin wrote:
> >>> Hi Olivier,
> >>>
>  -Original Message-
>  From: Olivier Matz [mailto:olivier.matz at 6wind.com]
>  Sent: Tuesday, March 31, 2015 8:23 PM
>  To: dev at dpdk.org
>  Cc: Ananyev, Konstantin; zoltan.kiss at linaro.org; Richardson, Bruce; 
>  Olivier Matz
>  Subject: [PATCH v3 1/5] mbuf: fix clone support when application uses 
>  private mbuf data
> 
>  From: Olivier Matz 
> 
>  Add a new private_size field in mbuf structure that should
>  be initialized at mbuf pool creation. This field contains the
>  size of the application private data in mbufs.
> 
>  Introduce new static inline functions rte_mbuf_from_indirect()
>  and rte_mbuf_to_baddr() to replace the existing macros, which
>  take the private size in account when attaching and detaching
>  mbufs.
> 
>  Signed-off-by: Olivier Matz 
>  ---
>    app/test-pmd/testpmd.c |  1 +
>    examples/vhost/main.c  |  4 +--
>    lib/librte_mbuf/rte_mbuf.c |  1 +
>    lib/librte_mbuf/rte_mbuf.h | 77 
>  +++---
>    4 files changed, 63 insertions(+), 20 deletions(-)
> 
>  diff --git a/app/test-pmd/testpmd.c b/app/test-pmd/testpmd.c
>  index 3057791..c5a195a 100644
>  --- a/app/test-pmd/testpmd.c
>  +++ b/app/test-pmd/testpmd.c
>  @@ -425,6 +425,7 @@ testpmd_mbuf_ctor(struct rte_mempool *mp,
>   mb->tx_offload   = 0;
>   mb->vlan_tci = 0;
>   mb->hash.rss = 0;
>  +mb->priv_size= 0;
>    }
> 
>    static void
>  diff --git a/examples/vhost/main.c b/examples/vhost/main.c
>  index c3fcb80..e44e82f 100644
>  --- a/examples/vhost/main.c
>  +++ b/examples/vhost/main.c
>  @@ -139,7 +139,7 @@
>    /* Number of descriptors per cacheline. */
>    #define DESC_PER_CACHELINE (RTE_CACHE_LINE_SIZE / sizeof(struct 
>  vring_desc))
> 
>  -#define MBUF_EXT_MEM(mb)   (RTE_MBUF_FROM_BADDR((mb)->buf_addr) != (mb))
>  +#define MBUF_EXT_MEM(mb)   (rte_mbuf_from_indirect(mb) != (mb))
> 
>    /* mask of enabled ports */
>    static uint32_t enabled_port_mask = 0;
>  @@ -1550,7 +1550,7 @@ attach_rxmbuf_zcp(struct virtio_net *dev)
>    static inline void pktmbuf_detach_zcp(struct rte_mbuf *m)
>    {
>   const struct rte_mempool *mp = m->pool;
>  -void *buf = RTE_MBUF_TO_BADDR(m);
>  +void *buf = rte_mbuf_to_baddr(m);
>   uint32_t buf_ofs;
>   uint32_t buf_len = mp->elt_size - sizeof(*m);
>   m->buf_physaddr = rte_mempool_virt2phy(mp, m) + sizeof(*m);
>  diff --git a/lib/librte_mbuf/rte_mbuf.c b/lib/librte_mbuf/rte_mbuf.c
>  index 526b18d..e095999 100644
>  --- a/lib/librte_mbuf/rte_mbuf.c
>  +++ b/lib/librte_mbuf/rte_mbuf.c
>  @@ -125,6 +125,7 @@ rte_pktmbuf_init(struct rte_mempool *mp,
>   m->pool = mp;
>   m->nb_segs = 1;
>   m->port = 0xff;
>  +m->priv_size = 0;
> >>>
> >>> Why it is 0?
> >>> Shouldn't it be the same calulations as in detach() below:
> >>> m->priv_size = /*get private size from mempool private*/;
> >>> m->buf_addr = (char *)m + sizeof(struct rte_mbuf) + m->priv_size;
> >>> m->buf_len = mp->elt_size - sizeof(struct rte_mbuf) - m->priv_size;
> >>> ?
> >>
> >> It's 0 because we also have in the function (not visible in the
> >> patch):
> >>
> >>m->buf_addr = (char *)m + sizeof(struct rte_mbuf);
> >
> > Yep, that's why as I wrote above, I think we need to setup here all 3 
> > fields:
> > priv_size, buf_addr, buf_len exactly in the same way as in detach().
> >
> >>
> >> It means that an application that wants to use a private area has
> >> to provide another init function derived from this default function.
> >
> > After your changes, attach/free and other functions from public mbuf API
> > rely on priv_size being set properly.
> > So I suppose 'official' pktmbuf_init() should also set it in a proper 
> > manner.
> >
> >> This was already t

[dpdk-dev] [PATCH 4/5] bond mode 4: allow external state machine

2015-04-07 Thread Eric Kinzie

On Tue Apr 07 16:18:08 +0200 2015, Pawel Wodkowski wrote:
> On 2015-04-06 19:01, Eric Kinzie wrote:
> 
> Interesting patch. I will closer look at this tomorrow.
> 
> For now I have first comments:
> 
> >+static void bond_mode_8023ad_ext_periodic_cb(void *arg);
> >+
> >  #ifdef RTE_LIBRTE_BOND_DEBUG_8023AD
> >  #define MODE4_DEBUG(fmt, ...) RTE_LOG(DEBUG, PMD, "%6u [Port %u: %s] " 
> > fmt, \
> > bond_dbg_get_time_diff_ms(), slave_id, \
> >@@ -1014,6 +1016,8 @@ bond_mode_8023ad_conf_get(struct rte_eth_dev *dev,
> > conf->tx_period_ms = mode4->tx_period_timeout / ms_ticks;
> > conf->update_timeout_ms = mode4->update_timeout_us / 1000;
> > conf->rx_marker_period_ms = mode4->rx_marker_timeout / ms_ticks;
> >+conf->slowrx_cb = mode4->slowrx_cb;
> >+conf->external_sm = mode4->external_sm;
> 
> mode4->external_sm flag realy needed? Why do not use
> mode4->slowrx_cb as external state machine indicator?

I'll remove the external_sm flag.  You're correct, it's not needed.


> >  }
> >
> >  void
> >@@ -1035,6 +1039,8 @@ bond_mode_8023ad_setup(struct rte_eth_dev *dev,
> > conf->tx_period_ms = BOND_8023AD_TX_MACHINE_PERIOD_MS;
> > conf->rx_marker_period_ms = BOND_8023AD_RX_MARKER_PERIOD_MS;
> > conf->update_timeout_ms = BOND_MODE_8023AX_UPDATE_TIMEOUT_MS;
> >+conf->slowrx_cb = NULL;
> >+conf->external_sm = 0;
> > }
> >
> > mode4->fast_periodic_timeout = conf->fast_periodic_ms * ms_ticks;
> >@@ -1045,6 +1051,8 @@ bond_mode_8023ad_setup(struct rte_eth_dev *dev,
> > mode4->tx_period_timeout = conf->tx_period_ms * ms_ticks;
> > mode4->rx_marker_timeout = conf->rx_marker_period_ms * ms_ticks;
> > mode4->update_timeout_us = conf->update_timeout_ms * 1000;
> >+mode4->slowrx_cb = conf->slowrx_cb;
> >+mode4->external_sm = conf->external_sm;
> >  }
> >
> >  int
> >@@ -1062,6 +1070,13 @@ bond_mode_8023ad_enable(struct rte_eth_dev *bond_dev)
> >  int
> >  bond_mode_8023ad_start(struct rte_eth_dev *bond_dev)
> >  {
> >+struct bond_dev_private *internals = bond_dev->data->dev_private;
> >+struct mode8023ad_private *mode4 = &internals->mode4;
> >+
> >+if (mode4->external_sm)
> >+return rte_eal_alarm_set(BOND_MODE_8023AX_UPDATE_TIMEOUT_MS * 
> >1000,
> >+&bond_mode_8023ad_ext_periodic_cb, bond_dev);
> >+
> > return rte_eal_alarm_set(BOND_MODE_8023AX_UPDATE_TIMEOUT_MS * 1000,
> > &bond_mode_8023ad_periodic_cb, bond_dev);
> >  }
> >@@ -1069,6 +1084,13 @@ bond_mode_8023ad_start(struct rte_eth_dev *bond_dev)
> >  void
> >  bond_mode_8023ad_stop(struct rte_eth_dev *bond_dev)
> >  {
> >+struct bond_dev_private *internals = bond_dev->data->dev_private;
> >+struct mode8023ad_private *mode4 = &internals->mode4;
> >+
> >+if (mode4->external_sm) {
> 
> This is bad idea. If bond_mode_8023ad_setup will be called you might
> have two handlers running for while. You should stop mode 4 by
> invoking bond_mode_8023ad_stop() before you set mode4->external_sm
> and then, if mode 4 was running, start it again.

Thanks, I'll make that change.



> Also, maybe a renaming "external_sm" to "state_machine_cb", set it
> to against default one and using it without "if()" will simplify
> code. It is no crucial but will eliminate couple of if's. In
> rte_eth_bond_8023ad_ext_slowtx() you can compare it against default
> one.
> 
> >+rte_eal_alarm_cancel(&bond_mode_8023ad_ext_periodic_cb, 
> >bond_dev);
> >+return;
> >+}
> > rte_eal_alarm_cancel(&bond_mode_8023ad_periodic_cb, bond_dev);
> >  }
> >
> >@@ -1215,3 +1237,156 @@ rte_eth_bond_8023ad_slave_info(uint8_t port_id, 
> >uint8_t slave_id,
> > info->agg_port_id = port->aggregator_port_id;
> > return 0;
> >  }
> 
> -- 
> Pawel

[dpdk-dev] [PATCH] enic: disable debug traces

2015-04-07 Thread Thomas Monjalon

The function name is printed in each enic_ethdev function.
Disable it by default with a new build option.

Signed-off-by: Thomas Monjalon 
---
 config/common_bsdapp  | 1 +
 config/common_linuxapp| 1 +
 lib/librte_pmd_enic/enic_ethdev.c | 4 
 3 files changed, 6 insertions(+)

diff --git a/config/common_bsdapp b/config/common_bsdapp
index a8ba484..c2374c0 100644
--- a/config/common_bsdapp
+++ b/config/common_bsdapp
@@ -214,6 +214,7 @@ CONFIG_RTE_LIBRTE_MLX4_SOFT_COUNTERS=1
 # Compile burst-oriented Cisco ENIC PMD driver
 #
 CONFIG_RTE_LIBRTE_ENIC_PMD=y
+CONFIG_RTE_LIBRTE_ENIC_DEBUG=n

 #
 # Compile burst-oriented VIRTIO PMD driver
diff --git a/config/common_linuxapp b/config/common_linuxapp
index 0b25f34..0078dc9 100644
--- a/config/common_linuxapp
+++ b/config/common_linuxapp
@@ -211,6 +211,7 @@ CONFIG_RTE_LIBRTE_MLX4_SOFT_COUNTERS=1
 # Compile burst-oriented Cisco ENIC PMD driver
 #
 CONFIG_RTE_LIBRTE_ENIC_PMD=y
+CONFIG_RTE_LIBRTE_ENIC_DEBUG=n

 #
 # Compile burst-oriented VIRTIO PMD driver
diff --git a/lib/librte_pmd_enic/enic_ethdev.c 
b/lib/librte_pmd_enic/enic_ethdev.c
index 4950ede..18fadfb 100644
--- a/lib/librte_pmd_enic/enic_ethdev.c
+++ b/lib/librte_pmd_enic/enic_ethdev.c
@@ -48,8 +48,12 @@
 #include "vnic_enet.h"
 #include "enic.h"

+#ifdef RTE_LIBRTE_ENIC_DEBUG
 #define ENICPMD_FUNC_TRACE() \
RTE_LOG(DEBUG, PMD, "ENICPMD trace: %s\n", __func__)
+#else
+#define ENICPMD_FUNC_TRACE() do {} while (0)
+#endif

 /*
  * The set of PCI devices this driver supports
-- 
2.2.2

[dpdk-dev] [PATCH] eth_dev: make ether dev_ops const

2015-04-07 Thread Stephen Hemminger

On Tue, Apr 7, 2015 at 10:21 AM, Don Provan  wrote:

> -Original Message-
> >From: Stephen Hemminger [mailto:stephen at networkplumber.org]
> >Sent: Monday, April 06, 2015 11:05 AM
> >To: dev at dpdk.org
> >Subject: [dpdk-dev] [PATCH] eth_dev: make ether dev_ops const
> >
> >Ethernet device function tables should be immutable for correctness and
> security. Special case for the test code driver.
> ...
> >diff --git a/app/test/virtual_pmd.c b/app/test/virtual_pmd.c index
> f163562..f579558 100644
> >--- a/app/test/virtual_pmd.c
> >+++ b/app/test/virtual_pmd.c
> ...
> >+/* This driver uses private mutable eth_dev_ops for each
> >+ * instance so it is safe to override const here.
> >+ */
> >+#pragma GCC diagnostic push
> >+#pragma GCC diagnostic ignored "-Wcast-qual"
> > void
> > virtual_ethdev_start_fn_set_success(uint8_t port_id, uint8_t success)  {
> >   struct rte_eth_dev *vrtl_eth_dev = &rte_eth_devices[port_id];
> >+  struct eth_dev_ops *dev_ops
> >+  = (struct eth_dev_ops *) vrtl_eth_dev->dev_ops;
>  ...
>
> If this is really safe, then you should be able to accomplish it
> without disabling a bunch of protection. I suggest adding a
> pointer that isn't const to the private data block and adjusting
> the allocated dispatch table through that instead of through
> the pointer to the immutable dispatch table you've established
> in struct rte_eth_dev. That reinforces the fact that modifying
> the dispatch table is a private matter within the driver while
> showing structurally exactly why it's safe to change it.
>
> And it's not nearly so ugly.
>
> -don provan
> dprovan at bivio.net
>
>
In this case it is safe, but only because this dummy driver used in testing
does non-standard things.
It copies a base template for ops into a allocated area of memory, then
modifies it.
Not the best design, but did not want to hold back the ethernet dev_ops.
Probably the private pointer is a better way.

[dpdk-dev] [PATCH v2 1/6] test: remove useless memset

2015-04-07 Thread Stephen Hemminger

Remove useless memset, since dev_private is created by rte_zmalloc
it must already be zero.

Signed-off-by: Stephen Hemminger 
---
 app/test/virtual_pmd.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/app/test/virtual_pmd.c b/app/test/virtual_pmd.c
index f163562..39ecf80 100644
--- a/app/test/virtual_pmd.c
+++ b/app/test/virtual_pmd.c
@@ -565,8 +565,6 @@ virtual_ethdev_create(const char *name, struct ether_addr 
*mac_addr,
if (dev_private == NULL)
goto err;

-   memset(dev_private, 0, sizeof(*dev_private));
-
snprintf(name_buf, sizeof(name_buf), "%s_rxQ", name);
dev_private->rx_queue = rte_ring_create(name_buf, MAX_PKT_BURST, 
socket_id,
0);
-- 
2.1.4

[dpdk-dev] [PATCH v2 2/6] test: remove useless check of NULL before rte_free

2015-04-07 Thread Stephen Hemminger

rte_free like Glibc free allows rte_free(NULL) as null operation.

Signed-off-by: Stephen Hemminger 
---
 app/test/test_hash_perf.c |  2 +-
 app/test/virtual_pmd.c| 18 ++
 2 files changed, 7 insertions(+), 13 deletions(-)

diff --git a/app/test/test_hash_perf.c b/app/test/test_hash_perf.c
index bd531ec..6eabb21 100644
--- a/app/test/test_hash_perf.c
+++ b/app/test/test_hash_perf.c
@@ -387,7 +387,7 @@ struct tbl_perf_test_params tbl_perf_params[] =
if (cond) { \
printf("ERROR line %d: " str "\n", __LINE__, ##__VA_ARGS__); \
if (handle) rte_fbk_hash_free(handle);  \
-   if (keys) rte_free(keys);   \
+   rte_free(keys); \
return -1;  \
}   \
 } while(0)
diff --git a/app/test/virtual_pmd.c b/app/test/virtual_pmd.c
index 39ecf80..1f4da96 100644
--- a/app/test/virtual_pmd.c
+++ b/app/test/virtual_pmd.c
@@ -635,18 +635,12 @@ virtual_ethdev_create(const char *name, struct ether_addr 
*mac_addr,
return eth_dev->data->port_id;

 err:
-   if (pci_dev)
-   rte_free(pci_dev);
-   if (pci_drv)
-   rte_free(pci_drv);
-   if (eth_drv)
-   rte_free(eth_drv);
-   if (dev_ops)
-   rte_free(dev_ops);
-   if (id_table)
-   rte_free(id_table);
-   if (dev_private)
-   rte_free(dev_private);
+   rte_free(pci_dev);
+   rte_free(pci_drv);
+   rte_free(eth_drv);
+   rte_free(dev_ops);
+   rte_free(id_table);
+   rte_free(dev_private);

return -1;
 }
-- 
2.1.4

[dpdk-dev] [PATCH v2 3/6] examples: get rid of unneeded null checks before rte_free

2015-04-07 Thread Stephen Hemminger

rte_free handles getting passed a NULL pointer.

Signed-off-by: Stephen Hemminger 
---
 examples/vhost/main.c   | 3 +--
 examples/vhost_xen/vhost_monitor.c  | 7 +++
 examples/vm_power_manager/channel_manager.c | 6 ++
 3 files changed, 6 insertions(+), 10 deletions(-)

diff --git a/examples/vhost/main.c b/examples/vhost/main.c
index c3fcb80..ad10f82 100644
--- a/examples/vhost/main.c
+++ b/examples/vhost/main.c
@@ -2747,8 +2747,7 @@ new_device (struct virtio_net *dev)
RTE_LOG(INFO, VHOST_DATA, "(%"PRIu64") Failed to add device to 
data core\n", dev->device_fh);
vdev->ready = DEVICE_SAFE_REMOVE;
destroy_device(dev);
-   if (vdev->regions_hpa)
-   rte_free(vdev->regions_hpa);
+   rte_free(vdev->regions_hpa);
rte_free(vdev);
return -1;
}
diff --git a/examples/vhost_xen/vhost_monitor.c 
b/examples/vhost_xen/vhost_monitor.c
index 9d99962..6455993 100644
--- a/examples/vhost_xen/vhost_monitor.c
+++ b/examples/vhost_xen/vhost_monitor.c
@@ -298,10 +298,9 @@ virtio_net_config_ll *new_device(unsigned int virtio_idx, 
struct xen_guest *gues
 err:
if (new_ll_dev)
free(new_ll_dev);
-   if (virtqueue_rx)
-   rte_free(virtqueue_rx);
-   if (virtqueue_tx)
-   rte_free(virtqueue_tx);
+   rte_free(virtqueue_rx);
+   rte_free(virtqueue_tx);
+
return NULL;
 }

diff --git a/examples/vm_power_manager/channel_manager.c 
b/examples/vm_power_manager/channel_manager.c
index 04344ae..7d892e2 100644
--- a/examples/vm_power_manager/channel_manager.c
+++ b/examples/vm_power_manager/channel_manager.c
@@ -800,9 +800,7 @@ channel_manager_exit(void)
rte_free(vm_info);
}

-   if (global_cpumaps != NULL)
-   rte_free(global_cpumaps);
-   if (global_vircpuinfo != NULL)
-   rte_free(global_vircpuinfo);
+   rte_free(global_cpumaps);
+   rte_free(global_vircpuinfo);
disconnect_hypervisor();
 }
-- 
2.1.4

[dpdk-dev] [PATCH v2 4/6] pmd: remove unnecessary if() before rte_free

2015-04-07 Thread Stephen Hemminger

Since rte_free accept NULL and does nothing, better to save
space and remove these useless checks.

Signed-off-by: Stephen Hemminger 
---
 lib/librte_pmd_af_packet/rte_eth_af_packet.c | 14 ++
 lib/librte_pmd_bond/rte_eth_bond_api.c   | 13 +
 lib/librte_pmd_enic/enic_main.c  |  3 +--
 lib/librte_pmd_mlx4/mlx4.c   | 15 +++
 lib/librte_pmd_null/rte_eth_null.c   | 13 +
 lib/librte_pmd_pcap/rte_eth_pcap.c   | 11 +--
 lib/librte_pmd_ring/rte_eth_ring.c   | 10 --
 lib/librte_pmd_xenvirt/rte_eth_xenvirt.c | 10 --
 8 files changed, 37 insertions(+), 52 deletions(-)

diff --git a/lib/librte_pmd_af_packet/rte_eth_af_packet.c 
b/lib/librte_pmd_af_packet/rte_eth_af_packet.c
index 2ac50ba..c10da6c 100644
--- a/lib/librte_pmd_af_packet/rte_eth_af_packet.c
+++ b/lib/librte_pmd_af_packet/rte_eth_af_packet.c
@@ -680,18 +680,16 @@ rte_pmd_init_internals(const char *name,
return 0;

 error:
-   if (data)
-   rte_free(data);
-   if (pci_dev)
-   rte_free(pci_dev);
+   rte_free(data);
+   rte_free(pci_dev);
+
if (*internals) {
for (q = 0; q < nb_queues; q++) {
munmap((*internals)->rx_queue[q].map,
   2 * req->tp_block_size * req->tp_block_nr);
-   if ((*internals)->rx_queue[q].rd)
-   rte_free((*internals)->rx_queue[q].rd);
-   if ((*internals)->tx_queue[q].rd)
-   rte_free((*internals)->tx_queue[q].rd);
+
+   rte_free((*internals)->rx_queue[q].rd);
+   rte_free((*internals)->tx_queue[q].rd);
if (((*internals)->rx_queue[q].sockfd != 0) &&
((*internals)->rx_queue[q].sockfd != qsockfd))
close((*internals)->rx_queue[q].sockfd);
diff --git a/lib/librte_pmd_bond/rte_eth_bond_api.c 
b/lib/librte_pmd_bond/rte_eth_bond_api.c
index 13f3941..f594fe1 100644
--- a/lib/librte_pmd_bond/rte_eth_bond_api.c
+++ b/lib/librte_pmd_bond/rte_eth_bond_api.c
@@ -318,14 +318,11 @@ rte_eth_bond_create(const char *name, uint8_t mode, 
uint8_t socket_id)
return eth_dev->data->port_id;

 err:
-   if (pci_dev)
-   rte_free(pci_dev);
-   if (pci_id_table)
-   rte_free(pci_id_table);
-   if (eth_drv)
-   rte_free(eth_drv);
-   if (internals)
-   rte_free(internals);
+   rte_free(pci_dev);
+   rte_free(pci_id_table);
+   rte_free(eth_drv);
+   rte_free(internals);
+
return -1;
 }

diff --git a/lib/librte_pmd_enic/enic_main.c b/lib/librte_pmd_enic/enic_main.c
index 0892b3e..0e40d46 100644
--- a/lib/librte_pmd_enic/enic_main.c
+++ b/lib/librte_pmd_enic/enic_main.c
@@ -983,8 +983,7 @@ static void enic_dev_deinit(struct enic *enic)
 {
struct rte_eth_dev *eth_dev = enic->rte_dev;

-   if (eth_dev->data->mac_addrs)
-   rte_free(eth_dev->data->mac_addrs);
+   rte_free(eth_dev->data->mac_addrs);
 }


diff --git a/lib/librte_pmd_mlx4/mlx4.c b/lib/librte_pmd_mlx4/mlx4.c
index fa749f4..0eca322 100644
--- a/lib/librte_pmd_mlx4/mlx4.c
+++ b/lib/librte_pmd_mlx4/mlx4.c
@@ -792,10 +792,10 @@ txq_alloc_elts(struct txq *txq, unsigned int elts_n)
 error:
if (mr_linear != NULL)
claim_zero(ibv_dereg_mr(mr_linear));
-   if (elts_linear != NULL)
-   rte_free(elts_linear);
-   if (elts != NULL)
-   rte_free(elts);
+
+   rte_free(elts_linear);
+   rte_free(elts);
+
DEBUG("%p: failed, freed everything", (void *)txq);
assert(ret > 0);
return ret;
@@ -823,8 +823,8 @@ txq_free_elts(struct txq *txq)
txq->mr_linear = NULL;
if (mr_linear != NULL)
claim_zero(ibv_dereg_mr(mr_linear));
-   if (elts_linear != NULL)
-   rte_free(elts_linear);
+
+   rte_free(elts_linear);
if (elts == NULL)
return;
for (i = 0; (i != elemof(*elts)); ++i) {
@@ -4602,8 +4602,7 @@ mlx4_pci_devinit(struct rte_pci_driver *pci_drv, struct 
rte_pci_device *pci_dev)
continue;

 port_error:
-   if (priv)
-   rte_free(priv);
+   rte_free(priv);
if (pd)
claim_zero(ibv_dealloc_pd(pd));
if (ctx)
diff --git a/lib/librte_pmd_null/rte_eth_null.c 
b/lib/librte_pmd_null/rte_eth_null.c
index 0e18502..036faaf 100644
--- a/lib/librte_pmd_null/rte_eth_null.c
+++ b/lib/librte_pmd_null/rte_eth_null.c
@@ -355,8 +355,7 @@ eth_queue_release(void *q)
return;

nq = q;
-   if (nq->dummy_packet)
-   rte_free(nq->dummy_packet);
+   rte_free(nq->dummy_packet);
 }

 static int
@@ -458,12 +457,10 @@ eth_dev_null_create(const char *n

[dpdk-dev] [PATCH v2 5/6] test: put dev_ops in private

2015-04-07 Thread Stephen Hemminger

The test PMD uses a special type of eth_dev_ops to test features.
Rather allocating this separately, just put in the private data area.
This allows for next change to make dev_ops const.

Signed-off-by: Stephen Hemminger 
---
v2 -- split into separate patch and put dev_ops in private

 app/test/virtual_pmd.c | 57 +-
 1 file changed, 28 insertions(+), 29 deletions(-)

diff --git a/app/test/virtual_pmd.c b/app/test/virtual_pmd.c
index 1f4da96..3ae8c90 100644
--- a/app/test/virtual_pmd.c
+++ b/app/test/virtual_pmd.c
@@ -45,6 +45,7 @@
 static const char *virtual_ethdev_driver_name = "Virtual PMD";

 struct virtual_ethdev_private {
+   struct eth_dev_ops dev_ops;
struct rte_eth_stats eth_stats;

struct rte_ring *rx_queue;
@@ -262,61 +263,67 @@ static struct eth_dev_ops virtual_ethdev_default_dev_ops 
= {
 void
 virtual_ethdev_start_fn_set_success(uint8_t port_id, uint8_t success)
 {
-   struct rte_eth_dev *vrtl_eth_dev = &rte_eth_devices[port_id];
+   struct rte_eth_dev *dev = &rte_eth_devices[port_id];
+   struct virtual_ethdev_private *dev_private = dev->data->dev_private;
+   struct eth_dev_ops *dev_ops = &dev_private->dev_ops;

if (success)
-   vrtl_eth_dev->dev_ops->dev_start = virtual_ethdev_start_success;
+   dev_ops->dev_start = virtual_ethdev_start_success;
else
-   vrtl_eth_dev->dev_ops->dev_start = virtual_ethdev_start_fail;
+   dev_ops->dev_start = virtual_ethdev_start_fail;

 }

 void
 virtual_ethdev_configure_fn_set_success(uint8_t port_id, uint8_t success)
 {
-   struct rte_eth_dev *vrtl_eth_dev = &rte_eth_devices[port_id];
+   struct rte_eth_dev *dev = &rte_eth_devices[port_id];
+   struct virtual_ethdev_private *dev_private = dev->data->dev_private;
+   struct eth_dev_ops *dev_ops = &dev_private->dev_ops;

if (success)
-   vrtl_eth_dev->dev_ops->dev_configure = 
virtual_ethdev_configure_success;
+   dev_ops->dev_configure = virtual_ethdev_configure_success;
else
-   vrtl_eth_dev->dev_ops->dev_configure = 
virtual_ethdev_configure_fail;
+   dev_ops->dev_configure = virtual_ethdev_configure_fail;
 }

 void
 virtual_ethdev_rx_queue_setup_fn_set_success(uint8_t port_id, uint8_t success)
 {
-   struct rte_eth_dev *vrtl_eth_dev = &rte_eth_devices[port_id];
+   struct rte_eth_dev *dev = &rte_eth_devices[port_id];
+   struct virtual_ethdev_private *dev_private = dev->data->dev_private;
+   struct eth_dev_ops *dev_ops = &dev_private->dev_ops;

if (success)
-   vrtl_eth_dev->dev_ops->rx_queue_setup =
-   virtual_ethdev_rx_queue_setup_success;
+   dev_ops->rx_queue_setup = virtual_ethdev_rx_queue_setup_success;
else
-   vrtl_eth_dev->dev_ops->rx_queue_setup =
-   virtual_ethdev_rx_queue_setup_fail;
+   dev_ops->rx_queue_setup = virtual_ethdev_rx_queue_setup_fail;
 }

 void
 virtual_ethdev_tx_queue_setup_fn_set_success(uint8_t port_id, uint8_t success)
 {
-   struct rte_eth_dev *vrtl_eth_dev = &rte_eth_devices[port_id];
+   struct rte_eth_dev *dev = &rte_eth_devices[port_id];
+   struct virtual_ethdev_private *dev_private = dev->data->dev_private;
+   struct eth_dev_ops *dev_ops = &dev_private->dev_ops;

if (success)
-   vrtl_eth_dev->dev_ops->tx_queue_setup =
-   virtual_ethdev_tx_queue_setup_success;
+   dev_ops->tx_queue_setup = virtual_ethdev_tx_queue_setup_success;
else
-   vrtl_eth_dev->dev_ops->tx_queue_setup =
-   virtual_ethdev_tx_queue_setup_fail;
+   dev_ops->tx_queue_setup = virtual_ethdev_tx_queue_setup_fail;
 }

 void
 virtual_ethdev_link_update_fn_set_success(uint8_t port_id, uint8_t success)
 {
-   struct rte_eth_dev *vrtl_eth_dev = &rte_eth_devices[port_id];
+   struct rte_eth_dev *dev = &rte_eth_devices[port_id];
+   struct virtual_ethdev_private *dev_private = dev->data->dev_private;
+   struct eth_dev_ops *dev_ops = &dev_private->dev_ops;

if (success)
-   vrtl_eth_dev->dev_ops->link_update = 
virtual_ethdev_link_update_success;
+   dev_ops->link_update = virtual_ethdev_link_update_success;
else
-   vrtl_eth_dev->dev_ops->link_update = 
virtual_ethdev_link_update_fail;
+   dev_ops->link_update = virtual_ethdev_link_update_fail;
 }


@@ -528,7 +535,6 @@ virtual_ethdev_create(const char *name, struct ether_addr 
*mac_addr,
struct rte_eth_dev *eth_dev = NULL;
struct eth_driver *eth_drv = NULL;
struct rte_pci_driver *pci_drv = NULL;
-   struct eth_dev_ops *dev_ops = NULL;
struct rte_pci_id *id_table = NULL;
struct virtual_ethdev_private *dev_private = NULL;
char name_buf

[dpdk-dev] [PATCH v2 6/6] eth: make dev_ops const

2015-04-07 Thread Stephen Hemminger

The ethernet device ops function table should be made const for
safety and security.

Signed-off-by: Stephen Hemminger 
---
v2 -- handle/fix virtual_pmd in earlier patches

 app/test/virtual_pmd.c   | 30 +-
 lib/librte_ether/rte_ethdev.h|  2 +-
 lib/librte_pmd_af_packet/rte_eth_af_packet.c |  2 +-
 lib/librte_pmd_e1000/em_ethdev.c |  2 +-
 lib/librte_pmd_e1000/igb_ethdev.c|  4 ++--
 lib/librte_pmd_enic/enic_ethdev.c|  2 +-
 lib/librte_pmd_fm10k/fm10k_ethdev.c  |  2 +-
 lib/librte_pmd_i40e/i40e_ethdev.c|  2 +-
 lib/librte_pmd_i40e/i40e_ethdev_vf.c |  2 +-
 lib/librte_pmd_ixgbe/ixgbe_ethdev.c  |  5 ++---
 lib/librte_pmd_mlx4/mlx4.c   |  2 +-
 lib/librte_pmd_null/rte_eth_null.c   | 24 ++---
 lib/librte_pmd_pcap/rte_eth_pcap.c   | 26 +++---
 lib/librte_pmd_ring/rte_eth_ring.c   | 32 ++--
 lib/librte_pmd_virtio/virtio_ethdev.c|  2 +-
 lib/librte_pmd_vmxnet3/vmxnet3_ethdev.c  |  2 +-
 lib/librte_pmd_xenvirt/rte_eth_xenvirt.c | 26 +++---
 17 files changed, 83 insertions(+), 84 deletions(-)

diff --git a/app/test/virtual_pmd.c b/app/test/virtual_pmd.c
index 3ae8c90..9581892 100644
--- a/app/test/virtual_pmd.c
+++ b/app/test/virtual_pmd.c
@@ -242,21 +242,21 @@ virtual_ethdev_promiscuous_mode_disable(struct 
rte_eth_dev *dev __rte_unused)
 {}


-static struct eth_dev_ops virtual_ethdev_default_dev_ops = {
-   .dev_configure = virtual_ethdev_configure_success,
-   .dev_start = virtual_ethdev_start_success,
-   .dev_stop = virtual_ethdev_stop,
-   .dev_close = virtual_ethdev_close,
-   .dev_infos_get = virtual_ethdev_info_get,
-   .rx_queue_setup = virtual_ethdev_rx_queue_setup_success,
-   .tx_queue_setup = virtual_ethdev_tx_queue_setup_success,
-   .rx_queue_release = virtual_ethdev_rx_queue_release,
-   .tx_queue_release = virtual_ethdev_tx_queue_release,
-   .link_update = virtual_ethdev_link_update_success,
-   .stats_get = virtual_ethdev_stats_get,
-   .stats_reset = virtual_ethdev_stats_reset,
-   .promiscuous_enable = virtual_ethdev_promiscuous_mode_enable,
-   .promiscuous_disable = virtual_ethdev_promiscuous_mode_disable
+static const struct eth_dev_ops virtual_ethdev_default_dev_ops = {
+   .dev_configure = virtual_ethdev_configure_success,
+   .dev_start = virtual_ethdev_start_success,
+   .dev_stop = virtual_ethdev_stop,
+   .dev_close = virtual_ethdev_close,
+   .dev_infos_get = virtual_ethdev_info_get,
+   .rx_queue_setup = virtual_ethdev_rx_queue_setup_success,
+   .tx_queue_setup = virtual_ethdev_tx_queue_setup_success,
+   .rx_queue_release = virtual_ethdev_rx_queue_release,
+   .tx_queue_release = virtual_ethdev_tx_queue_release,
+   .link_update = virtual_ethdev_link_update_success,
+   .stats_get = virtual_ethdev_stats_get,
+   .stats_reset = virtual_ethdev_stats_reset,
+   .promiscuous_enable = virtual_ethdev_promiscuous_mode_enable,
+   .promiscuous_disable = virtual_ethdev_promiscuous_mode_disable
 };


diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h
index e8df027..46a55ff 100644
--- a/lib/librte_ether/rte_ethdev.h
+++ b/lib/librte_ether/rte_ethdev.h
@@ -1472,7 +1472,7 @@ struct rte_eth_dev {
eth_tx_burst_t tx_pkt_burst; /**< Pointer to PMD transmit function. */
struct rte_eth_dev_data *data;  /**< Pointer to device data */
const struct eth_driver *driver;/**< Driver for this device */
-   struct eth_dev_ops *dev_ops;/**< Functions exported by PMD */
+   const struct eth_dev_ops *dev_ops; /**< Functions exported by PMD */
struct rte_pci_device *pci_dev; /**< PCI info. supplied by probing */
/** User application callbacks for NIC interrupts */
struct rte_eth_dev_cb_list link_intr_cbs;
diff --git a/lib/librte_pmd_af_packet/rte_eth_af_packet.c 
b/lib/librte_pmd_af_packet/rte_eth_af_packet.c
index c10da6c..f7e9ec9 100644
--- a/lib/librte_pmd_af_packet/rte_eth_af_packet.c
+++ b/lib/librte_pmd_af_packet/rte_eth_af_packet.c
@@ -384,7 +384,7 @@ eth_tx_queue_setup(struct rte_eth_dev *dev,
return 0;
 }

-static struct eth_dev_ops ops = {
+static const struct eth_dev_ops ops = {
.dev_start = eth_dev_start,
.dev_stop = eth_dev_stop,
.dev_close = eth_dev_close,
diff --git a/lib/librte_pmd_e1000/em_ethdev.c b/lib/librte_pmd_e1000/em_ethdev.c
index 76f45c9..12ecf5f 100644
--- a/lib/librte_pmd_e1000/em_ethdev.c
+++ b/lib/librte_pmd_e1000/em_ethdev.c
@@ -133,7 +133,7 @@ static struct rte_pci_id pci_id_em_map[] = {
 {.device_id = 0},
 };

-static struct eth_dev_ops eth_em_ops = {
+static const struct eth_dev_ops eth_em_o

[dpdk-dev] [RFC PATCHv3 0/7] add OSv support

[dpdk-dev] [RFC PATCHv3 1/7] mk: support compiling C++ code

[dpdk-dev] [RFC PATCHv3 2/7] eal: Add extern C on eal_hugepages.h

[dpdk-dev] [RFC PATCHv3 3/7] eal: Add extern C on eal_thread.h

[dpdk-dev] [RFC PATCHv3 4/7] eal: Add extern C on eal_private.h

[dpdk-dev] [RFC PATCHv3 5/7] add OSv support

[dpdk-dev] [RFC PATCHv3 6/7] app/test: support OSv

[dpdk-dev] [RFC PATCHv3 7/7] doc: Add Getting Started Guide for OSv

[dpdk-dev] [RFC PATCHv2 5/8] add OSv support

[dpdk-dev] CROSS_COMPILATION with custom toolchain

[dpdk-dev] [RFC] Cuckoo hash for DPDK 2.1

[dpdk-dev] Running DPDK Binaries on a different Target

[dpdk-dev] Running DPDK Binaries on a different Target

[dpdk-dev] [PATCH v3 1/5] mbuf: fix clone support when application uses private mbuf data

[dpdk-dev] [PATCH 4/5] bond mode 4: allow external state machine

[dpdk-dev] [snabb-devel] Re: memory barriers in virtq.lua?

[dpdk-dev] Running DPDK Binaries on a different Target

[dpdk-dev] [PATCH 4/5] bond mode 4: allow external state machine

[dpdk-dev] [snabb-devel] Re: memory barriers in virtq.lua?

[dpdk-dev] Crash related to virtio NICs in DPDK 2.0.0 on Freebsd 10.1 VM

[dpdk-dev] [PATCH v3 1/5] mbuf: fix clone support when application uses private mbuf data

[dpdk-dev] [PATCH] doc: fix code-block syntax

[dpdk-dev] [PATCH] eth_dev: make ether dev_ops const

[dpdk-dev] [PATCH v3 1/5] mbuf: fix clone support when application uses private mbuf data

[dpdk-dev] [PATCH 4/5] bond mode 4: allow external state machine

[dpdk-dev] [PATCH] enic: disable debug traces

[dpdk-dev] [PATCH] eth_dev: make ether dev_ops const

[dpdk-dev] [PATCH v2 1/6] test: remove useless memset

[dpdk-dev] [PATCH v2 2/6] test: remove useless check of NULL before rte_free

[dpdk-dev] [PATCH v2 3/6] examples: get rid of unneeded null checks before rte_free

[dpdk-dev] [PATCH v2 4/6] pmd: remove unnecessary if() before rte_free

[dpdk-dev] [PATCH v2 5/6] test: put dev_ops in private

[dpdk-dev] [PATCH v2 6/6] eth: make dev_ops const

33 matches

Site Navigation

Mail list logo

Footer information