Re: [dpdk-dev] [PATCH v3] app/testpmd: enable the heavyweight mode TCP/IPv4 GRO
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Jiayu Hu
> Sent: Sunday, September 3, 2017 2:30 PM
> To: dev@dpdk.org
> Cc: Yigit, Ferruh ; Ananyev, Konstantin
> ; Tan, Jianfeng ;
> Wu, Jingjing ; Hu, Jiayu
> Subject: [dpdk-dev] [PATCH v3] app/testpmd: enable the heavyweight
> mode TCP/IPv4 GRO
>
> The GRO library provides two modes to reassemble packets. Currently, the
> csum forwarding engine has supported to use the lightweight mode to
> reassemble TCP/IPv4 packets. This patch introduces the heavyweight mode
> for TCP/IPv4 GRO in the csum forwarding engine.
>
> With the command "set port gro on|off", users can enable
> TCP/IPv4 GRO for a given port. With the command "set gro flush ",
> users can determine when the GROed TCP/IPv4 packets are flushed from
> reassembly tables. With the command "show port gro", users can
> display GRO configuration.
>
> Signed-off-by: Jiayu Hu
Tested-by : Lei Yao
This patch has been tested on my bench, iperf test result is as following:
No-GRO: 8 Gbps
Kernel GRO: 14.3 Gbps
GRO flush 0 : 12.7 Gbps
GRO flush 1: 16.8 Gbps
But when I use 40G NIC and set GRO flush cycle as 2, sometimes
the iperf traffic will stall for several seconds. Still need investigate.
> ---
> changes in v3:
> - remove "heavyweight mode" and "lightweight mode" from GRO
> commands
> - combine two patches into one
> - use consistent help string for GRO commands
> - remove the unnecessary command "gro set (max_flow_num)
> (max_item_num_per_flow) (port_id)"
> changes in v2:
> - use "set" and "show" as the root level command
> - add a new command to show GRO configuration
> - fix l2_len/l3_len/l4_len unset etc. bugs
>
> app/test-pmd/cmdline.c | 206
>
> app/test-pmd/config.c | 67 +++--
> app/test-pmd/csumonly.c | 31 -
> app/test-pmd/testpmd.c | 18 ++-
> app/test-pmd/testpmd.h | 16 ++-
> doc/guides/testpmd_app_ug/testpmd_funcs.rst | 45 --
> 6 files changed, 263 insertions(+), 120 deletions(-)
>
> diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
> index cd8c358..d628250 100644
> --- a/app/test-pmd/cmdline.c
> +++ b/app/test-pmd/cmdline.c
> @@ -423,13 +423,16 @@ static void cmd_help_long_parsed(void
> *parsed_result,
> "tso show (portid)"
> "Display the status of TCP Segmentation
> Offload.\n\n"
>
> - "gro (on|off) (port_id)"
> + "set port (port_id) gro on|off\n"
> "Enable or disable Generic Receive Offload in"
> " csum forwarding engine.\n\n"
>
> - "gro set (max_flow_num)
> (max_item_num_per_flow) (port_id)\n"
> - "Set max flow number and max packet number
> per-flow"
> - " for GRO.\n\n"
> + "show port (port_id) gro\n"
> + "Display GRO configuration.\n\n"
> +
> + "set gro flush (cycles)\n"
> + "Set the cycle to flush GROed packets from"
> + " reassembly tables.\n\n"
>
> "set fwd (%s)\n"
> "Set packet forwarding mode.\n\n"
> @@ -3850,115 +3853,145 @@ cmdline_parse_inst_t cmd_tunnel_tso_show
> = {
> };
>
> /* *** SET GRO FOR A PORT *** */
> -struct cmd_gro_result {
> +struct cmd_gro_enable_result {
> + cmdline_fixed_string_t cmd_set;
> + cmdline_fixed_string_t cmd_port;
> cmdline_fixed_string_t cmd_keyword;
> - cmdline_fixed_string_t mode;
> - uint8_t port_id;
> + cmdline_fixed_string_t cmd_onoff;
> + uint8_t cmd_pid;
> };
>
> static void
> -cmd_enable_gro_parsed(void *parsed_result,
> +cmd_gro_enable_parsed(void *parsed_result,
> __attribute__((unused)) struct cmdline *cl,
> __attribute__((unused)) void *data)
> {
> - struct cmd_gro_result *res;
> + struct cmd_gro_enable_result *res;
>
> res = parsed_result;
> - setup_gro(res->mode, res->port_id);
> -}
> -
> -cmdline_parse_token_string_t cmd_gro_keyword =
> - TOKEN_STRING_INITIALIZER(struct cmd_gro_result,
> + if (!strcmp(res->cmd_keyword, "gro"))
> + setup_gro(res->cmd_onoff, res->cmd_pid);
> +}
> +
> +cmdline_parse_token_string_t cmd_gro_enable_set =
> + TOKEN_STRING_INITIALIZER(struct cmd_gro_enable_result,
> + cmd_set, "set");
> +cmdline_parse_token_string_t cmd_gro_enable_port =
> + TOKEN_STRING_INITIALIZER(struct cmd_gro_enable_result,
> + cmd_keyword, "port");
> +cmdline_parse_token_num_t cmd_gro_enable_pid =
> + TOKEN_NUM_INITIALIZER(struct cmd_gro_enable_result,
> + cmd_pid, UINT8);
> +cmdline_parse_token_string_t cmd_gro_enable_keyword =
> + TOKEN_STRING_INITIALIZER(struct cmd_gro_enable_result,
>
Re: [dpdk-dev] [PATCH v4 2/5] gso: add TCP/IPv4 GSO support
> -Original Message- > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Jiayu Hu > Sent: Tuesday, September 19, 2017 3:33 PM > To: dev@dpdk.org > Cc: Ananyev, Konstantin ; Kavanagh, Mark > B ; Tan, Jianfeng ; > Yigit, Ferruh ; tho...@monjalon.net; Hu, Jiayu > > Subject: [dpdk-dev] [PATCH v4 2/5] gso: add TCP/IPv4 GSO support > > This patch adds GSO support for TCP/IPv4 packets. Supported packets > may include a single VLAN tag. TCP/IPv4 GSO doesn't check if input > packets have correct checksums, and doesn't update checksums for > output packets (the responsibility for this lies with the application). > Additionally, TCP/IPv4 GSO doesn't process IP fragmented packets. > > TCP/IPv4 GSO uses two chained MBUFs, one direct MBUF and one indrect > MBUF, to organize an output packet. Note that we refer to these two > chained MBUFs as a two-segment MBUF. The direct MBUF stores the packet > header, while the indirect mbuf simply points to a location within the > original packet's payload. Consequently, use of the GSO library requires > multi-segment MBUF support in the TX functions of the NIC driver. > > If a packet is GSOed, TCP/IPv4 GSO reduces its MBUF refcnt by 1. As a > result, when all of its GSOed segments are freed, the packet is freed > automatically. > > TCP/IPv4 GSO clears the PKT_TX_TCP_SEG flag for the input packet and > GSO segments on the event of success. > > Signed-off-by: Jiayu Hu > Signed-off-by: Mark Kavanagh Tested-by: Lei Yao This patch is test on my bench. Iperf result as following: TSO : 18 Gbps DPDK GSO: 10 Gbps No TSO/GSO: 4 Gbps > --- > doc/guides/rel_notes/release_17_11.rst | 12 ++ > lib/librte_eal/common/include/rte_log.h | 1 + > lib/librte_gso/Makefile | 2 + > lib/librte_gso/gso_common.c | 202 > > lib/librte_gso/gso_common.h | 107 + > lib/librte_gso/gso_tcp4.c | 82 + > lib/librte_gso/gso_tcp4.h | 76 > lib/librte_gso/rte_gso.c| 52 +++- > 8 files changed, 531 insertions(+), 3 deletions(-) > create mode 100644 lib/librte_gso/gso_common.c > create mode 100644 lib/librte_gso/gso_common.h > create mode 100644 lib/librte_gso/gso_tcp4.c > create mode 100644 lib/librte_gso/gso_tcp4.h > > diff --git a/doc/guides/rel_notes/release_17_11.rst > b/doc/guides/rel_notes/release_17_11.rst > index 7508be7..7453bb0 100644 > --- a/doc/guides/rel_notes/release_17_11.rst > +++ b/doc/guides/rel_notes/release_17_11.rst > @@ -41,6 +41,18 @@ New Features > Also, make sure to start the actual text at the margin. > > = > > +* **Added the Generic Segmentation Offload Library.** > + > + Added the Generic Segmentation Offload (GSO) library to enable > + applications to split large packets (e.g. MSS is 64KB) into small > + ones (e.g. MTU is 1500B). Supported packet types are: > + > + * TCP/IPv4 packets, which may include a single VLAN tag. > + > + The GSO library doesn't check if the input packets have correct > + checksums, and doesn't update checksums for output packets. > + Additionally, the GSO library doesn't process IP fragmented packets. > + > > Resolved Issues > --- > diff --git a/lib/librte_eal/common/include/rte_log.h > b/lib/librte_eal/common/include/rte_log.h > index ec8dba7..2fa1199 100644 > --- a/lib/librte_eal/common/include/rte_log.h > +++ b/lib/librte_eal/common/include/rte_log.h > @@ -87,6 +87,7 @@ extern struct rte_logs rte_logs; > #define RTE_LOGTYPE_CRYPTODEV 17 /**< Log related to cryptodev. */ > #define RTE_LOGTYPE_EFD 18 /**< Log related to EFD. */ > #define RTE_LOGTYPE_EVENTDEV 19 /**< Log related to eventdev. */ > +#define RTE_LOGTYPE_GSO 20 /**< Log related to GSO. */ > > /* these log types can be used in an application */ > #define RTE_LOGTYPE_USER1 24 /**< User-defined log type 1. */ > diff --git a/lib/librte_gso/Makefile b/lib/librte_gso/Makefile > index aeaacbc..2be64d1 100644 > --- a/lib/librte_gso/Makefile > +++ b/lib/librte_gso/Makefile > @@ -42,6 +42,8 @@ LIBABIVER := 1 > > #source files > SRCS-$(CONFIG_RTE_LIBRTE_GSO) += rte_gso.c > +SRCS-$(CONFIG_RTE_LIBRTE_GSO) += gso_common.c > +SRCS-$(CONFIG_RTE_LIBRTE_GSO) += gso_tcp4.c > > # install this header file > SYMLINK-$(CONFIG_RTE_LIBRTE_GSO)-include += rte_gso.h > diff --git a/lib/librte_gso/gso_common.c b/lib/librte_gso/gso_common.c > new file mode 100644 > index 000..b2c84f6 > --- /dev/null > +++ b/lib/librte_gso/gso_common.c > @@ -0,0 +1,202 @@ > +/*- > + * BSD LICENSE > + * > + * Copyright(c) 2017 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
[dpdk-dev] [PATCH] examples/performance-thread: fix out-of-bounds read
Overrunning array schedcore of 128 8-byte elements at element index 128
using index lcore_id.
Fixed by correct check index lcoreid condition and
change type of lcoreid to unsigned.
Coverity issue: 143459
Fixes: 116819b9ed0d ("examples/performance-thread: add lthread subsystem")
Cc: ian.be...@intel.com
Cc: sta...@dpdk.org
Signed-off-by: Slawomir Mrozowicz
---
examples/performance-thread/common/lthread.h | 2 +-
examples/performance-thread/common/lthread_sched.c | 11 +++
2 files changed, 8 insertions(+), 5 deletions(-)
diff --git a/examples/performance-thread/common/lthread.h
b/examples/performance-thread/common/lthread.h
index 5c2c1a5f0..0cde5919b 100644
--- a/examples/performance-thread/common/lthread.h
+++ b/examples/performance-thread/common/lthread.h
@@ -87,7 +87,7 @@ int _lthread_desched_sleep(struct lthread *lt);
void _lthread_free(struct lthread *lt);
-struct lthread_sched *_lthread_sched_get(int lcore_id);
+struct lthread_sched *_lthread_sched_get(unsigned int lcore_id);
struct lthread_stack *_stack_alloc(void);
diff --git a/examples/performance-thread/common/lthread_sched.c
b/examples/performance-thread/common/lthread_sched.c
index 98291478e..3484387b4 100644
--- a/examples/performance-thread/common/lthread_sched.c
+++ b/examples/performance-thread/common/lthread_sched.c
@@ -562,11 +562,14 @@ void lthread_run(void)
* Return the scheduler for this lcore
*
*/
-struct lthread_sched *_lthread_sched_get(int lcore_id)
+struct lthread_sched *_lthread_sched_get(unsigned int lcore_id)
{
- if (lcore_id > LTHREAD_MAX_LCORES)
- return NULL;
- return schedcore[lcore_id];
+ struct lthread_sched *res = NULL;
+
+ if (lcore_id < LTHREAD_MAX_LCORES)
+ res = schedcore[lcore_id];
+
+ return res;
}
/*
--
2.11.0
[dpdk-dev] [PATCH] examples/performance-thread: fix out-of-bounds read
Overrunning array schedcore of 128 8-byte elements at element index 128
using index lcoreid.
Fixed by correct check index lcoreid condition.
Coverity issue: 143461
Fixes: 116819b9ed0d ("examples/performance-thread: add lthread subsystem")
Cc: ian.be...@intel.com
Cc: sta...@dpdk.org
Signed-off-by: Slawomir Mrozowicz
---
examples/performance-thread/common/lthread_sched.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/examples/performance-thread/common/lthread_sched.c
b/examples/performance-thread/common/lthread_sched.c
index 98291478e..138a87d32 100644
--- a/examples/performance-thread/common/lthread_sched.c
+++ b/examples/performance-thread/common/lthread_sched.c
@@ -578,10 +578,9 @@ int lthread_set_affinity(unsigned lcoreid)
struct lthread *lt = THIS_LTHREAD;
struct lthread_sched *dest_sched;
- if (unlikely(lcoreid > LTHREAD_MAX_LCORES))
+ if (unlikely(lcoreid >= LTHREAD_MAX_LCORES))
return POSIX_ERRNO(EINVAL);
-
DIAG_EVENT(lt, LT_DIAG_LTHREAD_AFFINITY, lcoreid, 0);
dest_sched = schedcore[lcoreid];
--
2.11.0
Re: [dpdk-dev] [PATCH v4 0/5] net/mlx5 multi-process support
On Tue, Sep 19, 2017 at 10:31:47PM +0800, Xueming Li wrote: > This patchset enhances Mellanox multi-process by supporting all multi-process > examples, also support reading ethdev (x)stats in secondary process. > > Start from V2, this patchset depends on upstream rdma-core enhancement > and l2fork example bug fix: > http://www.dpdk.org/ml/archives/dev/2017-August/073405.html > http://www.dpdk.org/ml/archives/dev/2017-September/075568.html > > V4: > * remove forked secondary mode For the V4 series Acked-by: Nelio Laranjeiro -- Nélio Laranjeiro 6WIND
[dpdk-dev] [PATCH] examples/performance-thread: fix out-of-bounds read
Overrunning array per_lcore_this_sched->current_lthread->tls->data of
1024 8-byte elements at element index 1024 using index k.
Fixed by correct check k condition.
Coverity issue: 143462
Fixes: 116819b9ed0d ("examples/performance-thread: add lthread subsystem")
Cc: ian.be...@intel.com
Cc: sta...@dpdk.org
Signed-off-by: Slawomir Mrozowicz
---
examples/performance-thread/common/lthread_tls.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/examples/performance-thread/common/lthread_tls.c
b/examples/performance-thread/common/lthread_tls.c
index 47505f2d4..58a9a8878 100644
--- a/examples/performance-thread/common/lthread_tls.c
+++ b/examples/performance-thread/common/lthread_tls.c
@@ -212,7 +212,7 @@ void
*/
int lthread_setspecific(unsigned int k, const void *data)
{
- if (k > LTHREAD_MAX_KEYS)
+ if (k >= LTHREAD_MAX_KEYS)
return POSIX_ERRNO(EINVAL);
int n = THIS_LTHREAD->tls->nb_keys_inuse;
--
2.11.0
Re: [dpdk-dev] [PATCH 3/3] maintainers: add Ferruh for main branch
On 9/19/2017 9:23 PM, Thomas Monjalon wrote: > Ferruh will co-maintain the main branch at git://dpdk.org/dpdk. > > Signed-off-by: Thomas Monjalon Acked-by: Ferruh Yigit
[dpdk-dev] [PATCH] net/ixgbe: fix VFIO interrupt mapping in PF
When a PF port is bound to VFIO-PIC, only miscellaneous interrupt
is mapped to VFIO vector 0 in ixgbe_dev_init( ).
In ixgbe_dev_start(), if previous VFIO interrupt mapping set in
ixgbe_dev_init( ) is not cleard, it will fail when calling
rte_intr_enable( ) tries to map Rx queue interrupt to other VFIO
vectors. This patch clears the VFIO interrupt mappings before
setting both miscellaneous and Rx queue interrupt mappings again
to avoid failure.
Fixes: 0a45657a6794 ("pci: rework interrupt handling")
Cc: sta...@dpdk.org
Signed-off-by: Wei Dai
---
drivers/net/ixgbe/ixgbe_ethdev.c | 9 +
1 file changed, 9 insertions(+)
diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c
index 9ca5cbc..9030d3d 100644
--- a/drivers/net/ixgbe/ixgbe_ethdev.c
+++ b/drivers/net/ixgbe/ixgbe_ethdev.c
@@ -2692,6 +2692,15 @@ ixgbe_dev_start(struct rte_eth_dev *dev)
" no intr multiplex");
}
+ /* When a PF port is bound to VFIO-PCI only miscellaneous interrupt
+ * is mapped to VFIO vector 0 in ixgbe_dev_init( ).
+ * If previous VFIO interrupt mapping set in ixgbe_dev_init( ) is
+ * not cleared, it will fail when following rte_intr_enable( ) tries
+ * to map Rx queue interrupt to other VFIO vectors.
+ * So clear uio/vfio intr/evevnfd first to avoid failure.
+ */
+ rte_intr_disable(intr_handle);
+
/* check if rxq interrupt is enabled */
if (dev->data->dev_conf.intr_conf.rxq != 0 &&
rte_intr_dp_is_en(intr_handle))
--
2.7.5
[dpdk-dev] [PATCH] examples/performance-thread: fix out-of-bounds read
Overrunning array per_lcore_this_sched->current_lthread->tls->data of
1024 8-byte elements at element index 1024 using index k.
Fixed by correct check k condition.
Coverity issue: 143463
Fixes: 116819b9ed0d ("examples/performance-thread: add lthread subsystem")
Cc: ian.be...@intel.com
Cc: sta...@dpdk.org
Signed-off-by: Slawomir Mrozowicz
---
examples/performance-thread/common/lthread_tls.c | 7 ---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/examples/performance-thread/common/lthread_tls.c
b/examples/performance-thread/common/lthread_tls.c
index 47505f2d4..4ae3c6c03 100644
--- a/examples/performance-thread/common/lthread_tls.c
+++ b/examples/performance-thread/common/lthread_tls.c
@@ -198,11 +198,12 @@ void _lthread_tls_destroy(struct lthread *lt)
void
*lthread_getspecific(unsigned int k)
{
+ void *res = NULL;
- if (k > LTHREAD_MAX_KEYS)
- return NULL;
+ if (k < LTHREAD_MAX_KEYS)
+ res = THIS_LTHREAD->tls->data[k];
- return THIS_LTHREAD->tls->data[k];
+ return res;
}
/*
--
2.11.0
[dpdk-dev] [PATCH] service: fix compilation error
CC rte_service.o
./dpdk/lib/librte_eal/common/rte_service.c: In function
‘rte_service_start_with_defaults’:
./dpdk/lib/librte_eal/common/rte_service.c:449:9:
error: ‘ids[0]’ may be used uninitialized in this
function [-Werror=maybe-uninitialized]
ret = rte_service_map_lcore_set(i, ids[lcore_iter], 1);
cc1: all warnings being treated as errors
./dpdk/mk/internal/rte.compile-pre.mk:138:
recipe for target 'rte_service.o' failed
Signed-off-by: Yi Yang
---
lib/librte_eal/common/rte_service.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/librte_eal/common/rte_service.c
b/lib/librte_eal/common/rte_service.c
index 43716bb..e598e16 100644
--- a/lib/librte_eal/common/rte_service.c
+++ b/lib/librte_eal/common/rte_service.c
@@ -431,7 +431,7 @@ rte_service_start_with_defaults(void)
uint32_t count = rte_service_get_count();
int32_t lcore_iter = 0;
- uint32_t ids[RTE_MAX_LCORE];
+ uint32_t ids[RTE_MAX_LCORE] = {0};
int32_t lcore_count = rte_service_lcore_list(ids, RTE_MAX_LCORE);
if (lcore_count == 0)
--
2.1.0
[dpdk-dev] [PATCH] vhost: fix vhost_user_set_mem_table error
Usually vhost_user message VHOST_USER_SET_MEM_TABLE
is only sent out during initialization and only sent
once, but it isn't so for memory hotplug and hotunplug
, for that case, vhost_user message VHOST_USER_SET_MEM_TABLE
will be resent with the old memory regions (not hotunplugged)
and the new memory regions (hotplugged), so current
vhost_user_set_mem_table implementation is wrong for
memory hotplug and hotunplug case, it will free current
memory regions and unmap hugepages no matter if they
are be using or not and if they are memory regions which
have been initialized and mapped in the previous vhost_user
message VHOST_USER_SET_MEM_TABLE or not.
This commit fixed this issue very well, it will keep them
intact for those old memory region it will unmap those
memroy regions if they have been hotunplugged, it will
initialize the new hotplugged memory regions and map
hugepages if they are hotplugged.
vhost_user message VHOST_USER_SET_MEM_TABLE will include
all the current effective memory regions, the hotunplugged
memory regions won't be included in VHOST_USER_SET_MEM_TABLE,
the hotplugged memroy regions will be included in
VHOST_USER_SET_MEM_TABLE.
This has been verified in OVS DPDK by memory hotplug and
hotunplug in qemu.
Signed-off-by: Yi Yang
---
lib/librte_vhost/vhost_user.c | 72 ++-
1 file changed, 65 insertions(+), 7 deletions(-)
diff --git a/lib/librte_vhost/vhost_user.c b/lib/librte_vhost/vhost_user.c
index ad2e8d3..1c475d1 100644
--- a/lib/librte_vhost/vhost_user.c
+++ b/lib/librte_vhost/vhost_user.c
@@ -509,17 +509,43 @@ vhost_user_set_mem_table(struct virtio_net *dev, struct
VhostUserMsg *pmsg)
uint64_t mmap_size;
uint64_t mmap_offset;
uint64_t alignment;
- uint32_t i;
+ uint32_t i, j;
int fd;
- if (dev->mem) {
- free_mem_region(dev);
- rte_free(dev->mem);
- dev->mem = NULL;
+ /* Handle hot unplug case */
+ if (dev->mem && dev->mem->nregions) {
+ i = 0;
+ j = 0;
+ while ((i < memory.nregions) && (j < dev->mem->nregions)) {
+ reg = &dev->mem->regions[j];
+ /* munmap this region if it is hot unplugged */
+ if (reg->guest_user_addr
+ != memory.regions[i].userspace_addr) {
+ if (reg->host_user_addr) {
+ munmap(reg->mmap_addr, reg->mmap_size);
+ close(reg->fd);
+ }
+ reg->guest_user_addr = 0;
+ j++;
+ } else {
+ i++;
+ j++;
+ }
+ }
+
+ /* munmap these regions because they have been hot unplugged */
+ for (; j < dev->mem->nregions; j++) {
+ reg = &dev->mem->regions[j];
+ if (reg->host_user_addr) {
+ munmap(reg->mmap_addr, reg->mmap_size);
+ close(reg->fd);
+ }
+ reg->guest_user_addr = 0;
+ }
}
- dev->nr_guest_pages = 0;
if (!dev->guest_pages) {
+ dev->nr_guest_pages = 0;
dev->max_guest_pages = 8;
dev->guest_pages = malloc(dev->max_guest_pages *
sizeof(struct guest_page));
@@ -532,7 +558,7 @@ vhost_user_set_mem_table(struct virtio_net *dev, struct
VhostUserMsg *pmsg)
}
}
- dev->mem = rte_zmalloc("vhost-mem-table", sizeof(struct
rte_vhost_memory) +
+ dev->mem = rte_realloc(dev->mem, sizeof(struct rte_vhost_memory) +
sizeof(struct rte_vhost_mem_region) * memory.nregions, 0);
if (dev->mem == NULL) {
RTE_LOG(ERR, VHOST_CONFIG,
@@ -546,6 +572,38 @@ vhost_user_set_mem_table(struct virtio_net *dev, struct
VhostUserMsg *pmsg)
fd = pmsg->fds[i];
reg = &dev->mem->regions[i];
+ /* This region should be skipped if it is initialized before */
+ if (reg->guest_user_addr == memory.regions[i].userspace_addr) {
+ alignment = get_blk_size(fd);
+ if (alignment == (uint64_t)-1) {
+ RTE_LOG(ERR, VHOST_CONFIG,
+ "couldn't get hugepage size "
+ "through fstat\n");
+ goto err_mmap;
+ }
+ RTE_LOG(INFO, VHOST_CONFIG,
+ "guest memory region(old)"
+ " %u, size: 0x%" PRIx64 "\n"
+ "\t gu
[dpdk-dev] [PATCH v2 2/2] net/i40e: add support for raw flow type for flow director
When addidng flow director filter for raw flow type, instead
of constructing packet use buffer with pre-constructed packet.
Signed-off-by: Kirill Rybalchenko
---
drivers/net/i40e/i40e_fdir.c | 31 +++
1 file changed, 23 insertions(+), 8 deletions(-)
diff --git a/drivers/net/i40e/i40e_fdir.c b/drivers/net/i40e/i40e_fdir.c
index 84c0a1f..d609bd5 100644
--- a/drivers/net/i40e/i40e_fdir.c
+++ b/drivers/net/i40e/i40e_fdir.c
@@ -1093,6 +1093,7 @@ i40e_add_del_fdir_filter(struct rte_eth_dev *dev,
struct i40e_fdir_filter *fdir_filter, *node;
struct i40e_fdir_filter check_filter; /* Check if the filter exists */
int ret = 0;
+ uint16_t flow_type = filter->input.flow_type;
if (dev->data->dev_conf.fdir_conf.mode != RTE_FDIR_MODE_PERFECT) {
PMD_DRV_LOG(ERR, "FDIR is not enabled, please"
@@ -1100,7 +1101,11 @@ i40e_add_del_fdir_filter(struct rte_eth_dev *dev,
return -ENOTSUP;
}
- if (!I40E_VALID_FLOW(filter->input.flow_type)) {
+ /*
+* This check against statically defined flow types will be removed
+* once we implement dynamic mappings of flow types to pctypes
+*/
+ if (flow_type != RTE_ETH_FLOW_RAW && !I40E_VALID_FLOW(flow_type)) {
PMD_DRV_LOG(ERR, "invalid flow_type input.");
return -EINVAL;
}
@@ -1132,20 +1137,30 @@ i40e_add_del_fdir_filter(struct rte_eth_dev *dev,
memset(pkt, 0, I40E_FDIR_PKT_LEN);
- ret = i40e_fdir_construct_pkt(pf, &filter->input, pkt);
- if (ret < 0) {
- PMD_DRV_LOG(ERR, "construct packet for fdir fails.");
- return ret;
+ if (flow_type == RTE_ETH_FLOW_RAW) {
+ if (filter->input.flow.raw_flow.length > I40E_FDIR_PKT_LEN ||
+ !filter->input.flow.raw_flow.packet ||
+ !I40E_VALID_FLOW(filter->input.flow.raw_flow.flow)) {
+ PMD_DRV_LOG(ERR, "Invalid raw flow filter parameters!");
+ }
+ memcpy(pkt, filter->input.flow.raw_flow.packet,
+ filter->input.flow.raw_flow.length);
+ flow_type = filter->input.flow.raw_flow.flow;
+ } else {
+ ret = i40e_fdir_construct_pkt(pf, &filter->input, pkt);
+ if (ret < 0) {
+ PMD_DRV_LOG(ERR, "construct packet for fdir fails.");
+ return ret;
+ }
}
if (hw->mac.type == I40E_MAC_X722) {
/* get translated pctype value in fd pctype register */
pctype = (enum i40e_filter_pctype)i40e_read_rx_ctl(
hw, I40E_GLQF_FD_PCTYPES(
- (int)i40e_flowtype_to_pctype(
- filter->input.flow_type)));
+ (int)i40e_flowtype_to_pctype(flow_type)));
} else
- pctype = i40e_flowtype_to_pctype(filter->input.flow_type);
+ pctype = i40e_flowtype_to_pctype(flow_type);
ret = i40e_fdir_filter_programming(pf, pctype, filter, add);
if (ret < 0) {
--
2.5.5
[dpdk-dev] [PATCH v2 1/2] ethdev: add support for raw flow type for flow director
Add new structure rte_eth_raw_flow to the union rte_eth_fdir_flow
to support filter for raw flow type
Signed-off-by: Kirill Rybalchenko
---
lib/librte_ether/rte_eth_ctrl.h | 10 ++
1 file changed, 10 insertions(+)
diff --git a/lib/librte_ether/rte_eth_ctrl.h b/lib/librte_ether/rte_eth_ctrl.h
index 8386904..213b408 100644
--- a/lib/librte_ether/rte_eth_ctrl.h
+++ b/lib/librte_ether/rte_eth_ctrl.h
@@ -525,6 +525,15 @@ struct rte_eth_tunnel_flow {
};
/**
+ * A structure used to define the input for raw flow
+ */
+struct rte_eth_raw_flow {
+ uint16_t flow; /**< flow type. */
+ void *packet;/**< pre-constructed packet buffer. */
+ uint16_t length; /**< buffer length. */
+};
+
+/**
* An union contains the inputs for all types of flow
* Items in flows need to be in big endian
*/
@@ -540,6 +549,7 @@ union rte_eth_fdir_flow {
struct rte_eth_ipv6_flow ipv6_flow;
struct rte_eth_mac_vlan_flow mac_vlan_flow;
struct rte_eth_tunnel_flow tunnel_flow;
+ struct rte_eth_raw_flowraw_flow;
};
/**
--
2.5.5
[dpdk-dev] [PATCH v2 0/2] ethdev: add support for raw flow type for flow director
For complex packets use raw flow type with pre-constructed packet buffer instead of creating a packet internally in PMD. v2: Fixed code style, comment added Kirill Rybalchenko (2): ethdev: add support for raw flow type for flow director net/i40e: add support for raw flow type for flow director drivers/net/i40e/i40e_fdir.c| 31 +++ lib/librte_ether/rte_eth_ctrl.h | 10 ++ 2 files changed, 33 insertions(+), 8 deletions(-) -- 2.5.5
Re: [dpdk-dev] [PATCH] service: fix compilation error
> From: Yang, Yi Y
> Sent: Wednesday, September 20, 2017 9:31 AM
> To: Van Haaren, Harry
> Cc: dev@dpdk.org; Yang, Yi Y
> Subject: [PATCH] service: fix compilation error
>
> CC rte_service.o
> ./dpdk/lib/librte_eal/common/rte_service.c: In function
> ‘rte_service_start_with_defaults’:
> ./dpdk/lib/librte_eal/common/rte_service.c:449:9:
> error: ‘ids[0]’ may be used uninitialized in this
> function [-Werror=maybe-uninitialized]
>ret = rte_service_map_lcore_set(i, ids[lcore_iter], 1);
>
> cc1: all warnings being treated as errors
> ./dpdk/mk/internal/rte.compile-pre.mk:138:
> recipe for target 'rte_service.o' failed
>
> Signed-off-by: Yi Yang
Hi Yi,
Would you provide a more descriptive commit title
and message please? The compilation output should not be
included in the commit message, instead a short description
of what the error is, and what steps can be taken to
reproduce the error are much more helpful.
The code changes look OK to me - but I'd like to be able
to catch these issues in future, so steps to reproduce are
valuable to me.
Thanks, -Harry
> ---
> lib/librte_eal/common/rte_service.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/lib/librte_eal/common/rte_service.c
> b/lib/librte_eal/common/rte_service.c
> index 43716bb..e598e16 100644
> --- a/lib/librte_eal/common/rte_service.c
> +++ b/lib/librte_eal/common/rte_service.c
> @@ -431,7 +431,7 @@ rte_service_start_with_defaults(void)
> uint32_t count = rte_service_get_count();
>
> int32_t lcore_iter = 0;
> - uint32_t ids[RTE_MAX_LCORE];
> + uint32_t ids[RTE_MAX_LCORE] = {0};
> int32_t lcore_count = rte_service_lcore_list(ids, RTE_MAX_LCORE);
>
> if (lcore_count == 0)
> --
> 2.1.0
Re: [dpdk-dev] [PATCH v8 2/9] eal/pci: get iommu class
Hi Santosh, On 19-Sep-17 6:29 PM, santosh wrote: Hi Anatoly, On Tuesday 19 September 2017 10:07 PM, Burakov, Anatoly wrote: On 18-Sep-17 11:42 AM, Santosh Shukla wrote: Introducing rte_pci_get_iommu_class API which helps to get iommu class of PCI device on the bus and returns preferred iova mapping mode for PCI bus. Patch also add rte_pci_get_iommu_class definition for bsdapp, in bsdapp case - api returns default iova mode. Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob Reviewed-by: Maxime Coquelin --- Hi Santosh, You have probably missed my comment on previous version of this patch, but for commit history reasons i really think you should add a linuxapp stub in this commit as well as a FreeBSD stub, even though you are adding a linuxapp function in the next commit. Any linuxapp application using that function will fail to compile with this commit, despite this API being already present and declared as public. First, apologies for not following up on your note: I prefer to keep less context in each patch and for [03/9], its already has _IOVA_AS_VA flag + whole autodetection algo inside (squashed per Aron suggestion). Now if I squash [2/9] into [3/9], then would be too much info for future reader to digest for (imo). Its a kind of trade-off. On any linuxapp appl breaking with this commit: This series exposes eal api for application to use and identify iova mode. If you still feel not convinced with my explanation then I'll spin v9 and squash [02/09], [03/09] in v9. No, i don't mean squashing these two patches into one. I mean, provide a stub like for FreeBSD, and then edit it to be a proper implementation in the next commit. I.e. in this commit, add a stub that just returns 0, like for FreeBSD. Next commit, instead of starting from scratch, start from this stub. Thanks, Anatoly Thanks. -- Thanks, Anatoly
[dpdk-dev] [PATCH v4] devtools: rework abi checker script
The initial version of the script had some limitations:
- cannot work on a non-clean workspace
- environment variables are not documented
- no compilation log in case of failure
- return success even it abi is incompatible
This patch addresses these issues and rework the code.
Signed-off-by: Olivier Matz
---
v3->v4:
- clarify logs on incompatible abi
- log when an error returned an error
- [really] fix the report path
- log the output of make config in the proper file
v2->v3:
- fix when not launched from dpdk root dir
- use "-Og -Wno-error" instead of "-O0"
- fix typo in commit log
v1->v2:
- use /usr/bin/env to find bash (which is required)
- fix displayed path to html reports
- reword help for -f option
devtools/validate-abi.sh | 397 ---
1 file changed, 205 insertions(+), 192 deletions(-)
diff --git a/devtools/validate-abi.sh b/devtools/validate-abi.sh
index 0accc99b1..8caf43e83 100755
--- a/devtools/validate-abi.sh
+++ b/devtools/validate-abi.sh
@@ -1,7 +1,8 @@
-#!/bin/sh
+#!/usr/bin/env bash
# BSD LICENSE
#
# Copyright(c) 2015 Neil Horman. All rights reserved.
+# Copyright(c) 2017 6WIND S.A.
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
@@ -27,236 +28,248 @@
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
-TAG1=$1
-TAG2=$2
-TARGET=$3
-ABI_DIR=`mktemp -d -p /tmp ABI.XX`
+set -e
-usage() {
- echo "$0 "
-}
+abicheck=abi-compliance-checker
+abidump=abi-dumper
+default_dst=abi-check
+default_target=x86_64-native-linuxapp-gcc
-log() {
- local level=$1
- shift
- echo "$*"
+# trap on error
+err_report() {
+echo "$0: error at line $1"
}
+trap 'err_report $LINENO' ERR
-validate_tags() {
+print_usage () {
+ cat <<- END_OF_HELP
+ $(basename $0) [options]
- if [ -z "$HASH1" ]
- then
- echo "invalid revision: $TAG1"
- return
- fi
- if [ -z "$HASH2" ]
- then
- echo "invalid revision: $TAG2"
- return
- fi
+ This script compares the ABI of 2 git revisions of the current
+ workspace. The output is a html report and a compilation log.
+
+ The objective is to make sure that applications built against
+ DSOs from the first revision can still run when executed using
+ the DSOs built from the second revision.
+
+and are git commit id or tags.
+
+ Options:
+ -hshow this help
+ -j enable parallel compilation with threads
+ -vshow compilation logs on the console
+ -d change working directory (default is ${default_dst})
+ -tthe dpdk target to use (default is ${default_target})
+ -foverwrite existing files in destination directory
+
+ The script returns 0 on success, or the value of last failing
+ call of ${abicheck} (incompatible abi or the tool has run with errors).
+ The errors returned by ${abidump} are ignored.
+
+ END_OF_HELP
}
-validate_args() {
- if [ -z "$TAG1" ]
- then
- echo "Must Specify REV1"
- return
- fi
- if [ -z "$TAG2" ]
- then
- echo "Must Specify REV2"
- return
- fi
- if [ -z "$TARGET" ]
- then
- echo "Must Specify a build target"
+# log in the file, and on stdout if verbose
+# $1: level string
+# $2: string to be logged
+log() {
+ echo "$1: $2"
+ if [ "${verbose}" != "true" ]; then
+ echo "$1: $2" >&3
fi
}
+# launch a command and log it, taking care of surrounding spaces with quotes
+cmd() {
+ local i s whitespace ret
+ s=""
+ whitespace="[[:space:]]"
+ for i in "$@"; do
+ if [[ $i =~ $whitespace ]]; then
+ i=\"$i\"
+ fi
+ if [ -z "$s" ]; then
+ s="$i"
+ else
+ s="$s $i"
+ fi
+ done
+
+ ret=0
+ log "CMD" "$s"
+ "$@" || ret=$?
+ if [ "$ret" != "0" ]; then
+ log "CMD" "previous command returned $ret"
+ fi
+
+ return $ret
+}
-cleanup_and_exit() {
- rm -rf $ABI_DIR
- git checkout $CURRENT_BRANCH
- exit $1
+# redirect or copy stderr/stdout to a file
+# the syntax is unfamiliar, but it makes the rest of the
+# code easier to read, avoiding the use of pipes
+set_log_file() {
+ # save original stdout and stderr in fd 3 and 4
+ exec 3>&1
+ exec 4>&2
+ # create a new fd 5 that send to a file
+ exec 5> >(cat > $1)
+ # send stdout and stderr to fd 5
+ if [ "${verbose}" = "true" ]; then
+ exec 1> >(tee /dev/fd/5 >&3)
+ exec 2> >(tee /dev
Re: [dpdk-dev] [PATCH] pmd_virtio: Unchecked return value from library
Hi,
Thank you for those fixes.
> -Original Message-
> From: Basierski, SebastianX
> Sent: Tuesday, September 19, 2017 7:47 PM
> To: skh...@vmware.com
> Cc: Basierski, SebastianX; Tan, Jianfeng; dev@dpdk.org
> Subject: [PATCH] pmd_virtio: Unchecked return value from library
>
> Check return value from library in order to prevent potential fail.
>
> Coverity issue: 143439
>
> Fixes: ef53b6030039 ("net/virtio-user: support LSC")
> Cc: jianfeng@intel.com
> cc: dev@dpdk.org
>
> Signed-off-by: SebastianX Basierski
> ---
> drivers/net/virtio/virtio_user_ethdev.c | 10 --
> 1 file changed, 8 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/net/virtio/virtio_user_ethdev.c
> b/drivers/net/virtio/virtio_user_ethdev.c
> index c961444..16aa350 100644
> --- a/drivers/net/virtio/virtio_user_ethdev.c
> +++ b/drivers/net/virtio/virtio_user_ethdev.c
> @@ -86,7 +86,10 @@ virtio_user_read_dev_config(struct virtio_hw *hw,
> size_t offset,
> int flags;
>
> flags = fcntl(dev->vhostfd, F_GETFL);
> - fcntl(dev->vhostfd, F_SETFL, flags | O_NONBLOCK);
> + if (fcntl(dev->vhostfd, F_SETFL,
> + flags | O_NONBLOCK) == -1)
> + return;
Actually, even it fails, I still prefer to continue instead of "return" here.
Maybe, we can report an error message here.
Thanks,
Jianfeng
Hi,
I think, we should at least report an error.
Also returning after fcntl failed could save us from operating on corrupted
data.
Regards,
Sebastian
Intel Technology Poland sp. z o.o.
ul. Slowackiego 173 | 80-298 Gdansk | Sad Rejonowy Gdansk Polnoc | VII Wydzial
Gospodarczy Krajowego Rejestru Sadowego - KRS 101882 | NIP 957-07-52-316 |
Kapital zakladowy 200.000 PLN.
Ta wiadomosc wraz z zalacznikami jest przeznaczona dla okreslonego adresata i
moze zawierac informacje poufne. W razie przypadkowego otrzymania tej
wiadomosci, prosimy o powiadomienie nadawcy oraz trwale jej usuniecie;
jakiekolwiek
przegladanie lub rozpowszechnianie jest zabronione.
This e-mail and any attachments may contain confidential material for the sole
use of the intended recipient(s). If you are not the intended recipient, please
contact the sender and delete all copies; any review or distribution by
others is strictly prohibited.
Re: [dpdk-dev] [PATCH] pmd_virtio: Buffer not null terminated
> -Original Message-
> From: Basierski, SebastianX
> Sent: Tuesday, September 19, 2017 7:41 PM
> To: skh...@vmware.com
> Cc: Basierski, SebastianX; Tan, Jianfeng; dev@dpdk.org
> Subject: [PATCH] pmd_virtio: Buffer not null terminated
>
> Fix calling strncpy with the a maximum size equal of destination array
> size.
>
> Coverity issue: 140732
>
> Fixes: e3b434818bbb ("net/virtio-user: support kernel vhost")
> Cc: jianfeng@intel.com
> cc: dev@dpdk.org
You want to Cc sta...@dpdk.org?
>
> Signed-off-by: SebastianX Basierski
Besides the above nit,
Acked-by: Jianfeng Tan
Hi,
I don't understand.
Are You asking me to send this patch as cc to sta...@dpdk.org, but without
lines below?
>
> Signed-off-by: SebastianX Basierski
Intel Technology Poland sp. z o.o.
ul. Slowackiego 173 | 80-298 Gdansk | Sad Rejonowy Gdansk Polnoc | VII Wydzial
Gospodarczy Krajowego Rejestru Sadowego - KRS 101882 | NIP 957-07-52-316 |
Kapital zakladowy 200.000 PLN.
Ta wiadomosc wraz z zalacznikami jest przeznaczona dla okreslonego adresata i
moze zawierac informacje poufne. W razie przypadkowego otrzymania tej
wiadomosci, prosimy o powiadomienie nadawcy oraz trwale jej usuniecie;
jakiekolwiek
przegladanie lub rozpowszechnianie jest zabronione.
This e-mail and any attachments may contain confidential material for the sole
use of the intended recipient(s). If you are not the intended recipient, please
contact the sender and delete all copies; any review or distribution by
others is strictly prohibited.
Re: [dpdk-dev] [PATCH v5] net/i40e: fix mirror rule reset when port is stopped
Hi Wei,
>
> When an i40e PF port is stopped, all mirror rules should be removed.
> All rule related software and hardware resources should also be
> removed.
Could you clarify why we have to remove all mirror rules when PF is stopped?
As I remember mirror rule can direct to VF, which still can be running, no?
Konstantin
>
> Fixes: a4def5edf0fc ("i40e: enable port mirroring")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Wei Dai
> Tested-by: Lijuan Tu
> ---
> drivers/net/i40e/i40e_ethdev.c | 21 -
> 1 file changed, 20 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/net/i40e/i40e_ethdev.c b/drivers/net/i40e/i40e_ethdev.c
> index f12aefa..14cf6c0 100644
> --- a/drivers/net/i40e/i40e_ethdev.c
> +++ b/drivers/net/i40e/i40e_ethdev.c
> @@ -361,6 +361,12 @@ static int i40e_dev_sync_phy_type(struct i40e_hw *hw);
> static void i40e_configure_registers(struct i40e_hw *hw);
> static void i40e_hw_init(struct rte_eth_dev *dev);
> static int i40e_config_qinq(struct i40e_hw *hw, struct i40e_vsi *vsi);
> +static enum i40e_status_code i40e_aq_del_mirror_rule(struct i40e_hw *hw,
> + uint16_t seid,
> + uint16_t rule_type,
> + uint16_t *entries,
> + uint16_t count,
> + uint16_t rule_id);
> static int i40e_mirror_rule_set(struct rte_eth_dev *dev,
> struct rte_eth_mirror_conf *mirror_conf,
> uint8_t sw_id, uint8_t on);
> @@ -2069,6 +2075,7 @@ i40e_dev_stop(struct rte_eth_dev *dev)
> struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev);
> struct rte_intr_handle *intr_handle = &pci_dev->intr_handle;
> int i;
> + int ret;
>
> if (hw->adapter_stopped == 1)
> return;
> @@ -2096,10 +2103,22 @@ i40e_dev_stop(struct rte_eth_dev *dev)
>
> /* Remove all mirror rules */
> while ((p_mirror = TAILQ_FIRST(&pf->mirror_list))) {
> + ret = i40e_aq_del_mirror_rule(hw,
> + pf->main_vsi->veb->seid,
> + p_mirror->rule_type,
> + p_mirror->entries,
> + p_mirror->num_entries,
> + p_mirror->id);
> + if (ret < 0)
> + PMD_DRV_LOG(ERR, "failed to remove mirror rule: "
> + "status = %d, aq_err = %d.", ret,
> + hw->aq.asq_last_status);
> +
> + /* remove mirror software resource any way */
> TAILQ_REMOVE(&pf->mirror_list, p_mirror, rules);
> rte_free(p_mirror);
> + pf->nb_mirror_rule--;
> }
> - pf->nb_mirror_rule = 0;
>
> if (!rte_intr_allow_others(intr_handle))
> /* resume to the default handler */
> --
> 2.7.5
[dpdk-dev] [PATCH] net/virtio: fix of untrusted scalar value
The unscrutinized value may be incorrectly assumed to be within a certain
range by later operations.
In vhost_user_read: An unscrutinized value from an untrusted source used
in a trusted context - the value of sz_payload may be harmfull and we need
limit them to the max value of payload.
Fixes: 6a84c37e3975 ("net/virtio-user: add vhost-user adapter layer")
Cc: jianfeng@intel.com
Signed-off-by: Daniel Mrzyglod
---
drivers/net/virtio/virtio_user/vhost_user.c | 4
1 file changed, 4 insertions(+)
diff --git a/drivers/net/virtio/virtio_user/vhost_user.c
b/drivers/net/virtio/virtio_user/vhost_user.c
index 4ad7b21..b490336 100644
--- a/drivers/net/virtio/virtio_user/vhost_user.c
+++ b/drivers/net/virtio/virtio_user/vhost_user.c
@@ -130,6 +130,10 @@ vhost_user_read(int fd, struct vhost_user_msg *msg)
}
sz_payload = msg->size;
+
+ if (sz_payload > sizeof(msg->payload))
+ goto fail;
+
if (sz_payload) {
ret = recv(fd, (void *)((char *)msg + sz_hdr), sz_payload, 0);
if (ret < sz_payload) {
--
2.7.4
Re: [dpdk-dev] [PATCH] net/virtio: fix of untrusted scalar value
This patch is for Coverity issue: 139601 Does it need v2 for updating Message Body ?
[dpdk-dev] [PATCH v3 0/3] Extend DES support
This patchset extends the support of DES algorithms to the AESNI MB and OpenSSL PMDs. For AESNI MB PMD, the following algorithms are added: - DES-CBC - DES-DOCSISBPI For OpenSSL PMD, DES-CBC is added, as DES-DOCSISBPI was already supported. Changes in v3: - Added release note for support for DES-CBC on OpenSSL PMD - Addressed Radu's comments Pablo de Lara (3): crypto/aesni_mb: add DES support crypto/openssl: add support for DES-CBC app/crypto-perf: fix packet length check app/test-crypto-perf/cperf_options_parsing.c | 18 +-- doc/guides/cryptodevs/aesni_mb.rst | 2 + doc/guides/cryptodevs/features/aesni_mb.ini| 3 ++ doc/guides/rel_notes/release_17_11.rst | 13 + drivers/crypto/aesni_mb/rte_aesni_mb_pmd.c | 72 ++ drivers/crypto/aesni_mb/rte_aesni_mb_pmd_ops.c | 42 +++ drivers/crypto/openssl/rte_openssl_pmd.c | 27 ++ drivers/crypto/openssl/rte_openssl_pmd_ops.c | 20 +++ test/test/test_cryptodev.c | 63 ++ test/test/test_cryptodev_des_test_vectors.h| 26 +++--- 10 files changed, 252 insertions(+), 34 deletions(-) -- 2.9.4
[dpdk-dev] [PATCH v3 1/3] crypto/aesni_mb: add DES support
The Multi-buffer library now supports DES-CBC
and DES-DOCSISBPI algorithms, so this commit
extends adds support for them in the PMD.
Signed-off-by: Pablo de Lara
Acked-by: Fan Zhang
Reviewed-by: Radu Nicolau
---
doc/guides/cryptodevs/aesni_mb.rst | 2 +
doc/guides/cryptodevs/features/aesni_mb.ini| 3 ++
doc/guides/rel_notes/release_17_11.rst | 7 +++
drivers/crypto/aesni_mb/rte_aesni_mb_pmd.c | 72 ++
drivers/crypto/aesni_mb/rte_aesni_mb_pmd_ops.c | 42 +++
test/test/test_cryptodev.c | 42 +++
test/test/test_cryptodev_des_test_vectors.h| 24 ++---
7 files changed, 161 insertions(+), 31 deletions(-)
diff --git a/doc/guides/cryptodevs/aesni_mb.rst
b/doc/guides/cryptodevs/aesni_mb.rst
index b3b937f..2e07e1e 100644
--- a/doc/guides/cryptodevs/aesni_mb.rst
+++ b/doc/guides/cryptodevs/aesni_mb.rst
@@ -52,6 +52,8 @@ Cipher algorithms:
* RTE_CRYPTO_CIPHER_AES192_CTR
* RTE_CRYPTO_CIPHER_AES256_CTR
* RTE_CRYPTO_CIPHER_AES_DOCSISBPI
+* RTE_CRYPTO_CIPHER_DES_CBC
+* RTE_CRYPTO_CIPHER_DES_DOCSISBPI
Hash algorithms:
diff --git a/doc/guides/cryptodevs/features/aesni_mb.ini
b/doc/guides/cryptodevs/features/aesni_mb.ini
index 03d8485..fab07cb 100644
--- a/doc/guides/cryptodevs/features/aesni_mb.ini
+++ b/doc/guides/cryptodevs/features/aesni_mb.ini
@@ -23,6 +23,9 @@ AES CTR (128) = Y
AES CTR (192) = Y
AES CTR (256) = Y
AES DOCSIS BPI = Y
+DES CBC= Y
+DES DOCSIS BPI = Y
+
;
; Supported authentication algorithms of the 'aesni_mb' crypto driver.
;
diff --git a/doc/guides/rel_notes/release_17_11.rst
b/doc/guides/rel_notes/release_17_11.rst
index 207c623..1ddc3cf 100644
--- a/doc/guides/rel_notes/release_17_11.rst
+++ b/doc/guides/rel_notes/release_17_11.rst
@@ -49,6 +49,13 @@ New Features
* Coalesce writes to HEAD CSR on response processing.
* Coalesce writes to TAIL CSR on request processing.
+* **Updated the AESNI MB PMD.**
+
+ The AESNI MB PMD has been updated with additional support for:
+
+ * DES CBC algorithm.
+ * DES DOCSIS BPI algorithm.
+
Resolved Issues
---
diff --git a/drivers/crypto/aesni_mb/rte_aesni_mb_pmd.c
b/drivers/crypto/aesni_mb/rte_aesni_mb_pmd.c
index f91504d..49242fc 100644
--- a/drivers/crypto/aesni_mb/rte_aesni_mb_pmd.c
+++ b/drivers/crypto/aesni_mb/rte_aesni_mb_pmd.c
@@ -30,6 +30,8 @@
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
+#include
+
#include
#include
#include
@@ -188,6 +190,7 @@ aesni_mb_set_session_cipher_parameters(const struct
aesni_mb_op_fns *mb_ops,
struct aesni_mb_session *sess,
const struct rte_crypto_sym_xform *xform)
{
+ uint8_t is_aes = 0;
aes_keyexp_t aes_keyexp_fn;
if (xform == NULL) {
@@ -217,45 +220,68 @@ aesni_mb_set_session_cipher_parameters(const struct
aesni_mb_op_fns *mb_ops,
switch (xform->cipher.algo) {
case RTE_CRYPTO_CIPHER_AES_CBC:
sess->cipher.mode = CBC;
+ is_aes = 1;
break;
case RTE_CRYPTO_CIPHER_AES_CTR:
sess->cipher.mode = CNTR;
+ is_aes = 1;
break;
case RTE_CRYPTO_CIPHER_AES_DOCSISBPI:
sess->cipher.mode = DOCSIS_SEC_BPI;
+ is_aes = 1;
break;
- default:
- MB_LOG_ERR("Unsupported cipher mode parameter");
- return -ENOTSUP;
- }
-
- /* Check key length and choose key expansion function */
- switch (xform->cipher.key.length) {
- case AES_128_BYTES:
- sess->cipher.key_length_in_bytes = AES_128_BYTES;
- aes_keyexp_fn = mb_ops->aux.keyexp.aes128;
+ case RTE_CRYPTO_CIPHER_DES_CBC:
+ sess->cipher.mode = DES;
break;
- case AES_192_BYTES:
- sess->cipher.key_length_in_bytes = AES_192_BYTES;
- aes_keyexp_fn = mb_ops->aux.keyexp.aes192;
- break;
- case AES_256_BYTES:
- sess->cipher.key_length_in_bytes = AES_256_BYTES;
- aes_keyexp_fn = mb_ops->aux.keyexp.aes256;
+ case RTE_CRYPTO_CIPHER_DES_DOCSISBPI:
+ sess->cipher.mode = DOCSIS_DES;
break;
default:
- MB_LOG_ERR("Invalid cipher key length");
- return -EINVAL;
+ MB_LOG_ERR("Unsupported cipher mode parameter");
+ return -ENOTSUP;
}
/* Set IV parameters */
sess->iv.offset = xform->cipher.iv.offset;
sess->iv.length = xform->cipher.iv.length;
- /* Expanded cipher keys */
- (*aes_keyexp_fn)(xform->cipher.key.data,
- sess->cipher.expanded_aes_keys.encode,
- sess->cipher.expanded_aes_keys.decode);
+ /* Check key length and choose key expansion function for AES */
+ if (is_aes) {
+ switch (x
Re: [dpdk-dev] [PATCH] net/virtio: fix of untrusted scalar value
Hi,
> -Original Message-
> From: Mrzyglod, DanielX T
> Sent: Wednesday, September 20, 2017 5:19 PM
> To: y...@fridaylinux.org
> Cc: dev@dpdk.org; Mrzyglod, DanielX T; Tan, Jianfeng
> Subject: [PATCH] net/virtio: fix of untrusted scalar value
>
> The unscrutinized value may be incorrectly assumed to be within a certain
> range by later operations.
>
> In vhost_user_read: An unscrutinized value from an untrusted source used
> in a trusted context - the value of sz_payload may be harmfull and we need
> limit them to the max value of payload.
Please add below line.
Coverity issue: 139601
>
> Fixes: 6a84c37e3975 ("net/virtio-user: add vhost-user adapter layer")
> Cc: jianfeng@intel.com
>
> Signed-off-by: Daniel Mrzyglod
Acked-by: Jianfeng Tan
Thanks,
Jianfeng
[dpdk-dev] [PATCH v3 3/3] app/crypto-perf: fix packet length check
When using DES-CBC, packet size has to be multiple
of 8 bytes, but if a list of packets is provided.
the check was not correct.
Fixes: fc4600fb2520 ("app/crypto-perf: add extra option checks")
Cc: sta...@dpdk.org
Signed-off-by: Pablo de Lara
Acked-by: Radu Nicolau
---
app/test-crypto-perf/cperf_options_parsing.c | 18 +++---
1 file changed, 15 insertions(+), 3 deletions(-)
diff --git a/app/test-crypto-perf/cperf_options_parsing.c
b/app/test-crypto-perf/cperf_options_parsing.c
index 663f53f..3f933bc 100644
--- a/app/test-crypto-perf/cperf_options_parsing.c
+++ b/app/test-crypto-perf/cperf_options_parsing.c
@@ -832,14 +832,26 @@ check_cipher_buffer_length(struct cperf_options *options)
if (options->cipher_algo == RTE_CRYPTO_CIPHER_DES_CBC ||
options->cipher_algo == RTE_CRYPTO_CIPHER_3DES_CBC ||
options->cipher_algo == RTE_CRYPTO_CIPHER_3DES_ECB) {
- for (buffer_size = options->min_buffer_size;
- buffer_size < options->max_buffer_size;
- buffer_size += options->inc_buffer_size) {
+ if (options->inc_buffer_size != 0)
+ buffer_size = options->min_buffer_size;
+ else
+ buffer_size = options->buffer_size_list[0];
+
+ while (buffer_size <= options->max_buffer_size) {
if ((buffer_size % DES_BLOCK_SIZE) != 0) {
RTE_LOG(ERR, USER1, "Some of the buffer sizes
are "
"not suitable for the algorithm
selected\n");
return -EINVAL;
}
+
+ if (options->inc_buffer_size != 0)
+ buffer_size += options->inc_buffer_size;
+ else {
+ if (++buffer_size_idx ==
options->buffer_size_count)
+ break;
+ buffer_size =
options->buffer_size_list[buffer_size_idx];
+ }
+
}
}
--
2.9.4
[dpdk-dev] [PATCH v3 2/3] crypto/openssl: add support for DES-CBC
Signed-off-by: Pablo de Lara
Acked-by: Radu Nicolau
---
doc/guides/rel_notes/release_17_11.rst | 6 ++
drivers/crypto/openssl/rte_openssl_pmd.c | 27 +++
drivers/crypto/openssl/rte_openssl_pmd_ops.c | 20
test/test/test_cryptodev.c | 21 +
test/test/test_cryptodev_des_test_vectors.h | 6 --
5 files changed, 78 insertions(+), 2 deletions(-)
diff --git a/doc/guides/rel_notes/release_17_11.rst
b/doc/guides/rel_notes/release_17_11.rst
index 1ddc3cf..683cc22 100644
--- a/doc/guides/rel_notes/release_17_11.rst
+++ b/doc/guides/rel_notes/release_17_11.rst
@@ -56,6 +56,12 @@ New Features
* DES CBC algorithm.
* DES DOCSIS BPI algorithm.
+* **Updated the OpenSSL PMD.**
+
+ The OpenSSL PMD has been updated with additional support for:
+
+ * DES CBC algorithm.
+
Resolved Issues
---
diff --git a/drivers/crypto/openssl/rte_openssl_pmd.c
b/drivers/crypto/openssl/rte_openssl_pmd.c
index 438535e..926caa6 100644
--- a/drivers/crypto/openssl/rte_openssl_pmd.c
+++ b/drivers/crypto/openssl/rte_openssl_pmd.c
@@ -336,6 +336,33 @@ openssl_set_session_cipher_parameters(struct
openssl_session *sess,
sess->cipher.key.data) != 0)
return -EINVAL;
break;
+
+ case RTE_CRYPTO_CIPHER_DES_CBC:
+ sess->cipher.algo = xform->cipher.algo;
+ sess->cipher.ctx = EVP_CIPHER_CTX_new();
+ sess->cipher.evp_algo = EVP_des_cbc();
+
+ get_cipher_key(xform->cipher.key.data, sess->cipher.key.length,
+ sess->cipher.key.data);
+ if (sess->cipher.direction == RTE_CRYPTO_CIPHER_OP_ENCRYPT) {
+ if (EVP_EncryptInit_ex(sess->cipher.ctx,
+ sess->cipher.evp_algo,
+ NULL, xform->cipher.key.data,
+ NULL) != 1) {
+ return -EINVAL;
+ }
+ } else if (sess->cipher.direction ==
+ RTE_CRYPTO_CIPHER_OP_DECRYPT) {
+ if (EVP_DecryptInit_ex(sess->cipher.ctx,
+ sess->cipher.evp_algo,
+ NULL, xform->cipher.key.data,
+ NULL) != 1) {
+ return -EINVAL;
+ }
+ }
+
+ break;
+
case RTE_CRYPTO_CIPHER_DES_DOCSISBPI:
sess->cipher.algo = xform->cipher.algo;
sess->chain_order = OPENSSL_CHAIN_CIPHER_BPI;
diff --git a/drivers/crypto/openssl/rte_openssl_pmd_ops.c
b/drivers/crypto/openssl/rte_openssl_pmd_ops.c
index 8cdd0b2..a197f79 100644
--- a/drivers/crypto/openssl/rte_openssl_pmd_ops.c
+++ b/drivers/crypto/openssl/rte_openssl_pmd_ops.c
@@ -427,6 +427,26 @@ static const struct rte_cryptodev_capabilities
openssl_pmd_capabilities[] = {
}, }
}, }
},
+ { /* DES CBC */
+ .op = RTE_CRYPTO_OP_TYPE_SYMMETRIC,
+ {.sym = {
+ .xform_type = RTE_CRYPTO_SYM_XFORM_CIPHER,
+ {.cipher = {
+ .algo = RTE_CRYPTO_CIPHER_DES_CBC,
+ .block_size = 8,
+ .key_size = {
+ .min = 8,
+ .max = 8,
+ .increment = 0
+ },
+ .iv_size = {
+ .min = 8,
+ .max = 8,
+ .increment = 0
+ }
+ }, }
+ }, }
+ },
{ /* DES DOCSIS BPI */
.op = RTE_CRYPTO_OP_TYPE_SYMMETRIC,
{.sym = {
diff --git a/test/test/test_cryptodev.c b/test/test/test_cryptodev.c
index c282f4d..0557152 100644
--- a/test/test/test_cryptodev.c
+++ b/test/test/test_cryptodev.c
@@ -4672,6 +4672,25 @@ test_DES_cipheronly_qat_all(void)
}
static int
+test_DES_cipheronly_openssl_all(void)
+{
+ struct crypto_testsuite_params *ts_params = &testsuite_params;
+ int status;
+
+ status = test_blockcipher_all_tests(ts_params->mbuf_pool,
+ ts_params->op_mpool,
+ ts_params->session_mpool,
+ ts_params->valid_devs[0],
+ rte_cryptodev_driver_id_get(
+ RTE_STR(CRYPTODEV_NAME_OPENSSL_PMD)),
+ BLKCIPHER_DES_CIPHERONLY_TYPE);
+
+ TEST_ASSERT_EQUAL(status, 0, "Test failed");
+
+ return TEST_SUCCESS;
+}
+
+static int
test_DES_docsis_openssl_all(void)
{
struct crypto_testsui
Re: [dpdk-dev] [PATCH] pmd_virtio: Buffer not null terminated
> -Original Message- > From: Basierski, SebastianX > Sent: Wednesday, September 20, 2017 5:20 PM > To: Tan, Jianfeng; skh...@vmware.com > Cc: dev@dpdk.org > Subject: RE: [PATCH] pmd_virtio: Buffer not null terminated > ... > > > Hi, > > I don't understand. > Are You asking me to send this patch as cc to sta...@dpdk.org, but without > lines below? Nope. We need to cc sta...@dpdk.org for bug fixes.
Re: [dpdk-dev] [PATCH 05/11] lib/librte_mbuf: add security crypto flags and mbuf fields
Hi Boris,
Some comments inline.
On Mon, Sep 18, 2017 at 07:54:03AM +, Boris Pismenny wrote:
> Hi Olivier,
>
> On 9/14/2017 11:27 AM, Akhil Goyal wrote:
> >
> > From: Boris Pismenny
> >
> > add security crypto flags and update mbuf fields to support
> > IPsec crypto offload for transmitted packets, and to indicate
> > crypto result for received packets.
> >
> > Signed-off-by: Aviad Yehezkel
> > Signed-off-by: Boris Pismenny
> > Signed-off-by: Radu Nicolau
> > ---
> > lib/librte_mbuf/rte_mbuf.c | 6 ++
> > lib/librte_mbuf/rte_mbuf.h | 32 +---
> > 2 files changed, 35 insertions(+), 3 deletions(-)
> >
> > diff --git a/lib/librte_mbuf/rte_mbuf.c b/lib/librte_mbuf/rte_mbuf.c
> > index 26a62b8..bbd42a6 100644
> > --- a/lib/librte_mbuf/rte_mbuf.c
> > +++ b/lib/librte_mbuf/rte_mbuf.c
> > @@ -323,6 +323,8 @@ const char *rte_get_rx_ol_flag_name(uint64_t mask)
> > case PKT_RX_QINQ_STRIPPED: return "PKT_RX_QINQ_STRIPPED";
> > case PKT_RX_LRO: return "PKT_RX_LRO";
> > case PKT_RX_TIMESTAMP: return "PKT_RX_TIMESTAMP";
> > + case PKT_RX_SEC_OFFLOAD: return "PKT_RX_SECURITY_OFFLOAD";
> > + case PKT_RX_SEC_OFFLOAD_FAILED: return
> > "PKT_RX_SECURITY_OFFLOAD_FAILED";
I think the string should be the same than the macro.
(SEC vs SECURITY)
> ...
> > +/**
> > + * Indicate that security offload processing was applied on the RX packet.
> > + */
> > +#define PKT_RX_SEC_OFFLOAD (1ULL << 18)
> > +
> > +/**
> > + * Indicate that security offload processing failed on the RX packet.
> > + */
> > +#define PKT_RX_SEC_OFFLOAD_FAILED (1ULL << 19)
> > +
If the presence of these flags implies that some fields are
valid (ex: inner_esp_next_proto), it should be specified in
the API comments.
> ...
> > @@ -456,8 +472,18 @@ struct rte_mbuf {
> > uint32_t l3_type:4; /**< (Outer) L3 type. */
> > uint32_t l4_type:4; /**< (Outer) L4 type. */
> > uint32_t tun_type:4; /**< Tunnel type. */
> > - uint32_t inner_l2_type:4; /**< Inner L2 type. */
> > - uint32_t inner_l3_type:4; /**< Inner L3 type. */
> > + RTE_STD_C11
> > + union {
> > + uint8_t inner_esp_next_proto;
> > +
> > + __extension__
> > + struct {
> > + uint8_t inner_l2_type:4;
> > + /**< Inner L2 type. */
> > + uint8_t inner_l3_type:4;
> > + /**< Inner L3 type. */
> > + };
> > + };
> > uint32_t inner_l4_type:4; /**< Inner L4 type. */
> > };
> > };
The (quite useless) API comment is missing. I think we should
have it for consistency.
Can you please also detail in which conditions inner_esp_next_proto is
valid, and when inner_l2/l3_type is valid?
> What do you think about this change to mbuf?
>
> It doesn't increase the mbuf size and it replaces some fields that have no
> meaning
> in IPsec encapsulations (inner L2 and L3) with a meaningful field of the
> correct
> size (inner_esp_next_proto - 8 bytes).
>
> We later use this for IPsec offload on both Tx and Rx to indicate the packet
> format.
>
Strangely, the abi-checker script finds an abi change. To me, it looks
like a false positive of abi-checker. The html output says:
Field inner_l2_type has been removed from this type.
Applications will access incorrect memory when attempting to access this
field.
Field inner_l3_type has been removed from this type.
Applications will access incorrect memory when attempting to access this
field.
[−] affected symbols: 3 (15.8%)
__rte_pktmbuf_read ( struct rte_mbuf const* m, uint32_t off, uint32_t len,
void* buf ) @@ DPDK_16.11
Field 'm.unnamed1.unnamed0' in 1st parameter 'm' (pointer) has type
'anon-struct-rte_mbuf.h-454'.
rte_mbuf_sanity_check ( struct rte_mbuf const* m, int is_header ) @@ DPDK_2.0
Field 'm.unnamed1.unnamed0' in 1st parameter 'm' (pointer) has type
'anon-struct-rte_mbuf.h-454'.
rte_pktmbuf_dump ( FILE* f, struct rte_mbuf const* m, unsigned int dump_len )
@@ DPDK_2.0
Field 'm.unnamed1.unnamed0' in 2nd parameter 'm' (pointer) has type
'anon-struct-rte_mbuf.h-454'.
If someone has a better explanation :)
You can reproduce it with the following patch:
http://dpdk.org/dev/patchwork/patch/28985/
However, with pahole, we can check that the sizes/offsets are
correct and also, the following test program behaves as expected:
#include
#include
#include
#include
#include
struct mbuf1 {
union {
uint32_t packet_type;
struct {
uint32_t l2_type:4;
uint32_t l3_type:4;
uint32_t l4_type:4;
uint32_t t
[dpdk-dev] [PATCH v3 1/5] app/testpmd: add traffic management forwarding mode
This commit extends the testpmd application with new forwarding engine
that demonstrates the use of ethdev traffic management APIs and softnic
PMD for QoS traffic management.
In this mode, 5-level hierarchical tree of the QoS scheduler is built
with the help of ethdev TM APIs such as shaper profile add/delete,
shared shaper add/update, node add/delete, hierarchy commit, etc.
The hierarchical tree has following nodes; root node(x1, level 0),
subport node(x1, level 1), pipe node(x4096, level 2),
tc node(x16348, level 3), queue node(x65536, level 4).
During runtime, each received packet is first classified by mapping the
packet fields information to 5-tuples (HQoS subport, pipe, traffic class,
queue within traffic class, and color) and storing it in the packet mbuf
sched field. After classification, each packet is sent to softnic port
which prioritizes the transmission of the received packets, and
accordingly sends them on to the output interface.
To enable traffic management mode, following testpmd command is used;
$ ./testpmd -c c -n 4 --vdev
'net_softnic0,hard_name=:06:00.1,soft_tm=on' -- -i
--forward-mode=tm
This patchset has dependency on following patch series;
http://dpdk.org/ml/archives/dev/2017-September/075655.html
http://dpdk.org/ml/archives/dev/2017-September/075656.html
http://dpdk.org/ml/archives/dev/2017-September/075657.html
http://dpdk.org/ml/archives/dev/2017-September/075658.html
Signed-off-by: Jasvinder Singh
---
v3 change:
- Implements feedback from Pablo[1]
- add flag to check required librte_sched lib and softnic pmd
- code cleanup
v2 change:
- change file name softnictm.c to tm.c
- change forward mode name to "tm"
- code clean up
[1] http://dpdk.org/ml/archives/dev/2017-September/075744.html
app/test-pmd/Makefile | 8 +
app/test-pmd/testpmd.c | 15 +
app/test-pmd/testpmd.h | 45 +++
app/test-pmd/tm.c | 863 +
4 files changed, 931 insertions(+)
create mode 100644 app/test-pmd/tm.c
diff --git a/app/test-pmd/Makefile b/app/test-pmd/Makefile
index c36be19..2c50f68 100644
--- a/app/test-pmd/Makefile
+++ b/app/test-pmd/Makefile
@@ -59,6 +59,10 @@ SRCS-y += csumonly.c
SRCS-y += icmpecho.c
SRCS-$(CONFIG_RTE_LIBRTE_IEEE1588) += ieee1588fwd.c
+ifeq ($(CONFIG_RTE_LIBRTE_PMD_SOFTNIC)$(CONFIG_RTE_LIBRTE_SCHED),yy)
+SRCS-y += tm.c
+endif
+
ifeq ($(CONFIG_RTE_BUILD_SHARED_LIB),y)
ifeq ($(CONFIG_RTE_LIBRTE_PMD_BOND),y)
@@ -81,6 +85,10 @@ ifeq ($(CONFIG_RTE_LIBRTE_PMD_XENVIRT),y)
LDLIBS += -lrte_pmd_xenvirt
endif
+ifeq ($(CONFIG_RTE_LIBRTE_PMD_SOFTNIC),y)
+LDLIBS += -lrte_pmd_softnic
+endif
+
endif
CFLAGS_cmdline.o := -D_GNU_SOURCE
diff --git a/app/test-pmd/testpmd.c b/app/test-pmd/testpmd.c
index e097ee0..d729267 100644
--- a/app/test-pmd/testpmd.c
+++ b/app/test-pmd/testpmd.c
@@ -167,6 +167,10 @@ struct fwd_engine * fwd_engines[] = {
&tx_only_engine,
&csum_fwd_engine,
&icmp_echo_engine,
+#if defined RTE_LIBRTE_PMD_SOFTNIC && defined RTE_LIBRTE_SCHED
+ &softnic_tm_engine,
+ &softnic_tm_bypass_engine,
+#endif
#ifdef RTE_LIBRTE_IEEE1588
&ieee1588_fwd_engine,
#endif
@@ -2085,6 +2089,17 @@ init_port_config(void)
(rte_eth_devices[pid].data->dev_flags &
RTE_ETH_DEV_INTR_RMV))
port->dev_conf.intr_conf.rmv = 1;
+
+#if defined RTE_LIBRTE_PMD_SOFTNIC && defined RTE_LIBRTE_SCHED
+ /* Detect softnic port */
+ if (!strcmp(port->dev_info.driver_name, "net_softnic")) {
+ port->softnic_enable = 1;
+ memset(&port->softport, 0, sizeof(struct softnic_port));
+
+ if (!strcmp(cur_fwd_eng->fwd_mode_name, "tm"))
+ port->softport.tm_flag = 1;
+ }
+#endif
}
}
diff --git a/app/test-pmd/testpmd.h b/app/test-pmd/testpmd.h
index 1d1ee75..973196d 100644
--- a/app/test-pmd/testpmd.h
+++ b/app/test-pmd/testpmd.h
@@ -84,6 +84,12 @@ typedef uint16_t streamid_t;
#define MAX_QUEUE_ID ((1 << (sizeof(queueid_t) * 8)) - 1)
+#if defined RTE_LIBRTE_PMD_SOFTNIC && defined RTE_LIBRTE_SCHED
+#define TM_MODE1
+#else
+#define TM_MODE0
+#endif
+
enum {
PORT_TOPOLOGY_PAIRED,
PORT_TOPOLOGY_CHAINED,
@@ -162,6 +168,37 @@ struct port_flow {
uint8_t data[]; /**< Storage for pattern/actions. */
};
+#ifdef TM_MODE
+/**
+ * Soft port tm related parameters
+ */
+struct softnic_port_tm {
+ uint32_t hierarchy_frozen; /**< flag for tm hierarchy status */
+
+ uint32_t n_subports_per_port; /**< Num of subport nodes per port */
+ uint32_t n_pipes_per_subport; /**< Num of pipe nodes per subport */
+
+ uint64_t tm_pktfield0_slabpos; /**< Pkt field position for subport */
+ uint64_t tm_pktfield0_slabmask; /**< Pkt field mask for the subport */
+ uint64_t tm_pktfield0_slabshr;
+
[dpdk-dev] [PATCH v3 3/5] app/test-pmd: add CLI for shaper and wred profiles
Add following CLIs in testpmd application;
- commands to add/del shaper profile for TM hieraqrchy nodes.
- commands to add/update shared shapers
- commands to add/del WRED profiles for TM hiearchy leaf nodes.
Signed-off-by: Jasvinder Singh
---
app/test-pmd/cmdline.c| 41 +++
app/test-pmd/cmdline_tm.c | 866 ++
2 files changed, 907 insertions(+)
diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
index 6012c3a..c41974b 100644
--- a/app/test-pmd/cmdline.c
+++ b/app/test-pmd/cmdline.c
@@ -653,7 +653,34 @@ static void cmd_help_long_parsed(void *parsed_result,
"ptype mapping update (port_id) (hw_ptype) (sw_ptype)\n"
"Update a ptype mapping item on a port\n\n"
+#ifdef TM_MODE
+ "add port tm node shaper profile (port_id)
(shaper_profile_id)"
+ " (tb_rate) (tb_size)\n"
+ " Add port tm node private shaper profile.\n\n"
+
+ "del port tm node shaper profile (port_id)
(shaper_profile_id)\n"
+ " Delete port tm node private shaper profile.\n\n"
+
+ "add port tm node shared shaper (port_id)
(shared_shaper_id)"
+ " (shaper_profile_id)\n"
+ " Add/update port tm node shared shaper.\n\n"
+
+ "del port tm node shared shaper (port_id)
(shared_shaper_id)\n"
+ " Delete port tm node shared shaper.\n\n"
+ "set port tm node shaper profile (port_id) (node_id)"
+ " (shaper_profile_id)\n"
+ " Set port tm node shaper profile.\n\n"
+
+ "add port tm node wred profile (port_id)
(wred_profile_id)"
+ " (color_g) (min_th_g) (max_th_g) (maxp_inv_g)
(wq_log2_g)"
+ " (color_y) (min_th_y) (max_th_y) (maxp_inv_y)
(wq_log2_y)"
+ " (color_r) (min_th_r) (max_th_r) (maxp_inv_r)
(wq_log2_r)\n"
+ " Add port tm node wred profile.\n\n"
+
+ "del port tm node wred profile (port_id)
(wred_profile_id)\n"
+ " Delete port tm node wred profile.\n\n"
+#endif
, list_pkt_forwarding_modes()
);
}
@@ -14212,6 +14239,13 @@ extern cmdline_parse_inst_t cmd_show_port_tm_level_cap;
extern cmdline_parse_inst_t cmd_show_port_tm_node_cap;
extern cmdline_parse_inst_t cmd_show_port_tm_node_type;
extern cmdline_parse_inst_t cmd_show_port_tm_node_stats;
+extern cmdline_parse_inst_t cmd_add_port_tm_node_shaper_profile;
+extern cmdline_parse_inst_t cmd_del_port_tm_node_shaper_profile;
+extern cmdline_parse_inst_t cmd_add_port_tm_node_shared_shaper;
+extern cmdline_parse_inst_t cmd_del_port_tm_node_shared_shaper;
+extern cmdline_parse_inst_t cmd_add_port_tm_node_wred_profile;
+extern cmdline_parse_inst_t cmd_del_port_tm_node_wred_profile;
+extern cmdline_parse_inst_t cmd_set_port_tm_node_shaper_profile;
#endif
/* *** */
@@ -14411,6 +14445,13 @@ cmdline_parse_ctx_t main_ctx[] = {
(cmdline_parse_inst_t *)&cmd_show_port_tm_node_cap,
(cmdline_parse_inst_t *)&cmd_show_port_tm_node_type,
(cmdline_parse_inst_t *)&cmd_show_port_tm_node_stats,
+ (cmdline_parse_inst_t *)&cmd_add_port_tm_node_shaper_profile,
+ (cmdline_parse_inst_t *)&cmd_del_port_tm_node_shaper_profile,
+ (cmdline_parse_inst_t *)&cmd_add_port_tm_node_shared_shaper,
+ (cmdline_parse_inst_t *)&cmd_del_port_tm_node_shared_shaper,
+ (cmdline_parse_inst_t *)&cmd_add_port_tm_node_wred_profile,
+ (cmdline_parse_inst_t *)&cmd_del_port_tm_node_wred_profile,
+ (cmdline_parse_inst_t *)&cmd_set_port_tm_node_shaper_profile,
#endif
NULL,
};
diff --git a/app/test-pmd/cmdline_tm.c b/app/test-pmd/cmdline_tm.c
index a20942e..7d78549 100644
--- a/app/test-pmd/cmdline_tm.c
+++ b/app/test-pmd/cmdline_tm.c
@@ -737,3 +737,869 @@ cmdline_parse_inst_t cmd_show_port_tm_node_type = {
NULL,
},
};
+
+/* *** Add Port TM Private Shaper Profile *** */
+struct cmd_add_port_tm_node_shaper_profile_result {
+ cmdline_fixed_string_t add;
+ cmdline_fixed_string_t port;
+ cmdline_fixed_string_t tm;
+ cmdline_fixed_string_t node;
+ cmdline_fixed_string_t shaper;
+ cmdline_fixed_string_t profile;
+ uint8_t port_id;
+ uint32_t shaper_id;
+ uint64_t tb_rate;
+ uint64_t tb_size;
+};
+
+cmdline_parse_token_string_t cmd_add_port_tm_node_shaper_profile_add =
+ TOKEN_STRING_INITIALIZER(
+ struct cmd_add_port_tm_node_shaper_profile_result, add, "add");
+cmdline_parse_token_string_t cmd_add_port_tm_node_shaper_profile_port =
+ TOKEN_STRING_INITIALIZER(
+ stru
[dpdk-dev] [PATCH v3 2/5] app/test-pmd: add CLI for TM capability and stats
Add following CLIs to testpmd application;
- commands to display TM capability information.
(per port, per hierarchy level and per hierarchy node)
- command to display hiearchy node type
- stats collection
Signed-off-by: Jasvinder Singh
---
v3 change:
- Implements feedback from Pablo[1]
- move TM API related CLIs into cmdline_tm.c
- split patch into small patches
- replace link status check with testpmd fn port_is_started()
[1]http://dpdk.org/ml/archives/dev/2017-September/075748.html
app/test-pmd/Makefile | 1 +
app/test-pmd/cmdline.c| 34 ++-
app/test-pmd/cmdline_tm.c | 739 ++
3 files changed, 773 insertions(+), 1 deletion(-)
create mode 100644 app/test-pmd/cmdline_tm.c
diff --git a/app/test-pmd/Makefile b/app/test-pmd/Makefile
index 2c50f68..d1252bc 100644
--- a/app/test-pmd/Makefile
+++ b/app/test-pmd/Makefile
@@ -61,6 +61,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_IEEE1588) += ieee1588fwd.c
ifeq ($(CONFIG_RTE_LIBRTE_PMD_SOFTNIC)$(CONFIG_RTE_LIBRTE_SCHED),yy)
SRCS-y += tm.c
+SRCS-y += cmdline_tm.c
endif
ifeq ($(CONFIG_RTE_BUILD_SHARED_LIB),y)
diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
index ccdf239..6012c3a 100644
--- a/app/test-pmd/cmdline.c
+++ b/app/test-pmd/cmdline.c
@@ -230,6 +230,23 @@ static void cmd_help_long_parsed(void *parsed_result,
"clear vf stats (port_id) (vf_id)\n"
"Reset a VF's statistics.\n\n"
+
+#ifdef TM_MODE
+ "show port tm cap (port_id)\n"
+ " Display the port TM capability.\n\n"
+
+ "show port tm level cap (port_id) (level_id)\n"
+ " Display the port TM hierarchical level
capability.\n\n"
+
+ "show port tm node cap (port_id) (node_id)\n"
+ " Display the port TM node capability.\n\n"
+
+ "show port tm node type (port_id) (node_id)\n"
+ " Display the port TM node type.\n\n"
+
+ "show port tm node stats (port_id) (node_id) (clear)\n"
+ " Display the port TM node stats.\n\n"
+#endif
);
}
@@ -14189,7 +14206,15 @@ cmdline_parse_inst_t cmd_load_from_file = {
},
};
-/*
*/
+#ifdef TM_MODE
+extern cmdline_parse_inst_t cmd_show_port_tm_cap;
+extern cmdline_parse_inst_t cmd_show_port_tm_level_cap;
+extern cmdline_parse_inst_t cmd_show_port_tm_node_cap;
+extern cmdline_parse_inst_t cmd_show_port_tm_node_type;
+extern cmdline_parse_inst_t cmd_show_port_tm_node_stats;
+#endif
+
+/* *** */
/* list of instructions */
cmdline_parse_ctx_t main_ctx[] = {
@@ -14380,6 +14405,13 @@ cmdline_parse_ctx_t main_ctx[] = {
(cmdline_parse_inst_t *)&cmd_ptype_mapping_replace,
(cmdline_parse_inst_t *)&cmd_ptype_mapping_reset,
(cmdline_parse_inst_t *)&cmd_ptype_mapping_update,
+#ifdef TM_MODE
+ (cmdline_parse_inst_t *)&cmd_show_port_tm_cap,
+ (cmdline_parse_inst_t *)&cmd_show_port_tm_level_cap,
+ (cmdline_parse_inst_t *)&cmd_show_port_tm_node_cap,
+ (cmdline_parse_inst_t *)&cmd_show_port_tm_node_type,
+ (cmdline_parse_inst_t *)&cmd_show_port_tm_node_stats,
+#endif
NULL,
};
diff --git a/app/test-pmd/cmdline_tm.c b/app/test-pmd/cmdline_tm.c
new file mode 100644
index 000..a20942e
--- /dev/null
+++ b/app/test-pmd/cmdline_tm.c
@@ -0,0 +1,739 @@
+/*-
+ * BSD LICENSE
+ *
+ * Copyright(c) 2017 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, PROCURE
[dpdk-dev] [PATCH v3 4/5] app/test-pmd: add CLI for TM nodes and hierarchy commit
Add following CLIs in testpmd application;
- commands to add TM hierarchy nodes (leaf and nonleaf).
- command for runtime update of node weight.
- command to commit the TM hierarchy
Signed-off-by: Jasvinder Singh
---
app/test-pmd/cmdline.c| 36 +++
app/test-pmd/cmdline_tm.c | 594 ++
2 files changed, 630 insertions(+)
diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
index c41974b..70a376a 100644
--- a/app/test-pmd/cmdline.c
+++ b/app/test-pmd/cmdline.c
@@ -680,6 +680,28 @@ static void cmd_help_long_parsed(void *parsed_result,
"del port tm node wred profile (port_id)
(wred_profile_id)\n"
" Delete port tm node wred profile.\n\n"
+
+ "add port tm nonleaf node (port_id) (node_id)
(parent_node_id)"
+ " (priority) (weight) (level_id) (shaper_profile_id)"
+ " (shared_shaper_id) (n_shared_shapers)
(n_sp_priorities)\n"
+ " Add port tm nonleaf node.\n\n"
+
+ "add port tm leaf node (port_id) (node_id)
(parent_node_id)"
+ " (priority) (weight) (level_id) (cman_mode)"
+ " (wred_profile_id)\n"
+ " Add port tm leaf node.\n\n"
+
+ "del port tm node (port_id) (node_id)\n"
+ " Delete port tm node.\n\n"
+
+#ifdef RTE_SCHED_SUBPORT_TC_OV
+ "set port tm node parent (port_id) (node_id)
(parent_node_id)"
+ " (priority) (weight)\n"
+ " Set port tm node parent.\n\n"
+#endif
+ "port tm hierarchy commit (port_id) (clean_on_fail)\n"
+ " Commit tm hierarchy.\n\n"
+
#endif
, list_pkt_forwarding_modes()
);
@@ -14246,6 +14268,13 @@ extern cmdline_parse_inst_t
cmd_del_port_tm_node_shared_shaper;
extern cmdline_parse_inst_t cmd_add_port_tm_node_wred_profile;
extern cmdline_parse_inst_t cmd_del_port_tm_node_wred_profile;
extern cmdline_parse_inst_t cmd_set_port_tm_node_shaper_profile;
+extern cmdline_parse_inst_t cmd_add_port_tm_nonleaf_node;
+extern cmdline_parse_inst_t cmd_add_port_tm_leaf_node;
+extern cmdline_parse_inst_t cmd_del_port_tm_node;
+#ifdef RTE_SCHED_SUBPORT_TC_OV
+extern cmdline_parse_inst_t cmd_set_port_tm_node_parent;
+#endif
+extern cmdline_parse_inst_t cmd_port_tm_hierarchy_commit;
#endif
/* *** */
@@ -14452,6 +14481,13 @@ cmdline_parse_ctx_t main_ctx[] = {
(cmdline_parse_inst_t *)&cmd_add_port_tm_node_wred_profile,
(cmdline_parse_inst_t *)&cmd_del_port_tm_node_wred_profile,
(cmdline_parse_inst_t *)&cmd_set_port_tm_node_shaper_profile,
+ (cmdline_parse_inst_t *)&cmd_add_port_tm_nonleaf_node,
+ (cmdline_parse_inst_t *)&cmd_add_port_tm_leaf_node,
+ (cmdline_parse_inst_t *)&cmd_del_port_tm_node,
+#ifdef RTE_SCHED_SUBPORT_TC_OV
+ (cmdline_parse_inst_t *)&cmd_set_port_tm_node_parent,
+#endif
+ (cmdline_parse_inst_t *)&cmd_port_tm_hierarchy_commit,
#endif
NULL,
};
diff --git a/app/test-pmd/cmdline_tm.c b/app/test-pmd/cmdline_tm.c
index 7d78549..b36326d 100644
--- a/app/test-pmd/cmdline_tm.c
+++ b/app/test-pmd/cmdline_tm.c
@@ -1603,3 +1603,597 @@ cmdline_parse_inst_t
cmd_set_port_tm_node_shaper_profile = {
NULL,
},
};
+
+/* *** Add Port TM nonleaf node *** */
+struct cmd_add_port_tm_nonleaf_node_result {
+ cmdline_fixed_string_t add;
+ cmdline_fixed_string_t port;
+ cmdline_fixed_string_t tm;
+ cmdline_fixed_string_t nonleaf;
+ cmdline_fixed_string_t node;
+ uint8_t port_id;
+ uint32_t node_id;
+ int32_t parent_node_id;
+ uint32_t priority;
+ uint32_t weight;
+ uint32_t level_id;
+ uint32_t shaper_profile_id;
+ uint32_t shared_shaper_id;
+ uint32_t n_shared_shapers;
+ uint32_t n_sp_priorities;
+};
+
+cmdline_parse_token_string_t cmd_add_port_tm_nonleaf_node_add =
+ TOKEN_STRING_INITIALIZER(
+ struct cmd_add_port_tm_nonleaf_node_result, add, "add");
+cmdline_parse_token_string_t cmd_add_port_tm_nonleaf_node_port =
+ TOKEN_STRING_INITIALIZER(
+ struct cmd_add_port_tm_nonleaf_node_result, port, "port");
+cmdline_parse_token_string_t cmd_add_port_tm_nonleaf_node_tm =
+ TOKEN_STRING_INITIALIZER(
+ struct cmd_add_port_tm_nonleaf_node_result, tm, "tm");
+cmdline_parse_token_string_t cmd_add_port_tm_nonleaf_node_nonleaf =
+ TOKEN_STRING_INITIALIZER(
+ struct cmd_add_port_tm_nonleaf_node_result, nonleaf, "nonleaf");
+cmdline_parse_token_string_t cmd_add_port_tm_nonleaf_node_node =
+ TOKEN_STRING_INITIALIZER(
+ struct cmd_add_port_tm_nonleaf_node_result, node, "node");
+c
[dpdk-dev] [PATCH v3 5/5] app/test-pmd: add CLI for TM packet classification
Add following CLIs in testpmd application;
- command to set the packet field mask and offset value for
classification.
- command to set traffic class translation table entry
Signed-off-by: Jasvinder Singh
---
app/test-pmd/cmdline.c| 11 ++
app/test-pmd/cmdline_tm.c | 305 ++
2 files changed, 316 insertions(+)
diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
index 70a376a..480fe4b 100644
--- a/app/test-pmd/cmdline.c
+++ b/app/test-pmd/cmdline.c
@@ -702,6 +702,13 @@ static void cmd_help_long_parsed(void *parsed_result,
"port tm hierarchy commit (port_id) (clean_on_fail)\n"
" Commit tm hierarchy.\n\n"
+ "set port tm pktfield (subport|pipe|tc) (port_id)
offset"
+ " (offset) mask (mask)\n"
+ " Set port tm packet field.\n\n"
+
+ "set port tm tc table (port_id) index (tb_index) tc
(tc_id)"
+ " queue (queue id)\n"
+ " Set port tm traffic class table entry.\n\n"
#endif
, list_pkt_forwarding_modes()
);
@@ -14275,6 +14282,8 @@ extern cmdline_parse_inst_t cmd_del_port_tm_node;
extern cmdline_parse_inst_t cmd_set_port_tm_node_parent;
#endif
extern cmdline_parse_inst_t cmd_port_tm_hierarchy_commit;
+extern cmdline_parse_inst_t cmd_set_port_tm_pktfield;
+extern cmdline_parse_inst_t cmd_set_port_tm_tc_table;
#endif
/* *** */
@@ -14488,6 +14497,8 @@ cmdline_parse_ctx_t main_ctx[] = {
(cmdline_parse_inst_t *)&cmd_set_port_tm_node_parent,
#endif
(cmdline_parse_inst_t *)&cmd_port_tm_hierarchy_commit,
+ (cmdline_parse_inst_t *)&cmd_set_port_tm_pktfield,
+ (cmdline_parse_inst_t *)&cmd_set_port_tm_tc_table,
#endif
NULL,
};
diff --git a/app/test-pmd/cmdline_tm.c b/app/test-pmd/cmdline_tm.c
index b36326d..829442f 100644
--- a/app/test-pmd/cmdline_tm.c
+++ b/app/test-pmd/cmdline_tm.c
@@ -2197,3 +2197,308 @@ cmdline_parse_inst_t cmd_port_tm_hierarchy_commit = {
NULL,
},
};
+
+static int
+port_tm_pktfield_validate_mask(uint64_t mask, uint64_t n)
+{
+ int count = __builtin_popcountll(mask);
+ int pos_lead = sizeof(uint64_t) * 8 - __builtin_clzll(mask);
+ int pos_trail = __builtin_ctzll(mask);
+ int count_expected = __builtin_popcount(n - 1);
+
+ /* Handle the exceptions */
+ if (n == 0)
+ return -1; /* Error */
+
+ if ((mask == 0) && (n == 1))
+ return 0; /* OK */
+
+ if (((mask == 0) && (n != 1)) || ((mask != 0) && (n == 1)))
+ return -2; /* Error */
+
+ /* Check that mask is contiguous */
+ if ((pos_lead - pos_trail) != count)
+ return -3; /* Error */
+
+ /* Check that mask contains the expected number of bits set */
+ if (count != count_expected)
+ return -4; /* Error */
+
+ return 0; /* OK */
+ }
+
+/* *** Set Port TM Packet Fields *** */
+struct cmd_set_port_tm_pktfield_result {
+ cmdline_fixed_string_t set;
+ cmdline_fixed_string_t port;
+ cmdline_fixed_string_t tm;
+ cmdline_fixed_string_t pktfield;
+ cmdline_fixed_string_t type;
+ uint8_t port_id;
+ cmdline_fixed_string_t offset_string;
+ uint32_t offset;
+ cmdline_fixed_string_t mask_string;
+ uint64_t mask;
+};
+
+cmdline_parse_token_string_t cmd_set_port_tm_pktfield_set =
+ TOKEN_STRING_INITIALIZER(
+ struct cmd_set_port_tm_pktfield_result, set, "set");
+cmdline_parse_token_string_t cmd_set_port_tm_pktfield_port =
+ TOKEN_STRING_INITIALIZER(
+ struct cmd_set_port_tm_pktfield_result, port, "port");
+cmdline_parse_token_string_t cmd_set_port_tm_pktfield_tm =
+ TOKEN_STRING_INITIALIZER(
+ struct cmd_set_port_tm_pktfield_result, tm, "tm");
+cmdline_parse_token_string_t cmd_set_port_tm_pktfield_pktfield =
+ TOKEN_STRING_INITIALIZER(
+ struct cmd_set_port_tm_pktfield_result,
+ pktfield, "pktfield");
+cmdline_parse_token_string_t cmd_set_port_tm_pktfield_type =
+ TOKEN_STRING_INITIALIZER(
+ struct cmd_set_port_tm_pktfield_result,
+ type, "subport#pipe#tc");
+cmdline_parse_token_num_t cmd_set_port_tm_pktfield_port_id =
+ TOKEN_NUM_INITIALIZER(
+ struct cmd_set_port_tm_pktfield_result,
+ port_id, UINT8);
+cmdline_parse_token_string_t cmd_set_port_tm_pktfield_offset_string =
+ TOKEN_STRING_INITIALIZER(
+ struct cmd_set_port_tm_pktfield_result,
+ offset_string, "offset");
+cmdline_pars
[dpdk-dev] [PATCH] crypto/openssl: fix compilation break with openssl 1.1
hmac APIs are changed in openssl version 1.1.
this patch handles both versions of openssl for hmac APIs
Fixes: d7174e8f368f ("crypto/openssl: replace evp APIs with HMAC APIs")
Signed-off-by: Akhil Goyal
---
drivers/crypto/openssl/rte_openssl_pmd.c | 27
drivers/crypto/openssl/rte_openssl_pmd_private.h | 2 +-
2 files changed, 24 insertions(+), 5 deletions(-)
diff --git a/drivers/crypto/openssl/rte_openssl_pmd.c
b/drivers/crypto/openssl/rte_openssl_pmd.c
index 438535e..280148e 100644
--- a/drivers/crypto/openssl/rte_openssl_pmd.c
+++ b/drivers/crypto/openssl/rte_openssl_pmd.c
@@ -48,6 +48,25 @@
static uint8_t cryptodev_driver_id;
+#if (OPENSSL_VERSION_NUMBER < 0x1010L)
+static HMAC_CTX *HMAC_CTX_new(void)
+{
+ HMAC_CTX *ctx = OPENSSL_malloc(sizeof(*ctx));
+
+ if (ctx != NULL)
+ HMAC_CTX_init(ctx);
+ return ctx;
+}
+
+static void HMAC_CTX_free(HMAC_CTX *ctx)
+{
+ if (ctx != NULL) {
+ HMAC_CTX_cleanup(ctx);
+ OPENSSL_free(ctx);
+ }
+}
+#endif
+
static int cryptodev_openssl_remove(struct rte_vdev_device *vdev);
/**/
@@ -437,12 +456,12 @@ openssl_set_session_auth_parameters(struct
openssl_session *sess,
case RTE_CRYPTO_AUTH_SHA384_HMAC:
case RTE_CRYPTO_AUTH_SHA512_HMAC:
sess->auth.mode = OPENSSL_AUTH_AS_HMAC;
- HMAC_CTX_init(&sess->auth.hmac.ctx);
+ sess->auth.hmac.ctx = HMAC_CTX_new();
if (get_auth_algo(xform->auth.algo,
&sess->auth.hmac.evp_algo) != 0)
return -EINVAL;
- if (HMAC_Init_ex(&sess->auth.hmac.ctx,
+ if (HMAC_Init_ex(sess->auth.hmac.ctx,
xform->auth.key.data,
xform->auth.key.length,
sess->auth.hmac.evp_algo, NULL) != 1)
@@ -585,7 +604,7 @@ openssl_reset_session(struct openssl_session *sess)
break;
case OPENSSL_AUTH_AS_HMAC:
EVP_PKEY_free(sess->auth.hmac.pkey);
- HMAC_CTX_cleanup(&sess->auth.hmac.ctx);
+ HMAC_CTX_free(sess->auth.hmac.ctx);
break;
default:
break;
@@ -1293,7 +1312,7 @@ process_openssl_auth_op(struct openssl_qp *qp, struct
rte_crypto_op *op,
case OPENSSL_AUTH_AS_HMAC:
status = process_openssl_auth_hmac(mbuf_src, dst,
op->sym->auth.data.offset, srclen,
- &sess->auth.hmac.ctx);
+ sess->auth.hmac.ctx);
break;
default:
status = -1;
diff --git a/drivers/crypto/openssl/rte_openssl_pmd_private.h
b/drivers/crypto/openssl/rte_openssl_pmd_private.h
index cbb8f8d..26bf862 100644
--- a/drivers/crypto/openssl/rte_openssl_pmd_private.h
+++ b/drivers/crypto/openssl/rte_openssl_pmd_private.h
@@ -172,7 +172,7 @@ struct openssl_session {
/**< pointer to EVP key */
const EVP_MD *evp_algo;
/**< pointer to EVP algorithm function */
- HMAC_CTX ctx;
+ HMAC_CTX *ctx;
/**< pointer to EVP context structure */
} hmac;
};
--
2.9.3
Re: [dpdk-dev] [PATCH v2 1/2] crypto/openssl: replace evp APIs with HMAC APIs
Hi Pablo, On 9/8/2017 7:33 PM, De Lara Guarch, Pablo wrote: Hi Akhil, -Original Message- From: Akhil Goyal [mailto:akhil.go...@nxp.com] Sent: Tuesday, August 29, 2017 7:59 AM To: dev@dpdk.org; De Lara Guarch, Pablo Cc: hemant.agra...@nxp.com; Doherty, Declan ; Akhil Goyal Subject: [PATCH v2 1/2] crypto/openssl: replace evp APIs with HMAC APIs in case of HMAC the openssl APIs HMAC_XXX give better performance for all HMAC cases as compared with EVP_XXX Signed-off-by: Akhil Goyal --- changes in v2: patch split in two patches as per Pablo's recommendations drivers/crypto/openssl/rte_openssl_pmd.c | 37 +--- I just come across an issue with this patch on openssl 1.1.0 (below). Unfortunately, I have already applied the patch in the subtree, but if you could send a patch to fix this, I can integrate as part of that patch. I have sent a patch to fix this. http://dpdk.org/dev/patchwork/patch/28995/ Thanks, Akhil
Re: [dpdk-dev] [PATCH v2] net/af_packet: make bypass configurable
On Tue, 2017-09-19 at 17:45 -0400, Chas Williams wrote:
> From: "Charles (Chas) Williams"
>
> In certain situations, low speed interfaces, it may be desirable to
> have the flow control provided by the kernel queueing disciplines.
>
> Signed-off-by: Chas Williams
> ---
> drivers/net/af_packet/rte_eth_af_packet.c | 22 +--
> ---
> 1 file changed, 17 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/net/af_packet/rte_eth_af_packet.c
> b/drivers/net/af_packet/rte_eth_af_packet.c
> index 7aa26e5..e3858fa 100644
> --- a/drivers/net/af_packet/rte_eth_af_packet.c
> +++ b/drivers/net/af_packet/rte_eth_af_packet.c
> @@ -59,6 +59,7 @@
> #define ETH_AF_PACKET_BLOCKSIZE_ARG "blocksz"
> #define ETH_AF_PACKET_FRAMESIZE_ARG "framesz"
> #define ETH_AF_PACKET_FRAMECOUNT_ARG "framecnt"
> +#define ETH_AF_PACKET_BYPASS_ARG "bypass"
>
> #define DFLT_BLOCK_SIZE (1 << 12)
> #define DFLT_FRAME_SIZE (1 << 11)
> @@ -115,6 +116,7 @@ static const char *valid_arguments[] = {
> ETH_AF_PACKET_BLOCKSIZE_ARG,
> ETH_AF_PACKET_FRAMESIZE_ARG,
> ETH_AF_PACKET_FRAMECOUNT_ARG,
> + ETH_AF_PACKET_BYPASS_ARG,
> NULL
> };
>
> @@ -559,6 +561,7 @@ rte_pmd_init_internals(struct rte_vdev_device
> *dev,
> unsigned int blockcnt,
> unsigned int framesize,
> unsigned int framecnt,
> + unsigned int bypass,
> struct pmd_internals **internals,
> struct rte_eth_dev **eth_dev,
> struct rte_kvargs *kvlist)
> @@ -580,9 +583,6 @@ rte_pmd_init_internals(struct rte_vdev_device
> *dev,
> #if defined(PACKET_FANOUT)
> int fanout_arg;
> #endif
> -#if defined(PACKET_QDISC_BYPASS)
> - int bypass;
> -#endif
>
> for (k_idx = 0; k_idx < kvlist->count; k_idx++) {
> pair = &kvlist->pairs[k_idx];
> @@ -698,7 +698,6 @@ rte_pmd_init_internals(struct rte_vdev_device
> *dev,
> }
>
> #if defined(PACKET_QDISC_BYPASS)
> - bypass = 1;
> rc = setsockopt(qsockfd, SOL_PACKET,
> PACKET_QDISC_BYPASS,
> &bypass, sizeof(bypass));
> if (rc == -1) {
> @@ -851,6 +850,7 @@ rte_eth_from_packet(struct rte_vdev_device *dev,
> unsigned int framesize = DFLT_FRAME_SIZE;
> unsigned int framecount = DFLT_FRAME_COUNT;
> unsigned int qpairs = 1;
> + unsigned int bypass = 1;
>
> /* do some parameter checking */
> if (*sockfd < 0)
> @@ -902,6 +902,16 @@ rte_eth_from_packet(struct rte_vdev_device *dev,
> }
> continue;
> }
> + if (strstr(pair->key, ETH_AF_PACKET_BYPASS_ARG) !=
> NULL) {
> + bypass = atoi(pair->value);
> + if (bypass > 2) {
> + RTE_LOG(ERR, PMD,
> + "%s: invalid bypass
> value\n",
> + name);
> + return -1;
> + }
> + continue;
> + }
> }
>
> if (framesize > blocksize) {
> @@ -927,6 +937,7 @@ rte_eth_from_packet(struct rte_vdev_device *dev,
> if (rte_pmd_init_internals(dev, *sockfd, qpairs,
> blocksize, blockcount,
> framesize, framecount,
> + bypass,
> &internals, ð_dev,
> kvlist) < 0)
> return -1;
> @@ -1021,4 +1032,5 @@ RTE_PMD_REGISTER_PARAM_STRING(net_af_packet,
> "qpairs= "
> "blocksz= "
> "framesz= "
> - "framecnt=");
> + "framecnt= "
> + "bypass=");
Acked-by: Luca Boccassi
--
Kind regards,
Luca Boccassi
[dpdk-dev] [PATCH] vfio: fix close unchecked file descriptor
Add file descriptor value check before calling close() function.
Coverity issue: 141297
Fixes: 811b6b25060f ("vfio: fix file descriptor leak in multi-process")
Cc: patr...@patrickmacarthur.net
Cc: sta...@dpdk.org
Signed-off-by: Kuba Kozak
---
lib/librte_eal/linuxapp/eal/eal_vfio_mp_sync.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/lib/librte_eal/linuxapp/eal/eal_vfio_mp_sync.c
b/lib/librte_eal/linuxapp/eal/eal_vfio_mp_sync.c
index 7e8095c..c04f548 100644
--- a/lib/librte_eal/linuxapp/eal/eal_vfio_mp_sync.c
+++ b/lib/librte_eal/linuxapp/eal/eal_vfio_mp_sync.c
@@ -301,7 +301,8 @@ vfio_mp_sync_thread(void __rte_unused * arg)
vfio_mp_sync_send_request(conn_sock,
SOCKET_ERR);
else
vfio_mp_sync_send_fd(conn_sock, fd);
- close(fd);
+ if (fd != -1)
+ close(fd);
break;
case SOCKET_REQ_GROUP:
/* wait for group number */
--
2.7.4
[dpdk-dev] [PATCH] acl: fix unchecked return value
Add return value check and error handling for fseek call.
Coverity issue: 143435
Fixes: 361b2e9559fc ("acl: new sample l3fwd-acl")
Cc: konstantin.anan...@intel.com
Cc: sta...@dpdk.org
Signed-off-by: Kuba Kozak
---
examples/l3fwd-acl/main.c | 7 ++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/examples/l3fwd-acl/main.c b/examples/l3fwd-acl/main.c
index 8eff4de..708e9eb 100644
--- a/examples/l3fwd-acl/main.c
+++ b/examples/l3fwd-acl/main.c
@@ -1026,6 +1026,7 @@ add_rules(const char *rule_path,
char buff[LINE_MAX];
FILE *fh = fopen(rule_path, "rb");
unsigned int i = 0;
+ int val;
if (fh == NULL)
rte_exit(EXIT_FAILURE, "%s: Open %s failed\n", __func__,
@@ -1042,7 +1043,11 @@ add_rules(const char *rule_path,
rte_exit(EXIT_FAILURE, "Not find any route entries in %s!\n",
rule_path);
- fseek(fh, 0, SEEK_SET);
+ val = fseek(fh, 0, SEEK_SET);
+ if (val < 0) {
+ rte_exit(EXIT_FAILURE, "%s: File seek operation failed\n",
+ __func__);
+ }
acl_rules = calloc(acl_num, rule_size);
--
2.7.4
Re: [dpdk-dev] [PATCH v3 01/17] build: add initial infrastructure for meson & ninja builds
On 09/13/2017 04:12 PM, Bruce Richardson wrote:
[...]
> +
> +# for static libs, treat the drivers as regular libraries, otherwise
> +# for shared libs, put them in a driver folder
> +if get_option('default_library') == 'static'
> + driver_install_path = get_option('libdir')
> +else
> + driver_install_path = join_paths(get_option('datadir'), 'dpdk/drivers')
> +endif
This is against hier, since:
"This directory contains subdirectories with specific application data,
that can be shared among different architectures of the same OS."
(from man 7 hier)
Please consider to use a subdirectory of libdir like the actual dpdk
instead.
Re: [dpdk-dev] [PATCH] vfio: fix close unchecked file descriptor
On 20-Sep-17 10:59 AM, Kuba Kozak wrote:
Add file descriptor value check before calling close() function.
Coverity issue: 141297
Fixes: 811b6b25060f ("vfio: fix file descriptor leak in multi-process")
Cc: patr...@patrickmacarthur.net
Cc: sta...@dpdk.org
Signed-off-by: Kuba Kozak
---
Acked-by: Anatoly Burakov
--
Thanks,
Anatoly
[dpdk-dev] [PATCH] linuxapp/eal: fix memory leak
In func rte_eal_init(), dynamic memory stored in 'logid' allocated
through func 'strdup' is lost when return. Fix it by freeing the memory
before return.
Signed-off-by: Yong Wang
---
lib/librte_eal/linuxapp/eal/eal.c | 19 ++-
1 file changed, 18 insertions(+), 1 deletion(-)
diff --git a/lib/librte_eal/linuxapp/eal/eal.c
b/lib/librte_eal/linuxapp/eal/eal.c
index 48f12f4..c2f9843 100644
--- a/lib/librte_eal/linuxapp/eal/eal.c
+++ b/lib/librte_eal/linuxapp/eal/eal.c
@@ -751,7 +751,7 @@ static void rte_eal_init_alert(const char *msg)
int i, fctret, ret;
pthread_t thread_id;
static rte_atomic32_t run_once = RTE_ATOMIC32_INIT(0);
- const char *logid;
+ char *logid;
char cpuset[RTE_CPU_AFFINITY_STR_LEN];
char thread_name[RTE_MAX_THREAD_NAME_LEN];
@@ -781,6 +781,7 @@ static void rte_eal_init_alert(const char *msg)
if (rte_eal_cpu_init() < 0) {
rte_eal_init_alert("Cannot detect lcores.");
rte_errno = ENOTSUP;
+ free(logid);
return -1;
}
@@ -789,6 +790,7 @@ static void rte_eal_init_alert(const char *msg)
rte_eal_init_alert("Invalid 'command line' arguments.");
rte_errno = EINVAL;
rte_atomic32_clear(&run_once);
+ free(logid);
return -1;
}
@@ -799,6 +801,7 @@ static void rte_eal_init_alert(const char *msg)
rte_eal_init_alert("Cannot get hugepage information.");
rte_errno = EACCES;
rte_atomic32_clear(&run_once);
+ free(logid);
return -1;
}
@@ -826,6 +829,7 @@ static void rte_eal_init_alert(const char *msg)
rte_eal_init_alert("Cannot init logging.");
rte_errno = ENOMEM;
rte_atomic32_clear(&run_once);
+ free(logid);
return -1;
}
@@ -834,6 +838,7 @@ static void rte_eal_init_alert(const char *msg)
rte_eal_init_alert("Cannot init VFIO\n");
rte_errno = EAGAIN;
rte_atomic32_clear(&run_once);
+ free(logid);
return -1;
}
#endif
@@ -841,6 +846,7 @@ static void rte_eal_init_alert(const char *msg)
if (rte_eal_memory_init() < 0) {
rte_eal_init_alert("Cannot init memory\n");
rte_errno = ENOMEM;
+ free(logid);
return -1;
}
@@ -850,24 +856,28 @@ static void rte_eal_init_alert(const char *msg)
if (rte_eal_memzone_init() < 0) {
rte_eal_init_alert("Cannot init memzone\n");
rte_errno = ENODEV;
+ free(logid);
return -1;
}
if (rte_eal_tailqs_init() < 0) {
rte_eal_init_alert("Cannot init tail queues for objects\n");
rte_errno = EFAULT;
+ free(logid);
return -1;
}
if (rte_eal_alarm_init() < 0) {
rte_eal_init_alert("Cannot init interrupt-handling thread\n");
/* rte_eal_alarm_init sets rte_errno on failure. */
+ free(logid);
return -1;
}
if (rte_eal_timer_init() < 0) {
rte_eal_init_alert("Cannot init HPET or TSC timers\n");
rte_errno = ENOTSUP;
+ free(logid);
return -1;
}
@@ -886,17 +896,20 @@ static void rte_eal_init_alert(const char *msg)
if (rte_eal_intr_init() < 0) {
rte_eal_init_alert("Cannot init interrupt-handling thread\n");
+ free(logid);
return -1;
}
if (eal_option_device_parse()) {
rte_errno = ENODEV;
+ free(logid);
return -1;
}
if (rte_bus_scan()) {
rte_eal_init_alert("Cannot scan the buses for devices\n");
rte_errno = ENODEV;
+ free(logid);
return -1;
}
@@ -941,6 +954,7 @@ static void rte_eal_init_alert(const char *msg)
if (ret) {
rte_eal_init_alert("rte_service_init() failed\n");
rte_errno = ENOEXEC;
+ free(logid);
return -1;
}
@@ -948,6 +962,7 @@ static void rte_eal_init_alert(const char *msg)
if (rte_bus_probe()) {
rte_eal_init_alert("Cannot probe devices\n");
rte_errno = ENOTSUP;
+ free(logid);
return -1;
}
@@ -957,10 +972,12 @@ static void rte_eal_init_alert(const char *msg)
ret = rte_service_start_with_defaults();
if (ret < 0 && ret != -ENOTSUP) {
rte_errno = ENOEXEC;
+ free(logid);
return -1;
}
rte_eal_mcfg_complete();
+ free(logid);
return fctret;
}
--
1.8.3.1
[dpdk-dev] [PATCH] net/ixgbe: fix VFIO interrupt mapping in VF
When a VF port is bound to VFIO-PIC, only miscellaneous interrupt
is mapped to VFIO vector 0 in ixgbevf_dev_init( ).
In ixgbevf_dev_start(), if previous VFIO interrupt mapping set in
ixgbevf_dev_init( ) is not cleard, it will fail when calling
rte_intr_enable( ) tries to map Rx queue interrupt to other VFIO
vectors. This patch clears the VFIO interrupt mappings before
setting both miscellaneous and Rx queue interrupt mappings again
to avoid failure.
Fixes: 77234603fba0 ("net/ixgbe: support VF mailbox interrupt for link up/down")
Cc: sta...@dpdk.org
Signed-off-by: Wei Dai
---
drivers/net/ixgbe/ixgbe_ethdev.c | 9 +
1 file changed, 9 insertions(+)
diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c
index 9ca5cbc..f7a8dc8 100644
--- a/drivers/net/ixgbe/ixgbe_ethdev.c
+++ b/drivers/net/ixgbe/ixgbe_ethdev.c
@@ -5046,6 +5046,15 @@ ixgbevf_dev_start(struct rte_eth_dev *dev)
}
ixgbevf_configure_msix(dev);
+ /* When a PF port is bound to VFIO-PCI, only miscellaneous interrupt
+* is mapped to VFIO vector 0 in ixgbe_dev_init( ).
+* If previous VFIO interrupt mapping set in ixgbe_dev_init( ) is
+* not cleared, it will fail when following rte_intr_enable( ) tries
+* to map Rx queue interrupt to other VFIO vectors.
+* So clear uio/vfio intr/evevnfd first to avoid failure.
+*/
+ rte_intr_disable(intr_handle);
+
rte_intr_enable(intr_handle);
/* Re-enable interrupt for VF */
--
2.7.5
Re: [dpdk-dev] [PATCH v3 01/17] build: add initial infrastructure for meson & ninja builds
On Wed, Sep 20, 2017 at 12:10:19PM +0200, Timothy M. Redaelli wrote:
> On 09/13/2017 04:12 PM, Bruce Richardson wrote:
> [...]
> > +
> > +# for static libs, treat the drivers as regular libraries, otherwise
> > +# for shared libs, put them in a driver folder
> > +if get_option('default_library') == 'static'
> > + driver_install_path = get_option('libdir')
> > +else
> > + driver_install_path = join_paths(get_option('datadir'), 'dpdk/drivers')
> > +endif
>
> This is against hier, since:
> "This directory contains subdirectories with specific application data,
> that can be shared among different architectures of the same OS."
> (from man 7 hier)
>
> Please consider to use a subdirectory of libdir like the actual dpdk
> instead.
Sure. Using "$libdir/dpdk/drivers" work well, or is there a better choice?
Re: [dpdk-dev] [PATCH v8 2/9] eal/pci: get iommu class
Hi Anatoly, On Wednesday 20 September 2017 02:39 PM, Burakov, Anatoly wrote: > Hi Santosh, > > On 19-Sep-17 6:29 PM, santosh wrote: >> Hi Anatoly, >> >> >> On Tuesday 19 September 2017 10:07 PM, Burakov, Anatoly wrote: >>> On 18-Sep-17 11:42 AM, Santosh Shukla wrote: Introducing rte_pci_get_iommu_class API which helps to get iommu class of PCI device on the bus and returns preferred iova mapping mode for PCI bus. Patch also add rte_pci_get_iommu_class definition for bsdapp, in bsdapp case - api returns default iova mode. Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob Reviewed-by: Maxime Coquelin --- >>> >>> Hi Santosh, >>> >>> You have probably missed my comment on previous version of this patch, but >>> for commit history reasons i really think you should add a linuxapp stub in >>> this commit as well as a FreeBSD stub, even though you are adding a >>> linuxapp function in the next commit. Any linuxapp application using that >>> function will fail to compile with this commit, despite this API being >>> already present and declared as public. >>> >> First, apologies for not following up on your note: >> >> I prefer to keep less context in each patch and >> for [03/9], its already has _IOVA_AS_VA flag + whole autodetection >> algo inside (squashed per Aron suggestion). >> Now if I squash [2/9] into [3/9], then would be too much info >> for future reader to digest for (imo). Its a kind of trade-off. >> >> On any linuxapp appl breaking with this commit: >> This series exposes eal api for application to use and identify iova mode. >> >> If you still feel not convinced with my explanation then I'll spin v9 and >> squash >> [02/09], [03/09] in v9. > > No, i don't mean squashing these two patches into one. I mean, provide a stub > like for FreeBSD, and then edit it to be a proper implementation in the next > commit. > > I.e. in this commit, add a stub that just returns 0, like for FreeBSD. Next > commit, instead of starting from scratch, start from this stub. > +1, Sending v9. Thanks. > Thanks, > Anatoly > >> >> Thanks. >> >> >> > >
[dpdk-dev] [PATCH] net/ixgbe: fix Rx queue interrupt mapping in VF
When a VF port is bound to VFIO-PCI, miscellaneous interrupt is
mapped to MSI-X vector 0 and Rx queues interrupt are mapped to
other vectors in vfio_enable_msix( ). To simplify implementation,
all VFIO-PCI bound ixgbe VF Rx queue interrupts can be mapped in
vector 1. And as current igb_uio only support only one vector,
ixgbe VF PMD should use vector 0 for igb_uio and vector 1 for
VFIO-PCI. Without this patch, VF Rx queue interrupt is mapped
to vector 0 in register settings and mapped to VFIO vector 1
in vfio_enable_msix( ), and then all Rx queue interrupts will
be missed.
Fixes: b13bfab4cdbe ("eal: reserve VFIO vector zero for misc interrupt")
Cc: sta...@dpdk.org
Signed-off-by: Wei Dai
---
drivers/net/ixgbe/ixgbe_ethdev.c | 25 ++---
1 file changed, 22 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c
index 9ca5cbc..39a3d8c 100644
--- a/drivers/net/ixgbe/ixgbe_ethdev.c
+++ b/drivers/net/ixgbe/ixgbe_ethdev.c
@@ -5029,7 +5029,10 @@ ixgbevf_dev_start(struct rte_eth_dev *dev)
/* check and configure queue intr-vector mapping */
if (dev->data->dev_conf.intr_conf.rxq != 0) {
- intr_vector = dev->data->nb_rx_queues;
+ /* According to datasheet, only vector 0/1/2 can be used,
+* now only one vector is used for Rx queue
+*/
+ intr_vector = 1;
if (rte_intr_efd_enable(intr_handle, intr_vector))
return -1;
}
@@ -5556,9 +5559,12 @@ ixgbevf_dev_rx_queue_intr_enable(struct rte_eth_dev
*dev, uint16_t queue_id)
uint32_t mask;
struct ixgbe_hw *hw =
IXGBE_DEV_PRIVATE_TO_HW(dev->data->dev_private);
+ uint32_t vec = IXGBE_MISC_VEC_ID;
mask = IXGBE_READ_REG(hw, IXGBE_VTEIMS);
- mask |= (1 << IXGBE_MISC_VEC_ID);
+ if (rte_intr_allow_others(intr_handle))
+ vec = IXGBE_RX_VEC_START;
+ mask |= (1 << vec);
RTE_SET_USED(queue_id);
IXGBE_WRITE_REG(hw, IXGBE_VTEIMS, mask);
@@ -5573,9 +5579,14 @@ ixgbevf_dev_rx_queue_intr_disable(struct rte_eth_dev
*dev, uint16_t queue_id)
uint32_t mask;
struct ixgbe_hw *hw =
IXGBE_DEV_PRIVATE_TO_HW(dev->data->dev_private);
+ struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev);
+ struct rte_intr_handle *intr_handle = &pci_dev->intr_handle;
+ uint32_t vec = IXGBE_MISC_VEC_ID;
mask = IXGBE_READ_REG(hw, IXGBE_VTEIMS);
- mask &= ~(1 << IXGBE_MISC_VEC_ID);
+ if (rte_intr_allow_others(intr_handle))
+ vec = IXGBE_RX_VEC_START;
+ mask &= ~(1 << vec);
RTE_SET_USED(queue_id);
IXGBE_WRITE_REG(hw, IXGBE_VTEIMS, mask);
@@ -5717,6 +5728,7 @@ ixgbevf_configure_msix(struct rte_eth_dev *dev)
IXGBE_DEV_PRIVATE_TO_HW(dev->data->dev_private);
uint32_t q_idx;
uint32_t vector_idx = IXGBE_MISC_VEC_ID;
+ uint32_t base = IXGBE_MISC_VEC_ID;
/* Configure VF other cause ivar */
ixgbevf_set_ivar_map(hw, -1, 1, vector_idx);
@@ -5727,6 +5739,11 @@ ixgbevf_configure_msix(struct rte_eth_dev *dev)
if (!rte_intr_dp_is_en(intr_handle))
return;
+ if (rte_intr_allow_others(intr_handle)) {
+ base = IXGBE_RX_VEC_START;
+ vector_idx = IXGBE_RX_VEC_START;
+ }
+
/* Configure all RX queues of VF */
for (q_idx = 0; q_idx < dev->data->nb_rx_queues; q_idx++) {
/* Force all queue use vector 0,
@@ -5734,6 +5751,8 @@ ixgbevf_configure_msix(struct rte_eth_dev *dev)
*/
ixgbevf_set_ivar_map(hw, 0, q_idx, vector_idx);
intr_handle->intr_vec[q_idx] = vector_idx;
+ if (vector_idx < base + intr_handle->nb_efd - 1)
+ vector_idx++;
}
}
--
2.7.5
Re: [dpdk-dev] [PATCH 1/2] net/i40e: queue region set and flush
On 9/20/2017 4:20 AM, Zhao1, Wei wrote: > Hi, Ferruh > >> -Original Message- >> From: Yigit, Ferruh >> Sent: Friday, September 15, 2017 7:01 PM >> To: Zhao1, Wei ; dev@dpdk.org >> Subject: Re: [dpdk-dev] [PATCH 1/2] net/i40e: queue region set and flush >> >> On 9/6/2017 10:11 AM, Ferruh Yigit wrote: >>> On 9/1/2017 3:38 AM, Zhao1, Wei wrote: HI, Ferruh > -Original Message- > From: Yigit, Ferruh > Sent: Friday, September 1, 2017 12:18 AM > To: Zhao1, Wei ; dev@dpdk.org > Subject: Re: [dpdk-dev] [PATCH 1/2] net/i40e: queue region set and > flush > > On 8/24/2017 4:26 AM, Wei Zhao wrote: >> This feature enable queue regions configuration for RSS in PF/VF, >> so that different traffic classes or different packet >> classification types can be separated to different queues in >> different queue regions.This patch can set queue region range, it >> include queue number in a region and the index of first queue. >> This patch enable mapping between different priorities (UP) and >> different traffic classes.It also enable mapping between a region >> index and a sepcific flowtype(PCTYPE).It also provide the solution >> of flush all configuration about queue region the above described. >> >> Signed-off-by: Wei Zhao > > Hi Wei, > > I wonder if this can be implemented using rte_flow, instead of PMD > specific API? > > And if not, what is missing in rte_flow for this? Yes, at first I have plan to use rte_flow to implement this feature. But there are many opposition for this, for example, this feature is only support by i40e but not igb/ixgbe From hardware capacity, not all NIC. >> So, private API is selected. >>> >>> I guess rte_flow headers needs to be updated for this support, how big >>> that update, what is missing in rte_flow for this? >> >> Hi Wei, >> >> Would you mind answering this? > > If we want rte_flow also to support this feature, > although there will be many new CLI command need to be Add to testpmd app, > and some code change in ethdev, for example add new member in enum > rte_filter_type and so on. new CLI part is side effect of the rte_flow, to demonstrate how to use flow modes. I was more interested in how much change required in rte_flow header. > But the biggest problem is to get recognize and support from expert and > authority in DPDK community for this new change. Is this is referring previous push back for rte_filter_type update, I think this can be evaluated case by case. As far as I remember previously it is rejected in favor of improving rte_flow, this time update is required to be able to use rte_flow, so I assume can be OK. I believe this can be discussed out of this patch scope. > By now, rte_flow has no support for rss config application, but only filter > configuration. But maybe this can be a work in next release. At least it is good that this effort can be re-used later for rte_flow, I still believe this option worth investigating for further releases. > > >> >> Thanks, >> ferruh >> >>> >>> Even this is only for i40e, rte_flow can be an option. I believe this >>> increases the usability of the feature for the driver. >>> > > Thanks, > ferruh >>> >
Re: [dpdk-dev] [PATCH v3 01/17] build: add initial infrastructure for meson & ninja builds
On September 20, 2017 12:24:15 PM GMT+02:00, Bruce Richardson
wrote:
>On Wed, Sep 20, 2017 at 12:10:19PM +0200, Timothy M. Redaelli wrote:
>> On 09/13/2017 04:12 PM, Bruce Richardson wrote:
>> [...]
>> > +
>> > +# for static libs, treat the drivers as regular libraries,
>otherwise
>> > +# for shared libs, put them in a driver folder
>> > +if get_option('default_library') == 'static'
>> > + driver_install_path = get_option('libdir')
>> > +else
>> > + driver_install_path = join_paths(get_option('datadir'),
>'dpdk/drivers')
>> > +endif
>>
>> This is against hier, since:
>> "This directory contains subdirectories with specific application
>data,
>> that can be shared among different architectures of the same OS."
>> (from man 7 hier)
>>
>> Please consider to use a subdirectory of libdir like the actual dpdk
>> instead.
>
>Sure. Using "$libdir/dpdk/drivers" work well, or is there a better
>choice?
Sounds good, thank you
Re: [dpdk-dev] [PATCH v3 1/2] net/i40e: queue region set and flush
On 9/15/2017 4:13 AM, Wei Zhao wrote:
> This feature enable queue regions configuration for RSS in PF/VF,
> so that different traffic classes or different packet
> classification types can be separated to different queues in
> different queue regions.This patch can set queue region range,
> it include queue number in a region and the index of first queue.
Is following correct:
So instead of distributing packets to the multiple queues, this will
distribute packets into queue reqions which may consists of multiple queues.
If so, is there a way to control how packets distributed within same
queue region to multiple queues?
And is this feature only supported with RSS? Can it be part of RSS
configuration instead of PMD specific API?
> This patch enable mapping between different priorities (UP) and
User priorities (UP)
> different traffic classes.It also enable mapping between a region
> index and a sepcific flowtype(PCTYPE).It also provide the solution
> of flush all configuration about queue region the above described.
>
> Signed-off-by: Wei Zhao
> ---
> drivers/net/i40e/i40e_ethdev.c| 19 +-
> drivers/net/i40e/i40e_ethdev.h| 30 ++
> drivers/net/i40e/rte_pmd_i40e.c | 482
> ++
> drivers/net/i40e/rte_pmd_i40e.h | 38 +++
> drivers/net/i40e/rte_pmd_i40e_version.map | 1 +
> 5 files changed, 566 insertions(+), 4 deletions(-)
>
<...>
> +static int
> +i40e_vsi_update_queue_region_mapping(struct i40e_hw *hw,
> + struct i40e_pf *pf)
> +{
> + uint16_t i;
> + struct i40e_vsi *vsi = pf->main_vsi;
> + uint16_t queue_offset, bsf, tc_index;
> + struct i40e_vsi_context ctxt;
> + struct i40e_aqc_vsi_properties_data *vsi_info;
> + struct i40e_queue_region_info *region_info =
> + &pf->queue_region;
> + uint32_t ret = -EINVAL;
> +
> + if (!region_info->queue_region_number) {
> + PMD_INIT_LOG(ERR, "there is no that region id been set before");
Can you please re-check the log message.
> + return ret;
> + }
> +
> + memset(&ctxt, 0, sizeof(struct i40e_vsi_context));
> +
> + /* Update Queue Pairs Mapping for currently enabled UPs */
> + ctxt.seid = vsi->seid;
> + ctxt.pf_num = hw->pf_id;
> + ctxt.vf_num = 0;
> + ctxt.uplink_seid = vsi->uplink_seid;
> + ctxt.info = vsi->info;
> + vsi_info = &ctxt.info;
> +
> + memset(vsi_info->tc_mapping, 0, sizeof(uint16_t) * 8);
> + memset(vsi_info->queue_mapping, 0, sizeof(uint16_t) * 16);
> +
> + /**
> + * Configure queue region and queue mapping parameters,
> + * for enabled queue region, allocate queues to this region.
> + */
> +
> + for (i = 0; i < region_info->queue_region_number; i++) {
> + tc_index = region_info->region[i].region_id;
> + bsf = rte_bsf32(region_info->region[i].queue_num);
> + queue_offset = region_info->region[i].queue_start_index;
> + vsi_info->tc_mapping[tc_index] = rte_cpu_to_le_16(
> + (queue_offset << I40E_AQ_VSI_TC_QUE_OFFSET_SHIFT) |
> + (bsf << I40E_AQ_VSI_TC_QUE_NUMBER_SHIFT));
> + }
> +
> + /* Associate queue number with VSI, Keep vsi->nb_qps unchanged */
> + if (vsi->type == I40E_VSI_SRIOV) {
> + vsi_info->mapping_flags |=
> + rte_cpu_to_le_16(I40E_AQ_VSI_QUE_MAP_NONCONTIG);
> + for (i = 0; i < vsi->nb_qps; i++)
> + vsi_info->queue_mapping[i] =
> + rte_cpu_to_le_16(vsi->base_queue + i);
> + } else {
> + vsi_info->mapping_flags |=
> + rte_cpu_to_le_16(I40E_AQ_VSI_QUE_MAP_CONTIG);
> + vsi_info->queue_mapping[0] = rte_cpu_to_le_16(vsi->base_queue);
> + }
> + vsi_info->valid_sections |=
> + rte_cpu_to_le_16(I40E_AQ_VSI_PROP_QUEUE_MAP_VALID);
> +
> + /* Update the VSI after updating the VSI queue-mapping information */
> + ret = i40e_aq_update_vsi_params(hw, &ctxt, NULL);
> + if (ret) {
> + PMD_INIT_LOG(ERR, "Failed to configure queue region "
> + "mapping = %d ", hw->aq.asq_last_status);
Please don't split log message.
> + return ret;
> + }
> + /* update the local VSI info with updated queue map */
> + (void)rte_memcpy(&vsi->info.tc_mapping, &ctxt.info.tc_mapping,
> + sizeof(vsi->info.tc_mapping));
> + (void)rte_memcpy(&vsi->info.queue_mapping,
> + &ctxt.info.queue_mapping,
> + sizeof(vsi->info.queue_mapping));
Please keep line allignment same with above line.
> + vsi->info.mapping_flags = ctxt.info.mapping_flags;
> + vsi->info.valid_sections = 0;
> +
> + return 0;
> +}
> +
> +
> +static int
> +i40e_set_queue_region(struct i40e_pf *pf,
> + struct rte_i40e_rss_
[dpdk-dev] [PATCH] net/ixgbe: fix VFIO interrupt mapping in PF
When a PF port is bound to VFIO-PIC, only miscellaneous interrupt
is mapped to VFIO vector 0 in ixgbe_dev_init( ).
In ixgbe_dev_start(), if previous VFIO interrupt mapping set in
ixgbe_dev_init( ) is not cleard, it will fail when calling
rte_intr_enable( ) tries to map Rx queue interrupt to other VFIO
vectors. This patch clears the VFIO interrupt mappings before
setting both miscellaneous and Rx queue interrupt mappings again
to avoid failure.
Fixes: 0a45657a6794 ("pci: rework interrupt handling")
Cc: sta...@dpdk.org
Signed-off-by: Wei Dai
---
drivers/net/ixgbe/ixgbe_ethdev.c | 9 +
1 file changed, 9 insertions(+)
diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c
index 9ca5cbc..5292694 100644
--- a/drivers/net/ixgbe/ixgbe_ethdev.c
+++ b/drivers/net/ixgbe/ixgbe_ethdev.c
@@ -2692,6 +2692,15 @@ ixgbe_dev_start(struct rte_eth_dev *dev)
" no intr multiplex");
}
+ /* When a PF port is bound to VFIO-PCI only miscellaneous interrupt
+* is mapped to VFIO vector 0 in ixgbe_dev_init( ).
+* If previous VFIO interrupt mapping set in ixgbe_dev_init( ) is
+* not cleared, it will fail when following rte_intr_enable( ) tries
+* to map Rx queue interrupt to other VFIO vectors.
+* So clear uio/vfio intr/evevnfd first to avoid failure.
+*/
+ rte_intr_disable(intr_handle);
+
/* check if rxq interrupt is enabled */
if (dev->data->dev_conf.intr_conf.rxq != 0 &&
rte_intr_dp_is_en(intr_handle))
--
2.7.5
Re: [dpdk-dev] [PATCH v3 2/2] app/testpmd: add API for configuration of queue region
On 9/15/2017 4:13 AM, Wei Zhao wrote:
> This patch add a API configuration of queue region in rss.
> It can parse the parameters of region index, queue number,
> queue start index, user priority, traffic classes and so on.
> According to commands from command line, it will call i40e
> private API and start the process of set or flush queue region
> configure. As this feature is specific for i40e, so private API
> will be used.
>
> Signed-off-by: Wei Zhao
> ---
> app/test-pmd/cmdline.c | 328
> +
Testpmd documentation also needs to be updated.
> 1 file changed, 328 insertions(+)
>
> diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
> index 0144191..060fcb1 100644
> --- a/app/test-pmd/cmdline.c
> +++ b/app/test-pmd/cmdline.c
> @@ -637,6 +637,21 @@ static void cmd_help_long_parsed(void *parsed_result,
> "ptype mapping update (port_id) (hw_ptype) (sw_ptype)\n"
> "Update a ptype mapping item on a port\n\n"
>
> + "queue-region set port (port_id) region_id (value) "
> + "queue_start_index (value) queue_num (value)\n"
> + "Set a queue region on a port\n\n"
> +
> + "queue-region set (pf|vf) port (port_id) region_id
> (value) "
> + "flowtype (value)\n"
> + "Set a flowtype region index on a port\n\n"
> +
> + "queue-region set port (port_id) UP (value) region_id
> (value)\n"
> + "Set the mapping of User Priority to "
> + "queue region on a port\n\n"
> +
> + "queue-region flush (on|off) port (port_id)\n"
> + "flush all queue region related configuration\n\n"
I keep doing same comment but I will do it again...
Each patch adding a new feature looking from its own context and adding
a new root level command and this is making overall testpmd confusing.
Since this is to set an option of the port, what do you think making
this command part of existing commands, like:
"port config #P queue-region "
OR
"set port #P queue-region ..." ?
> +
> , list_pkt_forwarding_modes()
> );
> }
> @@ -8224,6 +8239,315 @@ cmdline_parse_inst_t cmd_syn_filter = {
> NULL,
> },
> };
> +/* *** queue region set *** */
> +struct cmd_queue_region_result {
> + cmdline_fixed_string_t cmd;
> + cmdline_fixed_string_t set;
> + cmdline_fixed_string_t port;
> + uint8_t port_id;
> + cmdline_fixed_string_t region;
> + uint8_t region_id;
> + cmdline_fixed_string_t queue_start_index;
> + uint8_t queue_id;
> + cmdline_fixed_string_t queue_num;
> + uint8_t queue_num_value;
> +};
> +
> +static void
> +cmd_queue_region_parsed(void *parsed_result,
> + __attribute__((unused)) struct cmdline *cl,
> + __attribute__((unused)) void *data)
> +{
> + struct cmd_queue_region_result *res = parsed_result;
> + struct rte_i40e_rss_region_conf region_conf;
> + int ret = 0;
> +
> + memset(®ion_conf, 0, sizeof(region_conf));
> + region_conf.op = RTE_PMD_I40E_QUEUE_REGION_SET;
> + region_conf.region_id = res->region_id;
> + region_conf.queue_num = res->queue_num_value;
> + region_conf.queue_start_index = res->queue_id;
> +
> + ret = rte_pmd_i40e_queue_region_conf(res->port_id, ®ion_conf);
It is not safe to directly call PMD specific APIs from testpmd? What if
that PMD is not enabled? There are samples how to do this, can you
please check them?
<...>
[dpdk-dev] [PATCH 0/3] minor build enhancements and fixes
No major changes, just cleanups. Bruce Richardson (3): build: sort meson options alphabetically build: fix driver install path build/x86: add SSE cpuflags config/x86/meson.build | 10 ++ meson.build| 2 +- meson_options.txt | 24 3 files changed, 27 insertions(+), 9 deletions(-) -- 2.13.5
[dpdk-dev] [PATCH 1/3] build: sort meson options alphabetically
Wrap each entry at the description value to avoid really long lines also.
Signed-off-by: Bruce Richardson
---
meson_options.txt | 24
1 file changed, 16 insertions(+), 8 deletions(-)
diff --git a/meson_options.txt b/meson_options.txt
index fb945db91..29b548f5a 100644
--- a/meson_options.txt
+++ b/meson_options.txt
@@ -1,10 +1,18 @@
-option('machine', type: 'string', value: 'native', description: 'set the
target machine type')
-option('max_lcores', type: 'string', value: '128', description: 'maximum
number of cores/threads supported by EAL')
-option('max_numa_nodes', type: 'string', value: '4', description: 'maximum
number of NUMA nodes supported by EAL')
-option('use_hpet', type: 'boolean', value: false, description: 'use HPET timer
in EAL')
option('allow_invalid_socket_id', type: 'boolean', value: false,
description: 'allow out-of-range NUMA socket id\'s for platforms that
don\'t report the value correctly')
-option('enable_kmods', type: 'boolean', value: true, description: 'build
kernel modules')
-option('kernel_dir', type: 'string', value: '', description: 'path to the
kernel for building kernel modules')
-option('per_library_versions', type: 'boolean', value: true, description:
'true: each lib gets its own version number, false: DPDK version used for each
lib')
-option('include_subdir_arch', type: 'string', value: '', description:
'subdirectory where to install arch-dependent headers')
+option('enable_kmods', type: 'boolean', value: true,
+ description: 'build kernel modules')
+option('include_subdir_arch', type: 'string', value: '',
+ description: 'subdirectory where to install arch-dependent headers')
+option('kernel_dir', type: 'string', value: '',
+ description: 'path to the kernel for building kernel modules')
+option('machine', type: 'string', value: 'native',
+ description: 'set the target machine type')
+option('max_lcores', type: 'string', value: '128',
+ description: 'maximum number of cores/threads supported by EAL')
+option('max_numa_nodes', type: 'string', value: '4',
+ description: 'maximum number of NUMA nodes supported by EAL')
+option('per_library_versions', type: 'boolean', value: true,
+ description: 'true: each lib gets its own version number, false: DPDK
version used for each lib')
+option('use_hpet', type: 'boolean', value: false,
+ description: 'use HPET timer in EAL')
--
2.13.5
[dpdk-dev] [PATCH 2/3] build: fix driver install path
To comply with the "hier" standard [Ref: man 7 hier], the driver .so files
should not be placed in $datadir. Therefore we install them in a
sub-directory of $libdir instead.
Fixes: d123bba2dfbe ("build: add initial infrastructure for meson & ninja
builds")
Reported-by: Timothy M. Redaelli
Signed-off-by: Bruce Richardson
---
meson.build | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/meson.build b/meson.build
index 41426a706..6e4752562 100644
--- a/meson.build
+++ b/meson.build
@@ -48,7 +48,7 @@ dpdk_extra_ldflags = []
if get_option('default_library') == 'static'
driver_install_path = get_option('libdir')
else
- driver_install_path = join_paths(get_option('datadir'), 'dpdk/drivers')
+ driver_install_path = join_paths(get_option('libdir'), 'dpdk/drivers')
endif
# configure the build, and make sure configs here and in config folder are
--
2.13.5
[dpdk-dev] [PATCH 3/3] build/x86: add SSE cpuflags
Previous code only added in AVX, and a few other non-SSE flags to the
compile-time cpuflags because all SSE instruction set levels are now
required for an x86 build. However, some apps may still be checking for the
existing SSE ones in the legacy build system, so add them here for
completeness and compatibility.
Signed-off-by: Bruce Richardson
---
config/x86/meson.build | 10 ++
1 file changed, 10 insertions(+)
diff --git a/config/x86/meson.build b/config/x86/meson.build
index 0d1a532ad..c05c2f418 100644
--- a/config/x86/meson.build
+++ b/config/x86/meson.build
@@ -40,6 +40,12 @@ if cc.get_define('__SSE4_2__', args: march_opt) == ''
error(sse_errormsg)
endif
+base_flags = ['SSE', 'SSE2', 'SSE3','SSSE3', 'SSE4_1', 'SSE4_2']
+foreach f:base_flags
+ dpdk_conf.set('RTE_MACHINE_CPUFLAG_' + f, 1)
+ compile_time_cpuflags += ['RTE_CPUFLAG_' + f]
+endforeach
+
dpdk_conf.set('RTE_ARCH_X86', 1)
if (host_machine.cpu_family() == 'x86_64')
dpdk_conf.set('RTE_ARCH_X86_64', 1)
@@ -66,5 +72,9 @@ if cc.get_define('__AVX2__', args: march_opt) != ''
dpdk_conf.set('RTE_MACHINE_CPUFLAG_AVX2', 1)
compile_time_cpuflags += ['RTE_CPUFLAG_AVX2']
endif
+if cc.get_define('__AVX512F__', args: march_opt) != ''
+ dpdk_conf.set('RTE_MACHINE_CPUFLAG_AVX512F', 1)
+ compile_time_cpuflags += ['RTE_CPUFLAG_AVX512F']
+endif
dpdk_conf.set('RTE_CACHE_LINE_SIZE', 64)
--
2.13.5
Re: [dpdk-dev] [PATCH 02/11] doc: add details of rte security
Hi Jerin, On 9/18/2017 4:43 PM, Jerin Jacob wrote: -Original Message- Date: Thu, 14 Sep 2017 13:56:42 +0530 From: Akhil Goyal To: dev@dpdk.org CC: declan.dohe...@intel.com, pablo.de.lara.gua...@intel.com, hemant.agra...@nxp.com, radu.nico...@intel.com, bor...@mellanox.com, avia...@mellanox.com, tho...@monjalon.net, sandeep.ma...@nxp.com, jerin.ja...@caviumnetworks.com Subject: [PATCH 02/11] doc: add details of rte security X-Mailer: git-send-email 2.9.3 +Security Library + + +The security library provides a framework for management and provisioning +of security protocol operations offloaded to hardware based devices. The +library defines generic APIs to create and free security sessions which can +support complete protocol offload as well as inline crypto operation with +NIC or crypto devices. The framework currently only supports IPSEC protocol +and it's operations, other protocols will be added in future. + +Design Principles +- + +The security library provides an additional offload capability to existing +crypto device and/or ethernet device. + +.. code-block:: console + + +---+ + | rte_security | + +---+ + \/ ++---++--+ +| NIC PMD || CRYPTO PMD | ++---++--+ + +Following offload types can be supported: + +Inline Crypto +~ + +RTE_SECURITY_ACTION_TYPE_INLINE_CRYPTO: +The crypto processing for security protocol (e.g. IPSEC) is processed +inline during receive and transmission on NIC port. The flow based +security action should be configured on the port. + +Ingress Data path - The packet is decrypted in RX path and relevant +crypto status is set in rx descriptors. After the successful inline +crypto processing the packet is presented to host as a regular rx packet +but all security protocol related headers are still attached to the +packet. e.g. In case of IPSEC, the IPSEC tunnel headers (if any), +ESP/AH headers will remain in the packet but the received packet +contains the decrypted data where the encrypted data was when the packet +arrived. The driver rx path check the descriptos and and based on the s/descriptos/descriptors +crypto status sets additional flags in the rte_mbuf.ol_flags field. + +.. note:: + +The underlying device may not support crypto processing all ingress packet +matching to a particular flow (e.g. fragmented packets), such packets will +be passed as encrypted packets. It is the responsibility of driver to ^^^ Do you mean application here instead of driver? +process such encrypted packets using other crypto driver instance. + +Egress Data path - The software prepares the egress packet by adding +relevant security protocol headers in the packets. Only the data will not be +encryptoed by the software. The driver will accordingly configure the s/encryptoed/encrypted +tx descriptors. The HW device will encrypt the data before sending the +the packet out. + +.. note:: + +The underlying device may support post encryption TSO. + +.. code-block:: console + + Egress Data Path + | ++|+ +| egress IPsec | +||| +| +--V--+ | +| | SABD lookup | | +| +--|--+ | s/SABD/SADB +| +--V--+ | +| | Tunnel| | <-- Add tunnel header to packet +| +--|--+ | +| +--V--+ | +| | ESP | | <-- Add ESP header without trailer to packet +| | | | <-- Mark packet to be offloaded, add trailer +| +--|--+ |meta-data to mbuf ++V+ + | ++V+ +|L2 Stack | ++|+ + | ++V+ +| | +| NIC PMD | <-- Set hw context for inline crypto offload +| | ++|+ + | ++|+ +| HW ACCELERATED | <-- Packet Encryption/Decryption and Only packet Encryption here. Right? +|NIC | Authentication happens inline +| | ++|+ ^^^ I guess the "|" can be removed. + + +Inline protocol offload +~~~ + +RTE_SECURITY_ACTION_TYPE_INLINE_PROTOCOL: +The crypto and protocol processing for security protocol (e.g. IPSEC) +is processed inline during receive and transmission. The flow based +security action should be configured on the port. + +Ingress Data path - The packet is decrypted in RX path and relevant +crypto status is set in rx descriptors. After the successful
Re: [dpdk-dev] [PATCH 02/11] doc: add details of rte security
Hi John, On 9/18/2017 9:08 PM, Mcnamara, John wrote: -Original Message- From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Akhil Goyal Sent: Thursday, September 14, 2017 9:27 AM To: dev@dpdk.org Cc: Doherty, Declan ; De Lara Guarch, Pablo ; hemant.agra...@nxp.com; Nicolau, Radu ; bor...@mellanox.com; avia...@mellanox.com; tho...@monjalon.net; sandeep.ma...@nxp.com; jerin.ja...@caviumnetworks.com Subject: [dpdk-dev] [PATCH 02/11] doc: add details of rte security Signed-off-by: Hemant Agrawal Signed-off-by: Akhil Goyal I have a lot of minor corrections/suggestions. I've send you on the changed/review file and you can incorporated them as you wish. Reviewed-by: John McNamara I would include these in my next version. Thanks for reviewing. Regards, Akhil
Re: [dpdk-dev] [PATCH 00/53] net/qede/base: update PMD to 2.6.0.1
On 9/19/2017 2:29 AM, Rasesh Mody wrote: > Hi, > > This patch set adds support for new firmware 8.30.12.0, includes > enahncements, code cleanup and bug fixes. This patch set updates > PMD version to 2.6.0.1. > > Thanks! > Rasesh > > Rasesh Mody (53): > net/qede/base: add NVM config options > net/qede/base: update management FW supported features > net/qede/base: use crc32 OSAL macro > net/qede/base: allocate VF queues before PF > net/qede/base: convert device type to enum > net/qede/base: changes for VF queue zone > net/qede/base: interchangeably use SB between PF and VF > net/qede/base: add API to configure coalescing for VF queues > net/qede/base: restrict cache line size register padding > net/qede/base: fix to use a passed ptt handle > net/qede/base: add a sanity check > net/qede/base: add SmartAN support > net/qede/base: alter driver's force load behavior > net/qede/base: add mdump sub-commands > net/qede/base: add EEE support > net/qede/base: use passed ptt handler > net/qede/base: prevent re-assertions of parity errors > net/qede/base: avoid possible race condition > net/qede/base: revise management FW mbox access scheme > net/qede/base: remove helper functions/structures > net/qede/base: initialize resc lock/unlock params > net/qede/base: rename MFW get/set field defines > net/qede/base: allow clients to override VF MSI-X table size > net/qede/base: add API to send STAG config update to FW > net/qede/base: add support for doorbell overflow recovery > net/qede/base: block mbox command to unresponsive MFW > net/qede/base: prevent stop vport assert by malicious VF > net/qede/base: remove unused parameters > net/qede/base: fix macros to check chip revision/metal > net/qede/base: read per queue coalescing from HW > net/qede/base: refactor device's number of ports logic > net/qede/base: use proper units for rate limiting > net/qede/base: use available macro > net/qede/base: use function pointers for spq async callback > net/qede/base: fix API return types > net/qede/base: semantic changes > net/qede/base: handle the error condition properly > net/qede/base: add new macro for CMT mode > net/qede/base: change verbosity > net/qede/base: fix number of app table entries > net/qede/base: update firmware to 8.30.12.0 > net/qede/base: add UFP support > net/qede/base: add support for mapped doorbell Bars for VFs > net/qede/base: add support for driver attribute repository > net/qede/base: move define to header file > net/qede/base: dcbx dscp related extensions > net/qede/base: add feature support for per-PF virtual link > net/qede/base: catch an init command write failure > net/qede/base: retain dcbx config till actually applied > net/qede/base: disable aRFS for NPAR and 100G > net/qede/base: add support for WoL writes > net/qede/base: remove unused input parameter > net/qede/base: update PMD version to 2.6.0.1 Series applied to dpdk-next-net/master, thanks.
Re: [dpdk-dev] [PATCH] app/testpmd: port info prints dynamically mapped flow types
On 9/1/2017 5:23 PM, Kirill Rybalchenko wrote:
> Port info command prints information about all supported flow types,
> including dynamically mapped ones.
>
> Signed-off-by: Kirill Rybalchenko
> ---
> app/test-pmd/config.c | 9 ++---
> 1 file changed, 6 insertions(+), 3 deletions(-)
>
> diff --git a/app/test-pmd/config.c b/app/test-pmd/config.c
> index 3ae3e1c..bd5de92 100644
> --- a/app/test-pmd/config.c
> +++ b/app/test-pmd/config.c
> @@ -498,12 +498,15 @@ port_infos_display(portid_t port_id)
> char *p;
>
> printf("Supported flow types:\n");
> - for (i = RTE_ETH_FLOW_UNKNOWN + 1; i < RTE_ETH_FLOW_MAX;
> - i++) {
> + for (i = RTE_ETH_FLOW_UNKNOWN + 1;
> + i < sizeof(dev_info.flow_type_rss_offloads) * 8; i++) {
What do you think using "CHAR_BIT" instead of "8" to stress the intention?
If you are OK to send new version, please keep JingJing's Review tag.
> if (!(dev_info.flow_type_rss_offloads & (1ULL << i)))
> continue;
> p = flowtype_to_str(i);
> - printf(" %s\n", (p ? p : "unknown"));
> + if (p)
> + printf(" %s\n", p);
> + else
> + printf(" user defined %d\n", i);
> }
> }
>
>
Re: [dpdk-dev] [PATCH v3] mbuf: use refcnt = 0 when debugging
Hi Chas,
On Thu, Sep 07, 2017 at 05:21:51PM -0400, Charles (Chas) Williams wrote:
> After commit 8f094a9ac5d7 ("mbuf: set mbuf fields while in pool") is it
> much harder to detect a "double free". If the developer makes a copy
> of an mbuf pointer and frees it twice, this condition is never detected
> and the mbuf gets returned to the pool twice.
>
> Since this requires extra work to track, make this behavior conditional
> on CONFIG_RTE_LIBRTE_MBUF_DEBUG.
>
> Signed-off-by: Chas Williams
> ---
> lib/librte_mbuf/rte_mbuf.c | 2 +-
> lib/librte_mbuf/rte_mbuf.h | 40 ++--
> 2 files changed, 35 insertions(+), 7 deletions(-)
>
> diff --git a/lib/librte_mbuf/rte_mbuf.c b/lib/librte_mbuf/rte_mbuf.c
> index 26a62b8..b0d222c 100644
> --- a/lib/librte_mbuf/rte_mbuf.c
> +++ b/lib/librte_mbuf/rte_mbuf.c
> @@ -145,7 +145,7 @@ rte_pktmbuf_init(struct rte_mempool *mp,
> m->pool = mp;
> m->nb_segs = 1;
> m->port = 0xff;
> - rte_mbuf_refcnt_set(m, 1);
> + rte_mbuf_refcnt_set(m, RTE_MBUF_UNUSED_CNT);
> m->next = NULL;
> }
>
> diff --git a/lib/librte_mbuf/rte_mbuf.h b/lib/librte_mbuf/rte_mbuf.h
> index eaed7ee..1400b35 100644
> --- a/lib/librte_mbuf/rte_mbuf.h
> +++ b/lib/librte_mbuf/rte_mbuf.h
> @@ -671,11 +671,15 @@ struct rte_pktmbuf_pool_private {
>
> #ifdef RTE_LIBRTE_MBUF_DEBUG
>
> +#define RTE_MBUF_UNUSED_CNT 0
> +
> /** check mbuf type in debug mode */
> #define __rte_mbuf_sanity_check(m, is_h) rte_mbuf_sanity_check(m, is_h)
>
> #else /* RTE_LIBRTE_MBUF_DEBUG */
>
> +#define RTE_MBUF_UNUSED_CNT 1
> +
> /** check mbuf type in debug mode */
> #define __rte_mbuf_sanity_check(m, is_h) do { } while (0)
>
> @@ -721,6 +725,9 @@ rte_mbuf_refcnt_set(struct rte_mbuf *m, uint16_t
> new_value)
> static inline uint16_t
> rte_mbuf_refcnt_update(struct rte_mbuf *m, int16_t value)
> {
> +#ifdef RTE_LIBRTE_MBUF_DEBUG
> + RTE_ASSERT(rte_mbuf_refcnt_read(m) != 0);
> +#endif
> /*
>* The atomic_add is an expensive operation, so we don't want to
>* call it in the case where we know we are the uniq holder of
> @@ -791,10 +798,9 @@ void
> rte_mbuf_sanity_check(const struct rte_mbuf *m, int is_header);
>
> #define MBUF_RAW_ALLOC_CHECK(m) do { \
> - RTE_ASSERT(rte_mbuf_refcnt_read(m) == 1); \
> + RTE_ASSERT(rte_mbuf_refcnt_read(m) == RTE_MBUF_UNUSED_CNT); \
> RTE_ASSERT((m)->next == NULL); \
> RTE_ASSERT((m)->nb_segs == 1); \
> - __rte_mbuf_sanity_check(m, 0); \
> } while (0)
>
> /**
> @@ -825,6 +831,10 @@ static inline struct rte_mbuf *rte_mbuf_raw_alloc(struct
> rte_mempool *mp)
> return NULL;
> m = (struct rte_mbuf *)mb;
> MBUF_RAW_ALLOC_CHECK(m);
> +#ifdef RTE_LIBRTE_MBUF_DEBUG
> + rte_mbuf_refcnt_set(m, 1);
> +#endif
> + __rte_mbuf_sanity_check(m, 0);
> return m;
> }
>
> @@ -846,10 +856,9 @@ static __rte_always_inline void
> rte_mbuf_raw_free(struct rte_mbuf *m)
> {
> RTE_ASSERT(RTE_MBUF_DIRECT(m));
> - RTE_ASSERT(rte_mbuf_refcnt_read(m) == 1);
> + RTE_ASSERT(rte_mbuf_refcnt_read(m) == RTE_MBUF_UNUSED_CNT);
> RTE_ASSERT(m->next == NULL);
> RTE_ASSERT(m->nb_segs == 1);
> - __rte_mbuf_sanity_check(m, 0);
> rte_mempool_put(m->pool, m);
> }
>
> @@ -1159,21 +1168,37 @@ static inline int rte_pktmbuf_alloc_bulk(struct
> rte_mempool *pool,
> case 0:
> while (idx != count) {
> MBUF_RAW_ALLOC_CHECK(mbufs[idx]);
> +#ifdef RTE_LIBRTE_MBUF_DEBUG
> + rte_mbuf_refcnt_set(mbufs[idx], 1);
> +#endif
> + __rte_mbuf_sanity_check(mbufs[idx], 0);
> rte_pktmbuf_reset(mbufs[idx]);
> idx++;
> /* fall-through */
> case 3:
> MBUF_RAW_ALLOC_CHECK(mbufs[idx]);
> +#ifdef RTE_LIBRTE_MBUF_DEBUG
> + rte_mbuf_refcnt_set(mbufs[idx], 1);
> +#endif
> + __rte_mbuf_sanity_check(mbufs[idx], 0);
> rte_pktmbuf_reset(mbufs[idx]);
> idx++;
> /* fall-through */
> case 2:
> MBUF_RAW_ALLOC_CHECK(mbufs[idx]);
> +#ifdef RTE_LIBRTE_MBUF_DEBUG
> + rte_mbuf_refcnt_set(mbufs[idx], 1);
> +#endif
> + __rte_mbuf_sanity_check(mbufs[idx], 0);
> rte_pktmbuf_reset(mbufs[idx]);
> idx++;
> /* fall-through */
> case 1:
> MBUF_RAW_ALLOC_CHECK(mbufs[idx]);
> +#ifdef RTE_LIBRTE_MBUF_DEBUG
> + rte_mbuf_refcnt_set(mbufs[idx], 1);
> +#endif
> + __rte_mbuf_sanity_check(mbufs[idx], 0);
> rte_pktmbuf_reset(mbufs[idx]);
> idx++;
>
[dpdk-dev] [PATCH v9 2/9] eal/pci: get iommu class
Introducing rte_pci_get_iommu_class API which helps to get iommu class
of PCI device on the bus and returns preferred iova mapping mode for
PCI bus.
Patch also adds rte_pci_get_iommu_class definition for:
- bsdapp: api returns default iova mode.
- linuxapp: Has stub implementation, Followup patch has complete
implementation.
Signed-off-by: Santosh Shukla
Signed-off-by: Jerin Jacob
Reviewed-by: Maxime Coquelin
Tested-by: Hemant Agrawal
---
v8 --> v9:
- Added linuxapp iova stub definition (Suugested by Anatoly)
v6 --> v7:
- squashed v6 series patch [02/12] & [03/12] (Aaron comment).
lib/librte_eal/bsdapp/eal/eal_pci.c | 10 ++
lib/librte_eal/bsdapp/eal/rte_eal_version.map | 1 +
lib/librte_eal/common/include/rte_bus.h | 10 ++
lib/librte_eal/common/include/rte_pci.h | 11 +++
lib/librte_eal/linuxapp/eal/eal_pci.c | 9 +
lib/librte_eal/linuxapp/eal/rte_eal_version.map | 1 +
6 files changed, 42 insertions(+)
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c
b/lib/librte_eal/bsdapp/eal/eal_pci.c
index 04eacdcc7..e2c252320 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -403,6 +403,16 @@ rte_pci_scan(void)
return -1;
}
+/*
+ * Get iommu class of pci devices on the bus.
+ */
+enum rte_iova_mode
+rte_pci_get_iommu_class(void)
+{
+ /* Supports only RTE_KDRV_NIC_UIO */
+ return RTE_IOVA_PA;
+}
+
int
pci_update_device(const struct rte_pci_addr *addr)
{
diff --git a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
index cfbf8fbd0..c6ffd9399 100644
--- a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
+++ b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
@@ -243,5 +243,6 @@ DPDK_17.11 {
global:
rte_pci_match;
+ rte_pci_get_iommu_class;
} DPDK_17.08;
diff --git a/lib/librte_eal/common/include/rte_bus.h
b/lib/librte_eal/common/include/rte_bus.h
index c79368d3c..9e40687e5 100644
--- a/lib/librte_eal/common/include/rte_bus.h
+++ b/lib/librte_eal/common/include/rte_bus.h
@@ -55,6 +55,16 @@ extern "C" {
/** Double linked list of buses */
TAILQ_HEAD(rte_bus_list, rte_bus);
+
+/**
+ * IOVA mapping mode.
+ */
+enum rte_iova_mode {
+ RTE_IOVA_DC = 0,/* Don't care mode */
+ RTE_IOVA_PA = (1 << 0),
+ RTE_IOVA_VA = (1 << 1)
+};
+
/**
* Bus specific scan for devices attached on the bus.
* For each bus object, the scan would be responsible for finding devices and
diff --git a/lib/librte_eal/common/include/rte_pci.h
b/lib/librte_eal/common/include/rte_pci.h
index eab84c7a4..0e36de093 100644
--- a/lib/librte_eal/common/include/rte_pci.h
+++ b/lib/librte_eal/common/include/rte_pci.h
@@ -381,6 +381,17 @@ int
rte_pci_match(const struct rte_pci_driver *pci_drv,
const struct rte_pci_device *pci_dev);
+
+/**
+ * Get iommu class of PCI devices on the bus.
+ * And return their preferred iova mapping mode.
+ *
+ * @return
+ * - enum rte_iova_mode.
+ */
+enum rte_iova_mode
+rte_pci_get_iommu_class(void);
+
/**
* Map the PCI device resources in user space virtual memory address
*
diff --git a/lib/librte_eal/linuxapp/eal/eal_pci.c
b/lib/librte_eal/linuxapp/eal/eal_pci.c
index 8951ce742..26f2be822 100644
--- a/lib/librte_eal/linuxapp/eal/eal_pci.c
+++ b/lib/librte_eal/linuxapp/eal/eal_pci.c
@@ -487,6 +487,15 @@ rte_pci_scan(void)
return -1;
}
+/*
+ * Get iommu class of pci devices on the bus.
+ */
+enum rte_iova_mode
+rte_pci_get_iommu_class(void)
+{
+ return RTE_IOVA_PA;
+}
+
/* Read PCI config space. */
int rte_pci_read_config(const struct rte_pci_device *device,
void *buf, size_t len, off_t offset)
diff --git a/lib/librte_eal/linuxapp/eal/rte_eal_version.map
b/lib/librte_eal/linuxapp/eal/rte_eal_version.map
index 287cc75cd..a8c8ea4f4 100644
--- a/lib/librte_eal/linuxapp/eal/rte_eal_version.map
+++ b/lib/librte_eal/linuxapp/eal/rte_eal_version.map
@@ -248,5 +248,6 @@ DPDK_17.11 {
global:
rte_pci_match;
+ rte_pci_get_iommu_class;
} DPDK_17.08;
--
2.14.1
[dpdk-dev] [PATCH v9 3/9] linuxapp/eal_pci: get iommu class
Get iommu class of PCI device on the bus and returns preferred iova
mapping mode for that bus.
Patch also introduces RTE_PCI_DRV_IOVA_AS_VA drv flag.
Flag used when driver needs to operate in iova=va mode.
Algorithm for iova scheme selection for PCI bus:
0. If no device bound then return with RTE_IOVA_DC mapping mode,
else goto 1).
1. Look for device attached to vfio kdrv and has .drv_flag set
to RTE_PCI_DRV_IOVA_AS_VA.
2. Look for any device attached to UIO class of driver.
3. Check for vfio-noiommu mode enabled.
If 2) & 3) is false and 1) is true then select
mapping scheme as RTE_IOVA_VA. Otherwise use default
mapping scheme (RTE_IOVA_PA).
Signed-off-by: Santosh Shukla
Signed-off-by: Jerin Jacob
Reviewed-by: Maxime Coquelin
Acked-by: Hemant Agrawal
Reviewed-by: Anatoly Burakov
Tested-by: Hemant Agrawal
---
v7 --> v8:
- Replaced 0/1 with false/true boolean value (Suggested by Anatoly)
v6 --> v7:
- squashed v6 series patch no [01/12] & [05/12]..
i.e.. moved RTE_PCI_DRV_IOVA_AS_VA flag into this patch. (Aaron comment).
lib/librte_eal/common/include/rte_pci.h | 2 +
lib/librte_eal/linuxapp/eal/eal_pci.c | 89 -
lib/librte_eal/linuxapp/eal/eal_vfio.c | 19 +++
lib/librte_eal/linuxapp/eal/eal_vfio.h | 4 ++
4 files changed, 113 insertions(+), 1 deletion(-)
diff --git a/lib/librte_eal/common/include/rte_pci.h
b/lib/librte_eal/common/include/rte_pci.h
index 0e36de093..a67d77f22 100644
--- a/lib/librte_eal/common/include/rte_pci.h
+++ b/lib/librte_eal/common/include/rte_pci.h
@@ -202,6 +202,8 @@ struct rte_pci_bus {
#define RTE_PCI_DRV_INTR_RMV 0x0010
/** Device driver needs to keep mapped resources if unsupported dev detected */
#define RTE_PCI_DRV_KEEP_MAPPED_RES 0x0020
+/** Device driver supports iova as va */
+#define RTE_PCI_DRV_IOVA_AS_VA 0X0040
/**
* A structure describing a PCI mapping.
diff --git a/lib/librte_eal/linuxapp/eal/eal_pci.c
b/lib/librte_eal/linuxapp/eal/eal_pci.c
index 26f2be822..2971f1d4f 100644
--- a/lib/librte_eal/linuxapp/eal/eal_pci.c
+++ b/lib/librte_eal/linuxapp/eal/eal_pci.c
@@ -45,6 +45,7 @@
#include "eal_filesystem.h"
#include "eal_private.h"
#include "eal_pci_init.h"
+#include "eal_vfio.h"
/**
* @file
@@ -488,11 +489,97 @@ rte_pci_scan(void)
}
/*
- * Get iommu class of pci devices on the bus.
+ * Is pci device bound to any kdrv
+ */
+static inline int
+pci_device_is_bound(void)
+{
+ struct rte_pci_device *dev = NULL;
+ int ret = 0;
+
+ FOREACH_DEVICE_ON_PCIBUS(dev) {
+ if (dev->kdrv == RTE_KDRV_UNKNOWN ||
+ dev->kdrv == RTE_KDRV_NONE) {
+ continue;
+ } else {
+ ret = 1;
+ break;
+ }
+ }
+ return ret;
+}
+
+/*
+ * Any one of the device bound to uio
+ */
+static inline int
+pci_device_bound_uio(void)
+{
+ struct rte_pci_device *dev = NULL;
+
+ FOREACH_DEVICE_ON_PCIBUS(dev) {
+ if (dev->kdrv == RTE_KDRV_IGB_UIO ||
+ dev->kdrv == RTE_KDRV_UIO_GENERIC) {
+ return 1;
+ }
+ }
+ return 0;
+}
+
+/*
+ * Any one of the device has iova as va
+ */
+static inline int
+pci_device_has_iova_va(void)
+{
+ struct rte_pci_device *dev = NULL;
+ struct rte_pci_driver *drv = NULL;
+
+ FOREACH_DRIVER_ON_PCIBUS(drv) {
+ if (drv && drv->drv_flags & RTE_PCI_DRV_IOVA_AS_VA) {
+ FOREACH_DEVICE_ON_PCIBUS(dev) {
+ if (dev->kdrv == RTE_KDRV_VFIO &&
+ rte_pci_match(drv, dev))
+ return 1;
+ }
+ }
+ }
+ return 0;
+}
+
+/*
+ * Get iommu class of PCI devices on the bus.
*/
enum rte_iova_mode
rte_pci_get_iommu_class(void)
{
+ bool is_bound;
+ bool is_vfio_noiommu_enabled = true;
+ bool has_iova_va;
+ bool is_bound_uio;
+
+ is_bound = pci_device_is_bound();
+ if (!is_bound)
+ return RTE_IOVA_DC;
+
+ has_iova_va = pci_device_has_iova_va();
+ is_bound_uio = pci_device_bound_uio();
+#ifdef VFIO_PRESENT
+ is_vfio_noiommu_enabled = vfio_noiommu_is_enabled() == true ?
+ true : false;
+#endif
+
+ if (has_iova_va && !is_bound_uio && !is_vfio_noiommu_enabled)
+ return RTE_IOVA_VA;
+
+ if (has_iova_va) {
+ RTE_LOG(WARNING, EAL, "Some devices want iova as va but pa will
be used because.. ");
+ if (is_vfio_noiommu_enabled)
+ RTE_LOG(WARNING, EAL, "vfio-noiommu mode configured\n");
+ if (is_bound_uio)
+ RTE_LOG(WARNING, EAL, "few device bound to UIO\n");
+ }
+
return RTE_IOVA_PA;
}
diff --git a/lib/librte_eal/linuxapp/eal/eal_vfio.c
b/lib/librte_eal/linuxapp/eal/eal_vfio.c
[dpdk-dev] [PATCH v9 0/9] Infrastructure to detect iova mapping on the bus
v9: - Added Tested-By: to series. - Includes minor changes related to linuxapp api stub in [02/09] (Suggested by Anatoly) - Series rebased on tip commit : aee62e90 v8: Includes minor review changes per v7 review comment from Anatoly. Patches rebased on Tip commit:3d2e0448eb. v7: Includes no major change, minor change detailing: - patch sqashing (Aaron suggestion) - added run_once for device_parse() and bus_scan() in eal init (Aaron suggestion) - Moved rte_eal_device_parse() up in eal initialization order. - Patches rebased on top of version: 17.11-rc0 For v6 info refer [11]. v6: Sending v5 series rebased on top of version: 17.11-rc0. v5: Introducing RTE_PCI_DRV_IOVA_AS_VA flag for autodetection of iova va mapping. If a PCI driver demand for IOVA as VA scheme then the driver can add it in the PCI driver registration function. Algorithm to select IOVA as VA for PCI bus case: 0. If no device bound then return with RTE_IOVA_DC mapping mode, else goto 1). 1. Look for device attached to vfio kdrv and has .drv_flag set to RTE_PCI_DRV_IOVA_AS_VA. 2. Look for any device attached to UIO class of driver. 3. Check for vfio-noiommu mode enabled. If 2) & 3) is false and 1) is true then select mapping scheme as RTE_IOVA_VA. Otherwise use default mapping scheme (RTE_IOVA_PA). That way, Bus can truly autodetect the iova mapping mode for a device Or a set of the device. Change History: v8 --> v9: - Added Tested-by: signature of Hemant. - Added linuxapp stub api definition in [02/09] (Suggested by Amatoly) v7 --> v8: - Replace 0 / 1 with true/false boolean values (Suggested by Anatoly). v6 --> v7: - Patches squashed per v6. - Added run_once in eal per v6. - Moved rte_eal_device_parse() up in eal init oder. v5 --> v6: - Added api info in eal's versiom.map (release DPDK_v17.11). v4 --> v5: - Change DPDK_17.08 to DPDK_17.11 in _version.map. - Reworded bus api description (suggested by Hemant). - Added reviewed-by from Maxime in v5. - Added acked-by from Hemant for pci and bus patches. v3 --> v4: - Re-introduced RTE_IOVA_DEC mode (Suggested by Hemant [5]). - Renamed flag to RTE_PCI_DRV_IOVA_AS_VA (Suggested by Maxime). - Reworded WARNING message(suggested by Maxime[7]). - Created a separate patch for rte_pci_get_iommu_class (suggested by Maxime[]). - Added VFIO_PRESENT ifdef build fix. v2 --> v3: - Removed rte_mempool_virt2phy (suggested by Olivier [4]) v1 --> v2: - Removed override eal option i.e. (--iova-mode=<>) Because we have means to truly autodetect the iova mode. - Introduced RTE_PCI_DRV_NEED_IOVA_VA drv_flag (Suggested by Maxime [3]). - Using NEED_IOVA_VA drv_flag in autodetection logic. - Removed Linux version check macro in vfio code, As per Maxime feedback. - Moved rte_pci_match API from local to global. Patch Summary: 1) 1nd: declare rte_pci_match api in pci header. Required for autodetection in follow up patches. 2) 2nd - 3rd - 4th : autodetection mapping infrastructure for Linux/bsdapp. 3) 5th: iova mode helper API. 4) 6th: Infra to detect iova mode. 5) 7th: make vfio mapping iova aware. 6) 8th - 9th : Check for IOVA_VA mode in below APIs - rte_mem_virt2phy - rte_malloc_virt2phy Test History: - Tested for x86/XL710 40G NIC card for both modes (iova_va/pa). - Tested for arm64/thunderx vNIC Integrated NIC for both modes - Tested for arm64/Octeontx integrated NICs for only Iova_va mode(It supports only one mode.) - Ran standalone tests like mempool_autotest, mbuf_autotest. - Verified for Doxygen. Work History: For v1, Refer [1]. For v2, Refer [2]. For v3, Refer [9]. For v4, refer [10]. for v6, refer [11]. Checkpatch result: * None Thanks., [1] https://www.mail-archive.com/dev@dpdk.org/msg67438.html [2] https://www.mail-archive.com/dev@dpdk.org/msg70674.html [3] https://www.mail-archive.com/dev@dpdk.org/msg70279.html [4] https://www.mail-archive.com/dev@dpdk.org/msg70692.html [5] http://dpdk.org/ml/archives/dev/2017-July/071282.html [6] http://dpdk.org/ml/archives/dev/2017-July/070951.html [7] http://dpdk.org/ml/archives/dev/2017-July/070941.html [8] http://dpdk.org/ml/archives/dev/2017-July/070952.html [9] http://dpdk.org/ml/archives/dev/2017-July/070918.html [10] http://dpdk.org/ml/archives/dev/2017-July/071754.html [11] http://dpdk.org/ml/archives/dev/2017-August/072871.html Santosh Shukla (9): eal/pci: export match function eal/pci: get iommu class linuxapp/eal_pci: get iommu class bus: get iommu class eal: introduce helper API for iova mode eal: auto detect iova mode linuxapp/eal_vfio: honor iova mode before mapping linuxapp/eal_memory: honor iova mode in virt2phy eal/rte_malloc: honor iova mode in virt2phy lib/librte_eal/bsdapp/eal/eal.c | 33 ++--- lib/librte_eal/bsdapp/eal/eal_pci.c | 10 +++ lib/librte_eal/bsdapp/eal/rte_eal_version.map | 10 +++ lib/librte_eal/common/eal_common_bus.c | 23 ++ lib/librte_eal/common/eal_common_pci.c | 11
[dpdk-dev] [PATCH v9 4/9] bus: get iommu class
API(rte_bus_get_iommu_class) helps to automatically detect and select
appropriate iova mapping scheme for iommu capable device on that bus.
Algorithm for iova scheme selection for bus:
0. Iterate through bus_list.
1. Collect each bus iova mode value and update into 'mode' var.
2. Mode selection scheme is:
if mode == 0 then iova mode is _pa,
if mode == 1 then iova mode is _pa,
if mode == 2 then iova mode is _va,
if mode == 3 then iova mode ia _pa.
So mode !=2 will be default iova mode (_pa).
Signed-off-by: Santosh Shukla
Signed-off-by: Jerin Jacob
Reviewed-by: Maxime Coquelin
Acked-by: Hemant Agrawal
Reviewed-by: Anatoly Burakov
Tested-by: Hemant Agrawal
---
lib/librte_eal/bsdapp/eal/rte_eal_version.map | 1 +
lib/librte_eal/common/eal_common_bus.c | 23 +++
lib/librte_eal/common/eal_common_pci.c | 1 +
lib/librte_eal/common/include/rte_bus.h | 25 +
lib/librte_eal/linuxapp/eal/rte_eal_version.map | 1 +
5 files changed, 51 insertions(+)
diff --git a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
index c6ffd9399..3466eaf20 100644
--- a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
+++ b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
@@ -244,5 +244,6 @@ DPDK_17.11 {
rte_pci_match;
rte_pci_get_iommu_class;
+ rte_bus_get_iommu_class;
} DPDK_17.08;
diff --git a/lib/librte_eal/common/eal_common_bus.c
b/lib/librte_eal/common/eal_common_bus.c
index 08bec2d93..a30a8982e 100644
--- a/lib/librte_eal/common/eal_common_bus.c
+++ b/lib/librte_eal/common/eal_common_bus.c
@@ -222,3 +222,26 @@ rte_bus_find_by_device_name(const char *str)
c[0] = '\0';
return rte_bus_find(NULL, bus_can_parse, name);
}
+
+
+/*
+ * Get iommu class of devices on the bus.
+ */
+enum rte_iova_mode
+rte_bus_get_iommu_class(void)
+{
+ int mode = RTE_IOVA_DC;
+ struct rte_bus *bus;
+
+ TAILQ_FOREACH(bus, &rte_bus_list, next) {
+
+ if (bus->get_iommu_class)
+ mode |= bus->get_iommu_class();
+ }
+
+ if (mode != RTE_IOVA_VA) {
+ /* Use default IOVA mode */
+ mode = RTE_IOVA_PA;
+ }
+ return mode;
+}
diff --git a/lib/librte_eal/common/eal_common_pci.c
b/lib/librte_eal/common/eal_common_pci.c
index 3b7d0a0ee..0f0e4b93b 100644
--- a/lib/librte_eal/common/eal_common_pci.c
+++ b/lib/librte_eal/common/eal_common_pci.c
@@ -564,6 +564,7 @@ struct rte_pci_bus rte_pci_bus = {
.plug = pci_plug,
.unplug = pci_unplug,
.parse = pci_parse,
+ .get_iommu_class = rte_pci_get_iommu_class,
},
.device_list = TAILQ_HEAD_INITIALIZER(rte_pci_bus.device_list),
.driver_list = TAILQ_HEAD_INITIALIZER(rte_pci_bus.driver_list),
diff --git a/lib/librte_eal/common/include/rte_bus.h
b/lib/librte_eal/common/include/rte_bus.h
index 9e40687e5..70a291a4d 100644
--- a/lib/librte_eal/common/include/rte_bus.h
+++ b/lib/librte_eal/common/include/rte_bus.h
@@ -178,6 +178,20 @@ struct rte_bus_conf {
enum rte_bus_scan_mode scan_mode; /**< Scan policy. */
};
+
+/**
+ * Get common iommu class of the all the devices on the bus. The bus may
+ * check that those devices are attached to iommu driver.
+ * If no devices are attached to the bus. The bus may return with don't care
+ * (_DC) value.
+ * Otherwise, The bus will return appropriate _pa or _va iova mode.
+ *
+ * @return
+ * enum rte_iova_mode value.
+ */
+typedef enum rte_iova_mode (*rte_bus_get_iommu_class_t)(void);
+
+
/**
* A structure describing a generic bus.
*/
@@ -191,6 +205,7 @@ struct rte_bus {
rte_bus_unplug_t unplug; /**< Remove single device from driver */
rte_bus_parse_t parse; /**< Parse a device name */
struct rte_bus_conf conf;/**< Bus configuration */
+ rte_bus_get_iommu_class_t get_iommu_class; /**< Get iommu class */
};
/**
@@ -290,6 +305,16 @@ struct rte_bus *rte_bus_find_by_device(const struct
rte_device *dev);
*/
struct rte_bus *rte_bus_find_by_name(const char *busname);
+
+/**
+ * Get the common iommu class of devices bound on to buses available in the
+ * system. The default mode is PA.
+ *
+ * @return
+ * enum rte_iova_mode value.
+ */
+enum rte_iova_mode rte_bus_get_iommu_class(void);
+
/**
* Helper for Bus registration.
* The constructor has higher priority than PMD constructors.
diff --git a/lib/librte_eal/linuxapp/eal/rte_eal_version.map
b/lib/librte_eal/linuxapp/eal/rte_eal_version.map
index a8c8ea4f4..9115aa3e9 100644
--- a/lib/librte_eal/linuxapp/eal/rte_eal_version.map
+++ b/lib/librte_eal/linuxapp/eal/rte_eal_version.map
@@ -249,5 +249,6 @@ DPDK_17.11 {
rte_pci_match;
rte_pci_get_iommu_class;
+ rte_bus_get_iommu_class;
} DPDK_17.08;
--
2.14.1
[dpdk-dev] [PATCH v9 1/9] eal/pci: export match function
Export rte_pci_match() function as it needed in the followup patch.
Signed-off-by: Santosh Shukla
Signed-off-by: Jerin Jacob
Acked-by: Maxime Coquelin
Reviewed-by: Anatoly Burakov
Tested-by: Hemant Agrawal
---
lib/librte_eal/bsdapp/eal/rte_eal_version.map | 7 +++
lib/librte_eal/common/eal_common_pci.c | 10 +-
lib/librte_eal/common/include/rte_pci.h | 15 +++
lib/librte_eal/linuxapp/eal/rte_eal_version.map | 7 +++
4 files changed, 30 insertions(+), 9 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
index 47a09ea7f..cfbf8fbd0 100644
--- a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
+++ b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
@@ -238,3 +238,10 @@ EXPERIMENTAL {
rte_service_start_with_defaults;
} DPDK_17.08;
+
+DPDK_17.11 {
+ global:
+
+ rte_pci_match;
+
+} DPDK_17.08;
diff --git a/lib/librte_eal/common/eal_common_pci.c
b/lib/librte_eal/common/eal_common_pci.c
index 52fd38cdd..3b7d0a0ee 100644
--- a/lib/librte_eal/common/eal_common_pci.c
+++ b/lib/librte_eal/common/eal_common_pci.c
@@ -150,16 +150,8 @@ pci_unmap_resource(void *requested_addr, size_t size)
/*
* Match the PCI Driver and Device using the ID Table
- *
- * @param pci_drv
- * PCI driver from which ID table would be extracted
- * @param pci_dev
- * PCI device to match against the driver
- * @return
- * 1 for successful match
- * 0 for unsuccessful match
*/
-static int
+int
rte_pci_match(const struct rte_pci_driver *pci_drv,
const struct rte_pci_device *pci_dev)
{
diff --git a/lib/librte_eal/common/include/rte_pci.h
b/lib/librte_eal/common/include/rte_pci.h
index 8b123391c..eab84c7a4 100644
--- a/lib/librte_eal/common/include/rte_pci.h
+++ b/lib/librte_eal/common/include/rte_pci.h
@@ -366,6 +366,21 @@ int rte_pci_scan(void);
int
rte_pci_probe(void);
+/*
+ * Match the PCI Driver and Device using the ID Table
+ *
+ * @param pci_drv
+ * PCI driver from which ID table would be extracted
+ * @param pci_dev
+ * PCI device to match against the driver
+ * @return
+ * 1 for successful match
+ * 0 for unsuccessful match
+ */
+int
+rte_pci_match(const struct rte_pci_driver *pci_drv,
+ const struct rte_pci_device *pci_dev);
+
/**
* Map the PCI device resources in user space virtual memory address
*
diff --git a/lib/librte_eal/linuxapp/eal/rte_eal_version.map
b/lib/librte_eal/linuxapp/eal/rte_eal_version.map
index 8c08b8d1e..287cc75cd 100644
--- a/lib/librte_eal/linuxapp/eal/rte_eal_version.map
+++ b/lib/librte_eal/linuxapp/eal/rte_eal_version.map
@@ -243,3 +243,10 @@ EXPERIMENTAL {
rte_service_start_with_defaults;
} DPDK_17.08;
+
+DPDK_17.11 {
+ global:
+
+ rte_pci_match;
+
+} DPDK_17.08;
--
2.14.1
[dpdk-dev] [PATCH v9 5/9] eal: introduce helper API for iova mode
Introducing rte_eal_iova_mode() helper API. This API
used by non-eal library for detecting iova mode.
Signed-off-by: Santosh Shukla
Signed-off-by: Jerin Jacob
Reviewed-by: Maxime Coquelin
Tested-by: Hemant Agrawal
---
lib/librte_eal/bsdapp/eal/eal.c | 6 ++
lib/librte_eal/bsdapp/eal/rte_eal_version.map | 1 +
lib/librte_eal/common/include/rte_eal.h | 12
lib/librte_eal/linuxapp/eal/eal.c | 6 ++
lib/librte_eal/linuxapp/eal/rte_eal_version.map | 1 +
5 files changed, 26 insertions(+)
diff --git a/lib/librte_eal/bsdapp/eal/eal.c b/lib/librte_eal/bsdapp/eal/eal.c
index 5fa598842..07e72203f 100644
--- a/lib/librte_eal/bsdapp/eal/eal.c
+++ b/lib/librte_eal/bsdapp/eal/eal.c
@@ -119,6 +119,12 @@ rte_eal_get_configuration(void)
return &rte_config;
}
+enum rte_iova_mode
+rte_eal_iova_mode(void)
+{
+ return rte_eal_get_configuration()->iova_mode;
+}
+
/* parse a sysfs (or other) file containing one integer value */
int
eal_parse_sysfs_value(const char *filename, unsigned long *val)
diff --git a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
index 3466eaf20..6bed74dff 100644
--- a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
+++ b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
@@ -245,5 +245,6 @@ DPDK_17.11 {
rte_pci_match;
rte_pci_get_iommu_class;
rte_bus_get_iommu_class;
+ rte_eal_iova_mode;
} DPDK_17.08;
diff --git a/lib/librte_eal/common/include/rte_eal.h
b/lib/librte_eal/common/include/rte_eal.h
index 0e7363d77..932dc1a96 100644
--- a/lib/librte_eal/common/include/rte_eal.h
+++ b/lib/librte_eal/common/include/rte_eal.h
@@ -45,6 +45,7 @@
#include
#include
+#include
#ifdef __cplusplus
extern "C" {
@@ -87,6 +88,9 @@ struct rte_config {
/** Primary or secondary configuration */
enum rte_proc_type_t process_type;
+ /** PA or VA mapping mode */
+ enum rte_iova_mode iova_mode;
+
/**
* Pointer to memory configuration, which may be shared across multiple
* DPDK instances
@@ -287,6 +291,14 @@ static inline int rte_gettid(void)
return RTE_PER_LCORE(_thread_id);
}
+/**
+ * Get the iova mode
+ *
+ * @return
+ * enum rte_iova_mode value.
+ */
+enum rte_iova_mode rte_eal_iova_mode(void);
+
#define RTE_INIT(func) \
static void __attribute__((constructor, used)) func(void)
diff --git a/lib/librte_eal/linuxapp/eal/eal.c
b/lib/librte_eal/linuxapp/eal/eal.c
index 48f12f44c..febbafdb3 100644
--- a/lib/librte_eal/linuxapp/eal/eal.c
+++ b/lib/librte_eal/linuxapp/eal/eal.c
@@ -128,6 +128,12 @@ rte_eal_get_configuration(void)
return &rte_config;
}
+enum rte_iova_mode
+rte_eal_iova_mode(void)
+{
+ return rte_eal_get_configuration()->iova_mode;
+}
+
/* parse a sysfs (or other) file containing one integer value */
int
eal_parse_sysfs_value(const char *filename, unsigned long *val)
diff --git a/lib/librte_eal/linuxapp/eal/rte_eal_version.map
b/lib/librte_eal/linuxapp/eal/rte_eal_version.map
index 9115aa3e9..8e49bf5fa 100644
--- a/lib/librte_eal/linuxapp/eal/rte_eal_version.map
+++ b/lib/librte_eal/linuxapp/eal/rte_eal_version.map
@@ -250,5 +250,6 @@ DPDK_17.11 {
rte_pci_match;
rte_pci_get_iommu_class;
rte_bus_get_iommu_class;
+ rte_eal_iova_mode;
} DPDK_17.08;
--
2.14.1
[dpdk-dev] [PATCH v9 6/9] eal: auto detect iova mode
For auto detection purpose:
* Below calls moved up in the eal initialization order:
- eal_option_device_parse
- rte_bus_scan
Based on the result of rte_bus_scan_iommu_class - select iova
mapping mode.
Signed-off-by: Santosh Shukla
Signed-off-by: Jerin Jacob
Reviewed-by: Maxime Coquelin
Reviewed-by: Anatoly Burakov
Tested-by: Hemant Agrawal
---
v6 --> v7:
- Moved eal_option_device_parse() up in then order of eal init.
- Added run_once. (aaron suggestion).
- squashed v6 series patch no. [08/12] & [09/12] into one patch (Aaron
comment)
lib/librte_eal/bsdapp/eal/eal.c | 27 ---
lib/librte_eal/linuxapp/eal/eal.c | 27 ---
2 files changed, 32 insertions(+), 22 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal.c b/lib/librte_eal/bsdapp/eal/eal.c
index 07e72203f..f003f4c04 100644
--- a/lib/librte_eal/bsdapp/eal/eal.c
+++ b/lib/librte_eal/bsdapp/eal/eal.c
@@ -541,6 +541,22 @@ rte_eal_init(int argc, char **argv)
return -1;
}
+ if (eal_option_device_parse()) {
+ rte_errno = ENODEV;
+ rte_atomic32_clear(&run_once);
+ return -1;
+ }
+
+ if (rte_bus_scan()) {
+ rte_eal_init_alert("Cannot scan the buses for devices\n");
+ rte_errno = ENODEV;
+ rte_atomic32_clear(&run_once);
+ return -1;
+ }
+
+ /* autodetect the iova mapping mode (default is iova_pa) */
+ rte_eal_get_configuration()->iova_mode = rte_bus_get_iommu_class();
+
if (internal_config.no_hugetlbfs == 0 &&
internal_config.process_type != RTE_PROC_SECONDARY &&
eal_hugepage_info_init() < 0) {
@@ -620,17 +636,6 @@ rte_eal_init(int argc, char **argv)
rte_config.master_lcore, thread_id, cpuset,
ret == 0 ? "" : "...");
- if (eal_option_device_parse()) {
- rte_errno = ENODEV;
- return -1;
- }
-
- if (rte_bus_scan()) {
- rte_eal_init_alert("Cannot scan the buses for devices\n");
- rte_errno = ENODEV;
- return -1;
- }
-
RTE_LCORE_FOREACH_SLAVE(i) {
/*
diff --git a/lib/librte_eal/linuxapp/eal/eal.c
b/lib/librte_eal/linuxapp/eal/eal.c
index febbafdb3..f4901ffb6 100644
--- a/lib/librte_eal/linuxapp/eal/eal.c
+++ b/lib/librte_eal/linuxapp/eal/eal.c
@@ -798,6 +798,22 @@ rte_eal_init(int argc, char **argv)
return -1;
}
+ if (eal_option_device_parse()) {
+ rte_errno = ENODEV;
+ rte_atomic32_clear(&run_once);
+ return -1;
+ }
+
+ if (rte_bus_scan()) {
+ rte_eal_init_alert("Cannot scan the buses for devices\n");
+ rte_errno = ENODEV;
+ rte_atomic32_clear(&run_once);
+ return -1;
+ }
+
+ /* autodetect the iova mapping mode (default is iova_pa) */
+ rte_eal_get_configuration()->iova_mode = rte_bus_get_iommu_class();
+
if (internal_config.no_hugetlbfs == 0 &&
internal_config.process_type != RTE_PROC_SECONDARY &&
internal_config.xen_dom0_support == 0 &&
@@ -895,17 +911,6 @@ rte_eal_init(int argc, char **argv)
return -1;
}
- if (eal_option_device_parse()) {
- rte_errno = ENODEV;
- return -1;
- }
-
- if (rte_bus_scan()) {
- rte_eal_init_alert("Cannot scan the buses for devices\n");
- rte_errno = ENODEV;
- return -1;
- }
-
RTE_LCORE_FOREACH_SLAVE(i) {
/*
--
2.14.1
[dpdk-dev] [PATCH v9 7/9] linuxapp/eal_vfio: honor iova mode before mapping
Check iova mode and accordingly map iova to pa or va. Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob Reviewed-by: Maxime Coquelin Acked-by: Anatoly Burakov Tested-by: Hemant Agrawal --- lib/librte_eal/linuxapp/eal/eal_vfio.c | 10 -- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/lib/librte_eal/linuxapp/eal/eal_vfio.c b/lib/librte_eal/linuxapp/eal/eal_vfio.c index c8a97b7e7..b32cd09a2 100644 --- a/lib/librte_eal/linuxapp/eal/eal_vfio.c +++ b/lib/librte_eal/linuxapp/eal/eal_vfio.c @@ -706,7 +706,10 @@ vfio_type1_dma_map(int vfio_container_fd) dma_map.argsz = sizeof(struct vfio_iommu_type1_dma_map); dma_map.vaddr = ms[i].addr_64; dma_map.size = ms[i].len; - dma_map.iova = ms[i].phys_addr; + if (rte_eal_iova_mode() == RTE_IOVA_VA) + dma_map.iova = dma_map.vaddr; + else + dma_map.iova = ms[i].phys_addr; dma_map.flags = VFIO_DMA_MAP_FLAG_READ | VFIO_DMA_MAP_FLAG_WRITE; ret = ioctl(vfio_container_fd, VFIO_IOMMU_MAP_DMA, &dma_map); @@ -792,7 +795,10 @@ vfio_spapr_dma_map(int vfio_container_fd) dma_map.argsz = sizeof(struct vfio_iommu_type1_dma_map); dma_map.vaddr = ms[i].addr_64; dma_map.size = ms[i].len; - dma_map.iova = ms[i].phys_addr; + if (rte_eal_iova_mode() == RTE_IOVA_VA) + dma_map.iova = dma_map.vaddr; + else + dma_map.iova = ms[i].phys_addr; dma_map.flags = VFIO_DMA_MAP_FLAG_READ | VFIO_DMA_MAP_FLAG_WRITE; -- 2.14.1
[dpdk-dev] [PATCH v9 9/9] eal/rte_malloc: honor iova mode in virt2phy
Check iova mode and accordingly return phy addr.
Signed-off-by: Santosh Shukla
Signed-off-by: Jerin Jacob
Reviewed-by: Maxime Coquelin
Reviewed-by: Anatoly Burakov
Tested-by: Hemant Agrawal
---
lib/librte_eal/common/rte_malloc.c | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/lib/librte_eal/common/rte_malloc.c
b/lib/librte_eal/common/rte_malloc.c
index 5c0627bf4..d65c05a4d 100644
--- a/lib/librte_eal/common/rte_malloc.c
+++ b/lib/librte_eal/common/rte_malloc.c
@@ -251,10 +251,17 @@ rte_malloc_set_limit(__rte_unused const char *type,
phys_addr_t
rte_malloc_virt2phy(const void *addr)
{
+ phys_addr_t paddr;
const struct malloc_elem *elem = malloc_elem_from_data(addr);
if (elem == NULL)
return RTE_BAD_PHYS_ADDR;
if (elem->ms->phys_addr == RTE_BAD_PHYS_ADDR)
return RTE_BAD_PHYS_ADDR;
- return elem->ms->phys_addr + ((uintptr_t)addr -
(uintptr_t)elem->ms->addr);
+
+ if (rte_eal_iova_mode() == RTE_IOVA_VA)
+ paddr = (uintptr_t)addr;
+ else
+ paddr = elem->ms->phys_addr +
+ ((uintptr_t)addr - (uintptr_t)elem->ms->addr);
+ return paddr;
}
--
2.14.1
[dpdk-dev] [PATCH v9 8/9] linuxapp/eal_memory: honor iova mode in virt2phy
Check iova mode and accordingly return phy addr.
Signed-off-by: Santosh Shukla
Signed-off-by: Jerin Jacob
Reviewed-by: Maxime Coquelin
Reviewed-by: Anatoly Burakov
Tested-by: Hemant Agrawal
---
lib/librte_eal/linuxapp/eal/eal_memory.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/lib/librte_eal/linuxapp/eal/eal_memory.c
b/lib/librte_eal/linuxapp/eal/eal_memory.c
index 52791282f..2d9d7c2dc 100644
--- a/lib/librte_eal/linuxapp/eal/eal_memory.c
+++ b/lib/librte_eal/linuxapp/eal/eal_memory.c
@@ -139,6 +139,9 @@ rte_mem_virt2phy(const void *virtaddr)
int page_size;
off_t offset;
+ if (rte_eal_iova_mode() == RTE_IOVA_VA)
+ return (uintptr_t)virtaddr;
+
/* when using dom0, /proc/self/pagemap always returns 0, check in
* dpdk memory by browsing the memsegs */
if (rte_xen_dom0_supported()) {
--
2.14.1
Re: [dpdk-dev] [PATCH 2/3] build: fix driver install path
+Timothy, since git doesn't seem to automatically add reporters on CC.
On Wed, Sep 20, 2017 at 11:51:06AM +0100, Bruce Richardson wrote:
> To comply with the "hier" standard [Ref: man 7 hier], the driver .so files
> should not be placed in $datadir. Therefore we install them in a
> sub-directory of $libdir instead.
>
> Fixes: d123bba2dfbe ("build: add initial infrastructure for meson & ninja
> builds")
>
> Reported-by: Timothy M. Redaelli
> Signed-off-by: Bruce Richardson
> ---
> meson.build | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/meson.build b/meson.build
> index 41426a706..6e4752562 100644
> --- a/meson.build
> +++ b/meson.build
> @@ -48,7 +48,7 @@ dpdk_extra_ldflags = []
> if get_option('default_library') == 'static'
> driver_install_path = get_option('libdir')
> else
> - driver_install_path = join_paths(get_option('datadir'), 'dpdk/drivers')
> + driver_install_path = join_paths(get_option('libdir'), 'dpdk/drivers')
> endif
>
> # configure the build, and make sure configs here and in config folder are
> --
> 2.13.5
>
[dpdk-dev] [PATCH v2 2/2] test/ring: do not mask result of enqueue or dequeue
The define RTE_RING_SZ_MASK is the maximum size supported by the
rte_ring. The size is checked at ring creation.
There is no reason today to mask the result of
rte_ring_sp_enqueue_burst() or rte_ring_sc_dequeue_burst() with this
value. The flag RTE_RING_QUOT_EXCEED was previously included in the
returned value but it was removed in
commit 77dd3064270c ("ring: remove watermark support").
Signed-off-by: Olivier Matz
Reviewed-by: Anatoly Burakov
---
v1->v2
- fix checkpatch issues
test/test/test_ring.c | 59 +--
1 file changed, 29 insertions(+), 30 deletions(-)
diff --git a/test/test/test_ring.c b/test/test/test_ring.c
index d664b04b7..d26e75269 100644
--- a/test/test/test_ring.c
+++ b/test/test/test_ring.c
@@ -375,37 +375,37 @@ test_ring_burst_basic(void)
printf("enqueue 1 obj\n");
ret = rte_ring_sp_enqueue_burst(r, cur_src, 1, NULL);
cur_src += 1;
- if ((ret & RTE_RING_SZ_MASK) != 1)
+ if (ret != 1)
goto fail;
printf("enqueue 2 objs\n");
ret = rte_ring_sp_enqueue_burst(r, cur_src, 2, NULL);
cur_src += 2;
- if ((ret & RTE_RING_SZ_MASK) != 2)
+ if (ret != 2)
goto fail;
printf("enqueue MAX_BULK objs\n");
ret = rte_ring_sp_enqueue_burst(r, cur_src, MAX_BULK, NULL);
cur_src += MAX_BULK;
- if ((ret & RTE_RING_SZ_MASK) != MAX_BULK)
+ if (ret != MAX_BULK)
goto fail;
printf("dequeue 1 obj\n");
ret = rte_ring_sc_dequeue_burst(r, cur_dst, 1, NULL);
cur_dst += 1;
- if ((ret & RTE_RING_SZ_MASK) != 1)
+ if (ret != 1)
goto fail;
printf("dequeue 2 objs\n");
ret = rte_ring_sc_dequeue_burst(r, cur_dst, 2, NULL);
cur_dst += 2;
- if ((ret & RTE_RING_SZ_MASK) != 2)
+ if (ret != 2)
goto fail;
printf("dequeue MAX_BULK objs\n");
ret = rte_ring_sc_dequeue_burst(r, cur_dst, MAX_BULK, NULL);
cur_dst += MAX_BULK;
- if ((ret & RTE_RING_SZ_MASK) != MAX_BULK)
+ if (ret != MAX_BULK)
goto fail;
/* check data */
@@ -423,22 +423,21 @@ test_ring_burst_basic(void)
for (i = 0; i< (RING_SIZE/MAX_BULK - 1); i++) {
ret = rte_ring_sp_enqueue_burst(r, cur_src, MAX_BULK, NULL);
cur_src += MAX_BULK;
- if ((ret & RTE_RING_SZ_MASK) != MAX_BULK) {
+ if (ret != MAX_BULK)
goto fail;
- }
}
printf("Enqueue 2 objects, free entries = MAX_BULK - 2 \n");
ret = rte_ring_sp_enqueue_burst(r, cur_src, 2, NULL);
cur_src += 2;
- if ((ret & RTE_RING_SZ_MASK) != 2)
+ if (ret != 2)
goto fail;
printf("Enqueue the remaining entries = MAX_BULK - 2 \n");
/* Always one free entry left */
ret = rte_ring_sp_enqueue_burst(r, cur_src, MAX_BULK, NULL);
cur_src += MAX_BULK - 3;
- if ((ret & RTE_RING_SZ_MASK) != MAX_BULK - 3)
+ if (ret != MAX_BULK - 3)
goto fail;
printf("Test if ring is full \n");
@@ -447,26 +446,26 @@ test_ring_burst_basic(void)
printf("Test enqueue for a full entry \n");
ret = rte_ring_sp_enqueue_burst(r, cur_src, MAX_BULK, NULL);
- if ((ret & RTE_RING_SZ_MASK) != 0)
+ if (ret != 0)
goto fail;
printf("Test dequeue without enough objects \n");
for (i = 0; i
[dpdk-dev] [PATCH v2 1/2] ring: increase maximum ring size
There is no reason to prevent ring from being larger than 0x0FFF.
Increase the maximum size to 0x7FFF, which is the maximum possible
without changing the code and the structure definition (size is stored
on a uint32_t).
Link: http://dpdk.org/ml/archives/dev/2017-September/074701.html
Suggested-by: Venkatesh Nuthula
Signed-off-by: Olivier Matz
Reviewed-by: Anatoly Burakov
---
v1->v2
- fix checkpatch issues
lib/librte_ring/rte_ring.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/librte_ring/rte_ring.h b/lib/librte_ring/rte_ring.h
index 8f5a4937f..5e9b3b7b4 100644
--- a/lib/librte_ring/rte_ring.h
+++ b/lib/librte_ring/rte_ring.h
@@ -174,7 +174,7 @@ struct rte_ring {
* ring space will be wasted.
*/
#define RING_F_EXACT_SZ 0x0004
-#define RTE_RING_SZ_MASK (unsigned)(0x0fff) /**< Ring size mask */
+#define RTE_RING_SZ_MASK (0x7fffU) /**< Ring size mask */
/* @internal defines for passing to the enqueue dequeue worker functions */
#define __IS_SP 1
--
2.11.0
Re: [dpdk-dev] [PATCH 01/11] lib/rte_security: add security library
Hi Hemant,
On 9/15/2017 11:02 AM, Hemant Agrawal wrote:
Hi,
On 9/14/2017 1:56 PM, Akhil Goyal wrote:
..
diff --git a/lib/librte_security/rte_security.c
b/lib/librte_security/rte_security.c
new file mode 100644
index 000..5776246
--- /dev/null
+++ b/lib/librte_security/rte_security.c
@@ -0,0 +1,252 @@
+/*-
+ * BSD LICENSE
+ *
+ * Copyright 2017 NXP.
+ * Copyright(c) 2017 Intel Corporation. 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 NXP 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.
+ */
+
+#include
+#include
+
+#include "rte_security.h"
+#include "rte_security_driver.h"
+
+#define RTE_SECURITY_INSTANCES_BLOCK_ALLOC_SZ (8)
+
+struct rte_security_ctx {
+ uint16_t id;
+ enum {
+ RTE_SECURITY_INSTANCE_INVALID = 0,
+ RTE_SECURITY_INSTANCE_VALID
+ } state;
+ void *device;
+ struct rte_security_ops *ops;
+};
+
+static struct rte_security_ctx *security_instances;
+static uint16_t max_nb_security_instances;
+static uint16_t nb_security_instances;
+
+static int
+rte_security_is_valid_id(uint16_t id)
+{
+ if (id >= nb_security_instances ||
+ (security_instances[id].state != RTE_SECURITY_INSTANCE_VALID))
+ return 0;
+ else
+ return 1;
+}
+
+/* Macros to check for valid id */
+#define RTE_SEC_VALID_ID_OR_ERR_RET(id, retval) do { \
+ if (!rte_security_is_valid_id(id)) { \
+ RTE_PMD_DEBUG_TRACE("Invalid sec_id=%d\n", id); \
+ return retval; \
+ } \
+} while (0)
+
+#define RTE_SEC_VALID_ID_OR_RET(id) do { \
+ if (!rte_security_is_valid_id(id)) { \
+ RTE_PMD_DEBUG_TRACE("Invalid sec_id=%d\n", id); \
+ return; \
+ } \
+} while (0)
+
+int
+rte_security_register(uint16_t *id, void *device,
+ struct rte_security_ops *ops)
+{
+ if (max_nb_security_instances == 0) {
+ security_instances = rte_malloc(
+ "rte_security_instances_ops",
+ sizeof(*security_instances) *
+ RTE_SECURITY_INSTANCES_BLOCK_ALLOC_SZ, 0);
+
+ if (security_instances == NULL)
+ return -ENOMEM;
+ max_nb_security_instances =
+ RTE_SECURITY_INSTANCES_BLOCK_ALLOC_SZ;
+ } else if (nb_security_instances >= max_nb_security_instances) {
+ uint16_t *instances = rte_realloc(security_instances,
+ sizeof(struct rte_security_ops *) *
+ (max_nb_security_instances +
+ RTE_SECURITY_INSTANCES_BLOCK_ALLOC_SZ), 0);
I think "RTE_SECURITY_INSTANCES_BLOCK_ALLOC_SZ" value as 8 is relatively
small. you may want to keep it "64" or more.
you may change it into two parts
- Initial block size and incremental block size for realloc.
Also, do you want to make it a configurable variable. as some
implementation may need really large number of SAs.
Security Instances are not per SA, these are per eth/crypto device which
support security offload.
+
+ if (instances == NULL)
+ return -ENOMEM;
+
+ max_nb_security_instances +=
+ RTE_SECURITY_INSTANCES_BLOCK_ALLOC_SZ;
+ }
+
+ *id = nb_security_instances++;
+
+ security_instances[*id].id = *id;
+ security_instances[*id].state = RTE_SECURITY_INSTANCE_VALID;
+ security_instances[*id].device = device;
+ security_instances[*id].ops = ops;
+
+ return 0;
+}
+
+int
+rte_security_unregister(__rte_unused uint16_t *id)
+{
+ /* To be implemented */
+ return 0;
+}
+
+struct rte_security_session *
+rte_security_session_cr
Re: [dpdk-dev] [PATCH v9 2/9] eal/pci: get iommu class
On 20-Sep-17 12:23 PM, Santosh Shukla wrote: Introducing rte_pci_get_iommu_class API which helps to get iommu class of PCI device on the bus and returns preferred iova mapping mode for PCI bus. Patch also adds rte_pci_get_iommu_class definition for: - bsdapp: api returns default iova mode. - linuxapp: Has stub implementation, Followup patch has complete implementation. Signed-off-by: Santosh Shukla Signed-off-by: Jerin Jacob Reviewed-by: Maxime Coquelin Tested-by: Hemant Agrawal --- Reviewed-by: Anatoly Burakov -- Thanks, Anatoly
Re: [dpdk-dev] [PATCH] lib/librte_pipeline:fix the array index out of bound
> -Original Message-
> From: Rongqiang XIE [mailto:xie.rongqi...@zte.com.cn]
> Sent: Wednesday, August 23, 2017 8:06 AM
> To: Dumitrescu, Cristian
> Cc: dev@dpdk.org; Rongqiang XIE
> Subject: [PATCH] lib/librte_pipeline:fix the array index out of bound
>
> In function rte_pipeline_compute_masks(), the value pos equal
> p->entries[i]->action,type constraint p->entries[i]->action is
> [0,4],but array action_mask1 size is 4,it possible attempt to
> access element 4 of array action_mask1.And also in function
> rte_pipeline_run(),it possible attempt to access element 4 of
> array action_mask0.
>
> Signed-off-by: Rongqiang XIE
> ---
> lib/librte_pipeline/rte_pipeline.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/lib/librte_pipeline/rte_pipeline.c
> b/lib/librte_pipeline/rte_pipeline.c
> index 7f8fbac..2914445 100644
> --- a/lib/librte_pipeline/rte_pipeline.c
> +++ b/lib/librte_pipeline/rte_pipeline.c
> @@ -155,8 +155,8 @@ struct rte_pipeline {
> /* Pipeline run structures */
> struct rte_mbuf *pkts[RTE_PORT_IN_BURST_SIZE_MAX];
> struct rte_pipeline_table_entry
> *entries[RTE_PORT_IN_BURST_SIZE_MAX];
> - uint64_t action_mask0[RTE_PIPELINE_ACTIONS];
> - uint64_t action_mask1[RTE_PIPELINE_ACTIONS];
> + uint64_t action_mask0[RTE_PIPELINE_ACTIONS + 1];
> + uint64_t action_mask1[RTE_PIPELINE_ACTIONS + 1];
> uint64_t pkts_mask;
> uint64_t n_pkts_ah_drop;
> uint64_t pkts_drop_mask;
> --
> 1.8.3.1
>
>
NAK
Hi Rongqiang,
Thanks for your patch, but I think there is a confusion on your side here:
pos = p->entries[i]->action can only be 0 .. 3 (and not: 0 .. 4), as
the last value in the enum rte_pipeline_action, namely RTE_PIPELINE_ACTIONS
(equal to 4), is not a valid action, but the number of valid actions (which are
specified by enu values 0 ..3).
Makes sense?
Regards,
Cristian
Re: [dpdk-dev] [PATCH 0/3] minor build enhancements and fixes
On Wed, 2017-09-20 at 11:51 +0100, Bruce Richardson wrote: > No major changes, just cleanups. > > Bruce Richardson (3): > build: sort meson options alphabetically > build: fix driver install path > build/x86: add SSE cpuflags > > config/x86/meson.build | 10 ++ > meson.build| 2 +- > meson_options.txt | 24 > 3 files changed, 27 insertions(+), 9 deletions(-) Acked-by: Luca Boccassi Looks good, thanks! -- Kind regards, Luca Boccassi
Re: [dpdk-dev] [PATCH v5 1/6] librte_table: fix acl entry add and delete functions
> -Original Message-
> From: Singh, Jasvinder
> Sent: Monday, September 18, 2017 4:30 PM
> To: Iremonger, Bernard ; dev@dpdk.org;
> Yigit, Ferruh ; Ananyev, Konstantin
> ; Dumitrescu, Cristian
> ; adrien.mazarg...@6wind.com
> Cc: Iremonger, Bernard ; sta...@dpdk.org
> Subject: RE: [dpdk-dev] [PATCH v5 1/6] librte_table: fix acl entry add and
> delete functions
>
> Hi Bernard,
>
>
>
> > --- a/lib/librte_table/rte_table_acl.c
> > +++ b/lib/librte_table/rte_table_acl.c
> > @@ -316,8 +316,7 @@ struct rte_table_acl {
> > if (status == 0) {
> > *key_found = 1;
> > *entry_ptr = &acl->memory[i * acl->entry_size];
> > - memcpy(*entry_ptr, entry, acl->entry_size);
> > -
> > + memcpy(entry, *entry_ptr, acl->entry_size);
> > return 0;
> > }
> > }
>
> In this case, table entry which is being added already presents in the table.
> So, first the pointer to that entry from the memory[] that stores the
> pipeline
> table data(struct rte_pipeline_table_entry) associated with key is retrieved
> and the changes (action and metadara) are stored in the internal table
> pointed by action_table. So, above fix doesn't seem correct.
>
> > @@ -353,8 +352,8 @@ struct rte_table_acl {
> > rte_acl_free(acl->ctx);
> > acl->ctx = ctx;
> > *key_found = 0;
> > - *entry_ptr = &acl->memory[free_pos * acl->entry_size];
> > - memcpy(*entry_ptr, entry, acl->entry_size);
> > + *entry_ptr = &acl->acl_rule_memory[free_pos * acl->entry_size];
> > + memcpy(entry, *entry_ptr, acl->entry_size);
> >
> > return 0;
> > }
> > @@ -435,7 +434,7 @@ struct rte_table_acl {
> > acl->ctx = ctx;
> > *key_found = 1;
> > if (entry != NULL)
> > - memcpy(entry, &acl->memory[pos * acl->entry_size],
> > + memcpy(entry, &acl->acl_rule_memory[pos * acl-
> > >entry_size],
> > acl->entry_size);
>
>
> Above fixes also seems not correct. As per documentation, *entry_ptr is
> intended to store the handle to the pipeline table entry containing the data
> associated with the current key instead of pointing to memory used to store
> the acl rules, etc. Please refer rte_table_acl_create() where memory is
> initialized and organized to stores different types of internal tables
> (pointed
> by action_table, acl_rule_list and acl_rule_memory).
NACK
Fully agree with Jasvinder.
Existing code is correct, proposed code changes are wrong.
Re: [dpdk-dev] [PATCH v7] net/mlx5: support upstream rdma-core
On Mon, Sep 18, 2017 at 02:49:52PM +, Shachar Beiser wrote:
> This removes the dependency on specific Mellanox OFED libraries by
> using the upstream rdma-core and linux upstream community code.
> Both rdma-core upstream and Mellanox OFED are Linux user-space packages:
> 1. Rdma-core is Linux upstream user-space package.(Generic)
> 2. Mellanox OFED is Mellanox's Linux user-space package.(Proprietary)
> The difference between the two are the APIs towards the kernel.
>
> ---
> a. Compile with rdma-core commit f11292efd541 ("Merge pull request #202")
> b. Tested with linux kernel 4.13-rc4
> c. For performance testing recommended to wait till kernel 4.14
>
> Signed-off-by: Shachar Beiser
Acked-by: Nelio Laranjeiro
--
Nélio Laranjeiro
6WIND
Re: [dpdk-dev] [PATCH v5 2/6] librte_table: fix acl lookup function
> -Original Message-
> From: Iremonger, Bernard
> Sent: Thursday, September 7, 2017 5:43 PM
> To: dev@dpdk.org; Yigit, Ferruh ; Ananyev,
> Konstantin ; Dumitrescu, Cristian
> ; adrien.mazarg...@6wind.com
> Cc: Iremonger, Bernard ; sta...@dpdk.org
> Subject: [PATCH v5 2/6] librte_table: fix acl lookup function
>
> The rte_table_acl_lookup() function was returning data from acl_memory
> instead of acl_rule_memory.
>
> Fixes: 166923eb2f78 ("table: ACL")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Bernard Iremonger
> ---
> lib/librte_table/rte_table_acl.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/lib/librte_table/rte_table_acl.c
> b/lib/librte_table/rte_table_acl.c
> index e84b437..258916d 100644
> --- a/lib/librte_table/rte_table_acl.c
> +++ b/lib/librte_table/rte_table_acl.c
> @@ -794,7 +794,7 @@ struct rte_table_acl {
> if (action_table_pos != 0) {
> pkts_out_mask |= pkt_mask;
> entries[pkt_pos] = (void *)
> - &acl->memory[action_table_pos *
> + &acl->acl_rule_memory[action_table_pos *
> acl->entry_size];
> rte_prefetch0(entries[pkt_pos]);
> }
> --
> 1.9.1
NACK
Existing code is correct, proposed code changes are wrong (for same reasons
described in patch 1 of this patch set).
Re: [dpdk-dev] [PATCH 0/3] minor build enhancements and fixes
On Wed, Sep 20, 2017 at 01:05:19PM +0100, Luca Boccassi wrote: > On Wed, 2017-09-20 at 11:51 +0100, Bruce Richardson wrote: > > No major changes, just cleanups. > > > > Bruce Richardson (3): > > build: sort meson options alphabetically > > build: fix driver install path > > build/x86: add SSE cpuflags > > > > config/x86/meson.build | 10 ++ > > meson.build| 2 +- > > meson_options.txt | 24 > > 3 files changed, 27 insertions(+), 9 deletions(-) > > Acked-by: Luca Boccassi > Applied to dpdk-next-build. /Bruce
[dpdk-dev] [PATCH] doc: add Linux flower support check in TAP guide
The flow API is supported in TAP PMD if flower is supported in Linux.
Some commands are combined to suggest a convenient check of its support
by the running kernel.
Signed-off-by: Thomas Monjalon
---
doc/guides/nics/tap.rst | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/doc/guides/nics/tap.rst b/doc/guides/nics/tap.rst
index f3ee95d28..04086b110 100644
--- a/doc/guides/nics/tap.rst
+++ b/doc/guides/nics/tap.rst
@@ -113,7 +113,14 @@ Flow API support
The tap PMD supports major flow API pattern items and actions, when running on
-linux kernels above 4.2 ("Flower" classifier required). Supported items:
+linux kernels above 4.2 ("Flower" classifier required).
+The kernel support can be checked with this command::
+
+ zcat /proc/config.gz | ( grep 'CLS_FLOWER=' || echo 'not supported' ) |
+ tee -a /dev/stderr | grep -q '=m' &&
+ lsmod | ( grep cls_flower || echo 'try modprobe cls_flower' )
+
+Supported items:
- eth: src and dst (with variable masks), and eth_type (0x mask).
- vlan: vid, pcp, tpid, but not eid. (requires kernel 4.9)
--
2.14.1
Re: [dpdk-dev] [PATCH] doc: add Linux flower support check in TAP guide
Acked-by: Pascal Mazon
On 20/09/2017 15:03, Thomas Monjalon wrote:
> The flow API is supported in TAP PMD if flower is supported in Linux.
> Some commands are combined to suggest a convenient check of its support
> by the running kernel.
>
> Signed-off-by: Thomas Monjalon
> ---
> doc/guides/nics/tap.rst | 9 -
> 1 file changed, 8 insertions(+), 1 deletion(-)
>
> diff --git a/doc/guides/nics/tap.rst b/doc/guides/nics/tap.rst
> index f3ee95d28..04086b110 100644
> --- a/doc/guides/nics/tap.rst
> +++ b/doc/guides/nics/tap.rst
> @@ -113,7 +113,14 @@ Flow API support
>
>
> The tap PMD supports major flow API pattern items and actions, when running
> on
> -linux kernels above 4.2 ("Flower" classifier required). Supported items:
> +linux kernels above 4.2 ("Flower" classifier required).
> +The kernel support can be checked with this command::
> +
> + zcat /proc/config.gz | ( grep 'CLS_FLOWER=' || echo 'not supported' ) |
> + tee -a /dev/stderr | grep -q '=m' &&
> + lsmod | ( grep cls_flower || echo 'try modprobe cls_flower' )
> +
> +Supported items:
>
> - eth: src and dst (with variable masks), and eth_type (0x mask).
> - vlan: vid, pcp, tpid, but not eid. (requires kernel 4.9)
Re: [dpdk-dev] [PATCH v6 1/3] eal: introduce integer divide through reciprocal
Hi Pavan,
I think moving rte_reciprocal.[hc] to a common code area like EAL is a very
good idea, so thanks for doing this work!
One ask from my side: please do not change the existing code.
1. Keep the existing name for the 32-bit API functions and data structures (no
_u32 name suffix), add the _u64 suffix just for the new API functions that you
add.
- If you want, you can create aliases with _32 name suffix for the
existing 32-bit API
- The only change to rte_sched.c should be header include line:
#include "rte_reciprocal.h" -> #include
2. Do not do any cosmetic changes in existing code in rte_reciprocal.[hc] (such
as adding CR+LF), as they result in lots of code churn for no real value
Once these changes are done, the size of your patch set is reduced considerably.
> -Original Message-
> From: Pavan Nikhilesh [mailto:pbhagavat...@caviumnetworks.com]
> Sent: Wednesday, September 6, 2017 11:22 AM
> To: Dumitrescu, Cristian ;
> step...@networkplumber.org
> Cc: dev@dpdk.org; Pavan Bhagavatula
>
> Subject: [dpdk-dev] [PATCH v6 1/3] eal: introduce integer divide through
> reciprocal
>
> From: Pavan Bhagavatula
>
> In some use cases of integer division, denominator remains constant and
> numerator varies. It is possible to optimize division for such specific
> scenarios.
>
> The librte_sched uses rte_reciprocal to optimize division so, moving it to
> eal/common would allow other libraries and applications to use it.
>
> Signed-off-by: Pavan Nikhilesh
> Reviewed-by: Anatoly Burakov
> ---
>
> v6 changes:
> - remove cache alignment from rte_reciprocal_u{32/64}structures as they
> would
> be embedded in other structures.
>
> v5 changes:
> - fix test print strings
>
> v4 changes:
> - minor fix for test cases
> - fix u32 divisor generation
>
> v3 changes:
> - fix x86_32 compilation issue
> - fix improper licence in test
>
> v2 changes:
> - fix compilation issues with .map files
> - add test cases for correctness and performance
> - remove extra licence inclusion
> - fix coding style issues
>
> lib/librte_eal/bsdapp/eal/Makefile | 1 +
> lib/librte_eal/bsdapp/eal/rte_eal_version.map| 7 +++
> lib/librte_eal/common/Makefile | 1 +
> lib/{librte_sched => librte_eal/common/include}/rte_reciprocal.h | 6 --
> lib/{librte_sched => librte_eal/common}/rte_reciprocal.c | 6 --
> lib/librte_eal/linuxapp/eal/Makefile | 1 +
> lib/librte_eal/linuxapp/eal/rte_eal_version.map | 7 +++
> lib/librte_sched/Makefile| 2 --
> lib/librte_sched/rte_sched.c | 2 +-
> 9 files changed, 26 insertions(+), 7 deletions(-)
> rename lib/{librte_sched => librte_eal/common/include}/rte_reciprocal.h
> (87%)
> rename lib/{librte_sched => librte_eal/common}/rte_reciprocal.c (96%)
>
> diff --git a/lib/librte_eal/bsdapp/eal/Makefile
> b/lib/librte_eal/bsdapp/eal/Makefile
> index 005019e..56f9804 100644
> --- a/lib/librte_eal/bsdapp/eal/Makefile
> +++ b/lib/librte_eal/bsdapp/eal/Makefile
> @@ -88,6 +88,7 @@ SRCS-$(CONFIG_RTE_EXEC_ENV_BSDAPP) +=
> malloc_elem.c
> SRCS-$(CONFIG_RTE_EXEC_ENV_BSDAPP) += malloc_heap.c
> SRCS-$(CONFIG_RTE_EXEC_ENV_BSDAPP) += rte_keepalive.c
> SRCS-$(CONFIG_RTE_EXEC_ENV_BSDAPP) += rte_service.c
> +SRCS-$(CONFIG_RTE_EXEC_ENV_BSDAPP) += rte_reciprocal.c
>
> # from arch dir
> SRCS-$(CONFIG_RTE_EXEC_ENV_BSDAPP) += rte_cpuflags.c
> diff --git a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> index aac6fd7..90d7258 100644
> --- a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> +++ b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> @@ -237,3 +237,10 @@ EXPERIMENTAL {
> rte_service_unregister;
>
> } DPDK_17.08;
> +
> +DPDK_17.11 {
> + global:
> +
> + rte_reciprocal_value;
> +
> +} DPDK_17.08;
> diff --git a/lib/librte_eal/common/Makefile
> b/lib/librte_eal/common/Makefile
> index e8fd67a..a680b2d 100644
> --- a/lib/librte_eal/common/Makefile
> +++ b/lib/librte_eal/common/Makefile
> @@ -42,6 +42,7 @@ INC += rte_hexdump.h rte_devargs.h rte_bus.h
> rte_dev.h rte_vdev.h
> INC += rte_pci_dev_feature_defs.h rte_pci_dev_features.h
> INC += rte_malloc.h rte_keepalive.h rte_time.h
> INC += rte_service.h rte_service_component.h
> +INC += rte_reciprocal.h
>
> GENERIC_INC := rte_atomic.h rte_byteorder.h rte_cycles.h rte_prefetch.h
> GENERIC_INC += rte_spinlock.h rte_memcpy.h rte_cpuflags.h rte_rwlock.h
> diff --git a/lib/librte_sched/rte_reciprocal.h
> b/lib/librte_eal/common/include/rte_reciprocal.h
> similarity index 87%
> rename from lib/librte_sched/rte_reciprocal.h
> rename to lib/librte_eal/common/include/rte_reciprocal.h
> index 5e21f09..b6d752f 100644
> --- a/lib/librte_sched/rte_reciprocal.h
> +++ b/lib/librte_eal/common/include/rte_reciprocal.h
> @@ -29,13 +29,15 @@ struct
[dpdk-dev] [PATCH v3] sched: make RED scaling configurable
From: Alan Dewar
The RED code stores the maximum threshold is a 32-bit integer as a
pseudo fixed-point floating number with 10 fractional bits. Twelve
other bits are used to encode the filter weight, leaving just 10 bits
for the queue length. This limits the maximum queue length supported
by RED queues as 1024 packets.
Move the "hard" definitions from red.h into config/common_base so that
RED scaling can be configured during build.
Modified the RED unit-tests to use the new "soft" definition of
maximum-threshold from config/common_base in tests where it previously
used a hard coded limit of 1023.
The RED unit-tests all successfully pass when the maximum-threshold is
configured as 8191 and the RED scaling factor is dropped to seven.
Real-world testing has involved RED queue lengths of 8192 with multiple
different settings of the RED config parameters: min_th, max_th, wq_log2
and maxp_inv.
Signed-off-by: Alan Dewar
Acked-by: Tomasz Kantecki
---
config/common_base | 7 +++
lib/librte_sched/rte_red.h | 3 +--
test/test/test_red.c | 20 ++--
3 files changed, 18 insertions(+), 12 deletions(-)
diff --git a/config/common_base b/config/common_base
index 5e97a08..2626557 100644
--- a/config/common_base
+++ b/config/common_base
@@ -666,6 +666,13 @@ CONFIG_RTE_SCHED_COLLECT_STATS=n
CONFIG_RTE_SCHED_SUBPORT_TC_OV=n
CONFIG_RTE_SCHED_PORT_N_GRINDERS=8
CONFIG_RTE_SCHED_VECTOR=n
+#
+# RTE_RED_SCALING - number of fractional bits used for RED's moving average
+# For every bit that RTE_RED_SCALING is reduced, the max-queue size can doubled
+# RTE_RED_MAX_TH_MAX = (max-queue size - 1), max-queue size must be power of
two
+#
+CONFIG_RTE_RED_SCALING=10
+CONFIG_RTE_RED_MAX_TH_MAX=1023
#
# Compile the distributor library
diff --git a/lib/librte_sched/rte_red.h b/lib/librte_sched/rte_red.h
index ca12227..49d3379 100644
--- a/lib/librte_sched/rte_red.h
+++ b/lib/librte_sched/rte_red.h
@@ -51,10 +51,9 @@ extern "C" {
#include
#include
#include
+#include
-#define RTE_RED_SCALING 10 /**< Fraction size for
fixed-point */
#define RTE_RED_S (1 << 22) /**< Packet size
multiplied by number of leaf queues */
-#define RTE_RED_MAX_TH_MAX 1023 /**< Max threshold
limit in fixed point format */
#define RTE_RED_WQ_LOG2_MIN 1 /**< Min inverse filter
weight value */
#define RTE_RED_WQ_LOG2_MAX 12 /**< Max inverse filter
weight value */
#define RTE_RED_MAXP_INV_MIN1 /**< Min inverse mark
probability value */
diff --git a/test/test/test_red.c b/test/test/test_red.c
index 348075d..70e8cfe 100644
--- a/test/test/test_red.c
+++ b/test/test/test_red.c
@@ -653,14 +653,14 @@ static enum test_result func_test2(struct test_config
*tcfg)
/**
* Test F3: functional test 3
*/
-static uint32_t ft3_tlevel[] = {1022};
+static uint32_t ft3_tlevel[] = {RTE_RED_MAX_TH_MAX-1};
static struct test_rte_red_config ft3_tconfig = {
.rconfig = ft_wrconfig,
.num_cfg = RTE_DIM(ft_wrconfig),
.wq_log2 = ft_wq_log2,
.min_th = 32,
- .max_th = 1023,
+ .max_th = RTE_RED_MAX_TH_MAX,
.maxp_inv = ft_maxp_inv,
};
@@ -766,14 +766,14 @@ static enum test_result func_test3(struct test_config
*tcfg)
/**
* Test F4: functional test 4
*/
-static uint32_t ft4_tlevel[] = {1022};
+static uint32_t ft4_tlevel[] = {RTE_RED_MAX_TH_MAX-1};
static uint8_t ft4_wq_log2[] = {11};
static struct test_rte_red_config ft4_tconfig = {
.rconfig = ft_wrconfig,
.num_cfg = RTE_DIM(ft_wrconfig),
.min_th = 32,
- .max_th = 1023,
+ .max_th = RTE_RED_MAX_TH_MAX,
.wq_log2 = ft4_wq_log2,
.maxp_inv = ft_maxp_inv,
};
@@ -1048,7 +1048,7 @@ static enum test_result func_test5(struct test_config
*tcfg)
/**
* Test F6: functional test 6
*/
-static uint32_t ft6_tlevel[] = {1022};
+static uint32_t ft6_tlevel[] = {RTE_RED_MAX_TH_MAX-1};
static uint8_t ft6_wq_log2[] = {9, 8};
static uint8_t ft6_maxp_inv[] = {10, 20};
static struct rte_red_config ft6_config[2];
@@ -1059,7 +1059,7 @@ static struct test_rte_red_config ft6_tconfig = {
.rconfig = ft6_config,
.num_cfg = RTE_DIM(ft6_config),
.min_th = 32,
- .max_th = 1023,
+ .max_th = RTE_RED_MAX_TH_MAX,
.wq_log2 = ft6_wq_log2,
.maxp_inv = ft6_maxp_inv,
};
@@ -1547,7 +1547,7 @@ static uint32_t ovfl_qconfig[] = {0, 0, 1, 1};
static uint32_t ovfl_q[] ={0};
static uint32_t ovfl_dropped[] ={0};
static uint32_t ovfl_enqueued[] ={0};
-static uint32_t ovfl_tlevel[] = {1023};
+static uint32_t ovfl_tlevel[] = {RTE_RED_MAX_TH_MAX};
static uint8_t ovfl_wq_log2[] = {12};
static struct test_rte_red_config ovfl_tconfig = {
@@ -1555,7 +1555,7 @@ static struct test_rte_red_config ovfl_tconfig = {
.num_cfg = RTE_DIM(ovfl_wrconfig),
.wq_log2 = ovfl_wq_log2,
.min_th =
Re: [dpdk-dev] [PATCH v6 2/3] eal: add u64 bit variant for reciprocal
Hi Pavan,
Same ask as for the first patch:
-Do not change existing code in rte_reciprocal.[hc]: no _u32 suffix, please
-Do not add lots of CR+LF to existing code
> -Original Message-
> From: Pavan Nikhilesh [mailto:pbhagavat...@caviumnetworks.com]
> Sent: Wednesday, September 6, 2017 11:22 AM
> To: Dumitrescu, Cristian ;
> step...@networkplumber.org
> Cc: dev@dpdk.org; Pavan Bhagavatula
>
> Subject: [dpdk-dev] [PATCH v6 2/3] eal: add u64 bit variant for reciprocal
>
> From: Pavan Bhagavatula
>
> Currently, rte_reciprocal only supports unsigned 32bit divisors. This
> commit adds support for unsigned 64bit divisors.
>
> Rename unsigned 32bit specific functions appropriately and update
> librte_sched accordingly.
>
> Signed-off-by: Pavan Nikhilesh
> ---
> lib/librte_eal/bsdapp/eal/rte_eal_version.map | 3 +-
> lib/librte_eal/common/include/rte_reciprocal.h | 109
> --
> lib/librte_eal/common/rte_reciprocal.c | 116
> +---
> lib/librte_eal/linuxapp/eal/rte_eal_version.map | 3 +-
> lib/librte_sched/Makefile | 4 +-
> lib/librte_sched/rte_sched.c| 9 +-
> 6 files changed, 219 insertions(+), 25 deletions(-)
>
> diff --git a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> index 90d7258..59a85bb 100644
> --- a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> +++ b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> @@ -241,6 +241,7 @@ EXPERIMENTAL {
> DPDK_17.11 {
> global:
>
> - rte_reciprocal_value;
> + rte_reciprocal_value_u32;
> + rte_reciprocal_value_u64;
>
> } DPDK_17.08;
> diff --git a/lib/librte_eal/common/include/rte_reciprocal.h
> b/lib/librte_eal/common/include/rte_reciprocal.h
> index b6d752f..85599e6 100644
> --- a/lib/librte_eal/common/include/rte_reciprocal.h
> +++ b/lib/librte_eal/common/include/rte_reciprocal.h
> @@ -22,22 +22,117 @@
> #ifndef _RTE_RECIPROCAL_H_
> #define _RTE_RECIPROCAL_H_
>
> -#include
> +#include
>
> -struct rte_reciprocal {
> +/**
> + * Unsigned 32-bit divisor structure.
> + */
> +struct rte_reciprocal_u32 {
> uint32_t m;
> uint8_t sh1, sh2;
> };
>
Do not add _32 suffix for exiting function, no name change, please.
> +/**
> + * Unsigned 64-bit divisor structure.
> + */
> +struct rte_reciprocal_u64 {
> + uint64_t m;
> + uint8_t sh1;
> +};
> +
I am OK with adding +64 suffix for new API.
> +/**
> + * Divide given unsigned 32-bit integer with pre calculated divisor.
> + *
> + * @param a
> + * The 32-bit dividend.
> + * @param R
> + * The pointer to pre calculated divisor reciprocal structure.
> + *
> + * @return
> + * The result of the division
> + */
> static inline uint32_t
> -rte_reciprocal_divide(uint32_t a, struct rte_reciprocal R)
> +rte_reciprocal_divide_u32(uint32_t a, struct rte_reciprocal_u32 *R)
> {
> - uint32_t t = (uint32_t)(((uint64_t)a * R.m) >> 32);
> + uint32_t t = (((uint64_t)a * R->m) >> 32);
>
> - return (t + ((a - t) >> R.sh1)) >> R.sh2;
> + return (t + ((a - t) >> R->sh1)) >> R->sh2;
> }
Do not add _u32 suffix for existing API.
>
> -struct rte_reciprocal
> -rte_reciprocal_value(uint32_t d);
> +static inline uint64_t
> +mullhi_u64(uint64_t x, uint64_t y)
> +{
> +#ifdef __SIZEOF_INT128__
> + __uint128_t xl = x;
> + __uint128_t rl = xl * y;
> +
> + return (rl >> 64);
> +#else
> + uint64_t u0, u1, v0, v1, k, t;
> + uint64_t w1, w2;
> + uint64_t whi;
> +
> + u1 = x >> 32; u0 = x & 0x;
> + v1 = y >> 32; v0 = y & 0x;
> +
> + t = u0*v0;
> + k = t >> 32;
> +
> + t = u1*v0 + k;
> + w1 = t & 0x;
> + w2 = t >> 32;
> +
> + t = u0*v1 + w1;
> + k = t >> 32;
> +
> + whi = u1*v1 + w2 + k;
> +
> + return whi;
> +#endif
> +}
> +
> +/**
> + * Divide given unsigned 64-bit integer with pre calculated divisor.
> + *
> + * @param a
> + * The 64-bit dividend.
> + * @param R
> + * The pointer to pre calculated divisor reciprocal structure.
> + *
> + * @return
> + * The result of the division
> + */
> +static inline uint64_t
> +rte_reciprocal_divide_u64(uint64_t a, struct rte_reciprocal_u64 *R)
> +{
> + uint64_t q = mullhi_u64(R->m, a);
> + uint64_t t = ((a - q) >> 1) + q;
> +
> + return t >> R->sh1;
> +}
> +
> +/**
> + * Generate pre calculated divisor structure.
> + *
> + * @param d
> + * The unsigned 32-bit divisor.
> + *
> + * @return
> + * Divisor structure.
> + */
> +struct rte_reciprocal_u32
> +rte_reciprocal_value_u32(uint32_t d);
> +
Do not add _u32 suffix for existing API.
> +/**
> + * Generate pre calculated divisor structure.
> + *
> + * @param d
> + * The unsigned 64-bit divisor.
> + *
> + * @return
> + * Divisor structure.
> + */
> +struct rte_reciprocal_u64
> +rte_reciprocal_value_u64(uint64_t d);
>
> #endif /* _RTE_RECIPROCAL_H_ */
> diff --git a/lib/librte_eal/common/rte_reciprocal.c
> b/lib/librte_eal/co
Re: [dpdk-dev] [PATCH v2] net/af_packet: make bypass configurable
On 9/19/2017 10:45 PM, Chas Williams wrote:
> From: "Charles (Chas) Williams"
>
> In certain situations, low speed interfaces, it may be desirable to
> have the flow control provided by the kernel queueing disciplines.
Out of curiosity, do you have any compression of performance numbers
with and without qdisc?
>
> Signed-off-by: Chas Williams
> ---
> drivers/net/af_packet/rte_eth_af_packet.c | 22 +-
> 1 file changed, 17 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/net/af_packet/rte_eth_af_packet.c
> b/drivers/net/af_packet/rte_eth_af_packet.c
> index 7aa26e5..e3858fa 100644
> --- a/drivers/net/af_packet/rte_eth_af_packet.c
> +++ b/drivers/net/af_packet/rte_eth_af_packet.c
> @@ -59,6 +59,7 @@
> #define ETH_AF_PACKET_BLOCKSIZE_ARG "blocksz"
> #define ETH_AF_PACKET_FRAMESIZE_ARG "framesz"
> #define ETH_AF_PACKET_FRAMECOUNT_ARG "framecnt"
> +#define ETH_AF_PACKET_BYPASS_ARG "bypass"
I thinks argument name "bypass" on its own is not clear what is
bypassed, would you mind using something like "qdisc_bypass"?
>
> #define DFLT_BLOCK_SIZE (1 << 12)
> #define DFLT_FRAME_SIZE (1 << 11)
> @@ -115,6 +116,7 @@ static const char *valid_arguments[] = {
> ETH_AF_PACKET_BLOCKSIZE_ARG,
> ETH_AF_PACKET_FRAMESIZE_ARG,
> ETH_AF_PACKET_FRAMECOUNT_ARG,
> + ETH_AF_PACKET_BYPASS_ARG,
Same comment here, what about "ETH_AF_PACKET_QDISC_BYPASS_ARG" ?
> NULL
> };
>
> @@ -559,6 +561,7 @@ rte_pmd_init_internals(struct rte_vdev_device *dev,
> unsigned int blockcnt,
> unsigned int framesize,
> unsigned int framecnt,
> +unsigned int bypass,
> struct pmd_internals **internals,
> struct rte_eth_dev **eth_dev,
> struct rte_kvargs *kvlist)
> @@ -580,9 +583,6 @@ rte_pmd_init_internals(struct rte_vdev_device *dev,
> #if defined(PACKET_FANOUT)
> int fanout_arg;
> #endif
> -#if defined(PACKET_QDISC_BYPASS)
> - int bypass;
> -#endif
>
> for (k_idx = 0; k_idx < kvlist->count; k_idx++) {
> pair = &kvlist->pairs[k_idx];
> @@ -698,7 +698,6 @@ rte_pmd_init_internals(struct rte_vdev_device *dev,
> }
>
> #if defined(PACKET_QDISC_BYPASS)
> - bypass = 1;
> rc = setsockopt(qsockfd, SOL_PACKET, PACKET_QDISC_BYPASS,
> &bypass, sizeof(bypass));
> if (rc == -1) {
> @@ -851,6 +850,7 @@ rte_eth_from_packet(struct rte_vdev_device *dev,
> unsigned int framesize = DFLT_FRAME_SIZE;
> unsigned int framecount = DFLT_FRAME_COUNT;
> unsigned int qpairs = 1;
> + unsigned int bypass = 1;
>
> /* do some parameter checking */
> if (*sockfd < 0)
> @@ -902,6 +902,16 @@ rte_eth_from_packet(struct rte_vdev_device *dev,
> }
> continue;
> }
> + if (strstr(pair->key, ETH_AF_PACKET_BYPASS_ARG) != NULL) {
> + bypass = atoi(pair->value);
> + if (bypass > 2) {
This accepts "2" too .
As far as I can see kernel side checks if this value is zero or not, so
perhaps this check is not required at all. But if you want to keep it I
guess intention is to limit the value to "0" and "1".
> + RTE_LOG(ERR, PMD,
> + "%s: invalid bypass value\n",
> + name);
> + return -1;
> + }
> + continue;
> + }
> }
>
> if (framesize > blocksize) {
> @@ -927,6 +937,7 @@ rte_eth_from_packet(struct rte_vdev_device *dev,
> if (rte_pmd_init_internals(dev, *sockfd, qpairs,
> blocksize, blockcount,
> framesize, framecount,
> +bypass,
> &internals, ð_dev,
> kvlist) < 0)
> return -1;
> @@ -1021,4 +1032,5 @@ RTE_PMD_REGISTER_PARAM_STRING(net_af_packet,
> "qpairs= "
> "blocksz= "
> "framesz= "
> - "framecnt=");
> + "framecnt= "
> + "bypass="
Although storage is integer, does is make sense to document it as "0|1"
to clarify the usage?
);
>
Re: [dpdk-dev] [PATCH v6 1/3] eal: introduce integer divide through reciprocal
On Wed, Sep 20, 2017 at 01:10:53PM +, Dumitrescu, Cristian wrote:
> Hi Pavan,
Hi Cristian,
>
> I think moving rte_reciprocal.[hc] to a common code area like EAL is a very
> good idea, so thanks for doing this work!
>
> One ask from my side: please do not change the existing code.
> 1. Keep the existing name for the 32-bit API functions and data structures
> (no _u32 name suffix), add the _u64 suffix just for the new API functions
> that you add.
> - If you want, you can create aliases with _32 name suffix for the
> existing 32-bit API
> - The only change to rte_sched.c should be header include line:
> #include "rte_reciprocal.h" -> #include
> 2. Do not do any cosmetic changes in existing code in rte_reciprocal.[hc]
> (such as adding CR+LF), as they result in lots of code churn for no real value
>
> Once these changes are done, the size of your patch set is reduced
> considerably.
>
Will do the changes once the licencing issues get sorted out(v7).
-Pavan
>
> > -Original Message-
> > From: Pavan Nikhilesh [mailto:pbhagavat...@caviumnetworks.com]
> > Sent: Wednesday, September 6, 2017 11:22 AM
> > To: Dumitrescu, Cristian ;
> > step...@networkplumber.org
> > Cc: dev@dpdk.org; Pavan Bhagavatula
> >
> > Subject: [dpdk-dev] [PATCH v6 1/3] eal: introduce integer divide through
> > reciprocal
> >
> > From: Pavan Bhagavatula
> >
> > In some use cases of integer division, denominator remains constant and
> > numerator varies. It is possible to optimize division for such specific
> > scenarios.
> >
> > The librte_sched uses rte_reciprocal to optimize division so, moving it to
> > eal/common would allow other libraries and applications to use it.
> >
> > Signed-off-by: Pavan Nikhilesh
> > Reviewed-by: Anatoly Burakov
> > ---
> >
> > v6 changes:
> > - remove cache alignment from rte_reciprocal_u{32/64}structures as they
> > would
> > be embedded in other structures.
> >
> > v5 changes:
> > - fix test print strings
> >
> > v4 changes:
> > - minor fix for test cases
> > - fix u32 divisor generation
> >
> > v3 changes:
> > - fix x86_32 compilation issue
> > - fix improper licence in test
> >
> > v2 changes:
> > - fix compilation issues with .map files
> > - add test cases for correctness and performance
> > - remove extra licence inclusion
> > - fix coding style issues
> >
> > lib/librte_eal/bsdapp/eal/Makefile | 1 +
> > lib/librte_eal/bsdapp/eal/rte_eal_version.map| 7
> > +++
> > lib/librte_eal/common/Makefile | 1 +
> > lib/{librte_sched => librte_eal/common/include}/rte_reciprocal.h | 6 --
> > lib/{librte_sched => librte_eal/common}/rte_reciprocal.c | 6 --
> > lib/librte_eal/linuxapp/eal/Makefile | 1 +
> > lib/librte_eal/linuxapp/eal/rte_eal_version.map | 7
> > +++
> > lib/librte_sched/Makefile| 2 --
> > lib/librte_sched/rte_sched.c | 2 +-
> > 9 files changed, 26 insertions(+), 7 deletions(-)
> > rename lib/{librte_sched => librte_eal/common/include}/rte_reciprocal.h
> > (87%)
> > rename lib/{librte_sched => librte_eal/common}/rte_reciprocal.c (96%)
> >
> > diff --git a/lib/librte_eal/bsdapp/eal/Makefile
> > b/lib/librte_eal/bsdapp/eal/Makefile
> > index 005019e..56f9804 100644
> > --- a/lib/librte_eal/bsdapp/eal/Makefile
> > +++ b/lib/librte_eal/bsdapp/eal/Makefile
> > @@ -88,6 +88,7 @@ SRCS-$(CONFIG_RTE_EXEC_ENV_BSDAPP) +=
> > malloc_elem.c
> > SRCS-$(CONFIG_RTE_EXEC_ENV_BSDAPP) += malloc_heap.c
> > SRCS-$(CONFIG_RTE_EXEC_ENV_BSDAPP) += rte_keepalive.c
> > SRCS-$(CONFIG_RTE_EXEC_ENV_BSDAPP) += rte_service.c
> > +SRCS-$(CONFIG_RTE_EXEC_ENV_BSDAPP) += rte_reciprocal.c
> >
> > # from arch dir
> > SRCS-$(CONFIG_RTE_EXEC_ENV_BSDAPP) += rte_cpuflags.c
> > diff --git a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> > b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> > index aac6fd7..90d7258 100644
> > --- a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> > +++ b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
> > @@ -237,3 +237,10 @@ EXPERIMENTAL {
> > rte_service_unregister;
> >
> > } DPDK_17.08;
> > +
> > +DPDK_17.11 {
> > + global:
> > +
> > + rte_reciprocal_value;
> > +
> > +} DPDK_17.08;
> > diff --git a/lib/librte_eal/common/Makefile
> > b/lib/librte_eal/common/Makefile
> > index e8fd67a..a680b2d 100644
> > --- a/lib/librte_eal/common/Makefile
> > +++ b/lib/librte_eal/common/Makefile
> > @@ -42,6 +42,7 @@ INC += rte_hexdump.h rte_devargs.h rte_bus.h
> > rte_dev.h rte_vdev.h
> > INC += rte_pci_dev_feature_defs.h rte_pci_dev_features.h
> > INC += rte_malloc.h rte_keepalive.h rte_time.h
> > INC += rte_service.h rte_service_component.h
> > +INC += rte_reciprocal.h
> >
> > GENERIC_INC := rte_atomic.h rte_byteorder.h rte_cycles.h rte_prefetch.h
> > GENERIC_INC += rte_spinlock.h rte_memcpy.h rte_cpuflags.h rte_rwlock.h
>
Re: [dpdk-dev] [PATCH 1/2] eal: move bitmap from lib sched
> -Original Message-
> From: Pavan Nikhilesh [mailto:pbhagavat...@caviumnetworks.com]
> Sent: Thursday, September 7, 2017 3:40 PM
> To: Dumitrescu, Cristian ;
> step...@networkplumber.org
> Cc: dev@dpdk.org; Pavan Nikhilesh
> Subject: [dpdk-dev] [PATCH 1/2] eal: move bitmap from lib sched
>
> The librte_sched uses rte_bitmap to manage large arrays of bits in an
> optimized method so, moving it to eal/common would allow other libraries
> and applications to use it.
>
> Signed-off-by: Pavan Nikhilesh
> ---
> lib/librte_eal/common/Makefile | 1 +
> .../common/include}/rte_bitmap.h | 100
> +
> lib/librte_sched/Makefile | 5 +-
> lib/librte_sched/rte_sched.c | 2 +-
> 4 files changed, 70 insertions(+), 38 deletions(-)
> rename lib/{librte_sched => librte_eal/common/include}/rte_bitmap.h
> (85%)
>
> diff --git a/lib/librte_eal/common/Makefile
> b/lib/librte_eal/common/Makefile
> index e8fd67a..c2c6a7f 100644
> --- a/lib/librte_eal/common/Makefile
> +++ b/lib/librte_eal/common/Makefile
> @@ -42,6 +42,7 @@ INC += rte_hexdump.h rte_devargs.h rte_bus.h
> rte_dev.h rte_vdev.h
> INC += rte_pci_dev_feature_defs.h rte_pci_dev_features.h
> INC += rte_malloc.h rte_keepalive.h rte_time.h
> INC += rte_service.h rte_service_component.h
> +INC += rte_bitmap.h
>
> GENERIC_INC := rte_atomic.h rte_byteorder.h rte_cycles.h rte_prefetch.h
> GENERIC_INC += rte_spinlock.h rte_memcpy.h rte_cpuflags.h rte_rwlock.h
> diff --git a/lib/librte_sched/rte_bitmap.h
> b/lib/librte_eal/common/include/rte_bitmap.h
> similarity index 85%
> rename from lib/librte_sched/rte_bitmap.h
> rename to lib/librte_eal/common/include/rte_bitmap.h
> index 010d752..0938c63 100644
> --- a/lib/librte_sched/rte_bitmap.h
> +++ b/lib/librte_eal/common/include/rte_bitmap.h
> @@ -65,6 +65,7 @@ extern "C" {
> ***/
>
> #include
> +
> #include
> #include
> #include
> @@ -72,36 +73,46 @@ extern "C" {
> #include
>
> #ifndef RTE_BITMAP_OPTIMIZATIONS
> -#define RTE_BITMAP_OPTIMIZATIONS 1
> +#define RTE_BITMAP_OPTIMIZATIONS1
> #endif
>
> /* Slab */
> -#define RTE_BITMAP_SLAB_BIT_SIZE 64
> -#define RTE_BITMAP_SLAB_BIT_SIZE_LOG26
> -#define RTE_BITMAP_SLAB_BIT_MASK
> (RTE_BITMAP_SLAB_BIT_SIZE - 1)
> +#define RTE_BITMAP_SLAB_BIT_SIZE64
> +#define RTE_BITMAP_SLAB_BIT_SIZE_LOG2 6
> +#define RTE_BITMAP_SLAB_BIT_MASK(RTE_BITMAP_SLAB_BIT_SIZE -
> 1)
>
> /* Cache line (CL) */
> -#define RTE_BITMAP_CL_BIT_SIZE (RTE_CACHE_LINE_SIZE * 8)
> -#define RTE_BITMAP_CL_BIT_SIZE_LOG2
> (RTE_CACHE_LINE_SIZE_LOG2 + 3)
> -#define RTE_BITMAP_CL_BIT_MASK (RTE_BITMAP_CL_BIT_SIZE -
> 1)
> +#define RTE_BITMAP_CL_BIT_SIZE (RTE_CACHE_LINE_SIZE * 8)
> +#define RTE_BITMAP_CL_BIT_SIZE_LOG2 (RTE_CACHE_LINE_SIZE_LOG2
> + 3)
> +#define RTE_BITMAP_CL_BIT_MASK (RTE_BITMAP_CL_BIT_SIZE - 1)
>
> -#define RTE_BITMAP_CL_SLAB_SIZE (RTE_BITMAP_CL_BIT_SIZE /
> RTE_BITMAP_SLAB_BIT_SIZE)
> -#define RTE_BITMAP_CL_SLAB_SIZE_LOG2
> (RTE_BITMAP_CL_BIT_SIZE_LOG2 - RTE_BITMAP_SLAB_BIT_SIZE_LOG2)
> -#define RTE_BITMAP_CL_SLAB_MASK
> (RTE_BITMAP_CL_SLAB_SIZE - 1)
> +#define RTE_BITMAP_CL_SLAB_SIZE \
> + (RTE_BITMAP_CL_BIT_SIZE / RTE_BITMAP_SLAB_BIT_SIZE)
> +#define RTE_BITMAP_CL_SLAB_SIZE_LOG2\
> + (RTE_BITMAP_CL_BIT_SIZE_LOG2 -
> RTE_BITMAP_SLAB_BIT_SIZE_LOG2)
> +#define RTE_BITMAP_CL_SLAB_MASK (RTE_BITMAP_CL_SLAB_SIZE - 1)
>
> /** Bitmap data structure */
> struct rte_bitmap {
> /* Context for array1 and array2 */
> - uint64_t *array1;/**< Bitmap array1 */
> - uint64_t *array2;/**< Bitmap array2 */
> - uint32_t array1_size;/**< Number of 64-bit slabs in
> array1
> that are actually used */
> - uint32_t array2_size;/**< Number of 64-bit slabs in
> array2
> */
> + uint64_t *array1;
> + /**< Bitmap array1 */
> + uint64_t *array2;
> + /**< Bitmap array2 */
> + uint32_t array1_size;
> + /**< Number of 64-bit slabs in array1 that are actually used */
> + uint32_t array2_size;
> + /**< Number of 64-bit slabs in array2 */
>
> /* Context for the "scan next" operation */
> - uint32_t index1; /**< Bitmap scan: Index of current array1 slab */
> - uint32_t offset1; /**< Bitmap scan: Offset of current bit within
> current array1 slab */
> - uint32_t index2; /**< Bitmap scan: Index of current array2 slab */
> - uint32_t go2; /**< Bitmap scan: Go/stop condition for current
> array2 cache line */
> + uint32_t index1;
> + /**< Bitmap scan: Index of current array1 slab */
> + uint32_t offset1;
> + /**< Bitmap scan: Offset of current bit within current array1 slab */
> + uint32_t index2;
> + /**< Bitmap scan: Index of curre
[dpdk-dev] [PATCH v2] net/virtio: fix of untrusted scalar value
The unscrutinized value may be incorrectly assumed to be within a certain
range by later operations.
In vhost_user_read: An unscrutinized value from an untrusted source used
in a trusted context - the value of sz_payload may be harmfull and we need
limit them to the max value of payload.
Coverity issue: 139601
Fixes: 6a84c37e3975 ("net/virtio-user: add vhost-user adapter layer")
Cc: jianfeng@intel.com
Cc: stable@sta...@dpdk.org
Signed-off-by: Daniel Mrzyglod
---
v2:
* Add Cc for stable in gitlog massage
* Add Coverity line
* v1 was acked by Acked-by: Jianfeng Tan
drivers/net/virtio/virtio_user/vhost_user.c | 4
1 file changed, 4 insertions(+)
diff --git a/drivers/net/virtio/virtio_user/vhost_user.c
b/drivers/net/virtio/virtio_user/vhost_user.c
index 4ad7b21..b490336 100644
--- a/drivers/net/virtio/virtio_user/vhost_user.c
+++ b/drivers/net/virtio/virtio_user/vhost_user.c
@@ -130,6 +130,10 @@ vhost_user_read(int fd, struct vhost_user_msg *msg)
}
sz_payload = msg->size;
+
+ if (sz_payload > sizeof(msg->payload))
+ goto fail;
+
if (sz_payload) {
ret = recv(fd, (void *)((char *)msg + sz_hdr), sz_payload, 0);
if (ret < sz_payload) {
--
2.7.4
Re: [dpdk-dev] [PATCH 1/2] eal: move bitmap from lib sched
On Wed, Sep 20, 2017 at 01:27:41PM +, Dumitrescu, Cristian wrote:
> > -Original Message-
> > From: Pavan Nikhilesh [mailto:pbhagavat...@caviumnetworks.com]
> > Sent: Thursday, September 7, 2017 3:40 PM
> > To: Dumitrescu, Cristian ;
> > step...@networkplumber.org
> > Cc: dev@dpdk.org; Pavan Nikhilesh
> > Subject: [dpdk-dev] [PATCH 1/2] eal: move bitmap from lib sched
> >
> > The librte_sched uses rte_bitmap to manage large arrays of bits in an
> > optimized method so, moving it to eal/common would allow other libraries
> > and applications to use it.
> >
> > Signed-off-by: Pavan Nikhilesh
> > ---
> > lib/librte_eal/common/Makefile | 1 +
> > .../common/include}/rte_bitmap.h | 100
> > +
> > lib/librte_sched/Makefile | 5 +-
> > lib/librte_sched/rte_sched.c | 2 +-
> > 4 files changed, 70 insertions(+), 38 deletions(-)
> > rename lib/{librte_sched => librte_eal/common/include}/rte_bitmap.h
> > (85%)
> >
> > diff --git a/lib/librte_eal/common/Makefile
> > b/lib/librte_eal/common/Makefile
> > index e8fd67a..c2c6a7f 100644
> > --- a/lib/librte_eal/common/Makefile
> > +++ b/lib/librte_eal/common/Makefile
> > @@ -42,6 +42,7 @@ INC += rte_hexdump.h rte_devargs.h rte_bus.h
> > rte_dev.h rte_vdev.h
> > INC += rte_pci_dev_feature_defs.h rte_pci_dev_features.h
> > INC += rte_malloc.h rte_keepalive.h rte_time.h
> > INC += rte_service.h rte_service_component.h
> > +INC += rte_bitmap.h
> >
> > GENERIC_INC := rte_atomic.h rte_byteorder.h rte_cycles.h rte_prefetch.h
> > GENERIC_INC += rte_spinlock.h rte_memcpy.h rte_cpuflags.h rte_rwlock.h
> > diff --git a/lib/librte_sched/rte_bitmap.h
> > b/lib/librte_eal/common/include/rte_bitmap.h
> > similarity index 85%
> > rename from lib/librte_sched/rte_bitmap.h
> > rename to lib/librte_eal/common/include/rte_bitmap.h
> > index 010d752..0938c63 100644
> > --- a/lib/librte_sched/rte_bitmap.h
> > +++ b/lib/librte_eal/common/include/rte_bitmap.h
> > @@ -65,6 +65,7 @@ extern "C" {
> > ***/
> >
> > #include
> > +
> > #include
> > #include
> > #include
> > @@ -72,36 +73,46 @@ extern "C" {
> > #include
> >
> > #ifndef RTE_BITMAP_OPTIMIZATIONS
> > -#define RTE_BITMAP_OPTIMIZATIONS1
> > +#define RTE_BITMAP_OPTIMIZATIONS1
> > #endif
> >
> > /* Slab */
> > -#define RTE_BITMAP_SLAB_BIT_SIZE 64
> > -#define RTE_BITMAP_SLAB_BIT_SIZE_LOG26
> > -#define RTE_BITMAP_SLAB_BIT_MASK
> > (RTE_BITMAP_SLAB_BIT_SIZE - 1)
> > +#define RTE_BITMAP_SLAB_BIT_SIZE64
> > +#define RTE_BITMAP_SLAB_BIT_SIZE_LOG2 6
> > +#define RTE_BITMAP_SLAB_BIT_MASK(RTE_BITMAP_SLAB_BIT_SIZE -
> > 1)
> >
> > /* Cache line (CL) */
> > -#define RTE_BITMAP_CL_BIT_SIZE (RTE_CACHE_LINE_SIZE * 8)
> > -#define RTE_BITMAP_CL_BIT_SIZE_LOG2
> > (RTE_CACHE_LINE_SIZE_LOG2 + 3)
> > -#define RTE_BITMAP_CL_BIT_MASK (RTE_BITMAP_CL_BIT_SIZE -
> > 1)
> > +#define RTE_BITMAP_CL_BIT_SIZE (RTE_CACHE_LINE_SIZE * 8)
> > +#define RTE_BITMAP_CL_BIT_SIZE_LOG2 (RTE_CACHE_LINE_SIZE_LOG2
> > + 3)
> > +#define RTE_BITMAP_CL_BIT_MASK (RTE_BITMAP_CL_BIT_SIZE - 1)
> >
> > -#define RTE_BITMAP_CL_SLAB_SIZE (RTE_BITMAP_CL_BIT_SIZE /
> > RTE_BITMAP_SLAB_BIT_SIZE)
> > -#define RTE_BITMAP_CL_SLAB_SIZE_LOG2
> > (RTE_BITMAP_CL_BIT_SIZE_LOG2 - RTE_BITMAP_SLAB_BIT_SIZE_LOG2)
> > -#define RTE_BITMAP_CL_SLAB_MASK
> > (RTE_BITMAP_CL_SLAB_SIZE - 1)
> > +#define RTE_BITMAP_CL_SLAB_SIZE \
> > + (RTE_BITMAP_CL_BIT_SIZE / RTE_BITMAP_SLAB_BIT_SIZE)
> > +#define RTE_BITMAP_CL_SLAB_SIZE_LOG2\
> > + (RTE_BITMAP_CL_BIT_SIZE_LOG2 -
> > RTE_BITMAP_SLAB_BIT_SIZE_LOG2)
> > +#define RTE_BITMAP_CL_SLAB_MASK (RTE_BITMAP_CL_SLAB_SIZE - 1)
> >
> > /** Bitmap data structure */
> > struct rte_bitmap {
> > /* Context for array1 and array2 */
> > - uint64_t *array1;/**< Bitmap array1 */
> > - uint64_t *array2;/**< Bitmap array2 */
> > - uint32_t array1_size;/**< Number of 64-bit slabs in
> > array1
> > that are actually used */
> > - uint32_t array2_size;/**< Number of 64-bit slabs in
> > array2
> > */
> > + uint64_t *array1;
> > + /**< Bitmap array1 */
> > + uint64_t *array2;
> > + /**< Bitmap array2 */
> > + uint32_t array1_size;
> > + /**< Number of 64-bit slabs in array1 that are actually used */
> > + uint32_t array2_size;
> > + /**< Number of 64-bit slabs in array2 */
> >
> > /* Context for the "scan next" operation */
> > - uint32_t index1; /**< Bitmap scan: Index of current array1 slab */
> > - uint32_t offset1; /**< Bitmap scan: Offset of current bit within
> > current array1 slab */
> > - uint32_t index2; /**< Bitmap scan: Index of current array2 slab */
> > - uint32_t go2; /**< Bitmap scan: Go/stop condition for current
> > array2 cache
[dpdk-dev] [PATCH v5 0/5]
This patchset refactors the eventdev API to be more flexible and capable. In particular, the API is capable of returning an error value if an invalid device, port or attribute ID is passed in, which was not possible with the previous APIs. The implementation of this patchset is based on a v1 patch[1], and after some discussion this API was seen as the best solution. In terms of flexibility, the attribute id allows addition of new common eventdev layer attributes without breaking ABI or adding new functions. Note that these attributes are not data-path, and that PMDs should continue to use the xstats API for reporting any unique PMD statistics that are available. Regarding API/ABI compatibility, I have removed the functions from the .map files - please review the .map file changes for ABI issues carefully. The last patch of this series adds a started attribute to the device, allowing the application to query if a device is currently running. -Harry [1] http://dpdk.org/dev/patchwork/patch/27152/ --- v5: - Bump library version of Eventdev (Jerin) - http://dpdk.org/ml/archives/dev/2017-September/075551.html v4: - Rework based on review by Jerin - default: cases into switches - Remove old functions from .map file - Remove /* out */ parameters - Rework header file definitions to match logical order - Rework patch split - Cleaner removal of queue_count() function v3: - Fix checkpatch issues... somehow I broke my checkpatch script :/ v2: - New APIs design based on discussion of initial patch. Harry van Haaren (5): eventdev: add port attribute function eventdev: add dev attribute get function eventdev: add queue attribute function eventdev: add device started attribute eventdev: bump library version doc/guides/rel_notes/release_17_11.rst | 2 +- lib/librte_eventdev/Makefile | 2 +- lib/librte_eventdev/rte_eventdev.c | 97 -- lib/librte_eventdev/rte_eventdev.h | 115 +++-- lib/librte_eventdev/rte_eventdev_version.map | 14 ++- test/test/test_eventdev.c| 132 +++-- test/test/test_eventdev_octeontx.c | 143 --- 7 files changed, 347 insertions(+), 158 deletions(-) -- 2.7.4
[dpdk-dev] [PATCH v5 1/5] eventdev: add port attribute function
This commit reworks the port functions to retrieve information
about the port, like the enq or deq depths. Note that "port count"
is a device attribute, and is added in a later patch for dev attributes.
Signed-off-by: Harry van Haaren
---
v4: feedback from http://dpdk.org/dev/patchwork/patch/28520/
- Added default: case into switch, returns -EINVAL as per header file (Jerin)
- Reworked .map file, removed old functions (Jerin)
- Renamed EXPERIMENTAL section of map file to 17.11 (Jerin)
- Removed /* out */ comments in .h and .c (Jerin)
---
lib/librte_eventdev/rte_eventdev.c | 35
lib/librte_eventdev/rte_eventdev.h | 49
lib/librte_eventdev/rte_eventdev_version.map | 9 +++--
test/test/test_eventdev.c| 16 ++---
4 files changed, 61 insertions(+), 48 deletions(-)
diff --git a/lib/librte_eventdev/rte_eventdev.c
b/lib/librte_eventdev/rte_eventdev.c
index bbb3805..9eacc73 100644
--- a/lib/librte_eventdev/rte_eventdev.c
+++ b/lib/librte_eventdev/rte_eventdev.c
@@ -744,30 +744,37 @@ rte_event_port_setup(uint8_t dev_id, uint8_t port_id,
}
uint8_t
-rte_event_port_dequeue_depth(uint8_t dev_id, uint8_t port_id)
+rte_event_port_count(uint8_t dev_id)
{
struct rte_eventdev *dev;
dev = &rte_eventdevs[dev_id];
- return dev->data->ports_dequeue_depth[port_id];
+ return dev->data->nb_ports;
}
-uint8_t
-rte_event_port_enqueue_depth(uint8_t dev_id, uint8_t port_id)
+int
+rte_event_port_attr_get(uint8_t dev_id, uint8_t port_id, uint32_t attr_id,
+ uint32_t *attr_value)
{
struct rte_eventdev *dev;
-
+ RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL);
dev = &rte_eventdevs[dev_id];
- return dev->data->ports_enqueue_depth[port_id];
-}
-
-uint8_t
-rte_event_port_count(uint8_t dev_id)
-{
- struct rte_eventdev *dev;
+ if (!is_valid_port(dev, port_id)) {
+ RTE_EDEV_LOG_ERR("Invalid port_id=%" PRIu8, port_id);
+ return -EINVAL;
+ }
- dev = &rte_eventdevs[dev_id];
- return dev->data->nb_ports;
+ switch (attr_id) {
+ case RTE_EVENT_PORT_ATTR_ENQ_DEPTH:
+ *attr_value = dev->data->ports_enqueue_depth[port_id];
+ break;
+ case RTE_EVENT_PORT_ATTR_DEQ_DEPTH:
+ *attr_value = dev->data->ports_dequeue_depth[port_id];
+ break;
+ default:
+ return -EINVAL;
+ };
+ return 0;
}
int
diff --git a/lib/librte_eventdev/rte_eventdev.h
b/lib/librte_eventdev/rte_eventdev.h
index 128bc52..e1febfa 100644
--- a/lib/librte_eventdev/rte_eventdev.h
+++ b/lib/librte_eventdev/rte_eventdev.h
@@ -715,47 +715,40 @@ rte_event_port_setup(uint8_t dev_id, uint8_t port_id,
const struct rte_event_port_conf *port_conf);
/**
- * Get the number of dequeue queue depth configured for event port designated
- * by its *port_id* on a specific event device
+ * Get the number of ports on a specific event device
*
* @param dev_id
* Event device identifier.
- * @param port_id
- * Event port identifier.
* @return
- * - The number of configured dequeue queue depth
- *
- * @see rte_event_dequeue_burst()
+ * - The number of configured ports
*/
uint8_t
-rte_event_port_dequeue_depth(uint8_t dev_id, uint8_t port_id);
+rte_event_port_count(uint8_t dev_id);
/**
- * Get the number of enqueue queue depth configured for event port designated
- * by its *port_id* on a specific event device
- *
- * @param dev_id
- * Event device identifier.
- * @param port_id
- * Event port identifier.
- * @return
- * - The number of configured enqueue queue depth
- *
- * @see rte_event_enqueue_burst()
+ * The queue depth of the port on the enqueue side
*/
-uint8_t
-rte_event_port_enqueue_depth(uint8_t dev_id, uint8_t port_id);
+#define RTE_EVENT_PORT_ATTR_ENQ_DEPTH 0
+/**
+ * The queue depth of the port on the dequeue side
+ */
+#define RTE_EVENT_PORT_ATTR_DEQ_DEPTH 1
/**
- * Get the number of ports on a specific event device
+ * Get an attribute from a port.
*
- * @param dev_id
- * Event device identifier.
- * @return
- * - The number of configured ports
+ * @param dev_id Eventdev id
+ * @param port_id Eventdev port id
+ * @param attr_id The attribute ID to retrieve
+ * @param[out] attr_value A pointer that will be filled in with the attribute
+ * value if successful
+ *
+ * @retval 0 Successfully returned value
+ * -EINVAL Invalid device, port or attr_id, or attr_value was NULL
*/
-uint8_t
-rte_event_port_count(uint8_t dev_id);
+int
+rte_event_port_attr_get(uint8_t dev_id, uint8_t port_id, uint32_t attr_id,
+ uint32_t *attr_value);
/**
* Start an event device.
diff --git a/lib/librte_eventdev/rte_eventdev_version.map
b/lib/librte_eventdev/rte_eventdev_version.map
index 4c48e5f..57d0b72 100644
--- a/lib/librte_eventdev/rte_eventdev_version.map
++
[dpdk-dev] [PATCH v5 3/5] eventdev: add queue attribute function
This commit adds a generic queue attribute function. It also removes
the previous rte_event_queue_priority() and priority() functions, and
updates the map files and unit tests to use the new attr functions.
Signed-off-by: Harry van Haaren
---
v4: [feedback http://dpdk.org/dev/patchwork/patch/28522/ ]
- Add default case with -EINVAL (Jerin)
- Remove explicity /* out */ comments on parameters (Jerin)
- Move queue_attr function in header to queue area (Jerin)
- Remove queue_count() fixes, should be in 2/4 patch (Harry)
---
lib/librte_eventdev/rte_eventdev.c | 40 +++-
lib/librte_eventdev/rte_eventdev.h | 35 +++-
lib/librte_eventdev/rte_eventdev_version.map | 2 +-
test/test/test_eventdev.c| 6 -
4 files changed, 49 insertions(+), 34 deletions(-)
diff --git a/lib/librte_eventdev/rte_eventdev.c
b/lib/librte_eventdev/rte_eventdev.c
index ea76444..73e1df0 100644
--- a/lib/librte_eventdev/rte_eventdev.c
+++ b/lib/librte_eventdev/rte_eventdev.c
@@ -609,18 +609,6 @@ rte_event_queue_setup(uint8_t dev_id, uint8_t queue_id,
return (*dev->dev_ops->queue_setup)(dev, queue_id, queue_conf);
}
-uint8_t
-rte_event_queue_priority(uint8_t dev_id, uint8_t queue_id)
-{
- struct rte_eventdev *dev;
-
- dev = &rte_eventdevs[dev_id];
- if (dev->data->event_dev_cap & RTE_EVENT_DEV_CAP_QUEUE_QOS)
- return dev->data->queues_prio[queue_id];
- else
- return RTE_EVENT_DEV_PRIORITY_NORMAL;
-}
-
static inline int
is_valid_port(struct rte_eventdev *dev, uint8_t port_id)
{
@@ -789,6 +777,34 @@ rte_event_port_attr_get(uint8_t dev_id, uint8_t port_id,
uint32_t attr_id,
}
int
+rte_event_queue_attr_get(uint8_t dev_id, uint8_t queue_id, uint32_t attr_id,
+ uint32_t *attr_value)
+{
+ struct rte_eventdev *dev;
+
+ if (!attr_value)
+ return -EINVAL;
+
+ RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL);
+ dev = &rte_eventdevs[dev_id];
+ if (!is_valid_queue(dev, queue_id)) {
+ RTE_EDEV_LOG_ERR("Invalid queue_id=%" PRIu8, queue_id);
+ return -EINVAL;
+ }
+
+ switch (attr_id) {
+ case RTE_EVENT_QUEUE_ATTR_PRIORITY:
+ *attr_value = RTE_EVENT_DEV_PRIORITY_NORMAL;
+ if (dev->data->event_dev_cap & RTE_EVENT_DEV_CAP_QUEUE_QOS)
+ *attr_value = dev->data->queues_prio[queue_id];
+ break;
+ default:
+ return -EINVAL;
+ };
+ return 0;
+}
+
+int
rte_event_port_link(uint8_t dev_id, uint8_t port_id,
const uint8_t queues[], const uint8_t priorities[],
uint16_t nb_links)
diff --git a/lib/librte_eventdev/rte_eventdev.h
b/lib/librte_eventdev/rte_eventdev.h
index 0f7f10a..3fd447a 100644
--- a/lib/librte_eventdev/rte_eventdev.h
+++ b/lib/librte_eventdev/rte_eventdev.h
@@ -633,31 +633,26 @@ rte_event_queue_setup(uint8_t dev_id, uint8_t queue_id,
const struct rte_event_queue_conf *queue_conf);
/**
- * Get the number of event queues on a specific event device
- *
- * @param dev_id
- * Event device identifier.
- * @return
- * - The number of configured event queues
+ * The priority of the queue.
*/
-uint8_t
-rte_event_queue_count(uint8_t dev_id);
+#define RTE_EVENT_QUEUE_ATTR_PRIORITY 0
/**
- * Get the priority of the event queue on a specific event device
+ * Get an attribute from a queue.
*
- * @param dev_id
- * Event device identifier.
- * @param queue_id
- * Event queue identifier.
- * @return
- * - If the device has RTE_EVENT_DEV_CAP_QUEUE_QOS capability then the
- *configured priority of the event queue in
- *[RTE_EVENT_DEV_PRIORITY_HIGHEST, RTE_EVENT_DEV_PRIORITY_LOWEST] range
- *else the value RTE_EVENT_DEV_PRIORITY_NORMAL
+ * @param dev_id Eventdev id
+ * @param queue_id Eventdev queue id
+ * @param attr_id The attribute ID to retrieve
+ * @param[out] attr_value A pointer that will be filled in with the attribute
+ * value if successful
+ *
+ * @retval 0 Successfully returned value
+ * -EINVAL invalid device, queue or attr_id provided, or attr_value
+ * was NULL
*/
-uint8_t
-rte_event_queue_priority(uint8_t dev_id, uint8_t queue_id);
+int
+rte_event_queue_attr_get(uint8_t dev_id, uint8_t queue_id, uint32_t attr_id,
+ uint32_t *attr_value);
/* Event port specific APIs */
diff --git a/lib/librte_eventdev/rte_eventdev_version.map
b/lib/librte_eventdev/rte_eventdev_version.map
index 1353a6d..90266a8 100644
--- a/lib/librte_eventdev/rte_eventdev_version.map
+++ b/lib/librte_eventdev/rte_eventdev_version.map
@@ -25,7 +25,6 @@ DPDK_17.05 {
rte_event_queue_default_conf_get;
rte_event_queue_setup;
- rte_event_queue_priority;
rte_event_dequeue_timeout_ticks;
@@ -53,5 +52,6 @@ DPDK_17.11 {
rte_event_dev_attr_
[dpdk-dev] [PATCH v5 2/5] eventdev: add dev attribute get function
This commit adds a device attribute function, allowing flexible
fetching of device attributes, like port count or queue count.
The unit tests and .map file are updated to the new function.
Signed-off-by: Harry van Haaren
Acked-by: Jerin Jacob
---
v4: [feedback http://dpdk.org/dev/patchwork/patch/28521/ ]
- Fixed default case comment (Jerin)
- Removed /* out */ comment in .h and .c (Jerin)
- Move function declaration in .h file to dev area (Jerin)
- Fix inconsistent removal of queue_count() function (Harry)
---
lib/librte_eventdev/rte_eventdev.c | 35 ---
lib/librte_eventdev/rte_eventdev.h | 37 ---
lib/librte_eventdev/rte_eventdev_version.map | 3 +-
test/test/test_eventdev.c| 110 -
test/test/test_eventdev_octeontx.c | 143 ---
5 files changed, 241 insertions(+), 87 deletions(-)
diff --git a/lib/librte_eventdev/rte_eventdev.c
b/lib/librte_eventdev/rte_eventdev.c
index 9eacc73..ea76444 100644
--- a/lib/librte_eventdev/rte_eventdev.c
+++ b/lib/librte_eventdev/rte_eventdev.c
@@ -610,15 +610,6 @@ rte_event_queue_setup(uint8_t dev_id, uint8_t queue_id,
}
uint8_t
-rte_event_queue_count(uint8_t dev_id)
-{
- struct rte_eventdev *dev;
-
- dev = &rte_eventdevs[dev_id];
- return dev->data->nb_queues;
-}
-
-uint8_t
rte_event_queue_priority(uint8_t dev_id, uint8_t queue_id)
{
struct rte_eventdev *dev;
@@ -743,13 +734,29 @@ rte_event_port_setup(uint8_t dev_id, uint8_t port_id,
return 0;
}
-uint8_t
-rte_event_port_count(uint8_t dev_id)
+int
+rte_event_dev_attr_get(uint8_t dev_id, uint32_t attr_id,
+ uint32_t *attr_value)
{
struct rte_eventdev *dev;
+ if (!attr_value)
+ return -EINVAL;
+ RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL);
dev = &rte_eventdevs[dev_id];
- return dev->data->nb_ports;
+
+ switch (attr_id) {
+ case RTE_EVENT_DEV_ATTR_PORT_COUNT:
+ *attr_value = dev->data->nb_ports;
+ break;
+ case RTE_EVENT_DEV_ATTR_QUEUE_COUNT:
+ *attr_value = dev->data->nb_queues;
+ break;
+ default:
+ return -EINVAL;
+ }
+
+ return 0;
}
int
@@ -757,6 +764,10 @@ rte_event_port_attr_get(uint8_t dev_id, uint8_t port_id,
uint32_t attr_id,
uint32_t *attr_value)
{
struct rte_eventdev *dev;
+
+ if (!attr_value)
+ return -EINVAL;
+
RTE_EVENTDEV_VALID_DEVID_OR_ERR_RET(dev_id, -EINVAL);
dev = &rte_eventdevs[dev_id];
if (!is_valid_port(dev, port_id)) {
diff --git a/lib/librte_eventdev/rte_eventdev.h
b/lib/librte_eventdev/rte_eventdev.h
index e1febfa..0f7f10a 100644
--- a/lib/librte_eventdev/rte_eventdev.h
+++ b/lib/librte_eventdev/rte_eventdev.h
@@ -399,6 +399,32 @@ struct rte_event_dev_info {
int
rte_event_dev_info_get(uint8_t dev_id, struct rte_event_dev_info *dev_info);
+/**
+ * The count of ports.
+ */
+#define RTE_EVENT_DEV_ATTR_PORT_COUNT 0
+/**
+ * The count of queues.
+ */
+#define RTE_EVENT_DEV_ATTR_QUEUE_COUNT 1
+
+/**
+ * Get an attribute from a device.
+ *
+ * @param dev_id Eventdev id
+ * @param attr_id The attribute ID to retrieve
+ * @param[out] attr_value A pointer that will be filled in with the attribute
+ * value if successful.
+ *
+ * @retval 0 Successfully retrieved attribute value
+ * -EINVAL Invalid device or *attr_id* provided, or *attr_value*
+ * is NULL
+ */
+int
+rte_event_dev_attr_get(uint8_t dev_id, uint32_t attr_id,
+ uint32_t *attr_value);
+
+
/* Event device configuration bitmap flags */
#define RTE_EVENT_DEV_CFG_PER_DEQUEUE_TIMEOUT (1ULL << 0)
/**< Override the global *dequeue_timeout_ns* and use per dequeue timeout in
ns.
@@ -715,17 +741,6 @@ rte_event_port_setup(uint8_t dev_id, uint8_t port_id,
const struct rte_event_port_conf *port_conf);
/**
- * Get the number of ports on a specific event device
- *
- * @param dev_id
- * Event device identifier.
- * @return
- * - The number of configured ports
- */
-uint8_t
-rte_event_port_count(uint8_t dev_id);
-
-/**
* The queue depth of the port on the enqueue side
*/
#define RTE_EVENT_PORT_ATTR_ENQ_DEPTH 0
diff --git a/lib/librte_eventdev/rte_eventdev_version.map
b/lib/librte_eventdev/rte_eventdev_version.map
index 57d0b72..1353a6d 100644
--- a/lib/librte_eventdev/rte_eventdev_version.map
+++ b/lib/librte_eventdev/rte_eventdev_version.map
@@ -19,14 +19,12 @@ DPDK_17.05 {
rte_event_port_default_conf_get;
rte_event_port_setup;
- rte_event_port_count;
rte_event_port_link;
rte_event_port_unlink;
rte_event_port_links_get;
rte_event_queue_default_conf_get;
rte_event_queue_setup;
- rte_event_queue_count;
rte_event_queue_priority;
rte_event_dequeue_timeout_ticks;
@@ -53,6 +51,7
[dpdk-dev] [PATCH v5 4/5] eventdev: add device started attribute
This commit adds an attribute to the eventdev, allowing applications to retrieve if the eventdev is running or stopped. Note that no API or ABI changes were required in adding the statistic, and code changes are minimal. Signed-off-by: Harry van Haaren Acked-by: Jerin Jacob --- lib/librte_eventdev/rte_eventdev.c | 3 +++ lib/librte_eventdev/rte_eventdev.h | 4 2 files changed, 7 insertions(+) diff --git a/lib/librte_eventdev/rte_eventdev.c b/lib/librte_eventdev/rte_eventdev.c index 73e1df0..480b26e 100644 --- a/lib/librte_eventdev/rte_eventdev.c +++ b/lib/librte_eventdev/rte_eventdev.c @@ -740,6 +740,9 @@ rte_event_dev_attr_get(uint8_t dev_id, uint32_t attr_id, case RTE_EVENT_DEV_ATTR_QUEUE_COUNT: *attr_value = dev->data->nb_queues; break; + case RTE_EVENT_DEV_ATTR_STARTED: + *attr_value = dev->data->dev_started; + break; default: return -EINVAL; } diff --git a/lib/librte_eventdev/rte_eventdev.h b/lib/librte_eventdev/rte_eventdev.h index 3fd447a..1b629fc 100644 --- a/lib/librte_eventdev/rte_eventdev.h +++ b/lib/librte_eventdev/rte_eventdev.h @@ -407,6 +407,10 @@ rte_event_dev_info_get(uint8_t dev_id, struct rte_event_dev_info *dev_info); * The count of queues. */ #define RTE_EVENT_DEV_ATTR_QUEUE_COUNT 1 +/** + * The status of the device, zero for stopped, non-zero for started. + */ +#define RTE_EVENT_DEV_ATTR_STARTED 2 /** * Get an attribute from a device. -- 2.7.4
[dpdk-dev] [PATCH v5 5/5] eventdev: bump library version
This commit bumps the library version to refect the ABI change caused by removing the individual rte_event_port_count, queue_count, and other get functions. These functions are superseded by the get-attribute style API, which allows fetching values without API/ABI changes. Signed-off-by: Harry van Haaren --- doc/guides/rel_notes/release_17_11.rst | 2 +- lib/librte_eventdev/Makefile | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/doc/guides/rel_notes/release_17_11.rst b/doc/guides/rel_notes/release_17_11.rst index 170f4f9..c09b1a1 100644 --- a/doc/guides/rel_notes/release_17_11.rst +++ b/doc/guides/rel_notes/release_17_11.rst @@ -152,7 +152,7 @@ The libraries prepended with a plus sign were incremented in this version. librte_distributor.so.1 librte_eal.so.5 librte_ethdev.so.7 - librte_eventdev.so.2 + + librte_eventdev.so.3 librte_gro.so.1 librte_hash.so.2 librte_ip_frag.so.1 diff --git a/lib/librte_eventdev/Makefile b/lib/librte_eventdev/Makefile index 410578a..7d73ca9 100644 --- a/lib/librte_eventdev/Makefile +++ b/lib/librte_eventdev/Makefile @@ -34,7 +34,7 @@ include $(RTE_SDK)/mk/rte.vars.mk LIB = librte_eventdev.a # library version -LIBABIVER := 2 +LIBABIVER := 3 # build flags CFLAGS += -O3 -- 2.7.4
Re: [dpdk-dev] [PATCH 1/2] net/bonding: rename map file to standard name
On 9/14/2017 4:57 PM, Bruce Richardson wrote: > Naming convention for net drivers is "rte_pmd__version.map" > > Signed-off-by: Bruce Richardson Applied to dpdk-next-net/master, thanks.
Re: [dpdk-dev] [PATCH 2/2] net/xenvirt: rename map file to standard name
On 9/14/2017 4:57 PM, Bruce Richardson wrote: > Naming convention for net drivers is "rte_pmd__version.map" > > Signed-off-by: Bruce Richardson This one not get, because xenvirt will be removed this release.
Re: [dpdk-dev] [PATCH v5] ethdev: add return code to rte_eth_stats_reset()
> -Original Message-
> From: Ferruh Yigit [mailto:ferruh.yi...@intel.com]
>
> On 9/1/2017 3:26 AM, David Harton wrote:
> > Some devices do not support reset of eth stats. An application may
> > need to know not to clear shadow stats if the device cannot.
> >
> > rte_eth_stats_reset is updated to provide a return code to share
> > whether the device supports reset or not.
> >
> > Signed-off-by: David Harton
> > ---
> >
> > v5:
> > * squashed doc patch
> > * moved rel_note change from ABI to API section
> >
> > v4:
> > * commented return values
> >
> > v3:
> > * overcame noob errors and figured out patch challenged
> >
> > v2:
> > * fixed soft tab issue inserted while moving changes
> >
> > doc/guides/rel_notes/release_17_11.rst | 13 +
> > lib/librte_ether/rte_ethdev.c | 8 +---
> > lib/librte_ether/rte_ethdev.h | 6 +-
> > 3 files changed, 23 insertions(+), 4 deletions(-)
> >
> > diff --git a/doc/guides/rel_notes/release_17_11.rst
> > b/doc/guides/rel_notes/release_17_11.rst
> > index 22df4fd..6282667 100644
> > --- a/doc/guides/rel_notes/release_17_11.rst
> > +++ b/doc/guides/rel_notes/release_17_11.rst
> > @@ -110,6 +110,19 @@ API Changes
> > Also, make sure to start the actual text at the margin.
> > =
> >
> > +* **Modified the return type of rte_eth_stats_reset.**
> > +
> > + Changed return type of ``rte_eth_stats_reset`` from ``void`` to
> > + ``int`` so the caller may know whether a device supports the
> > + operation or not and if the operation was carried out.
> > +
>
> > +* **Modified the vlan_offload_set_t function prototype in the ethdev
> > +library.**
> > +
> > + Changed the function prototype of ``vlan_offload_set_t``. The
> > + return value has been changed from ``void`` to ``int`` so the
> > + caller to knows whether the backing device supports the operation
> > + or if the operation was successfully performed.
> > +
>
> Is this addition to the document related to this patch?
Good catch. No. :(
I must have mishandled the rebase I did to update this patch. V6 coming.
Would be great if you could re-ACK afterwards so this one can move.
Thanks,
Dave
>
> > * **Modified the vlan_offload_set_t function prototype in the ethdev
> > library.**
> >
> >Changed the function prototype of ``vlan_offload_set_t``. The
> > return value diff --git a/lib/librte_ether/rte_ethdev.c
> > b/lib/librte_ether/rte_ethdev.c index 05e52b8..f0f1775 100644
> > --- a/lib/librte_ether/rte_ethdev.c
> > +++ b/lib/librte_ether/rte_ethdev.c
> > @@ -1341,17 +1341,19 @@ struct rte_eth_dev *
> > return 0;
> > }
> >
> > -void
> > +int
> > rte_eth_stats_reset(uint8_t port_id)
> > {
> > struct rte_eth_dev *dev;
> >
> > - RTE_ETH_VALID_PORTID_OR_RET(port_id);
> > + RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
> > dev = &rte_eth_devices[port_id];
> >
> > - RTE_FUNC_PTR_OR_RET(*dev->dev_ops->stats_reset);
> > + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->stats_reset, -ENOTSUP);
> > (*dev->dev_ops->stats_reset)(dev);
> > dev->data->rx_mbuf_alloc_failed = 0;
> > +
> > + return 0;
> > }
> >
> > static int
> > diff --git a/lib/librte_ether/rte_ethdev.h
> > b/lib/librte_ether/rte_ethdev.h index 7254fd0..9110725 100644
> > --- a/lib/librte_ether/rte_ethdev.h
> > +++ b/lib/librte_ether/rte_ethdev.h
> > @@ -2246,8 +2246,12 @@ int rte_eth_tx_queue_setup(uint8_t port_id,
> uint16_t tx_queue_id,
> > *
> > * @param port_id
> > * The port identifier of the Ethernet device.
> > + * @return
> > + * - (0) if device notified to reset stats.
> > + * - (-ENOTSUP) if hardware doesn't support.
> > + * - (-ENODEV) if *port_id* invalid.
> > */
> > -void rte_eth_stats_reset(uint8_t port_id);
> > +int rte_eth_stats_reset(uint8_t port_id);
> >
> > /**
> > * Retrieve names of extended statistics of an Ethernet device.
> >
