[dpdk-dev] [PATCH v4 1/4] ethdev: rename rte_eth_vmdq_mirror_conf
Hi, Thomas
Any suggestions about this patch? Do I need rework it by using macro
RTE_NEXT_ABI?
Thanks a lot!
Jingjing
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Wu, Jingjing
> Sent: Friday, June 26, 2015 3:03 PM
> To: 'nhorman at tuxdriver.com'
> Cc: dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH v4 1/4] ethdev: rename
> rte_eth_vmdq_mirror_conf
>
> Hi, Neil
>
> About this patch I have an ABI concern about it.
> This patch just renamed a struct rte_eth_vmdq_mirror_conf to
> rte_eth_mirror_conf, the size and its elements don't change. As my
> understanding, it will not break the ABI. And I also tested it.
> But when I use the script ./scripts/validate-abi.sh to check. A low severity
> problem is reported in symbol "rte_eth_mirror_rule_set"
> - Change: "Base type of 2nd parameter mirror_conf has been changed from
> struct rte_eth_vmdq_mirror_conf to struct rte_eth_mirror_conf."
> - Effect: "Replacement of parameter base type may indicate a change in its
> semantic meaning."
>
> So, I'm not sure whether this patch meet the ABI policy?
>
> Additional, about the validate-abi.sh, does it mean we need to fix all the
> problems it reports? Or we can decide case by case. Can a Low Severity
> problem be acceptable?
>
> Look forward to your reply.
>
> Thanks
>
> Jingjing
>
> > -Original Message-
> > From: Wu, Jingjing
> > Sent: Wednesday, June 10, 2015 2:25 PM
> > To: dev at dpdk.org
> > Cc: Wu, Jingjing; Liu, Jijiang; Jiajia, SunX; Zhang, Helin
> > Subject: [PATCH v4 1/4] ethdev: rename rte_eth_vmdq_mirror_conf
> >
> > rename rte_eth_vmdq_mirror_conf to rte_eth_mirror_conf and move the
> > maximum rule id check from ethdev level to driver
> >
> > Signed-off-by: Jingjing Wu
> > ---
> > app/test-pmd/cmdline.c | 22 +++---
> > drivers/net/ixgbe/ixgbe_ethdev.c | 11 +++
> > drivers/net/ixgbe/ixgbe_ethdev.h | 4 +++-
> > lib/librte_ether/rte_ethdev.c| 18 ++
> > lib/librte_ether/rte_ethdev.h| 19 ++-
> > 5 files changed, 33 insertions(+), 41 deletions(-)
> >
> > diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c index
> > f01db2a..d693bde 100644
> > --- a/app/test-pmd/cmdline.c
> > +++ b/app/test-pmd/cmdline.c
> > @@ -6604,11 +6604,11 @@ cmd_set_mirror_mask_parsed(void
> > *parsed_result, {
> > int ret,nb_item,i;
> > struct cmd_set_mirror_mask_result *res = parsed_result;
> > - struct rte_eth_vmdq_mirror_conf mr_conf;
> > + struct rte_eth_mirror_conf mr_conf;
> >
> > - memset(&mr_conf,0,sizeof(struct rte_eth_vmdq_mirror_conf));
> > + memset(&mr_conf, 0, sizeof(struct rte_eth_mirror_conf));
> >
> > - unsigned int vlan_list[ETH_VMDQ_MAX_VLAN_FILTERS];
> > + unsigned int vlan_list[ETH_MIRROR_MAX_VLANS];
> >
> > mr_conf.dst_pool = res->dstpool_id;
> >
> > @@ -6618,11 +6618,11 @@ cmd_set_mirror_mask_parsed(void
> > *parsed_result,
> > } else if(!strcmp(res->what, "vlan-mirror")) {
> > mr_conf.rule_type_mask = ETH_VMDQ_VLAN_MIRROR;
> > nb_item = parse_item_list(res->value, "core",
> > -
> > ETH_VMDQ_MAX_VLAN_FILTERS,vlan_list,1);
> > + ETH_MIRROR_MAX_VLANS, vlan_list,
> > 1);
> > if (nb_item <= 0)
> > return;
> >
> > - for(i=0; i < nb_item; i++) {
> > + for (i = 0; i < nb_item; i++) {
> > if (vlan_list[i] > ETHER_MAX_VLAN_ID) {
> > printf("Invalid vlan_id: must be < 4096\n");
> > return;
> > @@ -6634,10 +6634,10 @@ cmd_set_mirror_mask_parsed(void
> > *parsed_result,
> > }
> >
> > if(!strcmp(res->on, "on"))
> > - ret = rte_eth_mirror_rule_set(res->port_id,&mr_conf,
> > + ret = rte_eth_mirror_rule_set(res->port_id, &mr_conf,
> > res->rule_id, 1);
> > else
> > - ret = rte_eth_mirror_rule_set(res->port_id,&mr_conf,
> > + ret = rte_eth_mirror_rule_set(res->port_id, &mr_conf,
> > res->rule_id, 0);
> > if(ret < 0)
> > printf("mirror rule add error: (%s)\n", strerror(-ret)); @@ -
> > 6711,9 +6711,9 @@ cmd_set_mirror_link_parsed(void *parsed_result, {
> > int ret;
> > struct cmd_set_mirror_link_result *res = parsed_result;
> > - struct rte_eth_vmdq_mirror_conf mr_conf;
> > + struct rte_eth_mirror_conf mr_conf;
> >
> > - memset(&mr_conf,0,sizeof(struct rte_eth_vmdq_mirror_conf));
> > + memset(&mr_conf, 0, sizeof(struct rte_eth_mirror_conf));
> > if(!strcmp(res->what, "uplink-mirror")) {
> > mr_conf.rule_type_mask = ETH_VMDQ_UPLINK_MIRROR;
> > }else if(!strcmp(res->what, "downlink-mirror")) @@ -6722,10
> > +6722,10 @@ cmd_set_mirror_link_parsed(void *parsed_result,
> > mr_conf.dst_pool = res->dstpool_id;
> >
> > if(!strcmp(res->on, "on"))
> > - ret = rte_eth_mirror_
[dpdk-dev] [PATCH v2 0/3] Fix vhost startup issue
The patch set fix vhost sample fails to start up in second time: It should call api to unregister vhost driver when sample exit/quit, then the socket file will be removed(by calling unlink), and thus make vhost sample work correctly in second time startup. It also adds/refines some log information. Changchun Ouyang (3): vhost: add log when failing to bind a socket vhost: fix the comments and log vhost: call api to unregister vhost driver examples/vhost/main.c| 16 ++-- lib/librte_vhost/vhost_user/vhost-net-user.c | 5 - 2 files changed, 18 insertions(+), 3 deletions(-) -- 1.8.4.2
[dpdk-dev] [PATCH v2 1/3] vhost: add log when failing to bind a socket
It adds more readable log info if a socket fails to bind to local socket file
name.
Signed-off-by: Changchun Ouyang
---
lib/librte_vhost/vhost_user/vhost-net-user.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/lib/librte_vhost/vhost_user/vhost-net-user.c
b/lib/librte_vhost/vhost_user/vhost-net-user.c
index 87a4711..f406a94 100644
--- a/lib/librte_vhost/vhost_user/vhost-net-user.c
+++ b/lib/librte_vhost/vhost_user/vhost-net-user.c
@@ -122,8 +122,11 @@ uds_socket(const char *path)
un.sun_family = AF_UNIX;
snprintf(un.sun_path, sizeof(un.sun_path), "%s", path);
ret = bind(sockfd, (struct sockaddr *)&un, sizeof(un));
- if (ret == -1)
+ if (ret == -1) {
+ RTE_LOG(ERR, VHOST_CONFIG, "fail to bind fd:%d, remove file:%s
and try again.\n",
+ sockfd, path);
goto err;
+ }
RTE_LOG(INFO, VHOST_CONFIG, "bind to %s\n", path);
ret = listen(sockfd, MAX_VIRTIO_BACKLOG);
--
1.8.4.2
[dpdk-dev] [PATCH v2 2/3] vhost: fix the comments and log
It fixes the wrong log info when failing to unregister vhost driver. Signed-off-by: Changchun Ouyang --- examples/vhost/main.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) change in v2: - refine the comment - fix checkpatch issue diff --git a/examples/vhost/main.c b/examples/vhost/main.c index 7863dcf..56a5c70 100644 --- a/examples/vhost/main.c +++ b/examples/vhost/main.c @@ -3051,10 +3051,10 @@ main(int argc, char *argv[]) if (mergeable == 0) rte_vhost_feature_disable(1ULL << VIRTIO_NET_F_MRG_RXBUF); - /* Register CUSE device to handle IOCTLs. */ + /* Register vhost(cuse or user) driver to handle vhost messages. */ ret = rte_vhost_driver_register((char *)&dev_basename); if (ret != 0) - rte_exit(EXIT_FAILURE,"CUSE device setup failure.\n"); + rte_exit(EXIT_FAILURE, "vhost driver register failure.\n"); rte_vhost_driver_callback_register(&virtio_net_device_ops); -- 1.8.4.2
[dpdk-dev] [PATCH v2 3/3] vhost: call api to unregister vhost driver
The following commit broke vhost sample when it runs in second time:
292959c71961acde0cda6e77e737bb0a4df1559c
It should call api to unregister vhost driver when sample exit/quit, then
the socket file will be removed(by calling unlink), and thus make vhost sample
work correctly in the second time startup.
Signed-off-by: Changchun Ouyang
---
examples/vhost/main.c | 12
1 file changed, 12 insertions(+)
change in v2:
- refine the signal handler name
diff --git a/examples/vhost/main.c b/examples/vhost/main.c
index 56a5c70..1b137b9 100644
--- a/examples/vhost/main.c
+++ b/examples/vhost/main.c
@@ -2871,6 +2871,16 @@ setup_mempool_tbl(int socket, uint32_t index, char
*pool_name,
}
}
+/* When we receive a INT signal, unregister vhost driver */
+static void
+sigint_handler(__rte_unused int signum)
+{
+ /* Unregister vhost driver. */
+ int ret = rte_vhost_driver_unregister((char *)&dev_basename);
+ if (ret != 0)
+ rte_exit(EXIT_FAILURE, "vhost driver unregister failure.\n");
+ exit(0);
+}
/*
* Main function, does initialisation and calls the per-lcore functions. The
CUSE
@@ -2887,6 +2897,8 @@ main(int argc, char *argv[])
uint16_t queue_id;
static pthread_t tid;
+ signal(SIGINT, sigint_handler);
+
/* init EAL */
ret = rte_eal_init(argc, argv);
if (ret < 0)
--
1.8.4.2
[dpdk-dev] [PATCH v8 00/12] Clean up pci uio implementations
Currently Linux implementation and BSD implementation have almost same code about pci uio. This patch series cleans up it. PATCH v8 changes: - Fix comments of header file. - Fix error handling logic of resource allocation functions. (Thanks to Bruce Richardson) - Add copyright. - Fix patch description. (Thanks to Iremonger, Bernard) PATCH v7 changes: - Add below patches. Also, the order of patches are changed. - eal: Add pci_uio_alloc_resource() - eal: Add pci_uio_map_resource_by_index() - eal: Consolidate pci_map and mapped_pci_resource of linuxapp and bsdapp - eal: Consolidate pci_map/unmap_resource() of linuxapp and bsdapp - eal: Consolidate pci uio functions of linuxapp and bsdapp - eal: Consolidate pci_map/unmap_device() of linuxapp and bsdapp - eal: Consolidate rte_eal_pci_probe/close_one_driver() of linuxapp and bsdapp (Thanks to Bruce Richardson) - While adding above, below patches are not changed at all. - eal: Fix coding style of eal_pci.c and eal_pci_uio.c - eal: Close file descriptor of uio configuration - eal: Fix memory leaks and needless increment of pci_map_addr - eal/bsdapp: Change names of pci related data structure - eal: Fix uio mapping differences between linuxapp and bsdapp - some function names are changed like below. - pci_uio_alloc_uio_resource() to pci_uio_alloc_resource(). - pci_uio_map_uio_resource_by_index() to pci_uio_map_resource_by_index(). (Thanks to Iremonger, Bernard) PATCH v6 changes: - Free mapped resources in pci_uio_map_resource(). - Fix error handling in pci_uio_map_resource(). (Thanks to David, Marchand) PATCH v5 changes: - Rebase to latest master branch. PATCH v4 changes: - Rebase to latest master branch. - Fix bug in pci_uio_map_resource() of BSD code. 'maps[i].path' shouldn't be freed. Fixed in below patch: [PATCH 3/5] eal: Fix memory leaks and needless increment of pci_map_addr - 'path' member of 'struct mapped_pci_resource' should not be removed because it will be used in BSD code. Fixed in below patch: [PATCH 5/5] eal: Fix uio mapping differences between linuxapp and bsdapp PATCH v3 changes: - Squash patches related with pci_map_resource(). - Free maps[].path to easy to understand. (Thanks to Iremonger, Bernard) - Close fds opened in this function. - Remove unused path variable from mapped_pci_resource structure. PATCH v2 changes: - Move 'if-condition' to later patch series. - Fix memory leaks of path. - Fix typos. (Thanks to David Marchand) - Fix commit title and body. - Fix pci_map_resource() to handle MAP_FAILED. (Thanks to Iremonger, Bernard) Changes: - This patch set is derived from below. "[PATCH v2] eal: Port Hotplug support for BSD" - Set cfg_fd as -1, when cfg_fd is closed. (Thanks to Iremonger, Bernard) - Remove needless coding style fixings. - Fix coding style of if-else condition. (Thanks to Richardson, Bruce) Tetsuya.Mukawa (12): eal: Fix coding style of eal_pci.c and eal_pci_uio.c eal: Close file descriptor of uio configuration eal: Fix memory leaks and needless increment of pci_map_addr eal/bsdapp: Change names of pci related data structure eal: Fix uio mapping differences between linuxapp and bsdapp eal: Add pci_uio_alloc_resource() eal: Add pci_uio_map_resource_by_index() eal: Consolidate pci_map and mapped_pci_resource of linuxapp and bsdapp eal: Consolidate pci_map/unmap_resource() of linuxapp and bsdapp eal: Consolidate pci uio functions of linuxapp and bsdapp eal: Consolidate pci_map/unmap_device() of linuxapp and bsdapp eal: Consolidate rte_eal_pci_probe/close_one_driver() of linuxapp and bsdapp lib/librte_eal/bsdapp/eal/Makefile | 2 + lib/librte_eal/bsdapp/eal/eal_pci.c| 293 ++--- .../bsdapp/eal/include/exec-env/rte_interrupts.h | 2 + lib/librte_eal/common/eal_common_pci.c | 226 lib/librte_eal/common/eal_common_pci_uio.c | 240 + lib/librte_eal/common/eal_private.h| 60 - lib/librte_eal/common/include/rte_pci.h| 62 + lib/librte_eal/linuxapp/eal/Makefile | 2 + lib/librte_eal/linuxapp/eal/eal_pci.c | 237 + lib/librte_eal/linuxapp/eal/eal_pci_init.h | 40 +-- lib/librte_eal/linuxapp/eal/eal_pci_uio.c | 293 +++-- lib/librte_ether/rte_ethdev.c | 2 + 12 files changed, 761 insertions(+), 698 deletions(-) create mode 100644 lib/librte_eal/common/eal_common_pci_uio.c -- 2.1.4
[dpdk-dev] [PATCH v8 01/12] eal: Fix coding style of eal_pci.c and eal_pci_uio.c
From: "Tetsuya.Mukawa"
This patch fixes coding style of below files in linuxapp and bsdapp.
- eal_pci.c
- eal_pci_uio.c
Signed-off-by: Tetsuya Mukawa
Acked-by: Stephen Hemminger
---
lib/librte_eal/bsdapp/eal/eal_pci.c | 13 -
lib/librte_eal/linuxapp/eal/eal_pci_uio.c | 13 +
2 files changed, 17 insertions(+), 9 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c
b/lib/librte_eal/bsdapp/eal/eal_pci.c
index 2df5c1c..a63d450 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -2,6 +2,7 @@
* BSD LICENSE
*
* Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ * Copyright(c) 2015 IGEL Co., Ltd.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -161,9 +162,10 @@ fail:
static int
pci_uio_map_secondary(struct rte_pci_device *dev)
{
-size_t i;
-struct uio_resource *uio_res;
- struct uio_res_list *uio_res_list = RTE_TAILQ_CAST(rte_uio_tailq.head,
uio_res_list);
+ size_t i;
+ struct uio_resource *uio_res;
+ struct uio_res_list *uio_res_list =
+ RTE_TAILQ_CAST(rte_uio_tailq.head, uio_res_list);
TAILQ_FOREACH(uio_res, uio_res_list, next) {
@@ -201,7 +203,8 @@ pci_uio_map_resource(struct rte_pci_device *dev)
uint64_t pagesz;
struct rte_pci_addr *loc = &dev->addr;
struct uio_resource *uio_res;
- struct uio_res_list *uio_res_list = RTE_TAILQ_CAST(rte_uio_tailq.head,
uio_res_list);
+ struct uio_res_list *uio_res_list =
+ RTE_TAILQ_CAST(rte_uio_tailq.head, uio_res_list);
struct uio_map *maps;
dev->intr_handle.fd = -1;
@@ -311,7 +314,7 @@ pci_scan_one(int dev_pci_fd, struct pci_conf *conf)
/* FreeBSD has no NUMA support (yet) */
dev->numa_node = 0;
-/* parse resources */
+ /* parse resources */
switch (conf->pc_hdr & PCIM_HDRTYPE) {
case PCIM_HDRTYPE_NORMAL:
max = PCIR_MAX_BAR_0;
diff --git a/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
b/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
index b5116a7..5915a84 100644
--- a/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
+++ b/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
@@ -2,6 +2,7 @@
* BSD LICENSE
*
* Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ * Copyright(c) 2015 IGEL Co., Ltd.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -92,7 +93,8 @@ pci_uio_map_secondary(struct rte_pci_device *dev)
{
int fd, i;
struct mapped_pci_resource *uio_res;
- struct mapped_pci_res_list *uio_res_list =
RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
+ struct mapped_pci_res_list *uio_res_list =
+ RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
TAILQ_FOREACH(uio_res, uio_res_list, next) {
@@ -272,7 +274,8 @@ pci_uio_map_resource(struct rte_pci_device *dev)
uint64_t phaddr;
struct rte_pci_addr *loc = &dev->addr;
struct mapped_pci_resource *uio_res;
- struct mapped_pci_res_list *uio_res_list =
RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
+ struct mapped_pci_res_list *uio_res_list =
+ RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
struct pci_map *maps;
dev->intr_handle.fd = -1;
@@ -417,7 +420,8 @@ static struct mapped_pci_resource *
pci_uio_find_resource(struct rte_pci_device *dev)
{
struct mapped_pci_resource *uio_res;
- struct mapped_pci_res_list *uio_res_list =
RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
+ struct mapped_pci_res_list *uio_res_list =
+ RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
if (dev == NULL)
return NULL;
@@ -436,7 +440,8 @@ void
pci_uio_unmap_resource(struct rte_pci_device *dev)
{
struct mapped_pci_resource *uio_res;
- struct mapped_pci_res_list *uio_res_list =
RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
+ struct mapped_pci_res_list *uio_res_list =
+ RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
if (dev == NULL)
return;
--
2.1.4
[dpdk-dev] [PATCH v8 02/12] eal: Close file descriptor of uio configuration
From: "Tetsuya.Mukawa" When pci_uio_unmap_resource() is called, a file descriptor that is used for uio configuration should be closed. Signed-off-by: Tetsuya Mukawa Acked-by: Stephen Hemminger --- lib/librte_eal/linuxapp/eal/eal_pci_uio.c | 6 +- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/lib/librte_eal/linuxapp/eal/eal_pci_uio.c b/lib/librte_eal/linuxapp/eal/eal_pci_uio.c index 5915a84..37dc936 100644 --- a/lib/librte_eal/linuxapp/eal/eal_pci_uio.c +++ b/lib/librte_eal/linuxapp/eal/eal_pci_uio.c @@ -465,8 +465,12 @@ pci_uio_unmap_resource(struct rte_pci_device *dev) /* close fd if in primary process */ close(dev->intr_handle.fd); - dev->intr_handle.fd = -1; + + /* close cfg_fd if in primary process */ + close(dev->intr_handle.uio_cfg_fd); + dev->intr_handle.uio_cfg_fd = -1; + dev->intr_handle.type = RTE_INTR_HANDLE_UNKNOWN; } #endif /* RTE_LIBRTE_EAL_HOTPLUG */ -- 2.1.4
[dpdk-dev] [PATCH v8 03/12] eal: Fix memory leaks and needless increment of pci_map_addr
From: "Tetsuya.Mukawa"
This patch fixes following memory leaks.
- When open() is failed, uio_res and fds won't be freed in
pci_uio_map_resource().
- When pci_map_resource() is failed but path is allocated correctly,
path and fds won't be freed in pci_uio_map_recource().
Also, some mapped resources should be freed.
- When pci_uio_unmap() is called, path should be freed.
Also, fixes below.
- When pci_map_resource() is failed, mapaddr will be MAP_FAILED.
In this case, pci_map_addr should not be incremented in
pci_uio_map_resource().
- To shrink code, move close().
- Remove fail variable.
Signed-off-by: Tetsuya Mukawa
---
lib/librte_eal/bsdapp/eal/eal_pci.c | 19 ++---
lib/librte_eal/linuxapp/eal/eal_pci_uio.c | 64 ---
2 files changed, 55 insertions(+), 28 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c
b/lib/librte_eal/bsdapp/eal/eal_pci.c
index a63d450..63758c7 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -202,7 +202,7 @@ pci_uio_map_resource(struct rte_pci_device *dev)
uint64_t offset;
uint64_t pagesz;
struct rte_pci_addr *loc = &dev->addr;
- struct uio_resource *uio_res;
+ struct uio_resource *uio_res = NULL;
struct uio_res_list *uio_res_list =
RTE_TAILQ_CAST(rte_uio_tailq.head, uio_res_list);
struct uio_map *maps;
@@ -228,7 +228,7 @@ pci_uio_map_resource(struct rte_pci_device *dev)
if (dev->intr_handle.fd < 0) {
RTE_LOG(ERR, EAL, "Cannot open %s: %s\n",
devname, strerror(errno));
- return -1;
+ goto error;
}
dev->intr_handle.type = RTE_INTR_HANDLE_UIO;
@@ -236,7 +236,7 @@ pci_uio_map_resource(struct rte_pci_device *dev)
if ((uio_res = rte_zmalloc("UIO_RES", sizeof (*uio_res), 0)) == NULL) {
RTE_LOG(ERR, EAL,
"%s(): cannot store uio mmap details\n", __func__);
- return -1;
+ goto error;
}
snprintf(uio_res->path, sizeof(uio_res->path), "%s", devname);
@@ -263,8 +263,7 @@ pci_uio_map_resource(struct rte_pci_device *dev)
(mapaddr = pci_map_resource(NULL, devname, (off_t)offset,
(size_t)maps[j].size)
) == NULL) {
- rte_free(uio_res);
- return -1;
+ goto error;
}
maps[j].addr = mapaddr;
@@ -275,6 +274,16 @@ pci_uio_map_resource(struct rte_pci_device *dev)
TAILQ_INSERT_TAIL(uio_res_list, uio_res, next);
return 0;
+
+error:
+ if (uio_res)
+ rte_free(uio_res);
+ if (dev->intr_handle.fd >= 0) {
+ close(dev->intr_handle.fd);
+ dev->intr_handle.fd = -1;
+ dev->intr_handle.type = RTE_INTR_HANDLE_UNKNOWN;
+ }
+ return -1;
}
/* Scan one pci sysfs entry, and fill the devices list from it. */
diff --git a/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
b/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
index 37dc936..4e50533 100644
--- a/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
+++ b/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
@@ -265,7 +265,7 @@ pci_get_uio_dev(struct rte_pci_device *dev, char *dstbuf,
int
pci_uio_map_resource(struct rte_pci_device *dev)
{
- int i, map_idx;
+ int i, map_idx = 0;
char dirname[PATH_MAX];
char cfgname[PATH_MAX];
char devname[PATH_MAX]; /* contains the /dev/uioX */
@@ -273,7 +273,7 @@ pci_uio_map_resource(struct rte_pci_device *dev)
int uio_num;
uint64_t phaddr;
struct rte_pci_addr *loc = &dev->addr;
- struct mapped_pci_resource *uio_res;
+ struct mapped_pci_resource *uio_res = NULL;
struct mapped_pci_res_list *uio_res_list =
RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
struct pci_map *maps;
@@ -300,7 +300,7 @@ pci_uio_map_resource(struct rte_pci_device *dev)
if (dev->intr_handle.fd < 0) {
RTE_LOG(ERR, EAL, "Cannot open %s: %s\n",
devname, strerror(errno));
- return -1;
+ goto error;
}
snprintf(cfgname, sizeof(cfgname),
@@ -309,7 +309,7 @@ pci_uio_map_resource(struct rte_pci_device *dev)
if (dev->intr_handle.uio_cfg_fd < 0) {
RTE_LOG(ERR, EAL, "Cannot open %s: %s\n",
cfgname, strerror(errno));
- return -1;
+ goto error;
}
if (dev->kdrv == RTE_KDRV_IGB_UIO)
@@ -320,7 +320,7 @@ pci_uio_map_resource(struct rte_pci_device *dev)
/* set bus master that is not done by uio_pci_generic */
if (pci_uio_set_bus_master(dev->intr_handle.uio_cfg_fd)) {
RTE_LOG(ERR, EAL, "Cannot set up bus mastering!\n
[dpdk-dev] [PATCH v8 04/12] eal/bsdapp: Change names of pci related data structure
From: "Tetsuya.Mukawa"
To merge pci code of linuxapp and bsdapp, this patch changes names
like below.
- uio_map to pci_map
- uio_resource to mapped_pci_resource
- uio_res_list to mapped_pci_res_list
Signed-off-by: Tetsuya Mukawa
---
lib/librte_eal/bsdapp/eal/eal_pci.c | 24
1 file changed, 12 insertions(+), 12 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c
b/lib/librte_eal/bsdapp/eal/eal_pci.c
index 63758c7..21d1e66 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -84,7 +84,7 @@
* enabling bus master.
*/
-struct uio_map {
+struct pci_map {
void *addr;
uint64_t offset;
uint64_t size;
@@ -95,16 +95,16 @@ struct uio_map {
* For multi-process we need to reproduce all PCI mappings in secondary
* processes, so save them in a tailq.
*/
-struct uio_resource {
- TAILQ_ENTRY(uio_resource) next;
+struct mapped_pci_resource {
+ TAILQ_ENTRY(mapped_pci_resource) next;
struct rte_pci_addr pci_addr;
char path[PATH_MAX];
size_t nb_maps;
- struct uio_map maps[PCI_MAX_RESOURCE];
+ struct pci_map maps[PCI_MAX_RESOURCE];
};
-TAILQ_HEAD(uio_res_list, uio_resource);
+TAILQ_HEAD(mapped_pci_res_list, mapped_pci_resource);
static struct rte_tailq_elem rte_uio_tailq = {
.name = "UIO_RESOURCE_LIST",
@@ -163,9 +163,9 @@ static int
pci_uio_map_secondary(struct rte_pci_device *dev)
{
size_t i;
- struct uio_resource *uio_res;
- struct uio_res_list *uio_res_list =
- RTE_TAILQ_CAST(rte_uio_tailq.head, uio_res_list);
+ struct mapped_pci_resource *uio_res;
+ struct mapped_pci_res_list *uio_res_list =
+ RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
TAILQ_FOREACH(uio_res, uio_res_list, next) {
@@ -202,10 +202,10 @@ pci_uio_map_resource(struct rte_pci_device *dev)
uint64_t offset;
uint64_t pagesz;
struct rte_pci_addr *loc = &dev->addr;
- struct uio_resource *uio_res = NULL;
- struct uio_res_list *uio_res_list =
- RTE_TAILQ_CAST(rte_uio_tailq.head, uio_res_list);
- struct uio_map *maps;
+ struct mapped_pci_resource *uio_res = NULL;
+ struct mapped_pci_res_list *uio_res_list =
+ RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
+ struct pci_map *maps;
dev->intr_handle.fd = -1;
dev->intr_handle.type = RTE_INTR_HANDLE_UNKNOWN;
--
2.1.4
[dpdk-dev] [PATCH v8 05/12] eal: Fix uio mapping differences between linuxapp and bsdapp
From: "Tetsuya.Mukawa"
This patch fixes below.
- bsdapp
- Use map_id in pci_uio_map_resource().
- Fix interface of pci_map_resource().
- Move path variable of mapped_pci_resource structure to pci_map.
- linuxapp
- Remove redundant error message of linuxapp.
'pci_uio_map_resource()' is implemented in both linuxapp and bsdapp,
but interface is different. The patch fixes the function of bsdapp
to do same as linuxapp. After applying it, file descriptor should be
opened and closed out of pci_map_resource().
Signed-off-by: Tetsuya Mukawa
---
lib/librte_eal/bsdapp/eal/eal_pci.c | 118 ++
lib/librte_eal/linuxapp/eal/eal_pci_uio.c | 20 ++---
2 files changed, 78 insertions(+), 60 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c
b/lib/librte_eal/bsdapp/eal/eal_pci.c
index 21d1e66..92d9886 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -86,6 +86,7 @@
struct pci_map {
void *addr;
+ char *path;
uint64_t offset;
uint64_t size;
uint64_t phaddr;
@@ -100,7 +101,7 @@ struct mapped_pci_resource {
struct rte_pci_addr pci_addr;
char path[PATH_MAX];
- size_t nb_maps;
+ int nb_maps;
struct pci_map maps[PCI_MAX_RESOURCE];
};
@@ -122,47 +123,30 @@ pci_unbind_kernel_driver(struct rte_pci_device *dev
__rte_unused)
/* map a particular resource from a file */
static void *
-pci_map_resource(void *requested_addr, const char *devname, off_t offset,
-size_t size)
+pci_map_resource(void *requested_addr, int fd, off_t offset, size_t size,
+int additional_flags)
{
- int fd;
void *mapaddr;
- /*
-* open devname, to mmap it
-*/
- fd = open(devname, O_RDWR);
- if (fd < 0) {
- RTE_LOG(ERR, EAL, "Cannot open %s: %s\n",
- devname, strerror(errno));
- goto fail;
- }
-
/* Map the PCI memory resource of device */
mapaddr = mmap(requested_addr, size, PROT_READ | PROT_WRITE,
- MAP_SHARED, fd, offset);
- close(fd);
- if (mapaddr == MAP_FAILED ||
- (requested_addr != NULL && mapaddr != requested_addr)) {
- RTE_LOG(ERR, EAL, "%s(): cannot mmap(%s(%d), %p, 0x%lx, 0x%lx):"
- " %s (%p)\n", __func__, devname, fd, requested_addr,
+ MAP_SHARED | additional_flags, fd, offset);
+ if (mapaddr == MAP_FAILED) {
+ RTE_LOG(ERR, EAL,
+ "%s(): cannot mmap(%d, %p, 0x%lx, 0x%lx): %s (%p)\n",
+ __func__, fd, requested_addr,
(unsigned long)size, (unsigned long)offset,
strerror(errno), mapaddr);
- goto fail;
- }
-
- RTE_LOG(DEBUG, EAL, " PCI memory mapped at %p\n", mapaddr);
+ } else
+ RTE_LOG(DEBUG, EAL, " PCI memory mapped at %p\n", mapaddr);
return mapaddr;
-
-fail:
- return NULL;
}
static int
pci_uio_map_secondary(struct rte_pci_device *dev)
{
- size_t i;
+ int i, fd;
struct mapped_pci_resource *uio_res;
struct mapped_pci_res_list *uio_res_list =
RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
@@ -170,19 +154,33 @@ pci_uio_map_secondary(struct rte_pci_device *dev)
TAILQ_FOREACH(uio_res, uio_res_list, next) {
/* skip this element if it doesn't match our PCI address */
- if (memcmp(&uio_res->pci_addr, &dev->addr, sizeof(dev->addr)))
+ if (rte_eal_compare_pci_addr(&uio_res->pci_addr, &dev->addr))
continue;
for (i = 0; i != uio_res->nb_maps; i++) {
- if (pci_map_resource(uio_res->maps[i].addr,
-uio_res->path,
-(off_t)uio_res->maps[i].offset,
-(size_t)uio_res->maps[i].size)
- != uio_res->maps[i].addr) {
+ /*
+* open devname, to mmap it
+*/
+ fd = open(uio_res->maps[i].path, O_RDWR);
+ if (fd < 0) {
+ RTE_LOG(ERR, EAL, "Cannot open %s: %s\n",
+ uio_res->maps[i].path, strerror(errno));
+ return -1;
+ }
+
+ void *mapaddr = pci_map_resource(uio_res->maps[i].addr,
+ fd, (off_t)uio_res->maps[i].offset,
+ (size_t)uio_res->maps[i].size, 0);
+ if (mapaddr != uio_res->maps[i].addr) {
RTE_LOG(ERR, EAL,
- "Cannot mmap
[dpdk-dev] [PATCH v8 06/12] eal: Add pci_uio_alloc_resource()
From: "Tetsuya.Mukawa"
This patch adds a new function called pci_uio_alloc_resource().
The function hides how to prepare uio resource in linuxapp and bsdapp.
With the function, pci_uio_map_resource() will be more abstracted.
Signed-off-by: Tetsuya Mukawa
---
lib/librte_eal/bsdapp/eal/eal_pci.c | 84 +-
lib/librte_eal/linuxapp/eal/eal_pci_uio.c | 98 +--
2 files changed, 123 insertions(+), 59 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c
b/lib/librte_eal/bsdapp/eal/eal_pci.c
index 92d9886..ce0ca07 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -189,28 +189,17 @@ pci_uio_map_secondary(struct rte_pci_device *dev)
return 1;
}
-/* map the PCI resource of a PCI device in virtual memory */
static int
-pci_uio_map_resource(struct rte_pci_device *dev)
+pci_uio_alloc_resource(struct rte_pci_device *dev,
+ struct mapped_pci_resource **uio_res)
{
- int i, map_idx = 0;
char devname[PATH_MAX]; /* contains the /dev/uioX */
- void *mapaddr;
- uint64_t phaddr;
- uint64_t offset;
- uint64_t pagesz;
- struct rte_pci_addr *loc = &dev->addr;
- struct mapped_pci_resource *uio_res = NULL;
- struct mapped_pci_res_list *uio_res_list =
- RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
- struct pci_map *maps;
+ struct rte_pci_addr *loc;
- dev->intr_handle.fd = -1;
- dev->intr_handle.type = RTE_INTR_HANDLE_UNKNOWN;
+ if ((dev == NULL) || (uio_res == NULL))
+ return -1;
- /* secondary processes - use already recorded details */
- if (rte_eal_process_type() != RTE_PROC_PRIMARY)
- return pci_uio_map_secondary(dev);
+ loc = &dev->addr;
snprintf(devname, sizeof(devname), "/dev/uio at pci:%u:%u:%u",
dev->addr.bus, dev->addr.devid, dev->addr.function);
@@ -231,18 +220,57 @@ pci_uio_map_resource(struct rte_pci_device *dev)
dev->intr_handle.type = RTE_INTR_HANDLE_UIO;
/* allocate the mapping details for secondary processes*/
- if ((uio_res = rte_zmalloc("UIO_RES", sizeof (*uio_res), 0)) == NULL) {
+ *uio_res = rte_zmalloc("UIO_RES", sizeof(**uio_res), 0);
+ if (*uio_res == NULL) {
RTE_LOG(ERR, EAL,
"%s(): cannot store uio mmap details\n", __func__);
goto error;
}
- snprintf(uio_res->path, sizeof(uio_res->path), "%s", devname);
- memcpy(&uio_res->pci_addr, &dev->addr, sizeof(uio_res->pci_addr));
+ snprintf((*uio_res)->path, sizeof((*uio_res)->path), "%s", devname);
+ memcpy(&(*uio_res)->pci_addr, &dev->addr, sizeof((*uio_res)->pci_addr));
+ return 0;
+
+error:
+ if (dev->intr_handle.fd) {
+ close(dev->intr_handle.fd);
+ dev->intr_handle.fd = -1;
+ dev->intr_handle.type = RTE_INTR_HANDLE_UNKNOWN;
+ }
+ return -1;
+}
+
+/* map the PCI resource of a PCI device in virtual memory */
+static int
+pci_uio_map_resource(struct rte_pci_device *dev)
+{
+ int i, map_idx = 0, ret;
+ char *devname;
+ void *mapaddr;
+ uint64_t phaddr;
+ uint64_t offset;
+ uint64_t pagesz;
+ struct mapped_pci_resource *uio_res = NULL;
+ struct mapped_pci_res_list *uio_res_list =
+ RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
+ struct pci_map *maps;
+
+ dev->intr_handle.fd = -1;
+ dev->intr_handle.type = RTE_INTR_HANDLE_UNKNOWN;
+
+ /* secondary processes - use already recorded details */
+ if (rte_eal_process_type() != RTE_PROC_PRIMARY)
+ return pci_uio_map_secondary(dev);
+
+ /* allocate uio resource */
+ ret = pci_uio_alloc_resource(dev, &uio_res);
+ if ((ret != 0) || (uio_res == NULL))
+ return ret;
/* Map all BARs */
pagesz = sysconf(_SC_PAGESIZE);
+ devname = uio_res->path;
maps = uio_res->maps;
for (i = 0; i != PCI_MAX_RESOURCE; i++) {
@@ -298,13 +326,15 @@ pci_uio_map_resource(struct rte_pci_device *dev)
error:
for (i = 0; i < map_idx; i++)
rte_free(maps[i].path);
- if (uio_res)
- rte_free(uio_res);
- if (dev->intr_handle.fd >= 0) {
- close(dev->intr_handle.fd);
- dev->intr_handle.fd = -1;
- dev->intr_handle.type = RTE_INTR_HANDLE_UNKNOWN;
- }
+
+ /* 'uio_res' has valid value here */
+ rte_free(uio_res);
+
+ /* 'fd' has valid value here */
+ close(dev->intr_handle.fd);
+ dev->intr_handle.fd = -1;
+ dev->intr_handle.type = RTE_INTR_HANDLE_UNKNOWN;
+
return -1;
}
diff --git a/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
b/lib/librte_eal/linuxapp/eal/eal_pci_uio.c
index 38a93c9..f408bd3 100644
--- a/lib/librte_ea
[dpdk-dev] [PATCH v8 07/12] eal: Add pci_uio_map_resource_by_index()
From: "Tetsuya.Mukawa"
This patch adds a new function called pci_uio_map_resource_by_index().
The function hides how to map uio resource in linuxapp and bsdapp.
With the function, pci_uio_map_resource() will be more abstracted.
Signed-off-by: Tetsuya Mukawa
---
lib/librte_eal/bsdapp/eal/eal_pci.c | 107 +++---
lib/librte_eal/linuxapp/eal/eal_pci_uio.c | 124 +-
2 files changed, 133 insertions(+), 98 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c
b/lib/librte_eal/bsdapp/eal/eal_pci.c
index ce0ca07..c76f936 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -241,20 +241,73 @@ error:
return -1;
}
+static int
+pci_uio_map_resource_by_index(struct rte_pci_device *dev, int res_idx,
+ struct mapped_pci_resource *uio_res, int map_idx)
+{
+ int fd;
+ char *devname;
+ void *mapaddr;
+ uint64_t offset;
+ uint64_t pagesz;
+ struct pci_map *maps;
+
+ if ((dev == NULL) || (uio_res == NULL) || (uio_res->path == NULL))
+ return -1;
+
+ maps = uio_res->maps;
+ devname = uio_res->path;
+ pagesz = sysconf(_SC_PAGESIZE);
+
+ /* allocate memory to keep path */
+ maps[map_idx].path = rte_malloc(NULL, strlen(devname) + 1, 0);
+ if (maps[map_idx].path == NULL) {
+ RTE_LOG(ERR, EAL, "Cannot allocate memory for path: %s\n",
+ strerror(errno));
+ return -1;
+ }
+
+ /*
+* open resource file, to mmap it
+*/
+ fd = open(devname, O_RDWR);
+ if (fd < 0) {
+ RTE_LOG(ERR, EAL, "Cannot open %s: %s\n",
+ devname, strerror(errno));
+ goto error;
+ }
+
+ /* if matching map is found, then use it */
+ offset = res_idx * pagesz;
+ mapaddr = pci_map_resource(NULL, fd, (off_t)offset,
+ (size_t)dev->mem_resource[res_idx].len, 0);
+ close(fd);
+ if (mapaddr == MAP_FAILED)
+ goto error;
+
+ maps[map_idx].phaddr = dev->mem_resource[res_idx].phys_addr;
+ maps[map_idx].size = dev->mem_resource[res_idx].len;
+ maps[map_idx].addr = mapaddr;
+ maps[map_idx].offset = offset;
+ strcpy(maps[map_idx].path, devname);
+ dev->mem_resource[res_idx].addr = mapaddr;
+
+ return 0;
+
+error:
+ rte_free(maps[map_idx].path);
+ return -1;
+}
+
/* map the PCI resource of a PCI device in virtual memory */
static int
pci_uio_map_resource(struct rte_pci_device *dev)
{
int i, map_idx = 0, ret;
- char *devname;
- void *mapaddr;
uint64_t phaddr;
- uint64_t offset;
- uint64_t pagesz;
struct mapped_pci_resource *uio_res = NULL;
struct mapped_pci_res_list *uio_res_list =
RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
- struct pci_map *maps;
dev->intr_handle.fd = -1;
dev->intr_handle.type = RTE_INTR_HANDLE_UNKNOWN;
@@ -269,53 +322,17 @@ pci_uio_map_resource(struct rte_pci_device *dev)
return ret;
/* Map all BARs */
- pagesz = sysconf(_SC_PAGESIZE);
- devname = uio_res->path;
-
- maps = uio_res->maps;
for (i = 0; i != PCI_MAX_RESOURCE; i++) {
- int fd;
-
/* skip empty BAR */
if ((phaddr = dev->mem_resource[i].phys_addr) == 0)
continue;
- /* allocate memory to keep path */
- maps[map_idx].path = rte_malloc(NULL, strlen(devname) + 1, 0);
- if (maps[map_idx].path == NULL) {
- RTE_LOG(ERR, EAL, "Cannot allocate memory for path:
%s\n",
- strerror(errno));
+ ret = pci_uio_map_resource_by_index(dev, i,
+ uio_res, map_idx);
+ if (ret != 0)
goto error;
- }
- /*
-* open resource file, to mmap it
-*/
- fd = open(devname, O_RDWR);
- if (fd < 0) {
- RTE_LOG(ERR, EAL, "Cannot open %s: %s\n",
- devname, strerror(errno));
- rte_free(maps[map_idx].path);
- goto error;
- }
-
- /* if matching map is found, then use it */
- offset = i * pagesz;
- mapaddr = pci_map_resource(NULL, fd, (off_t)offset,
- (size_t)dev->mem_resource[i].len, 0);
- close(fd);
- if (mapaddr == MAP_FAILED) {
- rte_free(maps[map_idx].path);
- goto error;
- }
-
- maps[map_idx].phaddr = dev->mem_resource[i].phys_addr;
- maps[m
[dpdk-dev] [PATCH v8 08/12] eal: Consolidate pci_map and mapped_pci_resource of linuxapp and bsdapp
From: "Tetsuya.Mukawa"
This patch consolidates below structures, and defines them in common code.
- struct pci_map
- strucy mapped_pci_resouces
Signed-off-by: Tetsuya Mukawa
---
lib/librte_eal/bsdapp/eal/eal_pci.c| 24
lib/librte_eal/common/include/rte_pci.h| 29 +
lib/librte_eal/linuxapp/eal/eal_pci_init.h | 23 ---
3 files changed, 29 insertions(+), 47 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c
b/lib/librte_eal/bsdapp/eal/eal_pci.c
index c76f936..6067e31 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -83,30 +83,6 @@
* network card, only providing access to PCI BAR to applications, and
* enabling bus master.
*/
-
-struct pci_map {
- void *addr;
- char *path;
- uint64_t offset;
- uint64_t size;
- uint64_t phaddr;
-};
-
-/*
- * For multi-process we need to reproduce all PCI mappings in secondary
- * processes, so save them in a tailq.
- */
-struct mapped_pci_resource {
- TAILQ_ENTRY(mapped_pci_resource) next;
-
- struct rte_pci_addr pci_addr;
- char path[PATH_MAX];
- int nb_maps;
- struct pci_map maps[PCI_MAX_RESOURCE];
-};
-
-TAILQ_HEAD(mapped_pci_res_list, mapped_pci_resource);
-
static struct rte_tailq_elem rte_uio_tailq = {
.name = "UIO_RESOURCE_LIST",
};
diff --git a/lib/librte_eal/common/include/rte_pci.h
b/lib/librte_eal/common/include/rte_pci.h
index 7801fa0..d95747e 100644
--- a/lib/librte_eal/common/include/rte_pci.h
+++ b/lib/librte_eal/common/include/rte_pci.h
@@ -2,6 +2,7 @@
* BSD LICENSE
*
* Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ * Copyright(c) 2015 IGEL Co., Ltd.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -220,6 +221,34 @@ struct rte_pci_driver {
/** Device driver supports detaching capability */
#define RTE_PCI_DRV_DETACHABLE 0x0010
+/**
+ * A structure describing a PCI mapping.
+ */
+struct pci_map {
+ void *addr;
+ char *path;
+ uint64_t offset;
+ uint64_t size;
+ uint64_t phaddr;
+};
+
+/**
+ * A structure describing a mapped PCI resource.
+ * For multi-process we need to reproduce all PCI mappings in secondary
+ * processes, so save them in a tailq.
+ */
+struct mapped_pci_resource {
+ TAILQ_ENTRY(mapped_pci_resource) next;
+
+ struct rte_pci_addr pci_addr;
+ char path[PATH_MAX];
+ int nb_maps;
+ struct pci_map maps[PCI_MAX_RESOURCE];
+};
+
+/** mapped pci device list */
+TAILQ_HEAD(mapped_pci_res_list, mapped_pci_resource);
+
/**< Internal use only - Macro used by pci addr parsing functions **/
#define GET_PCIADDR_FIELD(in, fd, lim, dlm) \
do { \
diff --git a/lib/librte_eal/linuxapp/eal/eal_pci_init.h
b/lib/librte_eal/linuxapp/eal/eal_pci_init.h
index aa7b755..d9d1878 100644
--- a/lib/librte_eal/linuxapp/eal/eal_pci_init.h
+++ b/lib/librte_eal/linuxapp/eal/eal_pci_init.h
@@ -36,29 +36,6 @@
#include "eal_vfio.h"
-struct pci_map {
- void *addr;
- char *path;
- uint64_t offset;
- uint64_t size;
- uint64_t phaddr;
-};
-
-/*
- * For multi-process we need to reproduce all PCI mappings in secondary
- * processes, so save them in a tailq.
- */
-struct mapped_pci_resource {
- TAILQ_ENTRY(mapped_pci_resource) next;
-
- struct rte_pci_addr pci_addr;
- char path[PATH_MAX];
- int nb_maps;
- struct pci_map maps[PCI_MAX_RESOURCE];
-};
-
-TAILQ_HEAD(mapped_pci_res_list, mapped_pci_resource);
-
/*
* Helper function to map PCI resources right after hugepages in virtual memory
*/
--
2.1.4
[dpdk-dev] [PATCH v8 09/12] eal: Consolidate pci_map/unmap_resource() of linuxapp and bsdapp
From: "Tetsuya.Mukawa"
The patch consolidates below functions, and implemented in common
eal code.
- pci_map_resource()
- pci_unmap_resource()
Signed-off-by: Tetsuya Mukawa
---
lib/librte_eal/bsdapp/eal/eal_pci.c| 22
lib/librte_eal/common/eal_common_pci.c | 40 +
lib/librte_eal/common/include/rte_pci.h| 32 +++
lib/librte_eal/linuxapp/eal/eal_pci.c | 41 --
lib/librte_eal/linuxapp/eal/eal_pci_init.h | 5
5 files changed, 72 insertions(+), 68 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c
b/lib/librte_eal/bsdapp/eal/eal_pci.c
index 6067e31..5bb93cd 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -97,28 +97,6 @@ pci_unbind_kernel_driver(struct rte_pci_device *dev
__rte_unused)
return -ENOTSUP;
}
-/* map a particular resource from a file */
-static void *
-pci_map_resource(void *requested_addr, int fd, off_t offset, size_t size,
-int additional_flags)
-{
- void *mapaddr;
-
- /* Map the PCI memory resource of device */
- mapaddr = mmap(requested_addr, size, PROT_READ | PROT_WRITE,
- MAP_SHARED | additional_flags, fd, offset);
- if (mapaddr == MAP_FAILED) {
- RTE_LOG(ERR, EAL,
- "%s(): cannot mmap(%d, %p, 0x%lx, 0x%lx): %s (%p)\n",
- __func__, fd, requested_addr,
- (unsigned long)size, (unsigned long)offset,
- strerror(errno), mapaddr);
- } else
- RTE_LOG(DEBUG, EAL, " PCI memory mapped at %p\n", mapaddr);
-
- return mapaddr;
-}
-
static int
pci_uio_map_secondary(struct rte_pci_device *dev)
{
diff --git a/lib/librte_eal/common/eal_common_pci.c
b/lib/librte_eal/common/eal_common_pci.c
index 4229aaf..eab147a 100644
--- a/lib/librte_eal/common/eal_common_pci.c
+++ b/lib/librte_eal/common/eal_common_pci.c
@@ -2,6 +2,7 @@
* BSD LICENSE
*
* Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ * Copyright(c) 2015 IGEL Co., Ltd.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -67,6 +68,7 @@
#include
#include
#include
+#include
#include
#include
@@ -98,6 +100,44 @@ static struct rte_devargs *pci_devargs_lookup(struct
rte_pci_device *dev)
return NULL;
}
+/* map a particular resource from a file */
+void *
+pci_map_resource(void *requested_addr, int fd, off_t offset, size_t size,
+int additional_flags)
+{
+ void *mapaddr;
+
+ /* Map the PCI memory resource of device */
+ mapaddr = mmap(requested_addr, size, PROT_READ | PROT_WRITE,
+ MAP_SHARED | additional_flags, fd, offset);
+ if (mapaddr == MAP_FAILED) {
+ RTE_LOG(ERR, EAL, "%s(): cannot mmap(%d, %p, 0x%lx, 0x%lx): %s
(%p)\n",
+ __func__, fd, requested_addr,
+ (unsigned long)size, (unsigned long)offset,
+ strerror(errno), mapaddr);
+ } else
+ RTE_LOG(DEBUG, EAL, " PCI memory mapped at %p\n", mapaddr);
+
+ return mapaddr;
+}
+
+/* unmap a particular resource */
+void
+pci_unmap_resource(void *requested_addr, size_t size)
+{
+ if (requested_addr == NULL)
+ return;
+
+ /* Unmap the PCI memory resource of device */
+ if (munmap(requested_addr, size)) {
+ RTE_LOG(ERR, EAL, "%s(): cannot munmap(%p, 0x%lx): %s\n",
+ __func__, requested_addr, (unsigned long)size,
+ strerror(errno));
+ } else
+ RTE_LOG(DEBUG, EAL, " PCI memory unmapped at %p\n",
+ requested_addr);
+}
+
/*
* If vendor/device ID match, call the devinit() function of all
* registered driver for the given device. Return -1 if initialization
diff --git a/lib/librte_eal/common/include/rte_pci.h
b/lib/librte_eal/common/include/rte_pci.h
index d95747e..4a4641c 100644
--- a/lib/librte_eal/common/include/rte_pci.h
+++ b/lib/librte_eal/common/include/rte_pci.h
@@ -365,6 +365,38 @@ int rte_eal_pci_scan(void);
*/
int rte_eal_pci_probe(void);
+/**
+ * @internal
+ * Map a particular resource from a file.
+ *
+ * @param requested_addr
+ * The starting address for the new mapping range.
+ * @param fd
+ * The file descriptor.
+ * @param offset
+ * The offset for the mapping range.
+ * @param size
+ * The size for the mapping range.
+ * @param additional_flags
+ * The additional flags for the mapping range.
+ * @return
+ * - On success, the function returns a pointer to the mapped area.
+ * - On error, the value MAP_FAILED is returned.
+ */
+void *pci_map_resource(void *requested_addr, int fd, off_t offset,
+ size_t size, int additional_flags);
+
+/**
+ * @internal
+ * Unmap a par
[dpdk-dev] [PATCH v8 10/12] eal: Consolidate pci uio functions of linuxapp and bsdapp
From: "Tetsuya.Mukawa"
The patch consolidates below functions, and implement these
in eal_common_pci_uio.c.
- pci_uio_map_secondary()
- pci_uio_map_resource()
- pci_uio_unmap()
- pci_uio_find_resource()
- pci_uio_unmap_resource()
Signed-off-by: Tetsuya Mukawa
---
lib/librte_eal/bsdapp/eal/Makefile | 2 +
lib/librte_eal/bsdapp/eal/eal_pci.c| 110 +-
.../bsdapp/eal/include/exec-env/rte_interrupts.h | 2 +
lib/librte_eal/common/eal_common_pci_uio.c | 240 +
lib/librte_eal/common/eal_private.h| 56 +
lib/librte_eal/linuxapp/eal/Makefile | 2 +
lib/librte_eal/linuxapp/eal/eal_pci_init.h | 12 +-
lib/librte_eal/linuxapp/eal/eal_pci_uio.c | 196 +
8 files changed, 311 insertions(+), 309 deletions(-)
create mode 100644 lib/librte_eal/common/eal_common_pci_uio.c
diff --git a/lib/librte_eal/bsdapp/eal/Makefile
b/lib/librte_eal/bsdapp/eal/Makefile
index c73ffb6..0d72330 100644
--- a/lib/librte_eal/bsdapp/eal/Makefile
+++ b/lib/librte_eal/bsdapp/eal/Makefile
@@ -1,6 +1,7 @@
# BSD LICENSE
#
# Copyright(c) 2010-2015 Intel Corporation. All rights reserved.
+# Copyright(c) 2015 IGEL Co., Ltd.
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
@@ -68,6 +69,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_memzone.c
SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_log.c
SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_launch.c
SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_pci.c
+SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_pci_uio.c
SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_memory.c
SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_tailqs.c
SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_errno.c
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c
b/lib/librte_eal/bsdapp/eal/eal_pci.c
index 5bb93cd..f85fc1e 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -83,10 +83,6 @@
* network card, only providing access to PCI BAR to applications, and
* enabling bus master.
*/
-static struct rte_tailq_elem rte_uio_tailq = {
- .name = "UIO_RESOURCE_LIST",
-};
-EAL_REGISTER_TAILQ(rte_uio_tailq)
/* unbind kernel driver for this device */
static int
@@ -97,53 +93,7 @@ pci_unbind_kernel_driver(struct rte_pci_device *dev
__rte_unused)
return -ENOTSUP;
}
-static int
-pci_uio_map_secondary(struct rte_pci_device *dev)
-{
- int i, fd;
- struct mapped_pci_resource *uio_res;
- struct mapped_pci_res_list *uio_res_list =
- RTE_TAILQ_CAST(rte_uio_tailq.head, mapped_pci_res_list);
-
- TAILQ_FOREACH(uio_res, uio_res_list, next) {
-
- /* skip this element if it doesn't match our PCI address */
- if (rte_eal_compare_pci_addr(&uio_res->pci_addr, &dev->addr))
- continue;
-
- for (i = 0; i != uio_res->nb_maps; i++) {
- /*
-* open devname, to mmap it
-*/
- fd = open(uio_res->maps[i].path, O_RDWR);
- if (fd < 0) {
- RTE_LOG(ERR, EAL, "Cannot open %s: %s\n",
- uio_res->maps[i].path, strerror(errno));
- return -1;
- }
-
- void *mapaddr = pci_map_resource(uio_res->maps[i].addr,
- fd, (off_t)uio_res->maps[i].offset,
- (size_t)uio_res->maps[i].size, 0);
- if (mapaddr != uio_res->maps[i].addr) {
- RTE_LOG(ERR, EAL,
- "Cannot mmap device resource file %s to
address: %p\n",
- uio_res->maps[i].path,
- uio_res->maps[i].addr);
- close(fd);
- return -1;
- }
- /* fd is not needed in slave process, close it */
- close(fd);
- }
- return 0;
- }
-
- RTE_LOG(ERR, EAL, "Cannot find resource for device\n");
- return 1;
-}
-
-static int
+int
pci_uio_alloc_resource(struct rte_pci_device *dev,
struct mapped_pci_resource **uio_res)
{
@@ -195,7 +145,7 @@ error:
return -1;
}
-static int
+int
pci_uio_map_resource_by_index(struct rte_pci_device *dev, int res_idx,
struct mapped_pci_resource *uio_res, int map_idx)
{
@@ -253,62 +203,6 @@ error:
return -1;
}
-/* map the PCI resource of a PCI device in virtual memory */
-static int
-pci_uio_map_resource(struct rte_pci_device *dev)
-{
- int i, map_idx = 0, ret;
- uint64_t phaddr;
- str
[dpdk-dev] [PATCH v8 11/12] eal: Consolidate pci_map/unmap_device() of linuxapp and bsdapp
From: "Tetsuya.Mukawa"
The patch consolidates below functions, and implemented in common
eal code.
- pci_map_device()
- pci_unmap_device()
Signed-off-by: Tetsuya Mukawa
---
lib/librte_eal/bsdapp/eal/eal_pci.c | 3 ++
lib/librte_eal/common/eal_common_pci.c | 57 +
lib/librte_eal/common/eal_private.h | 19 +++
lib/librte_eal/common/include/rte_pci.h | 1 +
lib/librte_eal/linuxapp/eal/eal_pci.c | 53 --
lib/librte_ether/rte_ethdev.c | 2 ++
6 files changed, 82 insertions(+), 53 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c
b/lib/librte_eal/bsdapp/eal/eal_pci.c
index f85fc1e..a27dca3 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -240,6 +240,9 @@ pci_scan_one(int dev_pci_fd, struct pci_conf *conf)
/* FreeBSD has no NUMA support (yet) */
dev->numa_node = 0;
+ /* FreeBSD has only one pass through driver */
+ dev->kdrv = RTE_KDRV_NIC_UIO;
+
/* parse resources */
switch (conf->pc_hdr & PCIM_HDRTYPE) {
case PCIM_HDRTYPE_NORMAL:
diff --git a/lib/librte_eal/common/eal_common_pci.c
b/lib/librte_eal/common/eal_common_pci.c
index eab147a..afd14ec 100644
--- a/lib/librte_eal/common/eal_common_pci.c
+++ b/lib/librte_eal/common/eal_common_pci.c
@@ -138,6 +138,63 @@ pci_unmap_resource(void *requested_addr, size_t size)
requested_addr);
}
+/* Map pci device */
+int
+pci_map_device(struct rte_pci_device *dev)
+{
+ int ret = -1;
+
+ /* try mapping the NIC resources using VFIO if it exists */
+ switch (dev->kdrv) {
+ case RTE_KDRV_VFIO:
+#ifdef VFIO_PRESENT
+ if (pci_vfio_is_enabled())
+ ret = pci_vfio_map_resource(dev);
+#endif
+ break;
+ case RTE_KDRV_IGB_UIO:
+ case RTE_KDRV_UIO_GENERIC:
+ case RTE_KDRV_NIC_UIO:
+ /* map resources for devices that use uio */
+ ret = pci_uio_map_resource(dev);
+ break;
+ default:
+ RTE_LOG(DEBUG, EAL, " Not managed by a supported kernel
driver,"
+ " skipped\n");
+ ret = 1;
+ break;
+ }
+
+ return ret;
+}
+
+#ifdef RTE_LIBRTE_EAL_HOTPLUG
+/* Unmap pci device */
+void
+pci_unmap_device(struct rte_pci_device *dev)
+{
+ if (dev == NULL)
+ return;
+
+ /* try unmapping the NIC resources using VFIO if it exists */
+ switch (dev->kdrv) {
+ case RTE_KDRV_VFIO:
+ RTE_LOG(ERR, EAL, "Hotplug doesn't support vfio yet\n");
+ break;
+ case RTE_KDRV_IGB_UIO:
+ case RTE_KDRV_UIO_GENERIC:
+ case RTE_KDRV_NIC_UIO:
+ /* unmap resources for devices that use uio */
+ pci_uio_unmap_resource(dev);
+ break;
+ default:
+ RTE_LOG(DEBUG, EAL, " Not managed by a supported kernel
driver,"
+ " skipped\n");
+ break;
+ }
+}
+#endif /* RTE_LIBRTE_EAL_HOTPLUG */
+
/*
* If vendor/device ID match, call the devinit() function of all
* registered driver for the given device. Return -1 if initialization
diff --git a/lib/librte_eal/common/eal_private.h
b/lib/librte_eal/common/eal_private.h
index 4125589..95b9e30 100644
--- a/lib/librte_eal/common/eal_private.h
+++ b/lib/librte_eal/common/eal_private.h
@@ -156,6 +156,25 @@ struct rte_pci_driver;
struct rte_pci_device;
/**
+ * Map this device
+ *
+ * This function is private to EAL.
+ *
+ * @return
+ * 0 on success, negative on error
+ */
+int pci_map_device(struct rte_pci_device *dev);
+
+#ifdef RTE_LIBRTE_EAL_HOTPLUG
+/**
+ * Unmap this device
+ *
+ * This function is private to EAL.
+ */
+void pci_unmap_device(struct rte_pci_device *dev);
+#endif /* RTE_LIBRTE_EAL_HOTPLUG */
+
+/**
* Map the PCI resource of a PCI device in virtual memory
*
* This function is private to EAL.
diff --git a/lib/librte_eal/common/include/rte_pci.h
b/lib/librte_eal/common/include/rte_pci.h
index 4a4641c..1f09252 100644
--- a/lib/librte_eal/common/include/rte_pci.h
+++ b/lib/librte_eal/common/include/rte_pci.h
@@ -149,6 +149,7 @@ enum rte_kernel_driver {
RTE_KDRV_IGB_UIO,
RTE_KDRV_VFIO,
RTE_KDRV_UIO_GENERIC,
+ RTE_KDRV_NIC_UIO,
};
/**
diff --git a/lib/librte_eal/linuxapp/eal/eal_pci.c
b/lib/librte_eal/linuxapp/eal/eal_pci.c
index fc99eaa..7e8df7d 100644
--- a/lib/librte_eal/linuxapp/eal/eal_pci.c
+++ b/lib/librte_eal/linuxapp/eal/eal_pci.c
@@ -522,59 +522,6 @@ pci_config_space_set(struct rte_pci_device *dev)
}
#endif
-static int
-pci_map_device(struct rte_pci_device *dev)
-{
- int ret = -1;
-
- /* try mapping the NIC resources using VFIO if it exists */
- switch (dev->kdrv) {
- case RTE_KDRV_VFIO:
-#ifdef VFIO_PRESENT
- if (pci_vfio_is_enabled())
-
[dpdk-dev] [PATCH v8 12/12] eal: Consolidate rte_eal_pci_probe/close_one_driver() of linuxapp and bsdapp
From: "Tetsuya.Mukawa"
This patch consolidates below functions, and implements these in common
eal code.
- rte_eal_pci_probe_one_driver()
- rte_eal_pci_close_one_driver()
Signed-off-by: Tetsuya Mukawa
---
lib/librte_eal/bsdapp/eal/eal_pci.c| 67 +--
lib/librte_eal/common/eal_common_pci.c | 133 +-
lib/librte_eal/common/eal_private.h| 39 +
lib/librte_eal/linuxapp/eal/eal_pci.c | 143 +
4 files changed, 136 insertions(+), 246 deletions(-)
diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c
b/lib/librte_eal/bsdapp/eal/eal_pci.c
index a27dca3..b7c34f4 100644
--- a/lib/librte_eal/bsdapp/eal/eal_pci.c
+++ b/lib/librte_eal/bsdapp/eal/eal_pci.c
@@ -85,7 +85,7 @@
*/
/* unbind kernel driver for this device */
-static int
+int
pci_unbind_kernel_driver(struct rte_pci_device *dev __rte_unused)
{
RTE_LOG(ERR, EAL, "RTE_PCI_DRV_FORCE_UNBIND flag is not implemented "
@@ -357,71 +357,6 @@ error:
return -1;
}
-/*
- * If vendor/device ID match, call the devinit() function of the
- * driver.
- */
-int
-rte_eal_pci_probe_one_driver(struct rte_pci_driver *dr, struct rte_pci_device
*dev)
-{
- const struct rte_pci_id *id_table;
- int ret;
-
- for (id_table = dr->id_table ; id_table->vendor_id != 0; id_table++) {
-
- /* check if device's identifiers match the driver's ones */
- if (id_table->vendor_id != dev->id.vendor_id &&
- id_table->vendor_id != PCI_ANY_ID)
- continue;
- if (id_table->device_id != dev->id.device_id &&
- id_table->device_id != PCI_ANY_ID)
- continue;
- if (id_table->subsystem_vendor_id !=
dev->id.subsystem_vendor_id &&
- id_table->subsystem_vendor_id != PCI_ANY_ID)
- continue;
- if (id_table->subsystem_device_id !=
dev->id.subsystem_device_id &&
- id_table->subsystem_device_id != PCI_ANY_ID)
- continue;
-
- struct rte_pci_addr *loc = &dev->addr;
-
- RTE_LOG(DEBUG, EAL, "PCI device "PCI_PRI_FMT" on NUMA socket
%i\n",
- loc->domain, loc->bus, loc->devid,
loc->function,
- dev->numa_node);
-
- RTE_LOG(DEBUG, EAL, " probe driver: %x:%x %s\n",
dev->id.vendor_id,
- dev->id.device_id, dr->name);
-
- /* no initialization when blacklisted, return without error */
- if (dev->devargs != NULL &&
- dev->devargs->type == RTE_DEVTYPE_BLACKLISTED_PCI) {
-
- RTE_LOG(DEBUG, EAL, " Device is blacklisted, not
initializing\n");
- return 0;
- }
-
- if (dr->drv_flags & RTE_PCI_DRV_NEED_MAPPING) {
- /* map resources for devices that use igb_uio */
- ret = pci_uio_map_resource(dev);
- if (ret != 0)
- return ret;
- } else if (dr->drv_flags & RTE_PCI_DRV_FORCE_UNBIND &&
- rte_eal_process_type() == RTE_PROC_PRIMARY) {
- /* unbind current driver */
- if (pci_unbind_kernel_driver(dev) < 0)
- return -1;
- }
-
- /* reference driver structure */
- dev->driver = dr;
-
- /* call the driver devinit() function */
- return dr->devinit(dr, dev);
- }
- /* return positive value if driver is not found */
- return 1;
-}
-
/* Init the PCI EAL subsystem */
int
rte_eal_pci_init(void)
diff --git a/lib/librte_eal/common/eal_common_pci.c
b/lib/librte_eal/common/eal_common_pci.c
index afd14ec..031c7a7 100644
--- a/lib/librte_eal/common/eal_common_pci.c
+++ b/lib/librte_eal/common/eal_common_pci.c
@@ -139,7 +139,7 @@ pci_unmap_resource(void *requested_addr, size_t size)
}
/* Map pci device */
-int
+static int
pci_map_device(struct rte_pci_device *dev)
{
int ret = -1;
@@ -170,7 +170,7 @@ pci_map_device(struct rte_pci_device *dev)
#ifdef RTE_LIBRTE_EAL_HOTPLUG
/* Unmap pci device */
-void
+static void
pci_unmap_device(struct rte_pci_device *dev)
{
if (dev == NULL)
@@ -196,6 +196,135 @@ pci_unmap_device(struct rte_pci_device *dev)
#endif /* RTE_LIBRTE_EAL_HOTPLUG */
/*
+ * If vendor/device ID match, call the devinit() function of the
+ * driver.
+ */
+static int
+rte_eal_pci_probe_one_driver(struct rte_pci_driver *dr, struct rte_pci_device
*dev)
+{
+ int ret;
+ const struct rte_pci_id *id_table;
+
+ for (id_table = dr->id_table; id_table->vendor_id != 0; id_table++) {
+
+ /* check if device's identifiers match the driver's ones */
+
[dpdk-dev] [PATCH v2 0/3] Fix vhost startup issue
Tested-by: Qian Xu - Test Commit: b283164694b6ed18be1856b949e91a80371e44d4 - OS: Fedora 21 - GCC: gcc (GCC) 4.9.2 20141101 (Red Hat 4.9.2-1) - CPU: Intel(R) Xeon(R) CPU E5-2699 v3 @ 2.30GHz - NIC: Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection (rev 01) - Target: Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection (rev 01) - Total 1 cases(Test Case7 plus rerun ), 1 passed, 0 failed. Test Case 1: test_perf_virtio_one_vm_dpdk_fwd_vhost-cuse_jumboframe On host: 1. Start up vhost-switch, mergeable 1 means the jubmo frame feature is enabled. vm2vm 0 means only one vm without vm to vm communication:: taskset -c 1-3 /examples/vhost/build/vhost-switch -c 0xf -n 4 --huge-dir /mnt/huge --socket-mem 1024,1024 -- -p 1 --mergeable 1 --zero-copy 0 --vm2vm 0 2. Start VM with vhost cuse as backend:: taskset -c 4-6 /home/qxu10/qemu-2.2.0/x86_64-softmmu/qemu-system-x86_64 -object memory-backend-file, id=mem,size=2048M,mem-path=/mnt/huge,share=on -numa node,memdev=mem -mem-prealloc \ -enable-kvm -m 2048 -smp 4 -cpu host -name dpdk1-vm1 \ -drive file=/home/img/dpdk1-vm1.img \ -netdev tap,id=vhost3,ifname=tap_vhost3,vhost=on,script=no \ -device virtio-net pci,netdev=vhost3,mac=52:54:00:00:00:01,id=net3,csum=off,gso=off,guest_csum=off,guest_tso4=off,guest_tso6=off,guest_ecn=off \ -netdev tap,id=vhost4,ifname=tap_vhost4,vhost=on,script=no \ -device virtio-net-pci,netdev=vhost4,mac=52:54:00:00:00:02,id=net4,csum=off,gso=off,guest_csum=off,guest_tso4=off,guest_tso6=off,guest_ecn=off \ -netdev tap,id=ipvm1,ifname=tap3,script=/etc/qemu-ifup -device rtl8139,netdev=ipvm1,id=net0,mac=00:00:00:00:00:01 \ -localtime -nographic On guest: 3. ensure the dpdk folder copied to the guest with the same config file and build process as host. Then bind 2 virtio devices to igb_uio and start testpmd, below is the step for reference:: .//tools/dpdk_nic_bind.py --bind igb_uio 00:03.0 00:04.0 .//x86_64-native-linuxapp-gcc/app/test-pmd/testpmd -c f -n 4 -- -i --txqflags 0x0f00 --max-pkt-len 9000 $ >set fwd mac $ >start tx_first 4. After typing start tx_first in testpmd, user can see there would be 2 virtio device with MAC and vlan id registered in vhost-user, the log would be shown in host's vhost-sample output. 5. Send traffic(30second) to virtio1 and virtio2, and set the packet size from 64 to 1518 as well as the jumbo frame 3000. Check the performance in Mpps. The traffic sent to virtio1 should have the DEST MAC of Virtio1's MAC, Vlan id of Virtio1. The traffic sent to virtio2 should have the DEST MAC of Virtio2's MAC, Vlan id of Virtio2. As to the functionality criteria, The received rate should not be zero. As to the performance criteria, need check it with developer or design doc/PRD. Test Case 7: test_perf_virtio_one_vm_dpdk_fwd_vhost-user_jumboframe This case is similar to TestCase1, just change the backend from vhost cuse to vhost-user, so need rebuild the dpdk in vhost-user on host, other steps are same as TestCase1. The command to launch vm is different, see below as example:: /x86_64-softmmu/qemu-system-x86_64 -name us-vhost-vm1 -cpu host -enable-kvm -m 2048 -object memory-backend-file,id=mem,size=2048M,mem-path=/mnt/huge,share=on -numa node,memdev=mem -mem-prealloc -smp 2 -drive file=/home/img/dpdk1-vm1.img -chardev socket,id=char0,path=
[dpdk-dev] [PATCH v3 00/12] Introducing the TILE-Gx platform
This series adds support for the EZchip TILE-Gx family of SoCs. The architecture port in itself is fairly straight forward due to its reliance on generics for the most part. In addition to adding TILE-Gx architecture specific code, this series includes a few cross-platform fixes for DPDK (cpuflags, SSE related, etc.), as well as minor extensions to to accomodate a wider range of hugepage sizes and configurable mempool element alignment boundaries. Changes in this series: v3: Renewed the Signed-off-by line. v2: Removed RTE_LIBNAME per Thomas' feedback. Cyril Chemparathy (12): test: limit x86 cpuflags checks to x86 builds hash: fix compilation on non-X86 platforms hash: check SSE flags only on x86 builds eal: allow empty compile time flags config: remove RTE_LIBNAME definition. memzone: refactor rte_memzone_reserve() variants memzone: allow multiple pagesizes to be requested mempool: allow config override on element alignment tile: add page sizes for TILE-Gx/Mx platforms tile: initial TILE-Gx support. tile: Add TILE-Gx mPIPE poll mode driver. maintainers: claim responsibility for TILE-Gx platform MAINTAINERS|4 + app/test/test_cpuflags.c |6 +- config/common_bsdapp |1 - config/common_linuxapp |1 - config/defconfig_ppc_64-power8-linuxapp-gcc|2 - config/defconfig_tile-tilegx-linuxapp-gcc | 70 + drivers/net/Makefile |1 + drivers/net/mpipe/Makefile | 46 + drivers/net/mpipe/mpipe_tilegx.c | 1637 lib/librte_eal/common/eal_common_cpuflags.c|5 +- lib/librte_eal/common/eal_common_memzone.c | 141 +- .../common/include/arch/tile/rte_atomic.h | 86 + .../common/include/arch/tile/rte_byteorder.h | 91 ++ .../common/include/arch/tile/rte_cpuflags.h| 85 + .../common/include/arch/tile/rte_cycles.h | 70 + .../common/include/arch/tile/rte_memcpy.h | 93 ++ .../common/include/arch/tile/rte_prefetch.h| 61 + .../common/include/arch/tile/rte_rwlock.h | 70 + .../common/include/arch/tile/rte_spinlock.h| 92 ++ lib/librte_eal/common/include/rte_memory.h | 16 +- lib/librte_eal/common/include/rte_memzone.h| 50 +- lib/librte_hash/rte_hash_crc.h |2 + lib/librte_hash/rte_jhash.h|3 +- lib/librte_mempool/rte_mempool.c | 16 +- lib/librte_mempool/rte_mempool.h |6 + mk/arch/tile/rte.vars.mk | 39 + mk/machine/tilegx/rte.vars.mk | 57 + mk/rte.app.mk |1 + mk/rte.vars.mk |5 +- 29 files changed, 2639 insertions(+), 118 deletions(-) create mode 100644 config/defconfig_tile-tilegx-linuxapp-gcc create mode 100644 drivers/net/mpipe/Makefile create mode 100644 drivers/net/mpipe/mpipe_tilegx.c create mode 100644 lib/librte_eal/common/include/arch/tile/rte_atomic.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_byteorder.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_cpuflags.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_cycles.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_memcpy.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_prefetch.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_rwlock.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_spinlock.h create mode 100644 mk/arch/tile/rte.vars.mk create mode 100644 mk/machine/tilegx/rte.vars.mk -- 2.1.2
[dpdk-dev] [PATCH v3 01/12] test: limit x86 cpuflags checks to x86 builds
The original code mistakenly defaulted to X86 when RTE_ARCH_PPC_64 was
left undefined. This did not accomodate other non-PPC/non-X86
architectures. This patch fixes this issue.
Change-Id: I5e8cf33c2eb917f7f6583dc95ed0f336066a285e
Signed-off-by: Zhigang Lu
---
app/test/test_cpuflags.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/app/test/test_cpuflags.c b/app/test/test_cpuflags.c
index 5aeba5d..5b92061 100644
--- a/app/test/test_cpuflags.c
+++ b/app/test/test_cpuflags.c
@@ -113,7 +113,9 @@ test_cpuflags(void)
printf("Check for ICACHE_SNOOP:\t\t");
CHECK_FOR_FLAG(RTE_CPUFLAG_ICACHE_SNOOP);
-#else
+#endif
+
+#if defined(RTE_ARCH_X86_64) || defined(RTE_ARCH_I686)
printf("Check for SSE:\t\t");
CHECK_FOR_FLAG(RTE_CPUFLAG_SSE);
@@ -149,8 +151,6 @@ test_cpuflags(void)
printf("Check for INVTSC:\t");
CHECK_FOR_FLAG(RTE_CPUFLAG_INVTSC);
-
-
#endif
/*
--
2.1.2
[dpdk-dev] [PATCH v3 02/12] hash: fix compilation on non-X86 platforms
The "hash: remove duplicated code" change unfortunately broke the
build for non-X86 platforms. This patch fixes this breakage.
Change-Id: Ie109d67e681b75b45320fab1bf9de4eb9c9701bf
Signed-off-by: Zhigang Lu
---
lib/librte_hash/rte_jhash.h | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/lib/librte_hash/rte_jhash.h b/lib/librte_hash/rte_jhash.h
index e230449..d1b6cf3 100644
--- a/lib/librte_hash/rte_jhash.h
+++ b/lib/librte_hash/rte_jhash.h
@@ -44,6 +44,7 @@
extern "C" {
#endif
+#include
#include
#include
@@ -122,7 +123,7 @@ __rte_jhash_2hashes(const void *key, uint32_t length,
uint32_t *pc,
const uint32_t *k = key;
const uint32_t s = 0;
#else
- const uint32_t *k = (uint32_t *)(uintptr_t)key & (uintptr_t)~3);
+ const uint32_t *k = (uint32_t *)((uintptr_t)key & (uintptr_t)~3);
const uint32_t s = ((uintptr_t)key & 3) * CHAR_BIT;
#endif
if (!check_align || s == 0) {
--
2.1.2
[dpdk-dev] [PATCH v3 03/12] hash: check SSE flags only on x86 builds
This is necessary because the required CPU flags may not be defined on
other architectures.
Change-Id: I14d3f9f625b2e7567123f1c97095f8d06abd674b
Signed-off-by: Zhigang Lu
---
lib/librte_hash/rte_hash_crc.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/lib/librte_hash/rte_hash_crc.h b/lib/librte_hash/rte_hash_crc.h
index abdbd9a..1f6f5bf 100644
--- a/lib/librte_hash/rte_hash_crc.h
+++ b/lib/librte_hash/rte_hash_crc.h
@@ -425,12 +425,14 @@ static inline void
rte_hash_crc_set_alg(uint8_t alg)
{
switch (alg) {
+#if defined(RTE_ARCH_I686) || defined(RTE_ARCH_X86_64)
case CRC32_SSE42_x64:
if (! rte_cpu_get_flag_enabled(RTE_CPUFLAG_EM64T))
alg = CRC32_SSE42;
case CRC32_SSE42:
if (! rte_cpu_get_flag_enabled(RTE_CPUFLAG_SSE4_2))
alg = CRC32_SW;
+#endif
case CRC32_SW:
crc32_alg = alg;
default:
--
2.1.2
[dpdk-dev] [PATCH v3 04/12] eal: allow empty compile time flags
The rte_cpu_check_supported() code breaks with a "comparison is always
false due to limited range of data type" when the compile_time_flags[]
array is empty. Assigning the array dimension to a local variable
apparently solves this.
Change-Id: I0ae21f529cf7b6dd9cf0f4532dce9198f4bf4230
Signed-off-by: Zhigang Lu
---
lib/librte_eal/common/eal_common_cpuflags.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/lib/librte_eal/common/eal_common_cpuflags.c
b/lib/librte_eal/common/eal_common_cpuflags.c
index 6fd360c..8ba7b30 100644
--- a/lib/librte_eal/common/eal_common_cpuflags.c
+++ b/lib/librte_eal/common/eal_common_cpuflags.c
@@ -30,6 +30,7 @@
* (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
/*
@@ -62,10 +63,10 @@ rte_cpu_check_supported(void)
static const enum rte_cpu_flag_t compile_time_flags[] = {
RTE_COMPILE_TIME_CPUFLAGS
};
- unsigned i;
+ unsigned count = RTE_DIM(compile_time_flags), i;
int ret;
- for (i = 0; i <
sizeof(compile_time_flags)/sizeof(compile_time_flags[0]); i++) {
+ for (i = 0; i < count; i++) {
ret = rte_cpu_get_flag_enabled(compile_time_flags[i]);
if (ret < 0) {
--
2.1.2
[dpdk-dev] [PATCH v3 05/12] config: remove RTE_LIBNAME definition.
The library name is now being pinned to "dpdk" instead of intel_dpdk, powerpc_dpdk, etc. As a result, we no longer need this config item. This patch removes it. Change-Id: I36f7cf6c18c3563c6f5ccdf01bb70579c7ccaa16 Signed-off-by: Zhigang Lu --- config/common_bsdapp| 1 - config/common_linuxapp | 1 - config/defconfig_ppc_64-power8-linuxapp-gcc | 2 -- mk/rte.vars.mk | 5 + 4 files changed, 1 insertion(+), 8 deletions(-) diff --git a/config/common_bsdapp b/config/common_bsdapp index 78754b2..ad1448f 100644 --- a/config/common_bsdapp +++ b/config/common_bsdapp @@ -87,7 +87,6 @@ CONFIG_RTE_BUILD_SHARED_LIB=n # Combine to one single library # CONFIG_RTE_BUILD_COMBINE_LIBS=n -CONFIG_RTE_LIBNAME=intel_dpdk # # Compile Environment Abstraction Layer diff --git a/config/common_linuxapp b/config/common_linuxapp index f5646e0..4cad71c 100644 --- a/config/common_linuxapp +++ b/config/common_linuxapp @@ -87,7 +87,6 @@ CONFIG_RTE_BUILD_SHARED_LIB=n # Combine to one single library # CONFIG_RTE_BUILD_COMBINE_LIBS=n -CONFIG_RTE_LIBNAME="intel_dpdk" # # Compile Environment Abstraction Layer diff --git a/config/defconfig_ppc_64-power8-linuxapp-gcc b/config/defconfig_ppc_64-power8-linuxapp-gcc index d97a885..f1af518 100644 --- a/config/defconfig_ppc_64-power8-linuxapp-gcc +++ b/config/defconfig_ppc_64-power8-linuxapp-gcc @@ -39,8 +39,6 @@ CONFIG_RTE_ARCH_64=y CONFIG_RTE_TOOLCHAIN="gcc" CONFIG_RTE_TOOLCHAIN_GCC=y -CONFIG_RTE_LIBNAME="powerpc_dpdk" - # Note: Power doesn't have this support CONFIG_RTE_LIBRTE_EAL_VMWARE_TSC_MAP_SUPPORT=n diff --git a/mk/rte.vars.mk b/mk/rte.vars.mk index d2f01b6..bb82395 100644 --- a/mk/rte.vars.mk +++ b/mk/rte.vars.mk @@ -73,10 +73,7 @@ ifneq ($(BUILDING_RTE_SDK),) endif endif -RTE_LIBNAME := $(CONFIG_RTE_LIBNAME:"%"=%) -ifeq ($(RTE_LIBNAME),) -RTE_LIBNAME := intel_dpdk -endif +RTE_LIBNAME := dpdk # RTE_TARGET is deducted from config when we are building the SDK. # Else, when building an external app, RTE_TARGET must be specified -- 2.1.2
[dpdk-dev] [PATCH v3 06/12] memzone: refactor rte_memzone_reserve() variants
The definitions of rte_memzone_reserve_aligned() and
rte_memzone_reserve_bounded() were identical with the exception of the
bound argument passed into rte_memzone_reserve_thread_safe().
This patch removes this replication of code by unifying it into
rte_memzone_reserve_thread_safe(), which is then called by all three
variants of rte_memzone_reserve().
Change-Id: Id26c25b4dd3d07861eaf35e72aaa2de555916fa3
Signed-off-by: Zhigang Lu
---
lib/librte_eal/common/eal_common_memzone.c | 77 +-
1 file changed, 33 insertions(+), 44 deletions(-)
diff --git a/lib/librte_eal/common/eal_common_memzone.c
b/lib/librte_eal/common/eal_common_memzone.c
index aee184a..1ea502b 100644
--- a/lib/librte_eal/common/eal_common_memzone.c
+++ b/lib/librte_eal/common/eal_common_memzone.c
@@ -77,18 +77,6 @@ memzone_lookup_thread_unsafe(const char *name)
}
/*
- * Return a pointer to a correctly filled memzone descriptor. If the
- * allocation cannot be done, return NULL.
- */
-const struct rte_memzone *
-rte_memzone_reserve(const char *name, size_t len, int socket_id,
- unsigned flags)
-{
- return rte_memzone_reserve_aligned(name,
- len, socket_id, flags, RTE_CACHE_LINE_SIZE);
-}
-
-/*
* Helper function for memzone_reserve_aligned_thread_unsafe().
* Calculate address offset from the start of the segment.
* Align offset in that way that it satisfy istart alignmnet and
@@ -307,13 +295,10 @@ memzone_reserve_aligned_thread_unsafe(const char *name,
size_t len,
return mz;
}
-/*
- * Return a pointer to a correctly filled memzone descriptor (with a
- * specified alignment). If the allocation cannot be done, return NULL.
- */
-const struct rte_memzone *
-rte_memzone_reserve_aligned(const char *name, size_t len,
- int socket_id, unsigned flags, unsigned align)
+static const struct rte_memzone *
+rte_memzone_reserve_thread_safe(const char *name, size_t len,
+ int socket_id, unsigned flags, unsigned align,
+ unsigned bound)
{
struct rte_mem_config *mcfg;
const struct rte_memzone *mz = NULL;
@@ -331,7 +316,7 @@ rte_memzone_reserve_aligned(const char *name, size_t len,
rte_rwlock_write_lock(&mcfg->mlock);
mz = memzone_reserve_aligned_thread_unsafe(
- name, len, socket_id, flags, align, 0);
+ name, len, socket_id, flags, align, bound);
rte_rwlock_write_unlock(&mcfg->mlock);
@@ -340,36 +325,40 @@ rte_memzone_reserve_aligned(const char *name, size_t len,
/*
* Return a pointer to a correctly filled memzone descriptor (with a
- * specified alignment and boundary).
- * If the allocation cannot be done, return NULL.
+ * specified alignment and boundary). If the allocation cannot be done,
+ * return NULL.
*/
const struct rte_memzone *
-rte_memzone_reserve_bounded(const char *name, size_t len,
- int socket_id, unsigned flags, unsigned align, unsigned bound)
+rte_memzone_reserve_bounded(const char *name, size_t len, int socket_id,
+ unsigned flags, unsigned align, unsigned bound)
{
- struct rte_mem_config *mcfg;
- const struct rte_memzone *mz = NULL;
-
- /* both sizes cannot be explicitly called for */
- if (((flags & RTE_MEMZONE_1GB) && (flags & RTE_MEMZONE_2MB))
- || ((flags & RTE_MEMZONE_16MB) && (flags & RTE_MEMZONE_16GB))) {
- rte_errno = EINVAL;
- return NULL;
- }
-
- /* get pointer to global configuration */
- mcfg = rte_eal_get_configuration()->mem_config;
-
- rte_rwlock_write_lock(&mcfg->mlock);
-
- mz = memzone_reserve_aligned_thread_unsafe(
- name, len, socket_id, flags, align, bound);
-
- rte_rwlock_write_unlock(&mcfg->mlock);
+ return rte_memzone_reserve_thread_safe(name, len, socket_id, flags,
+ align, bound);
+}
- return mz;
+/*
+ * Return a pointer to a correctly filled memzone descriptor (with a
+ * specified alignment). If the allocation cannot be done, return NULL.
+ */
+const struct rte_memzone *
+rte_memzone_reserve_aligned(const char *name, size_t len, int socket_id,
+ unsigned flags, unsigned align)
+{
+ return rte_memzone_reserve_thread_safe(name, len, socket_id, flags,
+ align, 0);
}
+/*
+ * Return a pointer to a correctly filled memzone descriptor. If the
+ * allocation cannot be done, return NULL.
+ */
+const struct rte_memzone *
+rte_memzone_reserve(const char *name, size_t len, int socket_id,
+ unsigned flags)
+{
+ return rte_memzone_reserve_thread_safe(name, len, socket_id,
+ flags, RTE_CACHE_LINE_SIZE, 0);
+}
/*
* Lookup for the memzone identified by the given name
--
2.1.2
[dpdk-dev] [PATCH v3 07/12] memzone: allow multiple pagesizes to be requested
This patch extends the memzone allocator to remove the restriction
that prevented callers from specifying multiple page sizes in the
flags argument.
In doing so, we also sanitize the free segment matching logic to get
rid of architecture specific disjunctions (2MB vs 1GB on x86, and 16MB
vs 16GB on PPC), thereby allowing for a broader range of hugepages on
architectures that support it.
Change-Id: Ic3713f61da49629a570fe4de34a8aaf5e2e0a19b
Signed-off-by: Zhigang Lu
---
lib/librte_eal/common/eal_common_memzone.c | 58 ++
1 file changed, 27 insertions(+), 31 deletions(-)
diff --git a/lib/librte_eal/common/eal_common_memzone.c
b/lib/librte_eal/common/eal_common_memzone.c
index 1ea502b..76bae72 100644
--- a/lib/librte_eal/common/eal_common_memzone.c
+++ b/lib/librte_eal/common/eal_common_memzone.c
@@ -113,7 +113,8 @@ align_phys_boundary(const struct rte_memseg *ms, size_t
len, size_t align,
static const struct rte_memzone *
memzone_reserve_aligned_thread_unsafe(const char *name, size_t len,
- int socket_id, unsigned flags, unsigned align, unsigned bound)
+ int socket_id, uint64_t size_mask, unsigned align,
+ unsigned bound)
{
struct rte_mem_config *mcfg;
unsigned i = 0;
@@ -201,18 +202,7 @@ memzone_reserve_aligned_thread_unsafe(const char *name,
size_t len,
if ((requested_len + addr_offset) > free_memseg[i].len)
continue;
- /* check flags for hugepage sizes */
- if ((flags & RTE_MEMZONE_2MB) &&
- free_memseg[i].hugepage_sz == RTE_PGSIZE_1G)
- continue;
- if ((flags & RTE_MEMZONE_1GB) &&
- free_memseg[i].hugepage_sz == RTE_PGSIZE_2M)
- continue;
- if ((flags & RTE_MEMZONE_16MB) &&
- free_memseg[i].hugepage_sz == RTE_PGSIZE_16G)
- continue;
- if ((flags & RTE_MEMZONE_16GB) &&
- free_memseg[i].hugepage_sz == RTE_PGSIZE_16M)
+ if ((size_mask & free_memseg[i].hugepage_sz) == 0)
continue;
/* this segment is the best until now */
@@ -244,16 +234,6 @@ memzone_reserve_aligned_thread_unsafe(const char *name,
size_t len,
/* no segment found */
if (memseg_idx == -1) {
- /*
-* If RTE_MEMZONE_SIZE_HINT_ONLY flag is specified,
-* try allocating again without the size parameter otherwise
-fail.
-*/
- if ((flags & RTE_MEMZONE_SIZE_HINT_ONLY) &&
- ((flags & RTE_MEMZONE_1GB) || (flags & RTE_MEMZONE_2MB)
- || (flags & RTE_MEMZONE_16MB) || (flags & RTE_MEMZONE_16GB)))
- return memzone_reserve_aligned_thread_unsafe(name,
- len, socket_id, 0, align, bound);
-
rte_errno = ENOMEM;
return NULL;
}
@@ -302,13 +282,18 @@ rte_memzone_reserve_thread_safe(const char *name, size_t
len,
{
struct rte_mem_config *mcfg;
const struct rte_memzone *mz = NULL;
-
- /* both sizes cannot be explicitly called for */
- if (((flags & RTE_MEMZONE_1GB) && (flags & RTE_MEMZONE_2MB))
- || ((flags & RTE_MEMZONE_16MB) && (flags & RTE_MEMZONE_16GB))) {
- rte_errno = EINVAL;
- return NULL;
- }
+ uint64_t size_mask = 0;
+
+ if (flags & RTE_MEMZONE_2MB)
+ size_mask |= RTE_PGSIZE_2M;
+ if (flags & RTE_MEMZONE_16MB)
+ size_mask |= RTE_PGSIZE_16M;
+ if (flags & RTE_MEMZONE_1GB)
+ size_mask |= RTE_PGSIZE_1G;
+ if (flags & RTE_MEMZONE_16GB)
+ size_mask |= RTE_PGSIZE_16G;
+ if (!size_mask)
+ size_mask = UINT64_MAX;
/* get pointer to global configuration */
mcfg = rte_eal_get_configuration()->mem_config;
@@ -316,7 +301,18 @@ rte_memzone_reserve_thread_safe(const char *name, size_t
len,
rte_rwlock_write_lock(&mcfg->mlock);
mz = memzone_reserve_aligned_thread_unsafe(
- name, len, socket_id, flags, align, bound);
+ name, len, socket_id, size_mask, align, bound);
+
+ /*
+* If we failed to allocate the requested page size, and the
+* RTE_MEMZONE_SIZE_HINT_ONLY flag is specified, try allocating
+* again.
+*/
+ if (!mz && rte_errno == ENOMEM && size_mask != UINT64_MAX &&
+ flags & RTE_MEMZONE_SIZE_HINT_ONLY) {
+ mz = memzone_reserve_aligned_thread_unsafe(
+ name, len, socket_id, UINT64_MAX, align, bound);
+ }
rte_rwlock_write_unlock(&mcfg->mlock);
--
2.1.2
[dpdk-dev] [PATCH v3 08/12] mempool: allow config override on element alignment
On TILE-Gx and TILE-Mx platforms, the buffers fed into the hardware
buffer manager require a 128-byte alignment. With this change, we
allow configuration based override of the element alignment, and
default to RTE_CACHE_LINE_SIZE if left unspecified.
Change-Id: I9cd789d92b0bc9c8f44a633de59bb04d45d927a7
Signed-off-by: Zhigang Lu
---
lib/librte_mempool/rte_mempool.c | 16 +---
lib/librte_mempool/rte_mempool.h | 6 ++
2 files changed, 15 insertions(+), 7 deletions(-)
diff --git a/lib/librte_mempool/rte_mempool.c b/lib/librte_mempool/rte_mempool.c
index 02699a1..8e185c5 100644
--- a/lib/librte_mempool/rte_mempool.c
+++ b/lib/librte_mempool/rte_mempool.c
@@ -120,10 +120,10 @@ static unsigned optimize_object_size(unsigned obj_size)
nrank = 1;
/* process new object size */
- new_obj_size = (obj_size + RTE_CACHE_LINE_MASK) / RTE_CACHE_LINE_SIZE;
+ new_obj_size = (obj_size + RTE_MEMPOOL_ALIGN_MASK) / RTE_MEMPOOL_ALIGN;
while (get_gcd(new_obj_size, nrank * nchan) != 1)
new_obj_size++;
- return new_obj_size * RTE_CACHE_LINE_SIZE;
+ return new_obj_size * RTE_MEMPOOL_ALIGN;
}
static void
@@ -267,7 +267,7 @@ rte_mempool_calc_obj_size(uint32_t elt_size, uint32_t flags,
#endif
if ((flags & MEMPOOL_F_NO_CACHE_ALIGN) == 0)
sz->header_size = RTE_ALIGN_CEIL(sz->header_size,
- RTE_CACHE_LINE_SIZE);
+ RTE_MEMPOOL_ALIGN);
/* trailer contains the cookie in debug mode */
sz->trailer_size = 0;
@@ -281,9 +281,9 @@ rte_mempool_calc_obj_size(uint32_t elt_size, uint32_t flags,
if ((flags & MEMPOOL_F_NO_CACHE_ALIGN) == 0) {
sz->total_size = sz->header_size + sz->elt_size +
sz->trailer_size;
- sz->trailer_size += ((RTE_CACHE_LINE_SIZE -
- (sz->total_size & RTE_CACHE_LINE_MASK)) &
-RTE_CACHE_LINE_MASK);
+ sz->trailer_size += ((RTE_MEMPOOL_ALIGN -
+ (sz->total_size & RTE_MEMPOOL_ALIGN_MASK)) &
+RTE_MEMPOOL_ALIGN_MASK);
}
/*
@@ -498,7 +498,7 @@ rte_mempool_xmem_create(const char *name, unsigned n,
unsigned elt_size,
* cache-aligned
*/
private_data_size = (private_data_size +
-RTE_CACHE_LINE_MASK) & (~RTE_CACHE_LINE_MASK);
+RTE_MEMPOOL_ALIGN_MASK) &
(~RTE_MEMPOOL_ALIGN_MASK);
if (! rte_eal_has_hugepages()) {
/*
@@ -525,6 +525,7 @@ rte_mempool_xmem_create(const char *name, unsigned n,
unsigned elt_size,
* enough to hold mempool header and metadata plus mempool objects.
*/
mempool_size = MEMPOOL_HEADER_SIZE(mp, pg_num) + private_data_size;
+ mempool_size = RTE_ALIGN_CEIL(mempool_size, RTE_MEMPOOL_ALIGN);
if (vaddr == NULL)
mempool_size += (size_t)objsz.total_size * n;
@@ -580,6 +581,7 @@ rte_mempool_xmem_create(const char *name, unsigned n,
unsigned elt_size,
/* calculate address of the first element for continuous mempool. */
obj = (char *)mp + MEMPOOL_HEADER_SIZE(mp, pg_num) +
private_data_size;
+ obj = RTE_PTR_ALIGN_CEIL(obj, RTE_MEMPOOL_ALIGN);
/* populate address translation fields. */
mp->pg_num = pg_num;
diff --git a/lib/librte_mempool/rte_mempool.h b/lib/librte_mempool/rte_mempool.h
index 6d4ce9a..ee67ce7 100644
--- a/lib/librte_mempool/rte_mempool.h
+++ b/lib/librte_mempool/rte_mempool.h
@@ -142,6 +142,12 @@ struct rte_mempool_objsz {
/** Mempool over one chunk of physically continuous memory */
#defineMEMPOOL_PG_NUM_DEFAULT 1
+#ifndef RTE_MEMPOOL_ALIGN
+#define RTE_MEMPOOL_ALIGN RTE_CACHE_LINE_SIZE
+#endif
+
+#define RTE_MEMPOOL_ALIGN_MASK (RTE_MEMPOOL_ALIGN - 1)
+
/**
* Mempool object header structure
*
--
2.1.2
[dpdk-dev] [PATCH v3 09/12] tile: add page sizes for TILE-Gx/Mx platforms
This patch adds a few new page sizes that are supported on the TILE-Gx
and TILE-Mx platforms.
Change-Id: I7da783703149c4f96b84d0017fa4f1191fc029c6
Signed-off-by: Zhigang Lu
---
lib/librte_eal/common/eal_common_memzone.c | 8 +
lib/librte_eal/common/include/rte_memory.h | 16 +
lib/librte_eal/common/include/rte_memzone.h | 50 +++--
3 files changed, 51 insertions(+), 23 deletions(-)
diff --git a/lib/librte_eal/common/eal_common_memzone.c
b/lib/librte_eal/common/eal_common_memzone.c
index 76bae72..dc39a79 100644
--- a/lib/librte_eal/common/eal_common_memzone.c
+++ b/lib/librte_eal/common/eal_common_memzone.c
@@ -284,12 +284,20 @@ rte_memzone_reserve_thread_safe(const char *name, size_t
len,
const struct rte_memzone *mz = NULL;
uint64_t size_mask = 0;
+ if (flags & RTE_MEMZONE_256KB)
+ size_mask |= RTE_PGSIZE_256K;
if (flags & RTE_MEMZONE_2MB)
size_mask |= RTE_PGSIZE_2M;
if (flags & RTE_MEMZONE_16MB)
size_mask |= RTE_PGSIZE_16M;
+ if (flags & RTE_MEMZONE_256MB)
+ size_mask |= RTE_PGSIZE_256M;
+ if (flags & RTE_MEMZONE_512MB)
+ size_mask |= RTE_PGSIZE_512M;
if (flags & RTE_MEMZONE_1GB)
size_mask |= RTE_PGSIZE_1G;
+ if (flags & RTE_MEMZONE_4GB)
+ size_mask |= RTE_PGSIZE_4G;
if (flags & RTE_MEMZONE_16GB)
size_mask |= RTE_PGSIZE_16G;
if (!size_mask)
diff --git a/lib/librte_eal/common/include/rte_memory.h
b/lib/librte_eal/common/include/rte_memory.h
index d948c0b..1bed415 100644
--- a/lib/librte_eal/common/include/rte_memory.h
+++ b/lib/librte_eal/common/include/rte_memory.h
@@ -53,12 +53,16 @@ extern "C" {
#endif
enum rte_page_sizes {
- RTE_PGSIZE_4K = 1ULL << 12,
- RTE_PGSIZE_2M = 1ULL << 21,
- RTE_PGSIZE_1G = 1ULL << 30,
- RTE_PGSIZE_64K = 1ULL << 16,
- RTE_PGSIZE_16M = 1ULL << 24,
- RTE_PGSIZE_16G = 1ULL << 34
+ RTE_PGSIZE_4K= 1ULL << 12,
+ RTE_PGSIZE_64K = 1ULL << 16,
+ RTE_PGSIZE_256K = 1ULL << 18,
+ RTE_PGSIZE_2M= 1ULL << 21,
+ RTE_PGSIZE_16M = 1ULL << 24,
+ RTE_PGSIZE_256M = 1ULL << 28,
+ RTE_PGSIZE_512M = 1ULL << 29,
+ RTE_PGSIZE_1G= 1ULL << 30,
+ RTE_PGSIZE_4G= 1ULL << 32,
+ RTE_PGSIZE_16G = 1ULL << 34,
};
#define SOCKET_ID_ANY -1/**< Any NUMA socket. */
diff --git a/lib/librte_eal/common/include/rte_memzone.h
b/lib/librte_eal/common/include/rte_memzone.h
index ee62680..de5ae55 100644
--- a/lib/librte_eal/common/include/rte_memzone.h
+++ b/lib/librte_eal/common/include/rte_memzone.h
@@ -60,8 +60,12 @@ extern "C" {
#define RTE_MEMZONE_2MB0x0001 /**< Use 2MB pages. */
#define RTE_MEMZONE_1GB0x0002 /**< Use 1GB pages. */
-#define RTE_MEMZONE_16MB0x0100 /**< Use 16MB pages. */
-#define RTE_MEMZONE_16GB0x0200 /**< Use 16GB pages. */
+#define RTE_MEMZONE_16MB 0x0100 /**< Use 16MB pages. */
+#define RTE_MEMZONE_16GB 0x0200 /**< Use 16GB pages. */
+#define RTE_MEMZONE_256KB 0x0001 /**< Use 256KB pages. */
+#define RTE_MEMZONE_256MB 0x0002 /**< Use 256MB pages. */
+#define RTE_MEMZONE_512MB 0x0004 /**< Use 512MB pages. */
+#define RTE_MEMZONE_4GB0x0008 /**< Use 4GB pages. */
#define RTE_MEMZONE_SIZE_HINT_ONLY 0x0004 /**< Use available page size */
/**
@@ -110,11 +114,15 @@ struct rte_memzone {
* constraint for the reserved zone.
* @param flags
* The flags parameter is used to request memzones to be
- * taken from 1GB or 2MB hugepages.
- * - RTE_MEMZONE_2MB - Reserve from 2MB pages
- * - RTE_MEMZONE_1GB - Reserve from 1GB pages
- * - RTE_MEMZONE_16MB - Reserve from 16MB pages
- * - RTE_MEMZONE_16GB - Reserve from 16GB pages
+ * taken from specifically sized hugepages.
+ * - RTE_MEMZONE_2MB - Reserved from 2MB pages
+ * - RTE_MEMZONE_1GB - Reserved from 1GB pages
+ * - RTE_MEMZONE_16MB - Reserved from 16MB pages
+ * - RTE_MEMZONE_16GB - Reserved from 16GB pages
+ * - RTE_MEMZONE_256KB - Reserved from 256KB pages
+ * - RTE_MEMZONE_256MB - Reserved from 256MB pages
+ * - RTE_MEMZONE_512MB - Reserved from 512MB pages
+ * - RTE_MEMZONE_4GB - Reserved from 4GB pages
* - RTE_MEMZONE_SIZE_HINT_ONLY - Allow alternative page size to be used if
* the requested page size is unavailable.
* If this flag is not set, the function
@@ -157,11 +165,15 @@ const struct rte_memzone *rte_memzone_reserve(const char
*name,
* constraint for the reserved zone.
* @param flags
* The flags parameter is used to request memzones to be
- * taken from 1GB or 2MB hugepages.
- * - RTE_MEMZONE_2MB - Reserve from 2MB pages
- * - RTE_MEMZONE_1GB - Reserve from 1G
[dpdk-dev] [PATCH v3 10/12] tile: initial TILE-Gx support.
This commit adds support for the TILE-Gx platform, as well as the TILE CPU architecture. This architecture port is fairly simple due to its reliance on generics for most arch stuff. Change-Id: I809fcf740e25ba5976a6b7736c1673515338cf80 Signed-off-by: Zhigang Lu --- config/defconfig_tile-tilegx-linuxapp-gcc | 69 .../common/include/arch/tile/rte_atomic.h | 86 .../common/include/arch/tile/rte_byteorder.h | 91 + .../common/include/arch/tile/rte_cpuflags.h| 85 .../common/include/arch/tile/rte_cycles.h | 70 .../common/include/arch/tile/rte_memcpy.h | 93 ++ .../common/include/arch/tile/rte_prefetch.h| 61 ++ .../common/include/arch/tile/rte_rwlock.h | 70 .../common/include/arch/tile/rte_spinlock.h| 92 + mk/arch/tile/rte.vars.mk | 39 + mk/machine/tilegx/rte.vars.mk | 57 + 11 files changed, 813 insertions(+) create mode 100644 config/defconfig_tile-tilegx-linuxapp-gcc create mode 100644 lib/librte_eal/common/include/arch/tile/rte_atomic.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_byteorder.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_cpuflags.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_cycles.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_memcpy.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_prefetch.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_rwlock.h create mode 100644 lib/librte_eal/common/include/arch/tile/rte_spinlock.h create mode 100644 mk/arch/tile/rte.vars.mk create mode 100644 mk/machine/tilegx/rte.vars.mk diff --git a/config/defconfig_tile-tilegx-linuxapp-gcc b/config/defconfig_tile-tilegx-linuxapp-gcc new file mode 100644 index 000..4023878 --- /dev/null +++ b/config/defconfig_tile-tilegx-linuxapp-gcc @@ -0,0 +1,69 @@ +# BSD LICENSE +# +# Copyright (C) EZchip Semiconductor 2015. +# +# 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 EZchip Semiconductor 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 "common_linuxapp" + +CONFIG_RTE_MACHINE="tilegx" + +CONFIG_RTE_ARCH="tile" +CONFIG_RTE_ARCH_TILE=y +CONFIG_RTE_ARCH_64=y +CONFIG_RTE_ARCH_STRICT_ALIGN=y +CONFIG_RTE_FORCE_INTRINSICS=y + +CONFIG_RTE_TOOLCHAIN="gcc" +CONFIG_RTE_TOOLCHAIN_GCC=y + +# Disable things that we don't support or need +CONFIG_RTE_LIBRTE_EAL_VMWARE_TSC_MAP_SUPPORT=n +CONFIG_RTE_EAL_IGB_UIO=n +CONFIG_RTE_EAL_VFIO=n +CONFIG_RTE_LIBRTE_KNI=n +CONFIG_RTE_LIBRTE_XEN_DOM0=n +CONFIG_RTE_LIBRTE_IGB_PMD=n +CONFIG_RTE_LIBRTE_EM_PMD=n +CONFIG_RTE_LIBRTE_IXGBE_PMD=n +CONFIG_RTE_LIBRTE_I40E_PMD=n +CONFIG_RTE_LIBRTE_FM10K_PMD=n +CONFIG_RTE_LIBRTE_VIRTIO_PMD=n +CONFIG_RTE_LIBRTE_VMXNET3_PMD=n +CONFIG_RTE_LIBRTE_ENIC_PMD=n + +# This following libraries are not available on the tile architecture. So +# they're turned off. +CONFIG_RTE_LIBRTE_LPM=n +CONFIG_RTE_LIBRTE_ACL=n +CONFIG_RTE_LIBRTE_SCHED=n +CONFIG_RTE_LIBRTE_PORT=n +CONFIG_RTE_LIBRTE_TABLE=n +CONFIG_RTE_LIBRTE_PIPELINE=n + +# Enable and override things that we need +CONFIG_RTE_MEMPOOL_ALIGN=128 diff --git a/lib/librte_eal/common/include/arch/tile/rte_atomic.h b/lib/librte_eal/common/include/arch/tile/rte_atomic.h new file mode 100644 index 000..3dc8eb8 --- /dev/null +++ b/lib/librte_eal/common/include/arc
[dpdk-dev] [PATCH v3 11/12] tile: Add TILE-Gx mPIPE poll mode driver.
This commit adds a poll mode driver for the mPIPE hardware present on TILE-Gx SoCs. Change-Id: I1b9a9ef2c9f1c96810ec58b4d2ae77b870a6ec94 Signed-off-by: Zhigang Lu --- config/defconfig_tile-tilegx-linuxapp-gcc |1 + drivers/net/Makefile |1 + drivers/net/mpipe/Makefile| 46 + drivers/net/mpipe/mpipe_tilegx.c | 1637 + mk/rte.app.mk |1 + 5 files changed, 1686 insertions(+) create mode 100644 drivers/net/mpipe/Makefile create mode 100644 drivers/net/mpipe/mpipe_tilegx.c diff --git a/config/defconfig_tile-tilegx-linuxapp-gcc b/config/defconfig_tile-tilegx-linuxapp-gcc index 4023878..e603d1b 100644 --- a/config/defconfig_tile-tilegx-linuxapp-gcc +++ b/config/defconfig_tile-tilegx-linuxapp-gcc @@ -66,4 +66,5 @@ CONFIG_RTE_LIBRTE_TABLE=n CONFIG_RTE_LIBRTE_PIPELINE=n # Enable and override things that we need +CONFIG_RTE_LIBRTE_MPIPE_PMD=y CONFIG_RTE_MEMPOOL_ALIGN=128 diff --git a/drivers/net/Makefile b/drivers/net/Makefile index 644cacb..ee77480 100644 --- a/drivers/net/Makefile +++ b/drivers/net/Makefile @@ -36,6 +36,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_BOND) += bonding DIRS-$(CONFIG_RTE_LIBRTE_CXGBE_PMD) += cxgbe DIRS-$(CONFIG_RTE_LIBRTE_E1000_PMD) += e1000 DIRS-$(CONFIG_RTE_LIBRTE_ENIC_PMD) += enic +DIRS-$(CONFIG_RTE_LIBRTE_MPIPE_PMD) += mpipe DIRS-$(CONFIG_RTE_LIBRTE_FM10K_PMD) += fm10k DIRS-$(CONFIG_RTE_LIBRTE_I40E_PMD) += i40e DIRS-$(CONFIG_RTE_LIBRTE_IXGBE_PMD) += ixgbe diff --git a/drivers/net/mpipe/Makefile b/drivers/net/mpipe/Makefile new file mode 100644 index 000..552b303 --- /dev/null +++ b/drivers/net/mpipe/Makefile @@ -0,0 +1,46 @@ +# +# Copyright 2015 EZchip Semiconductor Ltd. All rights reserved. +# +# Redistribution and use in source and binary forms, with or without +# modification, are permitted provided that the following conditions +# are met: +# +# 1. Redistributions of source code must retain the above copyright +# notice, this list of conditions and the following disclaimer. +# +# 2. 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. +# +# 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 HOLDER 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 $(RTE_SDK)/mk/rte.vars.mk + +# library name +LIB = librte_pmd_mpipe.a + +CFLAGS += $(WERROR_FLAGS) -O3 + +EXPORT_MAP := rte_pmd_mpipe_version.map + +LIBABIVER := 1 + +SRCS-$(CONFIG_RTE_LIBRTE_MPIPE_PMD) += mpipe_tilegx.c + +DEPDIRS-$(CONFIG_RTE_LIBRTE_MPIPE_PMD) += lib/librte_eal lib/librte_ether +DEPDIRS-$(CONFIG_RTE_LIBRTE_MPIPE_PMD) += lib/librte_mempool lib/librte_mbuf +DEPDIRS-$(CONFIG_RTE_LIBRTE_MPIPE_PMD) += lib/librte_net lib/librte_malloc + +include $(RTE_SDK)/mk/rte.lib.mk diff --git a/drivers/net/mpipe/mpipe_tilegx.c b/drivers/net/mpipe/mpipe_tilegx.c new file mode 100644 index 000..e222443 --- /dev/null +++ b/drivers/net/mpipe/mpipe_tilegx.c @@ -0,0 +1,1637 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2015 EZchip Semiconductor Ltd. 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 EZchip Semiconductor 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
[dpdk-dev] [PATCH v3 12/12] maintainers: claim responsibility for TILE-Gx platform
Change-Id: I6491108ff86c1249bf4ffa4d4624c01b4594805e Signed-off-by: Zhigang Lu --- MAINTAINERS | 4 1 file changed, 4 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 5476a73..6ffa01b 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -114,6 +114,10 @@ M: Bruce Richardson M: Konstantin Ananyev F: lib/librte_eal/common/include/arch/x86/ +EZchip TILE-Gx +M: Zhigang Lu +F: lib/librte_eal/common/include/arch/tile/ + Linux EAL (with overlaps) M: David Marchand F: lib/librte_eal/linuxapp/Makefile -- 2.1.2
[dpdk-dev] Using rte_ring_mp_xyz() across EAL and non-EAL threads ?
On Fri, Jul 03, 2015 at 08:13:47PM -0700, Gopakumar Choorakkot Edakkunni wrote: > Thanks for the clarification Bruce. But I find this link below in the > documentation which says it should not be used in cases where the > scheduling policy is SCHED_RR because guess it can lead to an endless > spin-lock-wait by the SCHED_RR thread waiting for the lower prio > thread which got pre-empted. I am assuming this is a very unlikely > situation 32 core Xeon like cpu where a low prio thread never gets > scheduled because there are always realtime threads with work to do ? > But since the warning is in bold and pretty loud :), I decided not to > use that and just do a handoff with a standard mutex-lock :(. If my > threads were not realtime (not SCHED_RR), I could have used this! > > http://dpdk.org/doc/guides/prog_guide/env_abstraction_layer.html section > 3.3.4. > > "The ?non-preemptive? constraint means: Bypassing this constraint it > may cause the 2nd pthread to spin until the 1st one is scheduled > again. Moreover, if the 1st pthread is preempted by a context that has > an higher priority, it may even cause a dead lock." > > > Rgds, > Gopa. Yes, the note is indeed correct. As a general statement, rte_rings should never be used to pass data between two threads on the same core. At worst, it can cause deadlock, at best it can just lead to lots of wasted time as one thread uses its entire scheduling quantum just waiting for the other thread which it is blocking. /Bruce > > On Thu, Jul 2, 2015 at 2:20 AM, Bruce Richardson > wrote: > > On Wed, Jul 01, 2015 at 10:50:49AM -0700, Gopakumar Choorakkot Edakkunni > > wrote: > >> rte_ring_create() needs a socket-id though and seems to be allocating > >> core-specific memory pools for the ring ? But my non-EAL app thread is > >> not bound to any core, so now I am wondering if that will work. > >> > >> Rgds, > >> Gopa. > > > > There are no core-specific elements for rte_rings, just for mempools. Yes, > > you > > need a NUMA node ID when creating the ring, so that DPDK knows where to > > allocate > > the memory for it. However, once that is done, the ring can safely be used > > from > > both EAL and non-EAL threads. There is no requirement to have an lcore-id > > for > > the thread. > > > > /Bruce > > > >> > >> On Wed, Jul 1, 2015 at 10:46 AM, Gopakumar Choorakkot Edakkunni > >> wrote: > >> > Hi, > >> > > >> > I have a requirement where one of my non-EAL app threads needs to > >> > handoff some packets to an EAL task. I was thinking of using > >> > rte_ring_mp_enqueue/dequeue for that purpose. I looked at the code for > >> > the rte_ring library and it doesnt look like it has any "EAL" > >> > dependencies, but I wanted to double confirm that there are no issues > >> > in using it that way. Dint find much yes/no info about that on the > >> > mailers/docs. Pls let me know your thoughts. > >> > > >> > Rgds, > >> > Gopa.
[dpdk-dev] [PATCH 1/2] ixgbe: add "cold" attribute to setup/teardown fns
On Fri, Jul 03, 2015 at 09:57:26PM +0200, Thomas Monjalon wrote: > 2015-07-03 16:56, Bruce Richardson: > > On Fri, Jul 03, 2015 at 05:45:34PM +0200, Thomas Monjalon wrote: > > > Hi Bruce, > > > > > > 2015-07-03 16:40, Bruce Richardson: > > > > As well as the fast-path functions in the rxtx code, there are also > > > > functions which set up and tear down the descriptor rings. Since these > > > > are not performance critical functions, there is no need to have them > > > > extensively optimized, so we add __attribute__((cold)) to their > > > > definitions. This has the side-effect of making debugging them easier as > > > > the compiler does not optimize them as heavily, so more variables are > > > > accessible by default in gdb. > > > > > > What is the benefit, compared to -O0? > > > > First off, it's per function, rather than having to use -O0 globally. > > Secondly, > > it doesn't disable optimization, it just tells the compiler that the code is > > not on the hotpath - whether or not the compiler optimizes it is up to the > > compiler itself. From GCC documentation: > > https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#Common-Function-Attributes > > > > "The cold attribute on functions is used to inform the compiler that the > > function is unlikely to be executed. The function is optimized for size > > rather > > than speed and on many targets it is placed into a special subsection of > > the > > text section so all cold functions appear close together, improving code > > locality of non-cold parts of program. The paths leading to calls of cold > > functions within code are marked as unlikely by the branch prediction > > mechanism. > > It is thus useful to mark functions used to handle unlikely conditions, > > such as > > perror, as cold to improve optimization of hot functions that do call marked > > functions in rare occasions." > > I know it may provide some optimization of the hot path. > I was asking compared to -O0 because you were justifying this change for > debug. > In other words, for debugging, -O0 is probably better. So the reason of this > change should be the optimization. And it would be interesting to know if you > have seen some performance improvement. For some cases, O0 will be necessary, but the advantage of this change is that for debugging of code that is not in the fast-path, the use of -O0 may be unnecessary - which is useful, since you don't always need to do a special debug build. As for performance impact: no, I have not seen any performance impact from this change. Personally, I view this as a low impact change that doesn't really have any negatives. Is there some concern in particular you have about it? It's really just providing some extra hints to the compiler. /Bruce
[dpdk-dev] [PATCH 1/2] ixgbe: add "cold" attribute to setup/teardown fns
2015-07-06 10:20, Bruce Richardson: > On Fri, Jul 03, 2015 at 09:57:26PM +0200, Thomas Monjalon wrote: > > 2015-07-03 16:56, Bruce Richardson: > > > On Fri, Jul 03, 2015 at 05:45:34PM +0200, Thomas Monjalon wrote: > > > > Hi Bruce, > > > > > > > > 2015-07-03 16:40, Bruce Richardson: > > > > > As well as the fast-path functions in the rxtx code, there are also > > > > > functions which set up and tear down the descriptor rings. Since these > > > > > are not performance critical functions, there is no need to have them > > > > > extensively optimized, so we add __attribute__((cold)) to their > > > > > definitions. This has the side-effect of making debugging them easier > > > > > as > > > > > the compiler does not optimize them as heavily, so more variables are > > > > > accessible by default in gdb. > > > > > > > > What is the benefit, compared to -O0? > > > > > > First off, it's per function, rather than having to use -O0 globally. > > > Secondly, > > > it doesn't disable optimization, it just tells the compiler that the code > > > is > > > not on the hotpath - whether or not the compiler optimizes it is up to > > > the > > > compiler itself. From GCC documentation: > > > https://gcc.gnu.org/onlinedocs/gcc/Common-Function-Attributes.html#Common-Function-Attributes > > > > > > "The cold attribute on functions is used to inform the compiler that the > > > function is unlikely to be executed. The function is optimized for size > > > rather > > > than speed and on many targets it is placed into a special subsection of > > > the > > > text section so all cold functions appear close together, improving code > > > locality of non-cold parts of program. The paths leading to calls of cold > > > functions within code are marked as unlikely by the branch prediction > > > mechanism. > > > It is thus useful to mark functions used to handle unlikely conditions, > > > such as > > > perror, as cold to improve optimization of hot functions that do call > > > marked > > > functions in rare occasions." > > > > I know it may provide some optimization of the hot path. > > I was asking compared to -O0 because you were justifying this change for > > debug. > > In other words, for debugging, -O0 is probably better. So the reason of this > > change should be the optimization. And it would be interesting to know if > > you > > have seen some performance improvement. > > For some cases, O0 will be necessary, but the advantage of this change is that > for debugging of code that is not in the fast-path, the use of -O0 may be > unnecessary - which is useful, since you don't always need to do a special > debug > build. > > As for performance impact: no, I have not seen any performance impact from > this > change. Personally, I view this as a low impact change that doesn't really > have > any negatives. Is there some concern in particular you have about it? It's > really > just providing some extra hints to the compiler. No concern. I was only interested to fully understand why you made this change. Thanks
[dpdk-dev] [PATCH v5 00/11] ip_pipeline: ip_pipeline application enhancements
This patchset enhances functionality of ip_pipeline application. New config file syntax is introduced, so parser is changed. Changed structure of the application. Now every global variable is stored in app_struct in app.h. Syntax of pipeline cli commands was changed. Implementation of cli commands for every pipeline is moved to the separate file. Changes in v2: - renamed some files - added more config files - reworked flow classification pipeline implementation - fixed some bugs Changes in v3: - fixed checkpatch errors - fixed bug with message queues - fixed bug with application log Changes in v4: - fixed build issue with gcc 5 - fixed bugs in flow classification and firewall pipelines Changes in v5: - fixed build issue with clang 3.6 Daniel Mrzyglod (1): ip_pipeline: added new implementation of firewall pipeline Jasvinder Singh (3): ip_pipeline: added config checks ip_pipeline: added master pipeline ip_pipeline: added new implementation of passthrough pipeline Maciej Gajdzica (6): ip_pipeline: modified init to match new params struct ip_pipeline: moved pipelines to separate folder ip_pipeline: added application thread ip_pipeline: moved config files to separate folder ip_pipeline: added new implementation of routing pipeline ip_pipeline: added new implementation of flow classification pipeline Pawel Wodkowski (1): ip_pipeline: add parsing for config files with new syntax examples/ip_pipeline/Makefile | 36 +- examples/ip_pipeline/app.h | 905 examples/ip_pipeline/cmdline.c | 1976 examples/ip_pipeline/config.c | 419 examples/ip_pipeline/config/ip_pipeline.cfg|9 + examples/ip_pipeline/config/ip_pipeline.sh |5 + examples/ip_pipeline/config/tm_profile.cfg | 105 + examples/ip_pipeline/config_check.c| 397 examples/ip_pipeline/config_parse.c| 2456 examples/ip_pipeline/config_parse_tm.c | 446 examples/ip_pipeline/cpu_core_map.c| 492 examples/ip_pipeline/cpu_core_map.h| 69 + examples/ip_pipeline/init.c| 1612 + examples/ip_pipeline/ip_pipeline.cfg | 56 - examples/ip_pipeline/ip_pipeline.sh| 18 - examples/ip_pipeline/main.c| 137 +- examples/ip_pipeline/main.h| 298 --- examples/ip_pipeline/pipeline.h| 87 + examples/ip_pipeline/pipeline/hash_func.h | 351 +++ .../ip_pipeline/pipeline/pipeline_actions_common.h | 119 + examples/ip_pipeline/pipeline/pipeline_common_be.c | 206 ++ examples/ip_pipeline/pipeline/pipeline_common_be.h | 163 ++ examples/ip_pipeline/pipeline/pipeline_common_fe.c | 1324 +++ examples/ip_pipeline/pipeline/pipeline_common_fe.h | 228 ++ examples/ip_pipeline/pipeline/pipeline_firewall.c | 1001 examples/ip_pipeline/pipeline/pipeline_firewall.h | 63 + .../ip_pipeline/pipeline/pipeline_firewall_be.c| 740 ++ .../ip_pipeline/pipeline/pipeline_firewall_be.h| 138 ++ .../pipeline/pipeline_flow_classification.c| 2057 .../pipeline/pipeline_flow_classification.h| 105 + .../pipeline/pipeline_flow_classification_be.c | 589 + .../pipeline/pipeline_flow_classification_be.h | 140 ++ examples/ip_pipeline/pipeline/pipeline_master.c| 47 + examples/ip_pipeline/pipeline/pipeline_master.h| 41 + examples/ip_pipeline/pipeline/pipeline_master_be.c | 150 ++ examples/ip_pipeline/pipeline/pipeline_master_be.h | 41 + .../ip_pipeline/pipeline/pipeline_passthrough.c| 47 + .../ip_pipeline/pipeline/pipeline_passthrough.h| 41 + .../ip_pipeline/pipeline/pipeline_passthrough_be.c | 772 ++ .../ip_pipeline/pipeline/pipeline_passthrough_be.h | 41 + examples/ip_pipeline/pipeline/pipeline_routing.c | 1541 examples/ip_pipeline/pipeline/pipeline_routing.h | 99 + .../ip_pipeline/pipeline/pipeline_routing_be.c | 869 +++ .../ip_pipeline/pipeline/pipeline_routing_be.h | 230 ++ examples/ip_pipeline/pipeline_be.h | 256 ++ examples/ip_pipeline/pipeline_firewall.c | 313 --- .../ip_pipeline/pipeline_flow_classification.c | 306 --- examples/ip_pipeline/pipeline_ipv4_frag.c | 184 -- examples/ip_pipeline/pipeline_ipv4_ras.c | 181 -- examples/ip_pipeline/pipeline_passthrough.c| 213 -- examples/ip_pipeline/pipeline_routing.c| 474 examples/ip_pipeline/pipeline_rx.c | 385 --- examples/ip_pipeline/pipeline_tx.c | 283 --- examples/ip_pipeline/thread.c | 110 + 54 files changed, 17702 insertions(+), 5669 deletions(-) create mode 100644 examples/ip_pipeline/app.h delete mode
[dpdk-dev] [PATCH v5 01/11] ip_pipeline: add parsing for config files with new syntax
From: Pawel Wodkowski
New syntax of config files is needed for ip_pipeline example
enhancements. Some old files are temporarily disabled in the Makefile.
It is part of a bigger change.
Signed-off-by: Pawel Wodkowski
---
examples/ip_pipeline/Makefile | 17 +-
examples/ip_pipeline/app.h | 905
examples/ip_pipeline/config.c | 419 --
examples/ip_pipeline/config_parse.c| 2456
examples/ip_pipeline/config_parse_tm.c | 446 ++
examples/ip_pipeline/cpu_core_map.c| 492 +++
examples/ip_pipeline/cpu_core_map.h| 69 +
examples/ip_pipeline/main.c| 130 +-
examples/ip_pipeline/main.h| 298
examples/ip_pipeline/pipeline.h| 87 ++
examples/ip_pipeline/pipeline_be.h | 256
11 files changed, 4722 insertions(+), 853 deletions(-)
create mode 100644 examples/ip_pipeline/app.h
delete mode 100644 examples/ip_pipeline/config.c
create mode 100644 examples/ip_pipeline/config_parse.c
create mode 100644 examples/ip_pipeline/config_parse_tm.c
create mode 100644 examples/ip_pipeline/cpu_core_map.c
create mode 100644 examples/ip_pipeline/cpu_core_map.h
delete mode 100644 examples/ip_pipeline/main.h
create mode 100644 examples/ip_pipeline/pipeline.h
create mode 100644 examples/ip_pipeline/pipeline_be.h
diff --git a/examples/ip_pipeline/Makefile b/examples/ip_pipeline/Makefile
index e70fdc7..b0feb4f 100644
--- a/examples/ip_pipeline/Makefile
+++ b/examples/ip_pipeline/Makefile
@@ -43,20 +43,9 @@ APP = ip_pipeline
# all source are stored in SRCS-y
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) := main.c
-SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config.c
-SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += init.c
-SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += cmdline.c
-SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_rx.c
-SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_tx.c
-SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_flow_classification.c
-SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_routing.c
-SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_passthrough.c
-SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_ipv4_frag.c
-SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_ipv4_ras.c
-
-ifeq ($(CONFIG_RTE_LIBRTE_ACL),y)
-SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_firewall.c
-endif
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_parse.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_parse_tm.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += cpu_core_map.c
CFLAGS += -O3
CFLAGS += $(WERROR_FLAGS)
diff --git a/examples/ip_pipeline/app.h b/examples/ip_pipeline/app.h
new file mode 100644
index 000..112473a
--- /dev/null
+++ b/examples/ip_pipeline/app.h
@@ -0,0 +1,905 @@
+/*-
+ * BSD LICENSE
+ *
+ * Copyright(c) 2010-2015 Intel Corporation. All rights reserved.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * * Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * * Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in
+ * the documentation and/or other materials provided with the
+ * distribution.
+ * * Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef __INCLUDE_APP_H__
+#define __INCLUDE_APP_H__
+
+#include
+#include
+
+#include
+#include
+#include
+#include
+#include
+
+#include
+
+#include "cpu_core_map.h"
+#include "pipeline.h"
+
+#define APP_PARAM_NAME_SIZE PIPELINE_NAME_SIZE
+
+struct app_mempool_params {
+ char *name;
+ uint32_t parsed;
+ uint32_t buffer_size;
+ uint32_t pool_size;
+ uint32_t cache_size;
+ uint32_t cpu_socket_id;
+};
+
+struct app_link_params {
+ char *name;
+ uint32_t parsed;
+ uint32_t pmd
[dpdk-dev] [PATCH v5 02/11] ip_pipeline: added config checks
From: Jasvinder Singh
After loading configuration from a file, data integrity is checked.
Signed-off-by: Jasvinder Singh
---
examples/ip_pipeline/Makefile |1 +
examples/ip_pipeline/config_check.c | 397 +++
examples/ip_pipeline/main.c |2 +
3 files changed, 400 insertions(+)
create mode 100644 examples/ip_pipeline/config_check.c
diff --git a/examples/ip_pipeline/Makefile b/examples/ip_pipeline/Makefile
index b0feb4f..bc50e71 100644
--- a/examples/ip_pipeline/Makefile
+++ b/examples/ip_pipeline/Makefile
@@ -45,6 +45,7 @@ APP = ip_pipeline
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) := main.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_parse.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_parse_tm.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_check.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += cpu_core_map.c
CFLAGS += -O3
diff --git a/examples/ip_pipeline/config_check.c
b/examples/ip_pipeline/config_check.c
new file mode 100644
index 000..b5fb7d8
--- /dev/null
+++ b/examples/ip_pipeline/config_check.c
@@ -0,0 +1,397 @@
+/*-
+ * BSD LICENSE
+ *
+ * Copyright(c) 2010-2015 Intel Corporation. All rights reserved.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * * Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * * Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in
+ * the documentation and/or other materials provided with the
+ * distribution.
+ * * Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include
+
+#include "app.h"
+
+static void
+check_mempools(struct app_params *app)
+{
+ uint32_t i;
+
+ for (i = 0; i < app->n_mempools; i++) {
+ struct app_mempool_params *p = &app->mempool_params[i];
+
+ APP_CHECK((p->pool_size > 0),
+ "Mempool %s size is 0\n", p->name);
+
+ APP_CHECK((p->cache_size > 0),
+ "Mempool %s cache size is 0\n", p->name);
+
+ APP_CHECK(rte_is_power_of_2(p->cache_size),
+ "Mempool %s cache size not a power of 2\n", p->name);
+ }
+}
+
+static void
+check_links(struct app_params *app)
+{
+ uint32_t n_links_port_mask = __builtin_popcountll(app->port_mask);
+ uint32_t i;
+
+ /* Check that number of links matches the port mask */
+ APP_CHECK((app->n_links == n_links_port_mask),
+ "Not enough links provided in the PORT_MASK\n");
+
+ for (i = 0; i < app->n_links; i++) {
+ struct app_link_params *link = &app->link_params[i];
+ uint32_t rxq_max, n_rxq, n_txq, link_id, i;
+
+ APP_PARAM_GET_ID(link, "LINK", link_id);
+
+ /* Check that link RXQs are contiguous */
+ rxq_max = 0;
+ if (link->arp_q > rxq_max)
+ rxq_max = link->arp_q;
+ if (link->tcp_syn_local_q > rxq_max)
+ rxq_max = link->tcp_syn_local_q;
+ if (link->ip_local_q > rxq_max)
+ rxq_max = link->ip_local_q;
+ if (link->tcp_local_q > rxq_max)
+ rxq_max = link->tcp_local_q;
+ if (link->udp_local_q > rxq_max)
+ rxq_max = link->udp_local_q;
+ if (link->sctp_local_q > rxq_max)
+ rxq_max = link->sctp_local_q;
+
+ for (i = 1; i <= rxq_max; i++)
+ APP_CHECK(((link->arp_q == i) ||
+ (link->tcp_syn_local_q == i) ||
+ (link->ip_local_q == i) ||
+ (link->tcp_local_q == i) ||
+
[dpdk-dev] [PATCH v5 03/11] ip_pipeline: modified init to match new params struct
After changes in config parser, app params struct is changed and
requires modifications in initialization procedures.
Signed-off-by: Maciej Gajdzica
---
examples/ip_pipeline/Makefile |1 +
examples/ip_pipeline/init.c | 1599 ++---
examples/ip_pipeline/main.c |3 +
3 files changed, 1172 insertions(+), 431 deletions(-)
diff --git a/examples/ip_pipeline/Makefile b/examples/ip_pipeline/Makefile
index bc50e71..59bea5b 100644
--- a/examples/ip_pipeline/Makefile
+++ b/examples/ip_pipeline/Makefile
@@ -46,6 +46,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) := main.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_parse.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_parse_tm.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_check.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += init.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += cpu_core_map.c
CFLAGS += -O3
diff --git a/examples/ip_pipeline/init.c b/examples/ip_pipeline/init.c
index d79762f..3bf446f 100644
--- a/examples/ip_pipeline/init.c
+++ b/examples/ip_pipeline/init.c
@@ -1,7 +1,7 @@
/*-
* BSD LICENSE
*
- * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ * Copyright(c) 2010-2015 Intel Corporation. All rights reserved.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -32,561 +32,1298 @@
*/
#include
-#include
-#include
-#include
-#include
#include
-#include
-#include
-#include
-#include
-
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
+
#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
#include
-#include
-#include
-#include
-#include
-#include
+#include
#include
-#include
-#include
-#include
-
-#include "main.h"
-
-#define NA APP_SWQ_INVALID
-
-struct app_params app = {
- /* CPU cores */
- .cores = {
- {0, APP_CORE_MASTER, {15, 16, 17, NA, NA, NA, NA, NA},
- {12, 13, 14, NA, NA, NA, NA, NA} },
- {0, APP_CORE_RX, {NA, NA, NA, NA, NA, NA, NA, 12},
- { 0, 1, 2, 3, NA, NA, NA, 15} },
- {0, APP_CORE_FC, { 0, 1, 2, 3, NA, NA, NA, 13},
- { 4, 5, 6, 7, NA, NA, NA, 16} },
- {0, APP_CORE_RT, { 4, 5, 6, 7, NA, NA, NA, 14},
- { 8, 9, 10, 11, NA, NA, NA, 17} },
- {0, APP_CORE_TX, { 8, 9, 10, 11, NA, NA, NA, NA},
- {NA, NA, NA, NA, NA, NA, NA, NA} },
- },
-
- /* Ports*/
- .n_ports = APP_MAX_PORTS,
- .rsz_hwq_rx = 128,
- .rsz_hwq_tx = 512,
- .bsz_hwq_rd = 64,
- .bsz_hwq_wr = 64,
-
- .port_conf = {
- .rxmode = {
- .split_hdr_size = 0,
- .header_split = 0, /* Header Split disabled */
- .hw_ip_checksum = 1, /* IP checksum offload enabled */
- .hw_vlan_filter = 0, /* VLAN filtering disabled */
- .jumbo_frame= 1, /* Jumbo Frame Support enabled */
- .max_rx_pkt_len = 9000, /* Jumbo Frame MAC pkt length */
- .hw_strip_crc = 0, /* CRC stripped by hardware */
- },
- .rx_adv_conf = {
- .rss_conf = {
- .rss_key = NULL,
- .rss_hf = ETH_RSS_IP,
- },
- },
- .txmode = {
- .mq_mode = ETH_MQ_TX_NONE,
- },
- },
-
- .rx_conf = {
- .rx_thresh = {
- .pthresh = 8,
- .hthresh = 8,
- .wthresh = 4,
- },
- .rx_free_thresh = 64,
- .rx_drop_en = 0,
- },
-
- .tx_conf = {
- .tx_thresh = {
- .pthresh = 36,
- .hthresh = 0,
- .wthresh = 0,
- },
- .tx_free_thresh = 0,
- .tx_rs_thresh = 0,
- },
-
- /* SWQs */
- .rsz_swq = 128,
- .bsz_swq_rd = 64,
- .bsz_swq_wr = 64,
-
- /* Buffer pool */
- .pool_buffer_size = RTE_MBUF_DEFAULT_BUF_SIZE,
- .pool_size = 32 * 1024,
- .pool_cache_size = 256,
-
- /* Message buffer pool */
- .msg_pool_buffer_size = 256,
- .msg_pool_size = 1024,
- .msg_pool_cache_size = 64,
-
- /* Rule tables */
- .max_arp_rules = 1 << 10,
- .max_firewall_rules = 1 << 5,
- .max_routing_rules = 1 << 24,
- .max_flow_rules = 1 << 24,
-
- /* Application processing */
- .ether_hdr_pop_push = 0,
-};
-
-struct app_core_params *
-app_get_core_params(uint32_t core_id)
-{
- uint32_t i;
+#include
+#include
- for (i = 0; i < RTE_MAX_LCORE; i++) {
- struct app_core
[dpdk-dev] [PATCH v5 04/11] ip_pipeline: moved pipelines to separate folder
Moved pipelines to separate folder, removed not needed pipelines and modified Makefile to match that change. Signed-off-by: Maciej Gajdzica --- examples/ip_pipeline/Makefile |9 +- examples/ip_pipeline/pipeline/pipeline_firewall.c | 313 + .../pipeline/pipeline_flow_classification.c| 306 + .../ip_pipeline/pipeline/pipeline_passthrough.c| 213 + examples/ip_pipeline/pipeline/pipeline_routing.c | 474 examples/ip_pipeline/pipeline_firewall.c | 313 - .../ip_pipeline/pipeline_flow_classification.c | 306 - examples/ip_pipeline/pipeline_ipv4_frag.c | 184 examples/ip_pipeline/pipeline_ipv4_ras.c | 181 examples/ip_pipeline/pipeline_passthrough.c| 213 - examples/ip_pipeline/pipeline_routing.c| 474 examples/ip_pipeline/pipeline_rx.c | 385 examples/ip_pipeline/pipeline_tx.c | 283 13 files changed, 1314 insertions(+), 2340 deletions(-) create mode 100644 examples/ip_pipeline/pipeline/pipeline_firewall.c create mode 100644 examples/ip_pipeline/pipeline/pipeline_flow_classification.c create mode 100644 examples/ip_pipeline/pipeline/pipeline_passthrough.c create mode 100644 examples/ip_pipeline/pipeline/pipeline_routing.c delete mode 100644 examples/ip_pipeline/pipeline_firewall.c delete mode 100644 examples/ip_pipeline/pipeline_flow_classification.c delete mode 100644 examples/ip_pipeline/pipeline_ipv4_frag.c delete mode 100644 examples/ip_pipeline/pipeline_ipv4_ras.c delete mode 100644 examples/ip_pipeline/pipeline_passthrough.c delete mode 100644 examples/ip_pipeline/pipeline_routing.c delete mode 100644 examples/ip_pipeline/pipeline_rx.c delete mode 100644 examples/ip_pipeline/pipeline_tx.c diff --git a/examples/ip_pipeline/Makefile b/examples/ip_pipeline/Makefile index 59bea5b..213e879 100644 --- a/examples/ip_pipeline/Makefile +++ b/examples/ip_pipeline/Makefile @@ -36,11 +36,17 @@ endif # Default target, can be overridden by command line or environment RTE_TARGET ?= x86_64-native-linuxapp-gcc +DIRS-(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline + include $(RTE_SDK)/mk/rte.vars.mk # binary name APP = ip_pipeline +VPATH += $(SRCDIR)/pipeline + +INC += $(wildcard *.h) $(wildcard pipeline/*.h) + # all source are stored in SRCS-y SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) := main.c SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_parse.c @@ -49,7 +55,8 @@ SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_check.c SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += init.c SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += cpu_core_map.c +CFLAGS += -I$(SRCDIR) -I$(SRCDIR)/pipeline CFLAGS += -O3 -CFLAGS += $(WERROR_FLAGS) +CFLAGS += $(WERROR_FLAGS) -Wno-error=unused-function -Wno-error=unused-variable include $(RTE_SDK)/mk/rte.extapp.mk diff --git a/examples/ip_pipeline/pipeline/pipeline_firewall.c b/examples/ip_pipeline/pipeline/pipeline_firewall.c new file mode 100644 index 000..b70260e --- /dev/null +++ b/examples/ip_pipeline/pipeline/pipeline_firewall.c @@ -0,0 +1,313 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include +#include +#include + +#include +#include +#include +#inc
[dpdk-dev] [PATCH v5 05/11] ip_pipeline: added master pipeline
From: Jasvinder Singh
Master pipeline is responsible for command line handling and
communicationg with all other pipelines via message queues. Removed
cmdline.c file, as its functionality will be split over multiple
pipeline files.
Signed-off-by: Jasvinder Singh
---
examples/ip_pipeline/Makefile |5 +
examples/ip_pipeline/cmdline.c | 1976
examples/ip_pipeline/init.c|5 +
examples/ip_pipeline/pipeline/pipeline_common_be.c | 206 ++
examples/ip_pipeline/pipeline/pipeline_common_be.h | 163 ++
examples/ip_pipeline/pipeline/pipeline_common_fe.c | 1324 +
examples/ip_pipeline/pipeline/pipeline_common_fe.h | 228 +++
examples/ip_pipeline/pipeline/pipeline_master.c| 47 +
examples/ip_pipeline/pipeline/pipeline_master.h| 41 +
examples/ip_pipeline/pipeline/pipeline_master_be.c | 150 ++
examples/ip_pipeline/pipeline/pipeline_master_be.h | 41 +
11 files changed, 2210 insertions(+), 1976 deletions(-)
delete mode 100644 examples/ip_pipeline/cmdline.c
create mode 100644 examples/ip_pipeline/pipeline/pipeline_common_be.c
create mode 100644 examples/ip_pipeline/pipeline/pipeline_common_be.h
create mode 100644 examples/ip_pipeline/pipeline/pipeline_common_fe.c
create mode 100644 examples/ip_pipeline/pipeline/pipeline_common_fe.h
create mode 100644 examples/ip_pipeline/pipeline/pipeline_master.c
create mode 100644 examples/ip_pipeline/pipeline/pipeline_master.h
create mode 100644 examples/ip_pipeline/pipeline/pipeline_master_be.c
create mode 100644 examples/ip_pipeline/pipeline/pipeline_master_be.h
diff --git a/examples/ip_pipeline/Makefile b/examples/ip_pipeline/Makefile
index 213e879..9ce80a8 100644
--- a/examples/ip_pipeline/Makefile
+++ b/examples/ip_pipeline/Makefile
@@ -55,6 +55,11 @@ SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_check.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += init.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += cpu_core_map.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_common_be.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_common_fe.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_master_be.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_master.c
+
CFLAGS += -I$(SRCDIR) -I$(SRCDIR)/pipeline
CFLAGS += -O3
CFLAGS += $(WERROR_FLAGS) -Wno-error=unused-function -Wno-error=unused-variable
diff --git a/examples/ip_pipeline/cmdline.c b/examples/ip_pipeline/cmdline.c
deleted file mode 100644
index 3173fd0..000
--- a/examples/ip_pipeline/cmdline.c
+++ /dev/null
@@ -1,1976 +0,0 @@
-/*-
- * BSD LICENSE
- *
- * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
- * All rights reserved.
- *
- * Redistribution and use in source and binary forms, with or without
- * modification, are permitted provided that the following conditions
- * are met:
- *
- * * Redistributions of source code must retain the above copyright
- * notice, this list of conditions and the following disclaimer.
- * * Redistributions in binary form must reproduce the above copyright
- * notice, this list of conditions and the following disclaimer in
- * the documentation and/or other materials provided with the
- * distribution.
- * * Neither the name of Intel Corporation nor the names of its
- * contributors may be used to endorse or promote products derived
- * from this software without specific prior written permission.
- *
- * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
- * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
- * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
- * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
- * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
- * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
- * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
- */
-
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-#include
-
-#include "main.h"
-
-#define IS_RULE_PRESENT(res, rule_key, table, type)\
-do { \
- struct app_rule *it;\
- \
- (res) = NULL; \
- TAILQ_FOREACH(it, &table, entries) {\
[dpdk-dev] [PATCH v5 06/11] ip_pipeline: added application thread
Application thread runs pipelines on assigned cores.
Signed-off-by: Maciej Gajdzica
---
examples/ip_pipeline/Makefile |1 +
examples/ip_pipeline/main.c |6 +++
examples/ip_pipeline/thread.c | 110 +
3 files changed, 117 insertions(+)
create mode 100644 examples/ip_pipeline/thread.c
diff --git a/examples/ip_pipeline/Makefile b/examples/ip_pipeline/Makefile
index 9ce80a8..f255338 100644
--- a/examples/ip_pipeline/Makefile
+++ b/examples/ip_pipeline/Makefile
@@ -53,6 +53,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_parse.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_parse_tm.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += config_check.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += init.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += thread.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += cpu_core_map.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_common_be.c
diff --git a/examples/ip_pipeline/main.c b/examples/ip_pipeline/main.c
index ef68c86..862e2f2 100644
--- a/examples/ip_pipeline/main.c
+++ b/examples/ip_pipeline/main.c
@@ -52,5 +52,11 @@ main(int argc, char **argv)
/* Init */
app_init(&app);
+ /* Run-time */
+ rte_eal_mp_remote_launch(
+ app_thread,
+ (void *) &app,
+ CALL_MASTER);
+
return 0;
}
diff --git a/examples/ip_pipeline/thread.c b/examples/ip_pipeline/thread.c
new file mode 100644
index 000..b2a8656
--- /dev/null
+++ b/examples/ip_pipeline/thread.c
@@ -0,0 +1,110 @@
+/*-
+ * BSD LICENSE
+ *
+ * Copyright(c) 2010-2015 Intel Corporation. All rights reserved.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * * Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * * Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in
+ * the documentation and/or other materials provided with the
+ * distribution.
+ * * Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include
+#include
+#include
+
+#include "pipeline_common_be.h"
+#include "app.h"
+
+int app_thread(void *arg)
+{
+ struct app_params *app = (struct app_params *) arg;
+ uint32_t core_id = rte_lcore_id(), i, j;
+ struct app_thread_data *t = &app->thread_data[core_id];
+ uint32_t n_regular = RTE_MIN(t->n_regular, RTE_DIM(t->regular));
+ uint32_t n_custom = RTE_MIN(t->n_custom, RTE_DIM(t->custom));
+
+ for (i = 0; ; i++) {
+ /* Run regular pipelines */
+ for (j = 0; j < n_regular; j++) {
+ struct app_thread_pipeline_data *data = &t->regular[j];
+ struct pipeline *p = data->be;
+
+ rte_pipeline_run(p->p);
+ }
+
+ /* Run custom pipelines */
+ for (j = 0; j < n_custom; j++) {
+ struct app_thread_pipeline_data *data = &t->custom[j];
+
+ data->f_run(data->be);
+ }
+
+ /* Timer */
+ if ((i & 0xF) == 0) {
+ uint64_t time = rte_get_tsc_cycles();
+ uint64_t t_deadline = UINT64_MAX;
+
+ if (time < t->deadline)
+ continue;
+
+ /* Timer for regular pipelines */
+ for (j = 0; j < n_regular; j++) {
+ struct app_thread_pipeline_data *data =
+ &t->regular[j];
+ uint64_t p_deadline = data->deadline;
+
+ if (p_deadline <= time) {
+ data->f_timer(data->be);
+
[dpdk-dev] [PATCH v5 07/11] ip_pipeline: moved config files to separate folder
Created new folder for config(.cfg) and script(.sh) files. Signed-off-by: Maciej Gajdzica --- examples/ip_pipeline/config/ip_pipeline.cfg |9 +++ examples/ip_pipeline/config/ip_pipeline.sh |5 ++ examples/ip_pipeline/config/tm_profile.cfg | 105 +++ examples/ip_pipeline/ip_pipeline.cfg| 56 -- examples/ip_pipeline/ip_pipeline.sh | 18 - 5 files changed, 119 insertions(+), 74 deletions(-) create mode 100644 examples/ip_pipeline/config/ip_pipeline.cfg create mode 100644 examples/ip_pipeline/config/ip_pipeline.sh create mode 100644 examples/ip_pipeline/config/tm_profile.cfg delete mode 100644 examples/ip_pipeline/ip_pipeline.cfg delete mode 100644 examples/ip_pipeline/ip_pipeline.sh diff --git a/examples/ip_pipeline/config/ip_pipeline.cfg b/examples/ip_pipeline/config/ip_pipeline.cfg new file mode 100644 index 000..095ed25 --- /dev/null +++ b/examples/ip_pipeline/config/ip_pipeline.cfg @@ -0,0 +1,9 @@ +[PIPELINE0] +type = MASTER +core = 0 + +[PIPELINE1] +type = PASS-THROUGH +core = 1 +pktq_in = RXQ0.0 RXQ1.0 RXQ2.0 RXQ3.0 +pktq_out = TXQ0.0 TXQ1.0 TXQ2.0 TXQ3.0 diff --git a/examples/ip_pipeline/config/ip_pipeline.sh b/examples/ip_pipeline/config/ip_pipeline.sh new file mode 100644 index 000..4fca259 --- /dev/null +++ b/examples/ip_pipeline/config/ip_pipeline.sh @@ -0,0 +1,5 @@ +# +#run config/ip_pipeline.sh +# + +p 1 ping diff --git a/examples/ip_pipeline/config/tm_profile.cfg b/examples/ip_pipeline/config/tm_profile.cfg new file mode 100644 index 000..53edb67 --- /dev/null +++ b/examples/ip_pipeline/config/tm_profile.cfg @@ -0,0 +1,105 @@ +; BSD LICENSE +; +; Copyright(c) 2010-2014 Intel Corporation. All rights reserved. +; All rights reserved. +; +; Redistribution and use in source and binary forms, with or without +; modification, are permitted provided that the following conditions +; are met: +; +; * Redistributions of source code must retain the above copyright +; notice, this list of conditions and the following disclaimer. +; * Redistributions in binary form must reproduce the above copyright +; notice, this list of conditions and the following disclaimer in +; the documentation and/or other materials provided with the +; distribution. +; * Neither the name of Intel Corporation nor the names of its +; contributors may be used to endorse or promote products derived +; from this software without specific prior written permission. +; +; THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +; "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +; LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +; A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +; OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +; SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +; LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +; DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +; THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +; (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +; OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +; This file enables the following hierarchical scheduler configuration for each +; 10GbE output port: +; * Single subport (subport 0): +; - Subport rate set to 100% of port rate +; - Each of the 4 traffic classes has rate set to 100% of port rate +; * 4K pipes per subport 0 (pipes 0 .. 4095) with identical configuration: +; - Pipe rate set to 1/4K of port rate +; - Each of the 4 traffic classes has rate set to 100% of pipe rate +; - Within each traffic class, the byte-level WRR weights for the 4 queues +; are set to 1:1:1:1 +; +; For more details, please refer to chapter "Quality of Service (QoS) Framework" +; of Intel Data Plane Development Kit (Intel DPDK) Programmer's Guide. + +; Port configuration +[port] +frame overhead = 24 ; frame overhead = Preamble (7) + SFD (1) + FCS (4) + IFG (12) +mtu = 1522; mtu = Q-in-Q MTU (FCS not included) +number of subports per port = 1 +number of pipes per subport = 4096 +queue sizes = 64 64 64 64 + +; Subport configuration +[subport 0] +tb rate = 125000 ; Bytes per second +tb size = 100 ; Bytes + +tc 0 rate = 125000 ; Bytes per second +tc 1 rate = 125000 ; Bytes per second +tc 2 rate = 125000 ; Bytes per second +tc 3 rate = 125000 ; Bytes per second +tc period = 10 ; Milliseconds + +pipe 0-4095 = 0; These pipes are configured with pipe profile 0 + +; Pipe configuration +[pipe profile 0] +tb rate = 305175 ; Bytes per second +tb size = 100 ; Bytes + +tc 0 rate = 30
[dpdk-dev] [PATCH v5 08/11] ip_pipeline: added new implementation of passthrough pipeline
From: Jasvinder Singh
Passthrough pipeline implementation is split to two files.
pipeline_passthrough.c file handles front-end functions (cli commands
parsing) pipeline_passthrough_ops.c contains implementation of functions
done by pipeline (back-end).
Signed-off-by: Jasvinder Singh
---
examples/ip_pipeline/Makefile |2 +
examples/ip_pipeline/init.c|2 +
examples/ip_pipeline/pipeline/hash_func.h | 351 +
.../ip_pipeline/pipeline/pipeline_actions_common.h | 119 +++
.../ip_pipeline/pipeline/pipeline_passthrough.c| 192 +
.../ip_pipeline/pipeline/pipeline_passthrough.h| 41 ++
.../ip_pipeline/pipeline/pipeline_passthrough_be.c | 772
.../ip_pipeline/pipeline/pipeline_passthrough_be.h | 41 ++
8 files changed, 1341 insertions(+), 179 deletions(-)
create mode 100644 examples/ip_pipeline/pipeline/hash_func.h
create mode 100644 examples/ip_pipeline/pipeline/pipeline_actions_common.h
create mode 100644 examples/ip_pipeline/pipeline/pipeline_passthrough.h
create mode 100644 examples/ip_pipeline/pipeline/pipeline_passthrough_be.c
create mode 100644 examples/ip_pipeline/pipeline/pipeline_passthrough_be.h
diff --git a/examples/ip_pipeline/Makefile b/examples/ip_pipeline/Makefile
index f255338..930dc61 100644
--- a/examples/ip_pipeline/Makefile
+++ b/examples/ip_pipeline/Makefile
@@ -60,6 +60,8 @@ SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_common_be.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_common_fe.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_master_be.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_master.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_passthrough_be.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_passthrough.c
CFLAGS += -I$(SRCDIR) -I$(SRCDIR)/pipeline
CFLAGS += -O3
diff --git a/examples/ip_pipeline/init.c b/examples/ip_pipeline/init.c
index 807993f..d80bede 100644
--- a/examples/ip_pipeline/init.c
+++ b/examples/ip_pipeline/init.c
@@ -45,6 +45,7 @@
#include "pipeline.h"
#include "pipeline_common_fe.h"
#include "pipeline_master.h"
+#include "pipeline_passthrough.h"
#define APP_NAME_SIZE 32
@@ -1241,6 +1242,7 @@ int app_init(struct app_params *app)
app_pipeline_common_cmd_push(app);
app_pipeline_type_register(app, &pipeline_master);
+ app_pipeline_type_register(app, &pipeline_passthrough);
app_init_pipelines(app);
app_init_threads(app);
diff --git a/examples/ip_pipeline/pipeline/hash_func.h
b/examples/ip_pipeline/pipeline/hash_func.h
new file mode 100644
index 000..7846300
--- /dev/null
+++ b/examples/ip_pipeline/pipeline/hash_func.h
@@ -0,0 +1,351 @@
+/*-
+ * BSD LICENSE
+ *
+ * Copyright(c) 2010-2015 Intel Corporation. All rights reserved.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * * Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * * Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in
+ * the documentation and/or other materials provided with the
+ * distribution.
+ * * Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+#ifndef __INCLUDE_HASH_FUNC_H__
+#define __INCLUDE_HASH_FUNC_H__
+
+static inline uint64_t
+hash_xor_key8(void *key, __rte_unused uint32_t key_size, uint64_t seed)
+{
+ uint64_t *k = key;
+ uint64_t xor0;
+
+ xor0 = seed ^ k[0];
+
+ return (xor0 >> 32) ^ xor0;
+}
+
+static inline uint64_t
+hash_xor_key16(void *key, __rte_unused uint32_t key_size, uint64_t seed)
+{
+ uint64_t *k = key;
+ uint64_t xor0;
+
+ xor0 = (k[0] ^ seed) ^ k[1];
+
+ return (xor0 >> 32) ^ xor0;
+}
+
+static inline uint64_t
+hash_xor_key24(void *
[dpdk-dev] [PATCH v5 09/11] ip_pipeline: added new implementation of firewall pipeline
From: Daniel Mrzyglod
Firewall pipeline implementation is split to two files.
pipeline_firewall.c file handles front-end functions (cli commands
parsing) pipeline_firewall_ops.c contains implementation of functions
done by pipeline (back-end).
Signed-off-by: Daniel Mrzyglod
---
examples/ip_pipeline/Makefile |2 +
examples/ip_pipeline/init.c|2 +
examples/ip_pipeline/pipeline/pipeline_firewall.c | 1162
examples/ip_pipeline/pipeline/pipeline_firewall.h | 63 ++
.../ip_pipeline/pipeline/pipeline_firewall_be.c| 740 +
.../ip_pipeline/pipeline/pipeline_firewall_be.h| 138 +++
6 files changed, 1870 insertions(+), 237 deletions(-)
create mode 100644 examples/ip_pipeline/pipeline/pipeline_firewall.h
create mode 100644 examples/ip_pipeline/pipeline/pipeline_firewall_be.c
create mode 100644 examples/ip_pipeline/pipeline/pipeline_firewall_be.h
diff --git a/examples/ip_pipeline/Makefile b/examples/ip_pipeline/Makefile
index 930dc61..382fee6 100644
--- a/examples/ip_pipeline/Makefile
+++ b/examples/ip_pipeline/Makefile
@@ -62,6 +62,8 @@ SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_master_be.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_master.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_passthrough_be.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_passthrough.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_firewall_be.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_firewall.c
CFLAGS += -I$(SRCDIR) -I$(SRCDIR)/pipeline
CFLAGS += -O3
diff --git a/examples/ip_pipeline/init.c b/examples/ip_pipeline/init.c
index d80bede..e8d33c7 100644
--- a/examples/ip_pipeline/init.c
+++ b/examples/ip_pipeline/init.c
@@ -46,6 +46,7 @@
#include "pipeline_common_fe.h"
#include "pipeline_master.h"
#include "pipeline_passthrough.h"
+#include "pipeline_firewall.h"
#define APP_NAME_SIZE 32
@@ -1243,6 +1244,7 @@ int app_init(struct app_params *app)
app_pipeline_common_cmd_push(app);
app_pipeline_type_register(app, &pipeline_master);
app_pipeline_type_register(app, &pipeline_passthrough);
+ app_pipeline_type_register(app, &pipeline_firewall);
app_init_pipelines(app);
app_init_threads(app);
diff --git a/examples/ip_pipeline/pipeline/pipeline_firewall.c
b/examples/ip_pipeline/pipeline/pipeline_firewall.c
index b70260e..40fbdc5 100644
--- a/examples/ip_pipeline/pipeline/pipeline_firewall.c
+++ b/examples/ip_pipeline/pipeline/pipeline_firewall.c
@@ -1,7 +1,7 @@
/*-
* BSD LICENSE
*
- * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ * Copyright(c) 2010-2015 Intel Corporation. All rights reserved.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -32,282 +32,970 @@
*/
#include
-#include
-#include
+#include
+#include
+#include
+#include
+#include
#include
-#include
-#include
-#include
-#include
-#include
+#include
+#include
+#include
+#include
+#include
+#include
+#include
+
+#include "app.h"
+#include "pipeline_common_fe.h"
+#include "pipeline_firewall.h"
+
+struct app_pipeline_firewall_rule {
+ struct pipeline_firewall_key key;
+ int32_t priority;
+ uint32_t port_id;
+ void *entry_ptr;
+
+ TAILQ_ENTRY(app_pipeline_firewall_rule) node;
+};
+
+struct app_pipeline_firewall {
+ /* parameters */
+ uint32_t n_ports_in;
+ uint32_t n_ports_out;
+
+ /* rules */
+ TAILQ_HEAD(, app_pipeline_firewall_rule) rules;
+ uint32_t n_rules;
+ uint32_t default_rule_present;
+ uint32_t default_rule_port_id;
+ void *default_rule_entry_ptr;
+};
-#include
-#include
-#include
+static void
+print_firewall_ipv4_rule(struct app_pipeline_firewall_rule *rule)
+{
+ printf("Prio = %d (SA = %u.%u.%u.%u/%u, "
+ "DA = %u.%u.%u.%u/%u, "
+ "SP = %u-%u, "
+ "DP = %u-%u, "
+ "Proto = %u / 0x%x) => "
+ "Port = %u (entry ptr = %p)\n",
+
+ rule->priority,
+
+ (rule->key.key.ipv4_5tuple.src_ip >> 24) & 0xFF,
+ (rule->key.key.ipv4_5tuple.src_ip >> 16) & 0xFF,
+ (rule->key.key.ipv4_5tuple.src_ip >> 8) & 0xFF,
+ rule->key.key.ipv4_5tuple.src_ip & 0xFF,
+ rule->key.key.ipv4_5tuple.src_ip_mask,
+
+ (rule->key.key.ipv4_5tuple.dst_ip >> 24) & 0xFF,
+ (rule->key.key.ipv4_5tuple.dst_ip >> 16) & 0xFF,
+ (rule->key.key.ipv4_5tuple.dst_ip >> 8) & 0xFF,
+ rule->key.key.ipv4_5tuple.dst_ip & 0xFF,
+ rule->key.key.ipv4_5tuple.dst_ip_mask,
+
+ rule->key.key.ipv4_5tuple.src_port_from,
+ rule->key.key.ipv4_5tuple.src_port_to,
+
+ rule->key.key.ipv4_5tuple.dst_port_from,
+ rule->key.key.ipv4_5tuple.dst_port_to,
+
+ rule->key.key.ipv4_5tuple.
[dpdk-dev] [PATCH v5 10/11] ip_pipeline: added new implementation of routing pipeline
Routing pipeline implementation is split to two files.
pipeline_routing.c file handles front-end functions (cli commands
parsing) pipeline_routing_ops.c contains implementation of functions
done by pipeline (back-end).
Signed-off-by: Pawel Wodkowski
---
examples/ip_pipeline/Makefile |2 +
examples/ip_pipeline/init.c|2 +
examples/ip_pipeline/pipeline/pipeline_routing.c | 1783
examples/ip_pipeline/pipeline/pipeline_routing.h | 99 ++
.../ip_pipeline/pipeline/pipeline_routing_be.c | 869 ++
.../ip_pipeline/pipeline/pipeline_routing_be.h | 230 +++
6 files changed, 2627 insertions(+), 358 deletions(-)
create mode 100644 examples/ip_pipeline/pipeline/pipeline_routing.h
create mode 100644 examples/ip_pipeline/pipeline/pipeline_routing_be.c
create mode 100644 examples/ip_pipeline/pipeline/pipeline_routing_be.h
diff --git a/examples/ip_pipeline/Makefile b/examples/ip_pipeline/Makefile
index 382fee6..a2881a6 100644
--- a/examples/ip_pipeline/Makefile
+++ b/examples/ip_pipeline/Makefile
@@ -64,6 +64,8 @@ SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) +=
pipeline_passthrough_be.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_passthrough.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_firewall_be.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_firewall.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_routing_be.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_routing.c
CFLAGS += -I$(SRCDIR) -I$(SRCDIR)/pipeline
CFLAGS += -O3
diff --git a/examples/ip_pipeline/init.c b/examples/ip_pipeline/init.c
index e8d33c7..88e1aa5 100644
--- a/examples/ip_pipeline/init.c
+++ b/examples/ip_pipeline/init.c
@@ -47,6 +47,7 @@
#include "pipeline_master.h"
#include "pipeline_passthrough.h"
#include "pipeline_firewall.h"
+#include "pipeline_routing.h"
#define APP_NAME_SIZE 32
@@ -1245,6 +1246,7 @@ int app_init(struct app_params *app)
app_pipeline_type_register(app, &pipeline_master);
app_pipeline_type_register(app, &pipeline_passthrough);
app_pipeline_type_register(app, &pipeline_firewall);
+ app_pipeline_type_register(app, &pipeline_routing);
app_init_pipelines(app);
app_init_threads(app);
diff --git a/examples/ip_pipeline/pipeline/pipeline_routing.c
b/examples/ip_pipeline/pipeline/pipeline_routing.c
index b1ce624..2415aa2 100644
--- a/examples/ip_pipeline/pipeline/pipeline_routing.c
+++ b/examples/ip_pipeline/pipeline/pipeline_routing.c
@@ -1,7 +1,7 @@
/*-
* BSD LICENSE
*
- * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ * Copyright(c) 2010-2015 Intel Corporation. All rights reserved.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -31,444 +31,1511 @@
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
-#include
-#include
-#include
+#include
+#include
+#include
+#include
+#include
-#include
-#include
-#include
-#include
-#include
-#include
+#include "app.h"
+#include "pipeline_common_fe.h"
+#include "pipeline_routing.h"
-#include
-#include
-#include
-#include
+struct app_pipeline_routing_route {
+ struct pipeline_routing_route_key key;
+ struct app_pipeline_routing_route_params params;
+ void *entry_ptr;
-#include "main.h"
+ TAILQ_ENTRY(app_pipeline_routing_route) node;
+};
-#include
+struct app_pipeline_routing_arp_entry {
+ struct pipeline_routing_arp_key key;
+ struct ether_addr macaddr;
+ void *entry_ptr;
-struct app_routing_table_entry {
- struct rte_pipeline_table_entry head;
- uint32_t nh_ip;
- uint32_t nh_iface;
+ TAILQ_ENTRY(app_pipeline_routing_arp_entry) node;
};
-struct app_arp_table_entry {
- struct rte_pipeline_table_entry head;
- struct ether_addr nh_arp;
+struct pipeline_routing {
+ /* Parameters */
+ uint32_t n_ports_in;
+ uint32_t n_ports_out;
+
+ /* Routes */
+ TAILQ_HEAD(, app_pipeline_routing_route) routes;
+ uint32_t n_routes;
+
+ uint32_t default_route_present;
+ uint32_t default_route_port_id;
+ void *default_route_entry_ptr;
+
+ /* ARP entries */
+ TAILQ_HEAD(, app_pipeline_routing_arp_entry) arp_entries;
+ uint32_t n_arp_entries;
+
+ uint32_t default_arp_entry_present;
+ uint32_t default_arp_entry_port_id;
+ void *default_arp_entry_ptr;
};
-static inline void
-app_routing_table_write_metadata(
- struct rte_mbuf *pkt,
- struct app_routing_table_entry *entry)
+static void *
+pipeline_routing_init(struct pipeline_params *params,
+ __rte_unused void *arg)
{
- struct app_pkt_metadata *c =
- (struct app_pkt_metadata *) RTE_MBUF_METADATA_UINT8_PTR(pkt, 0);
+ struct pipeline_routing *p;
+ uint32_t size;
+
+ /* Check input arguments */
+ if ((params == NULL) ||
+ (params->n_por
[dpdk-dev] [PATCH v5 11/11] ip_pipeline: added new implementation of flow classification pipeline
Flow classification pipeline implementation is split to two files.
pipeline_flow_classification.c file handles front-end functions (cli
commands parsing) pipeline_flow_classification_ops.c contains
implementation of functions done by pipeline (back-end).
Signed-off-by: Maciej Gajdzica
---
examples/ip_pipeline/Makefile |2 +
examples/ip_pipeline/init.c|2 +
.../pipeline/pipeline_flow_classification.c| 2205 ++--
.../pipeline/pipeline_flow_classification.h| 105 +
.../pipeline/pipeline_flow_classification_be.c | 589 ++
.../pipeline/pipeline_flow_classification_be.h | 140 ++
6 files changed, 2816 insertions(+), 227 deletions(-)
create mode 100644 examples/ip_pipeline/pipeline/pipeline_flow_classification.h
create mode 100644
examples/ip_pipeline/pipeline/pipeline_flow_classification_be.c
create mode 100644
examples/ip_pipeline/pipeline/pipeline_flow_classification_be.h
diff --git a/examples/ip_pipeline/Makefile b/examples/ip_pipeline/Makefile
index a2881a6..f3ff1ec 100644
--- a/examples/ip_pipeline/Makefile
+++ b/examples/ip_pipeline/Makefile
@@ -64,6 +64,8 @@ SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) +=
pipeline_passthrough_be.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_passthrough.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_firewall_be.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_firewall.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_flow_classification_be.c
+SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_flow_classification.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_routing_be.c
SRCS-$(CONFIG_RTE_LIBRTE_PIPELINE) += pipeline_routing.c
diff --git a/examples/ip_pipeline/init.c b/examples/ip_pipeline/init.c
index 88e1aa5..dd7bee4 100644
--- a/examples/ip_pipeline/init.c
+++ b/examples/ip_pipeline/init.c
@@ -47,6 +47,7 @@
#include "pipeline_master.h"
#include "pipeline_passthrough.h"
#include "pipeline_firewall.h"
+#include "pipeline_flow_classification.h"
#include "pipeline_routing.h"
#define APP_NAME_SIZE 32
@@ -1245,6 +1246,7 @@ int app_init(struct app_params *app)
app_pipeline_common_cmd_push(app);
app_pipeline_type_register(app, &pipeline_master);
app_pipeline_type_register(app, &pipeline_passthrough);
+ app_pipeline_type_register(app, &pipeline_flow_classification);
app_pipeline_type_register(app, &pipeline_firewall);
app_pipeline_type_register(app, &pipeline_routing);
diff --git a/examples/ip_pipeline/pipeline/pipeline_flow_classification.c
b/examples/ip_pipeline/pipeline/pipeline_flow_classification.c
index cc0cbf1..c3926fa 100644
--- a/examples/ip_pipeline/pipeline/pipeline_flow_classification.c
+++ b/examples/ip_pipeline/pipeline/pipeline_flow_classification.c
@@ -1,7 +1,7 @@
/*-
* BSD LICENSE
*
- * Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
+ * Copyright(c) 2010-2015 Intel Corporation. All rights reserved.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -32,275 +32,2026 @@
*/
#include
-#include
-#include
+#include
+#include
+#include
+#include
+#include
#include
-#include
-#include
-#include
-#include
-#include
+#include
+#include
+#include
+#include
+#include
+#include
-#include
-#include
-#include
+#include "app.h"
+#include "pipeline_common_fe.h"
+#include "pipeline_flow_classification.h"
+#include "hash_func.h"
-#include "main.h"
+/*
+ * Key conversion
+ */
+
+struct pkt_key_qinq {
+ uint16_t ethertype_svlan;
+ uint16_t svlan;
+ uint16_t ethertype_cvlan;
+ uint16_t cvlan;
+} __attribute__((__packed__));
+
+struct pkt_key_ipv4_5tuple {
+ uint8_t ttl;
+ uint8_t proto;
+ uint16_t checksum;
+ uint32_t ip_src;
+ uint32_t ip_dst;
+ uint16_t port_src;
+ uint16_t port_dst;
+} __attribute__((__packed__));
+
+struct pkt_key_ipv6_5tuple {
+ uint16_t payload_length;
+ uint8_t proto;
+ uint8_t hop_limit;
+ uint8_t ip_src[16];
+ uint8_t ip_dst[16];
+ uint16_t port_src;
+ uint16_t port_dst;
+} __attribute__((__packed__));
+
+static int
+app_pipeline_fc_key_convert(struct pipeline_fc_key *key_in,
+ uint8_t *key_out,
+ uint32_t *signature)
+{
+ uint8_t buffer[PIPELINE_FC_FLOW_KEY_MAX_SIZE];
+ void *key_buffer = (key_out) ? key_out : buffer;
+
+ switch (key_in->type) {
+ case FLOW_KEY_QINQ:
+ {
+ struct pkt_key_qinq *qinq = key_buffer;
+
+ qinq->ethertype_svlan = 0;
+ qinq->svlan = rte_bswap16(key_in->key.qinq.svlan);
+ qinq->ethertype_cvlan = 0;
+ qinq->cvlan = rte_bswap16(key_in->key.qinq.cvlan);
+
+ if (signature)
+ *signature = (uint32_t) hash_default_key8(qinq, 8, 0);
+ return 0;
+ }
+
+ case FLOW_KEY_IPV4_5TUPLE:
+ {
+
[dpdk-dev] [PATCH v5 00/11] ip_pipeline: ip_pipeline application enhancements
> -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Maciej Gajdzica > Sent: Monday, July 6, 2015 10:29 AM > To: dev at dpdk.org > Subject: [dpdk-dev] [PATCH v5 00/11] ip_pipeline: ip_pipeline application > enhancements > > Changes in v5: > - fixed build issue with clang 3.6 Acked-by: Cristian Dumitrescu
[dpdk-dev] [PATCH v5 00/11] ip_pipeline: ip_pipeline application enhancements
NACK
ICC Compilation Errors:
Pipeline/pipeline_routing_be.c:
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_routing_be.c(740):
error #188: enumerated type mixed with another type
.flags = 0,
^
ICC wants name like PIPELINE_ROUTING_ROUTE_LOCAL but got number instead.
Pipeline/Pipeline_routing.c:
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_routing.c(851):
error #188: enumerated type mixed with another type
rt_params.flags = 0; /* remote route */
^
The problem is that the value of PIPELINE_ROUTING_ROUTE_LOCAL is 1.
After this change it looks ok.
Solution:
We need add another enum:
PIPELINE_ROUTING_ROUTE_REMOTE
ICC_I686:
Init.c:
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/init.c(90): error #181:
argument of type "uint64_t={unsigned long long}" is incompatible with format
"%016lx", expecting argument of type "unsigned long"
APP_LOG(app, HIGH, "CPU core mask = 0x%016lx", app->core_mask);
^
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/init.c(103): error
#181: argument of type "uint64_t={unsigned long long}" is incompatible with
format "%lx", expecting argument of type "unsigned long"
snprintf(buffer, sizeof(buffer), "-c%lx", app->core_mask);
^
pipeline_common_fe:
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_common_fe.c(525):
error #181: argument of type "uint64_t={unsigned long long}" is incompatible
with format "%lu", expecting argument of type "unsigned long"
stats.stats.n_pkts_in,
^
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_common_fe.c(526):
error #181: argument of type "uint64_t={unsigned long long}" is incompatible
with format "%lu", expecting argument of type "unsigned long"
stats.n_pkts_dropped_by_ah,
^
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_common_fe.c(527):
error #181: argument of type "uint64_t={unsigned long long}" is incompatible
with format "%lu", expecting argument of type "unsigned long"
stats.stats.n_pkts_drop);
^
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_common_fe.c(611):
error #181: argument of type "uint64_t={unsigned long long}" is incompatible
with format "%lu", expecting argument of type "unsigned long"
stats.stats.n_pkts_in,
^
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_common_fe.c(612):
error #181: argument of type "uint64_t={unsigned long long}" is incompatible
with format "%lu", expecting argument of type "unsigned long"
stats.n_pkts_dropped_by_ah,
^
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_common_fe.c(613):
error #181: argument of type "uint64_t={unsigned long long}" is incompatible
with format "%lu", expecting argument of type "unsigned long"
stats.stats.n_pkts_drop);
^
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_common_fe.c(698):
error #181: argument of type "uint64_t={unsigned long long}" is incompatible
with format "%lu", expecting argument of type "unsigned long"
stats.stats.n_pkts_in,
^
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_common_fe.c(699):
error #181: argument of type "uint64_t={unsigned long long}" is incompatible
with format "%lu", expecting argument of type "unsigned long"
stats.stats.n_pkts_lookup_miss,
^
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_common_fe.c(700):
error #181: argument of type "uint64_t={unsigned long long}" is incompatible
with format "%lu", expecting argument of type "unsigned long"
stats.n_pkts_dropped_by_lkp_hit_ah,
^
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_common_fe.c(701):
error #181: argument of type "uint64_t={unsigned long long}" is incompatible
with format "%lu", expecting argument of type "unsigned long"
stats.n_pkts_dropped_lkp_hit,
^
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_common_fe.c(702):
error #181: argument of type "uint64_t={unsigned long long}" is incompatible
with format "%lu", expecting argument of type "unsigned long"
stats.n_pkts_dropped_by_lkp_miss_ah,
^
/mnt/shared/dtmrzglx/dpdk_pipeline/examples/ip_pipeline/pipeline/pipeline_common_fe.c(703):
error #181: argument of type "uint64_t={unsigned long long}" is incompatible
with format "%lu", expecting argument of type "unsigned long"
stats.n_pkts_dropped_lkp_miss);
^
pipeline_routing_be.c:
[dpdk-dev] [PATCH v3 01/12] test: limit x86 cpuflags checks to x86 builds
On Mon, Jul 06, 2015 at 04:51:26PM +0800, Zhigang Lu wrote: > The original code mistakenly defaulted to X86 when RTE_ARCH_PPC_64 was > left undefined. This did not accomodate other non-PPC/non-X86 > architectures. This patch fixes this issue. > > Change-Id: I5e8cf33c2eb917f7f6583dc95ed0f336066a285e We don't use gerrit on dpdk.org, so these change-ids are unnecessary on submissions. > Signed-off-by: Zhigang Lu Acked-by: Bruce Richardson
[dpdk-dev] [PATCH v3 02/12] hash: fix compilation on non-X86 platforms
On Mon, Jul 06, 2015 at 04:51:27PM +0800, Zhigang Lu wrote:
> The "hash: remove duplicated code" change unfortunately broke the
> build for non-X86 platforms. This patch fixes this breakage.
>
Fixes: 49361c3f3cfa ("hash: remove duplicated code")
> Change-Id: Ie109d67e681b75b45320fab1bf9de4eb9c9701bf
> Signed-off-by: Zhigang Lu
Acked-by: Bruce Richardson
[dpdk-dev] [PATCH v2] librte_ether: release memory in uninit function.
Hi, all
As we has gap on the memory release action to be done in which step, I
appreciate all your comments on this patch.
Currently, the correct quit sequence for testpmd is stop() --->
port_stop() --> port_close() --> quit(). This will lead lots of memory
not released by default, like queues.
We have 3 potential proposals(normal case means without hotplug):
1. Those memory release in close() other left in uninit()
This will work fine for both hotplug case and normal case.
2. leave close() just as the same before, all be done in uninit()
This will works fine for hotplug, for normal case maybe has
issue(memory not released?).
3. Combine close() and uninit(), only one will be left.
This will work fine for both hotplug case and normal case. But it
may change a lot, such as logic.
4. other.
For more details, you could go though this thread.
Thanks,
Michael
On 6/30/2015 9:32 AM, Qiu, Michael wrote:
> On 6/30/2015 12:42 AM, Iremonger, Bernard wrote:
>>> -Original Message-
>>> From: Qiu, Michael
>>> Sent: Monday, June 29, 2015 4:22 PM
>>> To: Iremonger, Bernard; dev at dpdk.org
>>> Cc: Zhang, Helin; Ananyev, Konstantin; mukawa at igel.co.jp; Stephen
>>> Hemminger
>>> Subject: Re: [PATCH v2] librte_ether: release memory in uninit function.
>>>
>>> On 2015/6/29 18:20, Iremonger, Bernard wrote:
> -Original Message-
> From: Qiu, Michael
> Sent: Monday, June 29, 2015 9:55 AM
> To: Iremonger, Bernard; dev at dpdk.org
> Cc: Zhang, Helin; Ananyev, Konstantin; mukawa at igel.co.jp; Stephen
> Hemminger
> Subject: Re: [PATCH v2] librte_ether: release memory in uninit function.
>
> On 6/26/2015 5:32 PM, Iremonger, Bernard wrote:
>> Changes in v2:
>> do not free mac_addrs and hash_mac_addrs here.
>>
>> Signed-off-by: Bernard Iremonger
>> ---
>> lib/librte_ether/rte_ethdev.c |6 +-
>> 1 files changed, 5 insertions(+), 1 deletions(-)
>>
>> diff --git a/lib/librte_ether/rte_ethdev.c
>> b/lib/librte_ether/rte_ethdev.c index e13fde5..7ae101a 100644
>> --- a/lib/librte_ether/rte_ethdev.c
>> +++ b/lib/librte_ether/rte_ethdev.c
>> @@ -369,8 +369,12 @@ rte_eth_dev_uninit(struct rte_pci_device
> *pci_dev)
>> /* free ether device */
>> rte_eth_dev_release_port(eth_dev);
>>
>> -if (rte_eal_process_type() == RTE_PROC_PRIMARY)
>> +if (rte_eal_process_type() == RTE_PROC_PRIMARY) {
>> +rte_free(eth_dev->data->rx_queues);
>> +rte_free(eth_dev->data->tx_queues);
>> rte_free(eth_dev->data->dev_private);
>> +memset(eth_dev->data, 0, sizeof(struct
> rte_eth_dev_data));
>> +}
>>
>> eth_dev->pci_dev = NULL;
>> eth_dev->driver = NULL;
> Actually, This could be put in rte_eth_dev_close() becasue queues
> should be released when closed.
>
> Also before free dev->data->rx_queues you should make sure
> dev->data->rx_queues[i] has been freed in PMD close() function, So
> dev->data->this
> two should be better done at the same time, ether in
> rte_eth_dev_close() or in PMD close() function. For hotplug in fm10k,
> I put it in PMD close() function.
>
> Thanks,
> Michael
Hi Michael,
The consensus is that the rx_queue and tx_queue memory should not be
>>> released in the PMD as it is not allocated by the PMD. The memory is
>>> allocated in rte_eth_dev_rx_queue_config() and
>>> rte_eth_dev_tx_queue_config(), which are both called from
>>> rte_eth_dev_configure() which is called by the application (for example
>>> test_pmd). So it seems to make sense to free this memory in
>>> rte_eth_dev_uninit().
>>>
>>> It really make sense to free memory in rte_ether level, but when close a
>>> port
>>> with out detached? just as stop --> close() --> quit(), the memory will not
>>> be
>>> released :)
>>>
>> In the above scenario lots of memory will not be released.
>>
>> This is why the detach() and the underlying dev_uninit() functions were
>> introduced.
> First detach is only for hotplug, for *users do not use hotplug*, that
> scenario is the right action. So "lots of memory will not be released"
> is issue need be fixed, actually, in fm10k driver, lots of memory has
> been released.
>
>> The dev_uninit() functions currently call dev_close() which in turn calls
>> dev_stop() which calls dev_clear_queues().
> Users do hotplug then must call stop() --> close() --> dev_uninit(), it
> works fine. But do you think it make sense to release memory when
> close() it?
>
>> The dev_clear_queues() function does not release the queue_memory or the
>> queue array memory. The queue memory is now released in the dev_uninit() and
>> the queue array memory is released in the rte_eth_dev_uninit() function.
> That's your implementation, make sure not all users will detach a
> device, but the r
[dpdk-dev] [PATCH v3 7/7] abi: announce mbuf addition for ieee1588 in DPDK 2.2
2015-07-02 16:16, John McNamara: > --- a/doc/guides/rel_notes/abi.rst > +++ b/doc/guides/rel_notes/abi.rst > Deprecation Notices > --- > + > +* In DPDK 2.1 the IEEE1588/802.1AS support in the i40e driver makes use of > the > + ``udata64`` field in the mbuf to pass the timesync register index to the > + user. In DPDK 2.2 this will be moved to a new field in the mbuf. We need more acknowledgements for this decision, as stated here: http://dpdk.org/browse/dpdk/tree/doc/guides/guidelines/versioning.rst#n51
[dpdk-dev] [PATCH] mk: enable next abi in static libs
Any comment or ack? 2015-07-03 00:05, Thomas Monjalon: > When a change makes really hard to keep ABI compatibility, > instead of waiting next release to break the ABI, it is smoother > to introduce the new code and enable it only for static libraries. > The flag RTE_NEXT_ABI may be used to "ifdef" the new code. > When the release is out, a dynamically linked application can use > the new shared libraries without rebuild while developpers can prepare > their application for the next ABI by reading the deprecation notice > and easily testing the new code. > When starting the next release cycle, the "ifdefs" will be removed > and the ABI break will be marked by incrementing LIBABIVER. > > The new option CONFIG_RTE_NEXT_ABI is not defined in the configuration > templates because it is deduced from CONFIG_RTE_BUILD_SHARED_LIB. > It is automatically enabled for static libraries and disabled for > shared libraries. > It can be forced to another value by editing the generated .config file. > It shouldn't be enabled for shared libraries because it would break the > ABI without changing the version number LIBABIVER. That's why a warning > is printed in this case. > > The guideline is also updated to integrate this new possibility. > > Signed-off-by: Thomas Monjalon
[dpdk-dev] [PATCH 0/2] Native uio-based PMD for Mellanox ConnectX-3 devices
This is a native UIO-based PMD for Mellanox ConnectX-3 devices. It uses a persistent memory library in order to provide a persistent scartch area for the mlx4 HCA driver. We release the driver itself under BSD license, but to use it for commercial products, you may have to re-implement the separated GPL sources. The GPL affected source codes reside in the mlnx_uio/kernel directory. leeopop (2): eal/persistent: new library to hold memory region after program exit mlnx_uio: new poll mode driver config/common_linuxapp | 10 + drivers/net/Makefile |1 + drivers/net/mlnx_uio/.gitignore|1 + drivers/net/mlnx_uio/LICENSE | 30 + drivers/net/mlnx_uio/Makefile | 139 + drivers/net/mlnx_uio/convert.py| 50 + drivers/net/mlnx_uio/include/autoconf.h| 10 + drivers/net/mlnx_uio/include/bitmap.h | 314 + drivers/net/mlnx_uio/include/bitops.h | 558 ++ drivers/net/mlnx_uio/include/dcbnl.h | 751 +++ drivers/net/mlnx_uio/include/etherdevice.h | 189 + drivers/net/mlnx_uio/include/ib_mad.h | 664 ++ drivers/net/mlnx_uio/include/ib_smi.h | 128 + drivers/net/mlnx_uio/include/ib_verbs.h| 806 +++ drivers/net/mlnx_uio/include/inline_functions.h| 307 + drivers/net/mlnx_uio/include/kcompat.h | 36 + drivers/net/mlnx_uio/include/kmod.h| 768 +++ drivers/net/mlnx_uio/include/list.h| 780 +++ drivers/net/mlnx_uio/include/log2.h| 229 + drivers/net/mlnx_uio/include/mlx4_dpdk.h | 17 + drivers/net/mlnx_uio/include/mlx4_uio.h| 24 + drivers/net/mlnx_uio/include/mlx4_uio_helper.h | 800 +++ drivers/net/mlnx_uio/include/module.h | 12 + drivers/net/mlnx_uio/include/netdev_features.h | 166 + drivers/net/mlnx_uio/include/post_kmod.h | 13 + drivers/net/mlnx_uio/include/radix-tree.h | 48 + drivers/net/mlnx_uio/include/rbtree.h | 105 + drivers/net/mlnx_uio/include/rbtree_augmented.h| 230 + drivers/net/mlnx_uio/kernel/LICENSE| 339 + drivers/net/mlnx_uio/kernel/bitmap.c | 831 +++ drivers/net/mlnx_uio/kernel/kcompat.c | 96 + drivers/net/mlnx_uio/kernel/radix-tree.c | 78 + drivers/net/mlnx_uio/kernel/rbtree.c | 561 ++ drivers/net/mlnx_uio/mlnx/include/mlx4/cmd.h | 309 + drivers/net/mlnx_uio/mlnx/include/mlx4/cq.h| 195 + drivers/net/mlnx_uio/mlnx/include/mlx4/device.h| 1744 + drivers/net/mlnx_uio/mlnx/include/mlx4/doorbell.h | 90 + drivers/net/mlnx_uio/mlnx/include/mlx4/driver.h| 175 + drivers/net/mlnx_uio/mlnx/include/mlx4/qp.h| 540 ++ drivers/net/mlnx_uio/mlnx/include/mlx4/srq.h | 50 + drivers/net/mlnx_uio/mlnx/include/mlx5/cmd.h | 56 + drivers/net/mlnx_uio/mlnx/include/mlx5/cq.h| 182 + drivers/net/mlnx_uio/mlnx/include/mlx5/device.h| 1204 drivers/net/mlnx_uio/mlnx/include/mlx5/doorbell.h | 85 + drivers/net/mlnx_uio/mlnx/include/mlx5/driver.h| 1063 +++ .../net/mlnx_uio/mlnx/include/mlx5/flow_table.h| 59 + drivers/net/mlnx_uio/mlnx/include/mlx5/mlx5_ifc.h | 6892 drivers/net/mlnx_uio/mlnx/include/mlx5/qp.h| 804 +++ drivers/net/mlnx_uio/mlnx/include/mlx5/srq.h | 46 + drivers/net/mlnx_uio/mlnx/include/mlx5/vport.h | 52 + drivers/net/mlnx_uio/mlnx/mlx4/Kconfig | 46 + drivers/net/mlnx_uio/mlnx/mlx4/Makefile| 19 + drivers/net/mlnx_uio/mlnx/mlx4/alloc.c | 872 +++ drivers/net/mlnx_uio/mlnx/mlx4/catas.c | 350 + drivers/net/mlnx_uio/mlnx/mlx4/cmd.c | 3456 ++ drivers/net/mlnx_uio/mlnx/mlx4/cq.c| 443 ++ drivers/net/mlnx_uio/mlnx/mlx4/en_clock.c | 330 + drivers/net/mlnx_uio/mlnx/mlx4/en_cq.c | 257 + drivers/net/mlnx_uio/mlnx/mlx4/en_dcb_nl.c | 613 ++ drivers/net/mlnx_uio/mlnx/mlx4/en_ethtool.c| 2582 drivers/net/mlnx_uio/mlnx/mlx4/en_main.c | 493 ++ drivers/net/mlnx_uio/mlnx/mlx4/en_netdev.c | 3786 +++ drivers/net/mlnx_uio/mlnx/mlx4/en_port.c | 493 ++ drivers/net/mlnx_uio/mlnx/mlx4/en_port.h | 593 ++ drivers/net/mlnx_uio/mlnx/mlx4/en_resources.c | 184 + drivers/net/mlnx_uio/mlnx/mlx4/en_rx.c | 1565 + drivers/net/mlnx_uio/mlnx/mlx4/en_rx_uio.c | 187 + drivers/net/mlnx_uio/mlnx/mlx4/en_selftest.c | 194 + drivers/net/mlnx_uio/mlnx/mlx4/en_sysfs.c | 623 ++ drivers/net/mlnx_uio/mlnx/mlx4/en_tx.c | 1143 drivers/net/mlnx_uio/mlnx/mlx4/en_tx_uio.c | 47 + drivers/net/mlnx_uio/mlnx/mlx4/eq.c| 1777 + drivers/ne
[dpdk-dev] [PATCH 1/2] eal/persistent: new library to hold memory region after program exit
Some NICs use host memory region as their scratch area.
When DPDK user applications terminate, all the memory regions are lost,
re-initialized (memzone), which causes HW faults.
This libraray maintains shared memory regions that is persistent across
multiple execution and termination of user level applications.
It also manages physically contiguous memory regions.
Signed-off-by: leeopop
---
drivers/net/mlnx_uio/LICENSE | 30
lib/Makefile | 1 +
lib/librte_eal/common/Makefile | 3 +
lib/librte_eal/common/include/rte_pci.h| 1 +
lib/librte_eal/common/include/rte_persistent_mem.h | 26 +++
lib/librte_eal/linuxapp/eal/Makefile | 6 +
lib/librte_eal/linuxapp/eal/eal.c | 9 +
lib/librte_eal/linuxapp/eal/eal_persistent_mem.c | 148 +++
.../eal/include/exec-env/rte_persistent_mem.h | 15 ++
lib/librte_eal/linuxapp/eal/rte_eal_version.map| 2 +
lib/librte_persistent/Makefile | 55 ++
lib/librte_persistent/rte_persistent.c | 198 +
lib/librte_persistent/rte_persistent.h | 20 +++
lib/librte_persistent/rte_persistent_version.map | 11 ++
14 files changed, 525 insertions(+)
create mode 100644 drivers/net/mlnx_uio/LICENSE
create mode 100644 lib/librte_eal/common/include/rte_persistent_mem.h
create mode 100644 lib/librte_eal/linuxapp/eal/eal_persistent_mem.c
create mode 100644
lib/librte_eal/linuxapp/eal/include/exec-env/rte_persistent_mem.h
create mode 100644 lib/librte_persistent/Makefile
create mode 100644 lib/librte_persistent/rte_persistent.c
create mode 100644 lib/librte_persistent/rte_persistent.h
create mode 100644 lib/librte_persistent/rte_persistent_version.map
diff --git a/drivers/net/mlnx_uio/LICENSE b/drivers/net/mlnx_uio/LICENSE
new file mode 100644
index 000..7ef5b4b
--- /dev/null
+++ b/drivers/net/mlnx_uio/LICENSE
@@ -0,0 +1,30 @@
+* Source code in kernel/ directory follows GPLv2 license.
+
+
+Copyright (c) 2015, Keunhong Lee
+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 bsd 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 HOLDER 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.
diff --git a/lib/Makefile b/lib/Makefile
index 5f480f9..7a491d3 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -57,6 +57,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PORT) += librte_port
DIRS-$(CONFIG_RTE_LIBRTE_TABLE) += librte_table
DIRS-$(CONFIG_RTE_LIBRTE_PIPELINE) += librte_pipeline
DIRS-$(CONFIG_RTE_LIBRTE_REORDER) += librte_reorder
+DIRS-$(CONFIG_RTE_LIBRTE_PERSISTENT) += librte_persistent
ifeq ($(CONFIG_RTE_EXEC_ENV_LINUXAPP),y)
DIRS-$(CONFIG_RTE_LIBRTE_KNI) += librte_kni
diff --git a/lib/librte_eal/common/Makefile b/lib/librte_eal/common/Makefile
index 38772d4..ce4b0a7 100644
--- a/lib/librte_eal/common/Makefile
+++ b/lib/librte_eal/common/Makefile
@@ -40,6 +40,9 @@ INC += rte_string_fns.h rte_version.h
INC += rte_eal_memconfig.h rte_malloc_heap.h
INC += rte_hexdump.h rte_devargs.h rte_dev.h
INC += rte_pci_dev_feature_defs.h rte_pci_dev_features.h
+ifeq ($(CONFIG_RTE_EAL_PERSISTENT_MEM),y)
+INC += rte_persistent_mem.h
+endif
ifeq ($(CONFIG_RTE_INSECURE_FUNCTION_WARNING),y)
INC += rte_warnings.h
diff --git a/lib/librte_eal/common/include/rte_pci.h
b/lib/librte_eal/common/include/rte_pci.h
index 7801fa0..a323e74 100644
--- a/lib/librte_eal/common/include/rte_pci.h
+++ b/lib/librte_eal/common/include/rte_pci.h
@@ -207,6 +207,7 @@ struct rte_pci_driver {
pci_devuninit_t *devuninit; /**< Device uninit function. */
const struct rte_pci_id *id_table; /**< ID ta
[dpdk-dev] [PATCH] mk: enable next abi in static libs
On Mon, Jul 06, 2015 at 03:18:51PM +0200, Thomas Monjalon wrote: > Any comment or ack? > > 2015-07-03 00:05, Thomas Monjalon: > > When a change makes really hard to keep ABI compatibility, > > instead of waiting next release to break the ABI, it is smoother > > to introduce the new code and enable it only for static libraries. > > The flag RTE_NEXT_ABI may be used to "ifdef" the new code. > > When the release is out, a dynamically linked application can use > > the new shared libraries without rebuild while developpers can prepare > > their application for the next ABI by reading the deprecation notice > > and easily testing the new code. > > When starting the next release cycle, the "ifdefs" will be removed > > and the ABI break will be marked by incrementing LIBABIVER. > > > > The new option CONFIG_RTE_NEXT_ABI is not defined in the configuration > > templates because it is deduced from CONFIG_RTE_BUILD_SHARED_LIB. > > It is automatically enabled for static libraries and disabled for > > shared libraries. > > It can be forced to another value by editing the generated .config file. > > It shouldn't be enabled for shared libraries because it would break the > > ABI without changing the version number LIBABIVER. That's why a warning > > is printed in this case. > > > > The guideline is also updated to integrate this new possibility. > > > > Signed-off-by: Thomas Monjalon > > Yeah, I'm not sure why this is necessecary. That is to say, if you want to introduce a new ABI operation prior to the old one being removed, that is precisely what the versioning macros are for, and can be used to map the static api to the new version. e.g, given function X that you want to enhance in an ABI breaking way: 1) Separate function X to X_v1 and X_v2 2) Map X_v2 to X at DPDK_v2, map X_v1 to X at DPDK_v1 3) Map the static symbol X to X_v2 4) Post the deprecation notice of X for release 3 immediately Splitting the static ABI from the shared ABI just means that applications will have the opportunity to isolate themselves to one kind of build, which is a bad idea. Neil
[dpdk-dev] [PATCH] mk: enable next abi in static libs
2015-07-06 09:35, Neil Horman: > On Mon, Jul 06, 2015 at 03:18:51PM +0200, Thomas Monjalon wrote: > > Any comment or ack? > > > > 2015-07-03 00:05, Thomas Monjalon: > > > When a change makes really hard to keep ABI compatibility, > > > instead of waiting next release to break the ABI, it is smoother > > > to introduce the new code and enable it only for static libraries. > > > The flag RTE_NEXT_ABI may be used to "ifdef" the new code. > > > When the release is out, a dynamically linked application can use > > > the new shared libraries without rebuild while developpers can prepare > > > their application for the next ABI by reading the deprecation notice > > > and easily testing the new code. > > > When starting the next release cycle, the "ifdefs" will be removed > > > and the ABI break will be marked by incrementing LIBABIVER. > > > > > > The new option CONFIG_RTE_NEXT_ABI is not defined in the configuration > > > templates because it is deduced from CONFIG_RTE_BUILD_SHARED_LIB. > > > It is automatically enabled for static libraries and disabled for > > > shared libraries. > > > It can be forced to another value by editing the generated .config file. > > > It shouldn't be enabled for shared libraries because it would break the > > > ABI without changing the version number LIBABIVER. That's why a warning > > > is printed in this case. > > > > > > The guideline is also updated to integrate this new possibility. > > > > > > Signed-off-by: Thomas Monjalon > > > > > Yeah, I'm not sure why this is necessecary. That is to say, if you want to It's explained at the beginning: "When a change makes really hard to keep ABI compatibility", e.g. mbuf change. > introduce a new ABI operation prior to the old one being removed, that is > precisely what > the versioning macros are for, and can be used to map the static api to the > new > version. e.g, given function X that you want to enhance in an ABI breaking > way: > > 1) Separate function X to X_v1 and X_v2 > 2) Map X_v2 to X at DPDK_v2, map X_v1 to X at DPDK_v1 > 3) Map the static symbol X to X_v2 > 4) Post the deprecation notice of X for release 3 immediately We cannot do that for mbuf change. > Splitting the static ABI from the shared ABI just means that applications will > have the opportunity to isolate themselves to one kind of build, which is a > bad > idea. It helps to be prepared for the next release by testing the app with static libraries. We agreed to allow API breaking for important changes like mbuf rework. This option NEXT_ABI is a step between announcement and effective ABI breaking. I think it's a reasonnable approach. But if nobody ack it, I'm perfectly OK to drop it and related features (unified packet type and interrupt mode).
[dpdk-dev] [PATCH 2/2] mlnx_uio: new poll mode driver
This PMD offers direct access to Mellanox ConnectX-3 NICs instead of using IB Verbs like mlx4 driver. Currrently it supports a limited set of features; it supports only physical functions (PF) and basic RX/TX functionalities such as RSS and scatter-gather I/O of mbufs. We are working on the missing features such as virtual functions, VLAN stripping, support for mlx5 NICs, and etc. We enable this PMD by default, as it only relies on the new eal/persistent library without external dependencies. Signed-off-by: leeopop --- config/common_linuxapp | 10 + drivers/net/Makefile |1 + drivers/net/mlnx_uio/.gitignore|1 + drivers/net/mlnx_uio/Makefile | 139 + drivers/net/mlnx_uio/convert.py| 50 + drivers/net/mlnx_uio/include/autoconf.h| 10 + drivers/net/mlnx_uio/include/bitmap.h | 314 + drivers/net/mlnx_uio/include/bitops.h | 558 ++ drivers/net/mlnx_uio/include/dcbnl.h | 751 +++ drivers/net/mlnx_uio/include/etherdevice.h | 189 + drivers/net/mlnx_uio/include/ib_mad.h | 664 ++ drivers/net/mlnx_uio/include/ib_smi.h | 128 + drivers/net/mlnx_uio/include/ib_verbs.h| 806 +++ drivers/net/mlnx_uio/include/inline_functions.h| 307 + drivers/net/mlnx_uio/include/kcompat.h | 36 + drivers/net/mlnx_uio/include/kmod.h| 768 +++ drivers/net/mlnx_uio/include/list.h| 780 +++ drivers/net/mlnx_uio/include/log2.h| 229 + drivers/net/mlnx_uio/include/mlx4_dpdk.h | 17 + drivers/net/mlnx_uio/include/mlx4_uio.h| 24 + drivers/net/mlnx_uio/include/mlx4_uio_helper.h | 800 +++ drivers/net/mlnx_uio/include/module.h | 12 + drivers/net/mlnx_uio/include/netdev_features.h | 166 + drivers/net/mlnx_uio/include/post_kmod.h | 13 + drivers/net/mlnx_uio/include/radix-tree.h | 48 + drivers/net/mlnx_uio/include/rbtree.h | 105 + drivers/net/mlnx_uio/include/rbtree_augmented.h| 230 + drivers/net/mlnx_uio/kernel/LICENSE| 339 + drivers/net/mlnx_uio/kernel/bitmap.c | 831 +++ drivers/net/mlnx_uio/kernel/kcompat.c | 96 + drivers/net/mlnx_uio/kernel/radix-tree.c | 78 + drivers/net/mlnx_uio/kernel/rbtree.c | 561 ++ drivers/net/mlnx_uio/mlnx/include/mlx4/cmd.h | 309 + drivers/net/mlnx_uio/mlnx/include/mlx4/cq.h| 195 + drivers/net/mlnx_uio/mlnx/include/mlx4/device.h| 1744 + drivers/net/mlnx_uio/mlnx/include/mlx4/doorbell.h | 90 + drivers/net/mlnx_uio/mlnx/include/mlx4/driver.h| 175 + drivers/net/mlnx_uio/mlnx/include/mlx4/qp.h| 540 ++ drivers/net/mlnx_uio/mlnx/include/mlx4/srq.h | 50 + drivers/net/mlnx_uio/mlnx/include/mlx5/cmd.h | 56 + drivers/net/mlnx_uio/mlnx/include/mlx5/cq.h| 182 + drivers/net/mlnx_uio/mlnx/include/mlx5/device.h| 1204 drivers/net/mlnx_uio/mlnx/include/mlx5/doorbell.h | 85 + drivers/net/mlnx_uio/mlnx/include/mlx5/driver.h| 1063 +++ .../net/mlnx_uio/mlnx/include/mlx5/flow_table.h| 59 + drivers/net/mlnx_uio/mlnx/include/mlx5/mlx5_ifc.h | 6892 drivers/net/mlnx_uio/mlnx/include/mlx5/qp.h| 804 +++ drivers/net/mlnx_uio/mlnx/include/mlx5/srq.h | 46 + drivers/net/mlnx_uio/mlnx/include/mlx5/vport.h | 52 + drivers/net/mlnx_uio/mlnx/mlx4/Kconfig | 46 + drivers/net/mlnx_uio/mlnx/mlx4/Makefile| 19 + drivers/net/mlnx_uio/mlnx/mlx4/alloc.c | 872 +++ drivers/net/mlnx_uio/mlnx/mlx4/catas.c | 350 + drivers/net/mlnx_uio/mlnx/mlx4/cmd.c | 3456 ++ drivers/net/mlnx_uio/mlnx/mlx4/cq.c| 443 ++ drivers/net/mlnx_uio/mlnx/mlx4/en_clock.c | 330 + drivers/net/mlnx_uio/mlnx/mlx4/en_cq.c | 257 + drivers/net/mlnx_uio/mlnx/mlx4/en_dcb_nl.c | 613 ++ drivers/net/mlnx_uio/mlnx/mlx4/en_ethtool.c| 2582 drivers/net/mlnx_uio/mlnx/mlx4/en_main.c | 493 ++ drivers/net/mlnx_uio/mlnx/mlx4/en_netdev.c | 3786 +++ drivers/net/mlnx_uio/mlnx/mlx4/en_port.c | 493 ++ drivers/net/mlnx_uio/mlnx/mlx4/en_port.h | 593 ++ drivers/net/mlnx_uio/mlnx/mlx4/en_resources.c | 184 + drivers/net/mlnx_uio/mlnx/mlx4/en_rx.c | 1565 + drivers/net/mlnx_uio/mlnx/mlx4/en_rx_uio.c | 187 + drivers/net/mlnx_uio/mlnx/mlx4/en_selftest.c | 194 + drivers/net/mlnx_uio/mlnx/mlx4/en_sysfs.c | 623 ++ drivers/net/mlnx_uio/mlnx/mlx4/en_tx.c | 1143 drivers/net/mlnx_uio/mlnx/mlx4/en_tx_uio.c | 47 + drivers/net/mlnx_uio/mlnx/mlx4/eq.c| 1777 + drivers/net/mlnx_uio/mlnx/mlx4/fw.c
[dpdk-dev] [PATCH 0/2] Native uio-based PMD for Mellanox ConnectX-3 devices
2015-07-06 22:28, leeopop: > This is a native UIO-based PMD for Mellanox ConnectX-3 devices. > It uses a persistent memory library in order to provide a persistent > scartch area for the mlx4 HCA driver. What is the benefit of this UIO approach compared to the OFED based driver? > We release the driver itself under BSD license, but to use it for > commercial products, you may have to re-implement the separated GPL sources. The GPL sources are not really separated. The DPDK libraries must be BSD-licensed. > The GPL affected source codes reside in the mlnx_uio/kernel directory. It seems that a large part of the GPL driver was also copied in mlnx_uio/mlnx/. Given that you are dropping a huge GPL codebase (whose you don't own the copyright) in a BSD library, and that you didn't give your real name in the signed-off line, it is NACK.
[dpdk-dev] [PATCH 1/2] eal/persistent: new library to hold memory region after program exit
On 07/06/2015 04:28 PM, leeopop wrote: > Some NICs use host memory region as their scratch area. > When DPDK user applications terminate, all the memory regions are lost, > re-initialized (memzone), which causes HW faults. > This libraray maintains shared memory regions that is persistent across > multiple execution and termination of user level applications. > It also manages physically contiguous memory regions. > > Signed-off-by: leeopop > Does dpdk accept anonymous signoffs? DCO usually requires a real name.
[dpdk-dev] [PATCH 1/2] eal/persistent: new library to hold memory region after program exit
2015-07-06 17:34, Avi Kivity: > On 07/06/2015 04:28 PM, leeopop wrote: > > Some NICs use host memory region as their scratch area. > > When DPDK user applications terminate, all the memory regions are lost, > > re-initialized (memzone), which causes HW faults. > > This libraray maintains shared memory regions that is persistent across > > multiple execution and termination of user level applications. > > It also manages physically contiguous memory regions. > > > > Signed-off-by: leeopop > > > > Does dpdk accept anonymous signoffs? > > DCO usually requires a real name. Exact. We follow Linux guidelines for DCO: http://dpdk.org/dev#send
[dpdk-dev] [PATCH v2] doc: fix minor sphinx build warning
2015-07-02 15:15, John McNamara: > Fix for a minor Sphinx build warning in the ABI guidelines docs: > > versioning.rst:126: WARNING: Bullet list ends without a > blank line; unexpected unindent. > > Signed-off-by: John McNamara Applied, thanks
[dpdk-dev] [PATCH 0/2] Fix crash with vpmd and mbuf debug
2015-07-03 16:40, Bruce Richardson: > When mbuf debug support is turned on in the build time config, crashes will > occur when clearing up the RX/TX rings, if the 10G vector PMD is in use. This > error can be reproduced using testpmd. > This patchset makes the setup/teardown code easier to debug by marking "cold" > code paths as such, and then fixes the errors by checking for already-freed > mbufs when clearing the rings. Applied, thanks
[dpdk-dev] [PATCH 4/4] ethdev: check support for rx_queue_count and descriptor_done fns
Neil, your ABI expertise is required for this patch. 2015-06-15 11:14, Bruce Richardson: > On Fri, Jun 12, 2015 at 01:32:56PM -0400, Roger B. Melton wrote: > > Hi Bruce, Comment in-line. Regards, Roger > > > > On 6/12/15 7:28 AM, Bruce Richardson wrote: > > >The functions rte_eth_rx_queue_count and rte_eth_descriptor_done are > > >supported by very few PMDs. Therefore, it is best to check for support > > >for the functions in the ethdev library, so as to avoid crashes > > >at run-time if the application goes to use those APIs. The performance > > >impact of this change should be very small as this is a predictable > > >branch in the function. > > > > > >Signed-off-by: Bruce Richardson > > >--- > > > lib/librte_ether/rte_ethdev.h | 8 ++-- > > > 1 file changed, 6 insertions(+), 2 deletions(-) > > > > > >diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h > > >index 827ca3e..9ad1b6a 100644 > > >--- a/lib/librte_ether/rte_ethdev.h > > >+++ b/lib/librte_ether/rte_ethdev.h > > >@@ -2496,6 +2496,8 @@ rte_eth_rx_burst(uint8_t port_id, uint16_t queue_id, > > > * The queue id on the specific port. > > > * @return > > > * The number of used descriptors in the specific queue. > > >+ * NOTE: if function is not supported by device this call > > >+ *returns (uint32_t)-ENOTSUP > > > */ > > > static inline uint32_t > > > > Why not change the return type to int32_t? > > In this way, the caller isn't required to make the assumption that a large > > queue count indicates an error. < 0 means error, other wise it's a valid > > queue count. > > > > This approach would be consistent with other APIs. > > > > Yes, good point, I should see about that. One thing I'm unsure of, though, is > does this count as ABI breakage? I don't see how it should break any older > apps, since the return type is the same size, but I'm not sure as we are > changing the return type of the function. > > Neil, can you perhaps comment here? Is changing uint32_t to int32_t ok, from > an ABI point of view? > > Regards, > /Bruce
[dpdk-dev] [PATCH v3 03/12] hash: check SSE flags only on x86 builds
On Mon, Jul 06, 2015 at 04:51:28PM +0800, Zhigang Lu wrote: > This is necessary because the required CPU flags may not be defined on > other architectures. > > Change-Id: I14d3f9f625b2e7567123f1c97095f8d06abd674b > Signed-off-by: Zhigang Lu Acked-by: Bruce Richardson
[dpdk-dev] [PATCH v3 04/12] eal: allow empty compile time flags
On Mon, Jul 06, 2015 at 04:51:29PM +0800, Zhigang Lu wrote: > The rte_cpu_check_supported() code breaks with a "comparison is always > false due to limited range of data type" when the compile_time_flags[] > array is empty. Assigning the array dimension to a local variable > apparently solves this. > > Change-Id: I0ae21f529cf7b6dd9cf0f4532dce9198f4bf4230 > Signed-off-by: Zhigang Lu Acked-by: Bruce Richardson
[dpdk-dev] [PATCH v3 05/12] config: remove RTE_LIBNAME definition.
On Mon, Jul 06, 2015 at 04:51:30PM +0800, Zhigang Lu wrote: > The library name is now being pinned to "dpdk" instead of intel_dpdk, > powerpc_dpdk, etc. As a result, we no longer need this config item. > This patch removes it. > > Change-Id: I36f7cf6c18c3563c6f5ccdf01bb70579c7ccaa16 > Signed-off-by: Zhigang Lu Acked-by: Bruce Richardson
[dpdk-dev] [PATCH v3 08/12] mempool: allow config override on element alignment
On Mon, Jul 06, 2015 at 04:51:33PM +0800, Zhigang Lu wrote: > On TILE-Gx and TILE-Mx platforms, the buffers fed into the hardware > buffer manager require a 128-byte alignment. With this change, we > allow configuration based override of the element alignment, and > default to RTE_CACHE_LINE_SIZE if left unspecified. > > Change-Id: I9cd789d92b0bc9c8f44a633de59bb04d45d927a7 > Signed-off-by: Zhigang Lu This looks an OK change. However, would it be worthwhile making this a runtime parameter rather than a compile-time one? Is it likely that we will ever have a case where someone wants two mempools with different alignments (and where using the larger of the two would be problematic)? /Bruce
[dpdk-dev] [PATCH v2] vfio: Fix overflow while assigning vfio BAR region offset and size
Hi all, >From the kernel VFIO maintainer: "I suppose in the short term, mmap should not be advertised as available on 32bit hosts. Thanks," So, as VFIO support for 32bit systems is broken, DPDK should not configure VFIO in that case. This is the complete email sent to the kernel maintainer and his answer: On Thu, 2015-07-02 at 14:42 +0100, Alejandro Lucero wrote: > Hi Alex, > > is VFIO expected to work in 32 bit systems? > > I know VFIO initial goal was a better control of device assignment to > virtual machines and it is based on IOMMU hardware support. I do not know > how likely is to have a 32 bit system with IOMMU but it seems to be > possible such hardware configuration. > > The problem with VFIO and 32 bit systems is the VFIO kernel code uses the > upper bits (VFIO_PCI_OFFSET_SHIFT) of a __u64 variable, offset field in > struct vfio_region_info, for saving info about the PCI BAR index to work > with. This is done inside the ioctl command VFIO_DEVICE_GET_REGION_INFO. > That vfio_region_info is got by the process doing the system call and the > offset is used as a parameter for mmap system call which expects such a > parameter as unsigned long. If I am not wrong, unsigned long in 32 bit > linux systems is a 32 bit type, so when the vfio_pci_mmap function is > executed, the index BAR to work with is obtained from the offset, which > turns to be always 0 as the value was "lost in translation". There is a > chance current implementation can work if all the PCI BARs are equal in > terms of size, but obviously this is not acceptable. > > So, if VFIO needs to work in 32 bits systems another way to map the device > PCI BARs is needed. Not necessarily, VFIO_PCI_OFFSET_SHIFT is an internal implementation detail, userspace should always use vfio_region_info.offset. We're therefore free to come up with other algorithms for handling this limitation. If we want to continue using a single device file descriptor, we could simply choose a smaller shift on 32bit systems. A 29 bit shift would give us 512MB regions support, which is sufficient for the vast majority of devices. We could also replace the macros with functions such that we pack regions as tightly as possible within the device file descriptor. It's reasonable to expect that we could support up to 2G BARs using such a method. A hybrid approach is also possible, for instance the config space region could also contain the ROM and VGA areas (or we could simply choose not to support VGA on 32bit hosts). If we need to support 4G BARs, our only choice is really to extend the vfio region support for a separate file descriptor per region. The only devices I'm aware of with 4G BARs are Nvidia Tesla. This is possible, but I would expect such devices would be extremely rare on 32bit hosts. I suppose in the short term, mmap should not be advertised as available on 32bit hosts. Thanks, Alex On Wed, Jul 1, 2015 at 11:00 AM, Burakov, Anatoly wrote: > Hi all, > > > The last patch from Rahul does not solve the problem. For those cases > where the MSI-X table is in one of the BARs to map, the memreg array is > still in use. > > Rahul's initial patch was pretty much what you have submitted, it just > didn't build on a 32-bit system. > > > My fix was using unsigned long instead of uint32_t for the memreg array > as this is used as a parameter for mmap system call which expects such a > type for the offset (and size). > > Maybe use off_t? That would at least be guaranteed to compile on any > system... > > > In a 32-bit system mmap system call and VFIO mmap implementation will > get an unsigned long offset, as it does the struct vma_area_struct for > vm_pgoff. > > VFIO will not be able to map the right BAR except for BAR 0. > > > > So, basically, VFIO kernel code does not work for 32 bit systems. > > > > I think we should define memreg as unsigned long and to report this > problem to the VFIO kernel maintainer. > > If that's the case, this should indeed be taken up with the kernel > maintainers. I don't have a 32-bit system handy to test it, unfortunately. > > Thanks, > Anatoly >
[dpdk-dev] [PATCH 1/2] eal:Introduce rte_dma_wmb/rte_dma_rmb.
Hi Konstantin,
Yes, it's better to use rte_rmb/rte_wmb instead of their source code.
Actually, the warning is a bug, it lose a "}" of rte_dma_rmb(), I didn't
test it on PowerPC, so the bug didn't be found out before I submit the
patch, sorry.
I'll resubmit this patch.
Dong
> Hi Dong,
>
>> -Original Message-
>> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of WangDong
>> Sent: Sunday, June 28, 2015 4:23 PM
>> To: dev at dpdk.org
>> Subject: [dpdk-dev] [PATCH 1/2] eal:Introduce rte_dma_wmb/rte_dma_rmb.
>>
>> These macro can be used to replace current PMD's compiler memory barrier
>> (volatile varible) and rte_wmb.
>> In x86, they implement to compiler memory barrier.
>> In power, they implement to processor memory barrier.
>>
>> ---
>> .../common/include/arch/ppc_64/rte_atomic.h| 4
>> .../common/include/arch/x86/rte_atomic.h | 4
>> lib/librte_eal/common/include/generic/rte_atomic.h | 25
>> ++
>> 3 files changed, 33 insertions(+)
>>
>> diff --git a/lib/librte_eal/common/include/arch/ppc_64/rte_atomic.h
>> b/lib/librte_eal/common/include/arch/ppc_64/rte_atomic.h
>> index fb7af2b..8f4129d 100644
>> --- a/lib/librte_eal/common/include/arch/ppc_64/rte_atomic.h
>> +++ b/lib/librte_eal/common/include/arch/ppc_64/rte_atomic.h
>> @@ -72,6 +72,10 @@ extern "C" {
>>*/
>> #definerte_rmb() {asm volatile("sync" : : : "memory"); }
>>
>> +#define rte_dma_wmb() {asm volatile("sync" : : : "memory"); }
>> +
>> +#define rte_dma_rmb() {asm volatile("sync" : : : "memory");
>
> As a nit, probably better:
>
> +#define rte_dma_wmb() rte_rmb()
> +#define rte_dma_rmb()rte_wmb()
>
> Here?
>
> Konstantin
>
>> +
>> /*- 16 bit atomic operations
>> -*/
>> /* To be compatible with Power7, use GCC built-in functions for 16 bit
>>* operations */
>> diff --git a/lib/librte_eal/common/include/arch/x86/rte_atomic.h
>> b/lib/librte_eal/common/include/arch/x86/rte_atomic.h
>> index e93e8ee..7cfbe8f 100644
>> --- a/lib/librte_eal/common/include/arch/x86/rte_atomic.h
>> +++ b/lib/librte_eal/common/include/arch/x86/rte_atomic.h
>> @@ -53,6 +53,10 @@ extern "C" {
>>
>> #definerte_rmb() _mm_lfence()
>>
>> +#define rte_dma_wmb() rte_compiler_barrier()
>> +
>> +#define rte_dma_rmb() rte_compiler_barrier()
>> +
>> /*- 16 bit atomic operations
>> -*/
>>
>> #ifndef RTE_FORCE_INTRINSICS
>> diff --git a/lib/librte_eal/common/include/generic/rte_atomic.h
>> b/lib/librte_eal/common/include/generic/rte_atomic.h
>> index 6c7581a..a51 100644
>> --- a/lib/librte_eal/common/include/generic/rte_atomic.h
>> +++ b/lib/librte_eal/common/include/generic/rte_atomic.h
>> @@ -72,6 +72,31 @@ static inline void rte_wmb(void);
>>*/
>> static inline void rte_rmb(void);
>>
>> +/**
>> + * Write memory barrier for DMA.
>> + *
>> + * Be used in PMD, unlike rte_wmb() which use processor memory barrier,
>> + * this memory barrier focus on performance, if compiler memory barrier
>> + * is sufficient for guarantee memory ordering, this function will
>> + * use compiler memory barrier.
>> + *
>> + * This function is architecture dependent.
>> + */
>> +static inline void rte_dma_wmb(void);
>> +
>> +/**
>> + * Read memory barrier for DMA.
>> + *
>> + * Be used in PMD, unlike rte_rmb() which use processor memory barrier,
>> + * this memory barrier focus on performance, if compiler memory barrier
>> + * is sufficient for guarantee memory ordering, this function will
>> + * use compiler memory barrier.
>> + *
>> + * This function is architecture dependent.
>> + */
>> +static inline void rte_dma_rmb(void);
>> +
>> +
>> #endif /* __DOXYGEN__ */
>>
>> /**
>> --
>> 2.1.0
>
[dpdk-dev] [PATCH] ethdev: fix checking for tx_free_thresh
> > This parameter is not consistent between the drivers: some use it as > > rte_eth_tx_burst() requires, some release buffers when the number of free > > descriptors drop below this value. > > Let's use it as most fast-path code does, which is the latter, and update > > comments throughout the code to reflect that. > > > > Signed-off-by: Zoltan Kiss > > Acked-by: Konstantin Ananyev Applied, thanks
[dpdk-dev] [PATCH 0/2] Native uio-based PMD for Mellanox ConnectX-3 devices
2015-07-07 00:57, Keunhong Lee: > Answer 1. UIO based driver is faster then ib based driver. > It can saturate 40G link with MTU sized packets using a single thread while > ib wrapper cannot. OK, interesting. Do you have numbers and details about your testbed/scenario? > Answer 2. Sorry, I missed that. I'll make a new patch email with my real > name. > > Question 1. Is it OK if I separate GPL-based and BSD-based codes into > separated patches? > mlx4 kernel driver itself is dual licenses, so I think they are considered > as BSD in my source code. > The only source code under GPL is bitmap, integer logarithm, and red-black > tree contained in mlnx_uio/kernel directory. These parts will be built in the user-space driver library, right? It would change the license, which is not desirable. Technically, your approach may be interesting. But from a maintenance point of view, this huge codebase may be a nightmare. > 2015-07-06 23:17 GMT+09:00 Thomas Monjalon : > > > 2015-07-06 22:28, leeopop: > > > This is a native UIO-based PMD for Mellanox ConnectX-3 devices. > > > It uses a persistent memory library in order to provide a persistent > > > scartch area for the mlx4 HCA driver. > > > > What is the benefit of this UIO approach compared to the OFED based driver? > > > > > We release the driver itself under BSD license, but to use it for > > > commercial products, you may have to re-implement the separated GPL > > > sources. > > > > The GPL sources are not really separated. > > The DPDK libraries must be BSD-licensed. > > > > > The GPL affected source codes reside in the mlnx_uio/kernel directory. > > > > It seems that a large part of the GPL driver was also copied in > > mlnx_uio/mlnx/. > > > > Given that you are dropping a huge GPL codebase (whose you don't own the > > copyright) in a BSD library, and that you didn't give your real name in > > the signed-off line, it is NACK.
[dpdk-dev] UIO RTE_INTR_MODE_NONE issue.
On Sat, 04 Jul 2015 14:36:15 +0200 Thomas Monjalon wrote: > Hi Stephen, > > 2015-07-03 12:58, Stephen Hemminger: > > PPS: With DPDK 2.2 it would be good to move the minimum kernel version up > > to 3.0 > > since that is what current stable distro's are using. > > Yes, this is something to consider. > Please send a doc patch to open the discussion. Something like?? diff --git a/doc/guides/rel_notes/faq.rst b/doc/guides/rel_notes/faq.rst index d87230a..715a03c 100644 --- a/doc/guides/rel_notes/faq.rst +++ b/doc/guides/rel_notes/faq.rst @@ -175,6 +175,9 @@ Any kernel greater than version 2.6.33 can be used without any patches applied. (not mandatory, but recommended on a NUMA system to support per-NUMA node hugepages allocation) +If starting a new project it worth using the latest stable version of a distribution. +The next version of the DPDK 2.2 will require kernel version 3.0 or later. + .. note:: Blue text in the lists above are direct links to the patch downloads.
[dpdk-dev] [PATCH] mk: enable next abi in static libs
On Mon, Jul 06, 2015 at 03:49:50PM +0200, Thomas Monjalon wrote: > 2015-07-06 09:35, Neil Horman: > > On Mon, Jul 06, 2015 at 03:18:51PM +0200, Thomas Monjalon wrote: > > > Any comment or ack? > > > > > > 2015-07-03 00:05, Thomas Monjalon: > > > > When a change makes really hard to keep ABI compatibility, > > > > instead of waiting next release to break the ABI, it is smoother > > > > to introduce the new code and enable it only for static libraries. > > > > The flag RTE_NEXT_ABI may be used to "ifdef" the new code. > > > > When the release is out, a dynamically linked application can use > > > > the new shared libraries without rebuild while developpers can prepare > > > > their application for the next ABI by reading the deprecation notice > > > > and easily testing the new code. > > > > When starting the next release cycle, the "ifdefs" will be removed > > > > and the ABI break will be marked by incrementing LIBABIVER. > > > > > > > > The new option CONFIG_RTE_NEXT_ABI is not defined in the configuration > > > > templates because it is deduced from CONFIG_RTE_BUILD_SHARED_LIB. > > > > It is automatically enabled for static libraries and disabled for > > > > shared libraries. > > > > It can be forced to another value by editing the generated .config file. > > > > It shouldn't be enabled for shared libraries because it would break the > > > > ABI without changing the version number LIBABIVER. That's why a warning > > > > is printed in this case. > > > > > > > > The guideline is also updated to integrate this new possibility. > > > > > > > > Signed-off-by: Thomas Monjalon > > > > > > > > Yeah, I'm not sure why this is necessecary. That is to say, if you want to > > It's explained at the beginning: > "When a change makes really hard to keep ABI compatibility", e.g. mbuf change. > Thats not what I was referring to. I was referring to the need to split out ABI's based on a separate config item only for static libraries. I understand that sometimes you want a 'preview' of the next abi. > > introduce a new ABI operation prior to the old one being removed, that is > > precisely what > > the versioning macros are for, and can be used to map the static api to the > > new > > version. e.g, given function X that you want to enhance in an ABI breaking > > way: > > > > 1) Separate function X to X_v1 and X_v2 > > 2) Map X_v2 to X at DPDK_v2, map X_v1 to X at DPDK_v1 > > 3) Map the static symbol X to X_v2 > > 4) Post the deprecation notice of X for release 3 immediately > > We cannot do that for mbuf change. > You can actually, its just alot of work. Also, I know this doesn't relate very closely to the subject, and I apologize, I was really just reacting to the immediately preceding sentence in the origional post. > > Splitting the static ABI from the shared ABI just means that applications > > will > > have the opportunity to isolate themselves to one kind of build, which is a > > bad > > idea. > > It helps to be prepared for the next release by testing the app with static > libraries. > We agreed to allow API breaking for important changes like mbuf rework. > This option NEXT_ABI is a step between announcement and effective ABI > breaking. > > I think it's a reasonnable approach. But if nobody ack it, I'm perfectly OK to > drop it and related features (unified packet type and interrupt mode). > I'd be ok with it iff: 1) It applies to static and shared ABI's together. That is to say that setting the NEXT_ABI config flag creates the same ABI changes regardless of other build configuration. It needs to be used in such a way that a consistent ABI is presented when set, otherwise it won't be useful. 2) It only applies to the next ABI. That is to say, it can't be a hodgepodge of the next ABI and the one after that, and the one after that, or it won't provide an appropriate preview for anyone. If we can meet those two standards, it would likely be a useful feature to have. Neil >
[dpdk-dev] [PATCH 1/2] eal/persistent: new library to hold memory region after program exit
Am I right, this library is using Unix shared memory to get
persistent storage. That is going to be slower and more worrying,
the kernel makes no guarantees that virtual to physical mapping
will not change.
There is also an ABI issue in this patch. You are introducing a new
API here, so the symbol should go in a new 2.1 section
> + if((len / RTE_PGSIZE_4K) > 1)
> + {
> + shmget_flag |= SHM_HUGETLB;
> + }
Please follow kernel coding style. This is one of many examples that
needs to be fixed before accepting.
ERROR: "foo* bar" should be "foo *bar"
#166: FILE: lib/librte_eal/common/include/rte_pci.h:210:
+ void* priv; /**< Private data. */
ERROR: "foo* bar" should be "foo *bar"
#195: FILE: lib/librte_eal/common/include/rte_persistent_mem.h:20:
+extern void*
persistent_allocated_memory[RTE_MAX_NUMA_NODES][RTE_EAL_PERSISTENT_MEM_COUNT];
ERROR: "foo* bar" should be "foo *bar"
#302: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:44:
+static void* reserve_shared_zone(int subindex, uint32_t len, int socket_id)
ERROR: do not use C99 // comments
#307: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:49:
+ int shmget_flag = IPC_CREAT | SHM_R | SHM_W | IPC_EXCL; // | SHM_LOCKED;
WARNING: Missing a blank line after declarations
#310: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:52:
+ int err;
+ if((len / RTE_PGSIZE_4K) > 1)
ERROR: that open brace { should be on the previous line
#310: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:52:
+ if((len / RTE_PGSIZE_4K) > 1)
+ {
ERROR: space required before the open parenthesis '('
#310: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:52:
+ if((len / RTE_PGSIZE_4K) > 1)
ERROR: "foo* bar" should be "foo *bar"
#316: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:58:
+ void* addr = 0;
WARNING: Missing a blank line after declarations
#318: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:60:
+ int clear = 1;
+ if(shmid < 0)
ERROR: that open brace { should be on the previous line
#318: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:60:
+ if(shmid < 0)
+ {
ERROR: space required before the open parenthesis '('
#318: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:60:
+ if(shmid < 0)
ERROR: do not use C99 // comments
#320: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:62:
+ //Reuse existing
ERROR: that open brace { should be on the previous line
#328: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:70:
+ if(socket_id != SOCKET_ID_ANY)
+ {
ERROR: space required before the open parenthesis '('
#328: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:70:
+ if(socket_id != SOCKET_ID_ANY)
ERROR: "foo * bar" should be "foo *bar"
#330: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:72:
+ struct bitmask * mask = numa_bitmask_alloc(RTE_MAX_NUMA_NODES);
WARNING: Missing a blank line after declarations
#331: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:73:
+ struct bitmask * mask = numa_bitmask_alloc(RTE_MAX_NUMA_NODES);
+ mask = numa_bitmask_clearall(mask);
ERROR: that open brace { should be on the previous line
#336: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:78:
+ if(ret < 0)
+ {
ERROR: space required before the open parenthesis '('
#336: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:78:
+ if(ret < 0)
ERROR: that open brace { should be on the previous line
#344: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:86:
+ if(clear)
+ {
ERROR: space required before the open parenthesis '('
#344: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:86:
+ if(clear)
WARNING: Missing a blank line after declarations
#350: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:92:
+ size_t size;
+ volatile uint8_t reader = 0; //this prevents from being optimized out
ERROR: do not use C99 // comments
#350: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:92:
+ volatile uint8_t reader = 0; //this prevents from being optimized out
WARNING: Use of volatile is usually wrong: see
Documentation/volatile-considered-harmful.txt
#350: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:92:
+ volatile uint8_t reader = 0; //this prevents from being optimized out
ERROR: "(foo*)" should be "(foo *)"
#351: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:93:
+ volatile uint8_t* readp = (uint8_t*)addr;
ERROR: "foo* bar" should be "foo *bar"
#351: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:93:
+ volatile uint8_t* readp = (uint8_t*)addr;
WARNING: Use of volatile is usually wrong: see
Documentation/volatile-considered-harmful.txt
#351: FILE: lib/librte_eal/linuxapp/eal/eal_persistent_mem.c:93:
+ volatile uint8_t* readp = (uint8_t*)addr;
WARNING: Missing a blan
[dpdk-dev] [dpdk-announce] DPDK Summit, San Francisco 2015 is open for registration
You are invited to DPDK Summit, San Francisco 2015! Date:Monday, August 17, 2015 Time:8:30 AM - 5:30 PM (Summit) 5:30 PM - 7:30 PM (Reception) Location: The Westin St. Francis 335 Powell Street San Francisco, CA 94102 DPDK Summit, San Francisco 2015 is announced to commence on August 17th. If you are new in the community, these events make a great occasion for our DPDK open source community to get together face-to-face around the world for a forward looking dialogue. Based on your feedback last year, we added an evening reception for you all to continue your dialogue throughout the day and into the evening over drinks. Register today at https://dpdksummit.com/ Seating is limited and offered on a first-come basis. Register now to reserve your complimentary seat. Learn more about DPDK Summit at https://dpdksummit.com/us/en/about or visit http://dpdk.org/events For more information regarding this event, please contact support at dpdksummit.com. Thanks, -Pravin
[dpdk-dev] [PATCH] mk: enable next abi in static libs
2015-07-06 14:22, Neil Horman: > On Mon, Jul 06, 2015 at 03:49:50PM +0200, Thomas Monjalon wrote: > > 2015-07-06 09:35, Neil Horman: > > > On Mon, Jul 06, 2015 at 03:18:51PM +0200, Thomas Monjalon wrote: > > > > Any comment or ack? > > > > > > > > 2015-07-03 00:05, Thomas Monjalon: > > > > > When a change makes really hard to keep ABI compatibility, > > > > > instead of waiting next release to break the ABI, it is smoother > > > > > to introduce the new code and enable it only for static libraries. > > > > > The flag RTE_NEXT_ABI may be used to "ifdef" the new code. > > > > > When the release is out, a dynamically linked application can use > > > > > the new shared libraries without rebuild while developpers can prepare > > > > > their application for the next ABI by reading the deprecation notice > > > > > and easily testing the new code. > > > > > When starting the next release cycle, the "ifdefs" will be removed > > > > > and the ABI break will be marked by incrementing LIBABIVER. > > > > > > > > > > The new option CONFIG_RTE_NEXT_ABI is not defined in the configuration > > > > > templates because it is deduced from CONFIG_RTE_BUILD_SHARED_LIB. > > > > > It is automatically enabled for static libraries and disabled for > > > > > shared libraries. > > > > > It can be forced to another value by editing the generated .config > > > > > file. > > > > > It shouldn't be enabled for shared libraries because it would break > > > > > the > > > > > ABI without changing the version number LIBABIVER. That's why a > > > > > warning > > > > > is printed in this case. > > > > > > > > > > The guideline is also updated to integrate this new possibility. [...] > I'd be ok with it iff: > > 1) It applies to static and shared ABI's together. That is to say that > setting > the NEXT_ABI config flag creates the same ABI changes regardless of other > build > configuration. It needs to be used in such a way that a consistent ABI is > presented when set, otherwise it won't be useful. Yes the option trigger exactly the same ABI for static and shared libraries. But it's too complicated (at least for 2.1) to make LIBABIVER and version map dependant of this build-time option. That's why, it should not be enabled to deploy shared libraries, though it can be used for tests and development. As static libraries are almost never packaged, they will be built and linked at the same time. That's why users of static libraries tend to prefer the newest ABI, which is the default in this case. > 2) It only applies to the next ABI. That is to say, it can't be a hodgepodge > of > the next ABI and the one after that, and the one after that, or it won't > provide > an appropriate preview for anyone. If you mean the next ABI must be promoted as the standard ABI in the next release, yes: ifdefs will be cleaned when starting a new release. Thanks, I learnt the english word hodgepodge :)
