From: "Li, Xiaoyun" <xiaoyun.li@intel.com>
To: "Xia, Chenbo" <chenbo.xia@intel.com>, "dev@dpdk.org" <dev@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH 2/8] app/testpmd: use PCI memory resource access APIs
Date: Thu, 16 Sep 2021 06:10:23 +0000 [thread overview]
Message-ID: <DM4PR11MB55340583E99443679B8132C899DC9@DM4PR11MB5534.namprd11.prod.outlook.com> (raw)
In-Reply-To: <20210910022402.26620-3-chenbo.xia@intel.com>
Hi
> -----Original Message-----
> From: Xia, Chenbo <chenbo.xia@intel.com>
> Sent: Friday, September 10, 2021 10:24
> To: dev@dpdk.org
> Cc: Li, Xiaoyun <xiaoyun.li@intel.com>
> Subject: [PATCH 2/8] app/testpmd: use PCI memory resource access APIs
>
> Currently testpmd uses struct rte_pci_device to access PCI memory resource.
> Since this structure will be internal later, this patch replaces use of
> rte_pci_device with new PCI memory resource access APIs to read/write BAR 0.
>
> Signed-off-by: Chenbo Xia <chenbo.xia@intel.com>
> ---
> app/test-pmd/config.c | 38 +++++++++++-----------------------
> app/test-pmd/testpmd.h | 46 +++++++++++++++++++++---------------------
> 2 files changed, 35 insertions(+), 49 deletions(-)
>
<snip>
>
> @@ -1081,7 +1063,9 @@ port_reg_bit_set(portid_t port_id, uint32_t reg_off,
> uint8_t bit_pos,
> (int) bit_v);
> return;
> }
> - reg_v = port_id_pci_reg_read(port_id, reg_off);
> + if (port_id_pci_reg_read(port_id, reg_off, ®_v))
> + return;
> +
> if (bit_v == 0)
> reg_v &= ~(1 << bit_pos);
> else
> @@ -1123,7 +1107,9 @@ port_reg_bit_field_set(portid_t port_id, uint32_t
> reg_off,
> (unsigned)max_v, (unsigned)max_v);
> return;
> }
> - reg_v = port_id_pci_reg_read(port_id, reg_off);
> + if (port_id_pci_reg_read(port_id, reg_off, ®_v))
> + return;
> +
> reg_v &= ~(max_v << l_bit); /* Keep unchanged bits */
> reg_v |= (value << l_bit); /* Set changed bits */
> port_id_pci_reg_write(port_id, reg_off, reg_v); diff --git a/app/test-
In your implementation, port_id_pci_reg_write() can fail. Then the following display shouldn't be called.
So change this function to have return value and check it before use it will be better.
> pmd/testpmd.h b/app/test-pmd/testpmd.h index 16a3598e48..7922807a6e
> 100644
> --- a/app/test-pmd/testpmd.h
> +++ b/app/test-pmd/testpmd.h
<snip>
> static inline void
> -port_pci_reg_write(struct rte_port *port, uint32_t reg_off, uint32_t reg_v)
> +port_id_pci_reg_write(portid_t pt_id, uint32_t reg_off, uint32_t reg_v)
> {
> - const struct rte_pci_device *pci_dev;
> + struct rte_port *port = &ports[(pt_id)];
> + char name[RTE_ETH_NAME_MAX_LEN];
> const struct rte_bus *bus;
> - void *reg_addr;
>
> if (!port->dev_info.device) {
> fprintf(stderr, "Invalid device\n");
> @@ -721,19 +721,19 @@ port_pci_reg_write(struct rte_port *port, uint32_t
> reg_off, uint32_t reg_v)
>
> bus = rte_bus_find_by_device(port->dev_info.device);
> if (bus && !strcmp(bus->name, "pci")) {
> - pci_dev = RTE_DEV_TO_PCI(port->dev_info.device);
> + rte_eth_dev_get_name_by_port(pt_id, name);
> } else {
> fprintf(stderr, "Not a PCI device\n");
> return;
> }
>
> - reg_addr = ((char *)pci_dev->mem_resource[0].addr + reg_off);
> - *((volatile uint32_t *)reg_addr) = rte_cpu_to_le_32(reg_v);
> + reg_v = rte_cpu_to_le_32(reg_v);
> + if (rte_pci_mem_wr32(name, 0, ®_v, reg_off)) {
> + fprintf(stderr, "Failed to write register\n");
> + return;
> + }
> }
>
> -#define port_id_pci_reg_write(pt_id, reg_off, reg_value) \
> - port_pci_reg_write(&ports[(pt_id)], (reg_off), (reg_value))
> -
> static inline void
> get_start_cycles(uint64_t *start_tsc)
> {
> --
> 2.17.1
next prev parent reply other threads:[~2021-09-16 6:10 UTC|newest]
Thread overview: 59+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-09-10 2:23 [dpdk-dev] [PATCH 0/8] Removal of PCI bus ABIs Chenbo Xia
2021-09-10 2:23 ` [dpdk-dev] [PATCH 1/8] bus/pci: add new memory resource access APIs Chenbo Xia
2021-09-13 11:59 ` Kinsella, Ray
2021-09-10 2:23 ` [dpdk-dev] [PATCH 2/8] app/testpmd: use PCI " Chenbo Xia
2021-09-16 6:10 ` Li, Xiaoyun [this message]
2021-09-16 6:38 ` Xia, Chenbo
2021-09-10 2:23 ` [dpdk-dev] [PATCH 3/8] examples/ethtool: use PCI library API to get PCI address Chenbo Xia
2021-09-10 2:23 ` [dpdk-dev] [PATCH 4/8] examples/kni: remove unused PCI bus header Chenbo Xia
2021-09-17 15:38 ` Ferruh Yigit
2021-09-10 2:23 ` [dpdk-dev] [PATCH 5/8] test/kni: remove setting of PCI ID and address Chenbo Xia
2021-09-10 7:12 ` David Marchand
2021-09-17 15:38 ` Ferruh Yigit
2021-09-10 2:24 ` [dpdk-dev] [PATCH 6/8] examples/ip_pipeline: " Chenbo Xia
2021-09-10 7:18 ` David Marchand
2021-09-10 8:21 ` Xia, Chenbo
2021-09-17 3:09 ` Xia, Chenbo
2021-09-17 11:55 ` David Marchand
2021-09-17 15:37 ` Ferruh Yigit
2021-09-10 2:24 ` [dpdk-dev] [PATCH 7/8] kni: replace unused variable definition with reserved bytes Chenbo Xia
2021-09-10 2:24 ` [dpdk-dev] [PATCH 8/8] bus/pci: remove ABIs in PCI bus Chenbo Xia
2021-09-13 12:06 ` Kinsella, Ray
2021-09-14 8:15 ` Xu, Rosen
2021-09-18 2:24 ` [dpdk-dev] [PATCH v2 0/7] Removal of PCI bus ABIs Chenbo Xia
2021-09-18 2:24 ` [dpdk-dev] [PATCH v2 1/7] bus/pci: add new memory resource access APIs Chenbo Xia
2021-09-18 2:24 ` [dpdk-dev] [PATCH v2 2/7] app/testpmd: use PCI " Chenbo Xia
2021-09-18 2:44 ` Li, Xiaoyun
2021-09-18 2:24 ` [dpdk-dev] [PATCH v2 3/7] examples/ethtool: use PCI library API to get PCI address Chenbo Xia
2021-09-18 2:24 ` [dpdk-dev] [PATCH v2 4/7] examples/kni: remove unused PCI bus header Chenbo Xia
2021-09-18 2:24 ` [dpdk-dev] [PATCH v2 5/7] kni: remove unused PCI info from test and example Chenbo Xia
2021-09-18 2:24 ` [dpdk-dev] [PATCH v2 6/7] kni: replace unused variable definition with reserved bytes Chenbo Xia
2021-09-18 2:24 ` [dpdk-dev] [PATCH v2 7/7] bus/pci: remove ABIs in PCI bus Chenbo Xia
2021-09-29 7:38 ` [dpdk-dev] [PATCH v2 0/7] Removal of PCI bus ABIs Xia, Chenbo
2021-09-30 8:45 ` David Marchand
2021-10-04 13:37 ` David Marchand
2021-10-04 15:56 ` Harris, James R
2021-10-06 4:25 ` Xia, Chenbo
2021-10-08 6:15 ` Liu, Changpeng
2021-10-08 7:08 ` David Marchand
2021-10-08 7:44 ` Liu, Changpeng
2021-10-11 6:58 ` Xia, Chenbo
2021-10-11 12:55 ` Thomas Monjalon
2021-10-12 0:35 ` Harris, James R
2021-10-12 7:04 ` Thomas Monjalon
2021-10-12 16:59 ` Walker, Benjamin
2021-10-12 18:43 ` Thomas Monjalon
2021-10-12 19:26 ` Walker, Benjamin
2021-10-12 21:50 ` Thomas Monjalon
2021-10-13 17:56 ` Walker, Benjamin
2021-10-13 18:59 ` Thomas Monjalon
2021-10-13 22:48 ` Walker, Benjamin
2021-10-14 6:41 ` Thomas Monjalon
2022-07-11 12:11 ` Thomas Monjalon
2021-10-14 2:21 ` Xia, Chenbo
2021-10-14 6:41 ` Thomas Monjalon
2021-10-14 7:00 ` Xia, Chenbo
2021-10-14 7:07 ` Thomas Monjalon
2021-10-14 8:07 ` Xia, Chenbo
2021-10-14 8:25 ` Thomas Monjalon
2021-10-27 12:03 ` Xia, Chenbo
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=DM4PR11MB55340583E99443679B8132C899DC9@DM4PR11MB5534.namprd11.prod.outlook.com \
--to=xiaoyun.li@intel.com \
--cc=chenbo.xia@intel.com \
--cc=dev@dpdk.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).