Re: [PATCH v1] net/axgbe: use PCI root complex device to distinguish AMD hardware

[Awan, Arsalan]

Hi Chandu,

I have tested this in the following configurations and it works very well in 
all cases:

* V1000 with Yocto v3.3.2 & Linux v5.10.47
* E3000 with Yocto v3.3.2 & Linux v5.10.47
* V1000 with Ubuntu Server v20.04.3 LTS
* E3000 with Ubuntu Server v20.04.3 LTS

I think this is a much better approach than scanning the whole PCI bus and 
comparing
each and every device against the PCI Root Complex Device ID of the platform.

This solution looks good to me!

Acked by: Arsalan H. Awan 

From: Chandubabu Namburu 
Sent: Friday, November 26, 2021 3:24 PM
To: dev@dpdk.org
Cc: selwin.sebast...@amd.com; Awan, Arsalan; Chandubabu Namburu
Subject: [PATCH v1] net/axgbe: use PCI root complex device to distinguish AMD 
hardware

"bus/pci: optimize bus scan" broke axgbe on V1000/R1000.
RV root complex pci device does not have any kernel driver assigned
so it is removed from pci scan list which is used in
"net/axgbe: add a HW quirk for register definitions"

Get root complex device id directly from pci sysfs instead of
pci scan list

Signed-off-by: Chandubabu Namburu 
---
 drivers/net/axgbe/axgbe_ethdev.c | 39 
 1 file changed, 20 insertions(+), 19 deletions(-)

diff --git a/drivers/net/axgbe/axgbe_ethdev.c b/drivers/net/axgbe/axgbe_ethdev.c
index 9cd056d04a..26babde354 100644
--- a/drivers/net/axgbe/axgbe_ethdev.c
+++ b/drivers/net/axgbe/axgbe_ethdev.c
@@ -10,6 +10,8 @@
 #include "axgbe_regs.h"
 #include "rte_time.h"

+#include "eal_filesystem.h"
+
 static int eth_axgbe_dev_init(struct rte_eth_dev *eth_dev);
 static int  axgbe_dev_configure(struct rte_eth_dev *dev);
 static int  axgbe_dev_start(struct rte_eth_dev *dev);
@@ -1923,28 +1925,27 @@ static void axgbe_default_config(struct axgbe_port 
*pdata)
pdata->power_down = 0;
 }

-static int
-pci_device_cmp(const struct rte_device *dev, const void *_pci_id)
+/*
+ * Return PCI root complex device id on success else 0
+ */
+static uint16_t
+get_pci_rc_devid(void)
 {
-   const struct rte_pci_device *pdev = RTE_DEV_TO_PCI_CONST(dev);
-   const struct rte_pci_id *pcid = _pci_id;
+   char pci_sysfs[PATH_MAX];
+   const struct rte_pci_addr pci_rc_addr = {0, 0, 0, 0};
+   unsigned long device_id;

-   if (pdev->id.vendor_id == AMD_PCI_VENDOR_ID &&
-   pdev->id.device_id == pcid->device_id)
-   return 0;
-   return 1;
-}
+   snprintf(pci_sysfs, sizeof(pci_sysfs), "%s/" PCI_PRI_FMT "/device",
+rte_pci_get_sysfs_path(), pci_rc_addr.domain,
+pci_rc_addr.bus, pci_rc_addr.devid, pci_rc_addr.function);

-static bool
-pci_search_device(int device_id)
-{
-   struct rte_bus *pci_bus;
-   struct rte_pci_id dev_id;
+   /* get device id */
+   if (eal_parse_sysfs_value(pci_sysfs, &device_id) < 0) {
+   PMD_INIT_LOG(ERR, "Error in reading PCI sysfs\n");
+   return 0;
+   }

-   dev_id.device_id = device_id;
-   pci_bus = rte_bus_find_by_name("pci");
-   return (pci_bus != NULL) &&
-   (pci_bus->find_device(NULL, pci_device_cmp, &dev_id) != NULL);
+   return (uint16_t)device_id;
 }

 /*
@@ -1986,7 +1987,7 @@ eth_axgbe_dev_init(struct rte_eth_dev *eth_dev)
/*
 * Use root complex device ID to differentiate RV AXGBE vs SNOWY AXGBE
 */
-   if (pci_search_device(AMD_PCI_RV_ROOT_COMPLEX_ID)) {
+   if ((get_pci_rc_devid()) == AMD_PCI_RV_ROOT_COMPLEX_ID) {
pdata->xpcs_window_def_reg = PCS_V2_RV_WINDOW_DEF;
pdata->xpcs_window_sel_reg = PCS_V2_RV_WINDOW_SELECT;
} else {
--
2.25.1

Re: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

[Awan, Arsalan]

Hi Chandu,

I have provided my comments on your patch acknowledging your solution as a 
better approach than this one.

Thanks and regards,
Arsalan

From: Namburu, Chandu-babu 
Sent: Friday, November 26, 2021 3:35 PM
To: Awan, Arsalan; dev@dpdk.org
Cc: Somalapuram, Amaranath; Sebastian, Selwin; Giriyapura, Maheshwaramurthy
Subject: RE: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

[Public]

Hi Arsalan,

I have submitted new patch to community review and looped you in. Please share 
your comments.

https://patches.dpdk.org/project/dpdk/list/?series=20788

Regards,
Chandu

-Original Message-
From: Awan, Arsalan 
Sent: Thursday, November 25, 2021 2:37 PM
To: Namburu, Chandu-babu ; dev@dpdk.org
Cc: Somalapuram, Amaranath ; Sebastian, Selwin 
; Giriyapura, Maheshwaramurthy 

Subject: Re: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

Thanks, Chandu!

This will help a lot!

Regards,
Arsalan

From: Namburu, Chandu-babu 
Sent: Thursday, November 25, 2021 1:23 PM
To: Awan, Arsalan; dev@dpdk.org
Cc: Somalapuram, Amaranath; Sebastian, Selwin; Giriyapura, Maheshwaramurthy
Subject: RE: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

[Public]

Hi Arsalan,

We have identified fix for this issue and will submit the patch for community 
review asap.

Thanks,
Chandu

-Original Message-
From: Awan, Arsalan 
Sent: Thursday, November 25, 2021 1:42 PM
To: Namburu, Chandu-babu ; dev@dpdk.org
Cc: Somalapuram, Amaranath ; Sebastian, Selwin 
; Giriyapura, Maheshwaramurthy 

Subject: Re: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

Hi Chandu,

Do we have an update on this?

Can we go ahead with this patch for now as it does not break anything else, but 
fixes the problem.

DPDK will be working fine as is on the e3000.

This patch only adjusts the code for the v1000 and it works fine and there 
isn't a difference in the approach of fixing the problem compared to that being 
done currently to distinguish between the two platforms i.e. e3000 & v1000.

Hoping to hear from you soon.

Thanks,
Arsalan

From: Awan, Arsalan
Sent: Monday, November 15, 2021 3:15 PM
To: Namburu, Chandu-babu; dev@dpdk.org
Cc: Somalapuram, Amaranath; Sebastian, Selwin; Giriyapura, Maheshwaramurthy
Subject: Re: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

Thanks, Chandu!

Appreciate you guys looking into this!

Looking forward to hear from you soon.

Thanks and regards,
Arsalan

From: Namburu, Chandu-babu 
Sent: Friday, November 12, 2021 7:48 PM
To: Awan, Arsalan; dev@dpdk.org
Cc: Somalapuram, Amaranath; Sebastian, Selwin; Giriyapura, Maheshwaramurthy
Subject: RE: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

[AMD Official Use Only]

Hi Arsalan Awan,

Thank you for sharing the patch. DPDK not working on V1000 is regression due to 
"bus/pci: optimize bus scan" patch. As RV Root Complex device does not have any 
Linux kernel driver assigned, this device is removed from PCI scan list and not 
found using pcie_search_device() routine.

Patch submitted is potential fix for the issue, but we are checking if this is 
the best way to fix regression. Will update you soon on the fix shared.

Regards,
Chandu

-Original Message-
From: Arsalan H. Awan 
Sent: Wednesday, November 10, 2021 5:15 PM
To: dev@dpdk.org
Cc: Namburu, Chandu-babu ; Somalapuram, Amaranath 
; Sebastian, Selwin ; 
Giriyapura, Maheshwaramurthy ; Arsalan H. 
Awan 
Subject: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

The RV PCI Root Complex Device is not visible to DPDK on the PCI bus via 
pci_search_device on some OSs including Ubuntu and Yocto. This makes it 
impossible to determine which machine DPDK/axgbe is running on. As a result, 
DPDK/axgbe does not work on v1000.

Let's use the Raven Internal PCIe GPP Bridge ID that DPDK can see on the PCI 
bus that exists on the v1000 to differentiate between RV AXGBE and SNOWY AXGBE, 
and set the registers accordingly. This fixes DPDK not working on v1000 
platform.

Signed-off-by: Arsalan H. Awan 
---
 drivers/net/axgbe/axgbe_ethdev.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/net/axgbe/axgbe_ethdev.c b/drivers/net/axgbe/axgbe_ethdev.c
index 7d40c18a86..7b1b46990f 100644
--- a/drivers/net/axgbe/axgbe_ethdev.c
+++ b/drivers/net/axgbe/axgbe_ethdev.c
@@ -170,7 +170,7 @@ static const struct axgbe_xstats axgbe_xstats_strings[] = {

 /* The set of PCI devices this driver supports */
 #define AMD_PCI_VENDOR_ID   0x1022
-#define AMD_PCI_RV_ROOT_COMPLEX_ID 0x15d0
+#define AMD_PCI_RV_INTERNAL_PCIE_GPP_BRIDGE_ID 0x15db
 #define AMD_PCI_AXGBE_DEVICE_V2A 0x1458  #define AMD_PCI_AXGBE_DEVICE_V2B 
0x1459

@@ -2178,9 +2178,9 @@ eth_axgbe_dev_init(struct rte_eth_dev *eth_dev)
pdata->pci_dev = pci_dev;

/*
-* Use root complex de

Re: [PATCH v1] net/axgbe: use PCI root complex device to distinguish AMD hardware

[David Marchand]

On Fri, Nov 26, 2021 at 11:24 AM Chandubabu Namburu  wrote:
>
> "bus/pci: optimize bus scan" broke axgbe on V1000/R1000.
> RV root complex pci device does not have any kernel driver assigned
> so it is removed from pci scan list which is used in
> "net/axgbe: add a HW quirk for register definitions"
>
> Get root complex device id directly from pci sysfs instead of
> pci scan list

Please end the sentences with a '.'.

As you described, this change is a fix.
Please add a Fixes: tag and Cc: sta...@dpdk.org in the commitlog if
you think it should be backported to LTS releases.
https://doc.dpdk.org/guides/contributing/patches.html#patch-for-stable-releases


>
> Signed-off-by: Chandubabu Namburu 
> ---
>  drivers/net/axgbe/axgbe_ethdev.c | 39 
>  1 file changed, 20 insertions(+), 19 deletions(-)
>
> diff --git a/drivers/net/axgbe/axgbe_ethdev.c 
> b/drivers/net/axgbe/axgbe_ethdev.c
> index 9cd056d04a..26babde354 100644
> --- a/drivers/net/axgbe/axgbe_ethdev.c
> +++ b/drivers/net/axgbe/axgbe_ethdev.c
> @@ -10,6 +10,8 @@
>  #include "axgbe_regs.h"
>  #include "rte_time.h"
>
> +#include "eal_filesystem.h"
> +
>  static int eth_axgbe_dev_init(struct rte_eth_dev *eth_dev);
>  static int  axgbe_dev_configure(struct rte_eth_dev *dev);
>  static int  axgbe_dev_start(struct rte_eth_dev *dev);
> @@ -1923,28 +1925,27 @@ static void axgbe_default_config(struct axgbe_port 
> *pdata)
> pdata->power_down = 0;
>  }
>
> -static int
> -pci_device_cmp(const struct rte_device *dev, const void *_pci_id)
> +/*
> + * Return PCI root complex device id on success else 0
> + */
> +static uint16_t
> +get_pci_rc_devid(void)
>  {
> -   const struct rte_pci_device *pdev = RTE_DEV_TO_PCI_CONST(dev);
> -   const struct rte_pci_id *pcid = _pci_id;
> +   char pci_sysfs[PATH_MAX];
> +   const struct rte_pci_addr pci_rc_addr = {0, 0, 0, 0};
> +   unsigned long device_id;
>
> -   if (pdev->id.vendor_id == AMD_PCI_VENDOR_ID &&
> -   pdev->id.device_id == pcid->device_id)
> -   return 0;
> -   return 1;
> -}
> +   snprintf(pci_sysfs, sizeof(pci_sysfs), "%s/" PCI_PRI_FMT "/device",
> +rte_pci_get_sysfs_path(), pci_rc_addr.domain,
> +pci_rc_addr.bus, pci_rc_addr.devid, pci_rc_addr.function);

The use of rte_pci_get_sysfs_path() and in general is ugly because it
only works for Linux.
We could come up with a better API in the pci bus, but I don't think
it is worth it atm.

We can live with your implementation since this driver is only
compiled on Linux.


-- 
David Marchand

[PATCH] net/cnxk: fix promiscuous mode in multicast enable flow

[skori]

From: Asaf Ravid 

When multicast promisc was being enabled it caused the unicast promisc
to be disabled. This fix resolves this by setting NIX_RX_MODE_PROMISC
when eth_dev->data->promiscuous is set, regardless.

ci: skip_checkpatch skip_roc_check

Fixes: 325d79c00a5a ("net/cnxk: support all multicast")

Signed-off-by: Asaf Ravid 
---
 drivers/common/cnxk/roc_nix_npc.c  | 2 +-
 drivers/net/cnxk/cnxk_ethdev_ops.c | 3 ++-
 2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/common/cnxk/roc_nix_npc.c 
b/drivers/common/cnxk/roc_nix_npc.c
index c0666c87d5..ad8839dde8 100644
--- a/drivers/common/cnxk/roc_nix_npc.c
+++ b/drivers/common/cnxk/roc_nix_npc.c
@@ -96,7 +96,7 @@ roc_nix_npc_mcast_config(struct roc_nix *roc_nix, bool 
mcast_enable,
 
if (mcast_enable)
req->mode = NIX_RX_MODE_ALLMULTI;
-   else if (prom_enable)
+   if (prom_enable)
req->mode = NIX_RX_MODE_PROMISC;
 
return mbox_process(mbox);
diff --git a/drivers/net/cnxk/cnxk_ethdev_ops.c 
b/drivers/net/cnxk/cnxk_ethdev_ops.c
index ce5f1f7240..34e4809650 100644
--- a/drivers/net/cnxk/cnxk_ethdev_ops.c
+++ b/drivers/net/cnxk/cnxk_ethdev_ops.c
@@ -517,7 +517,8 @@ cnxk_nix_allmulticast_enable(struct rte_eth_dev *eth_dev)
 {
struct cnxk_eth_dev *dev = cnxk_eth_pmd_priv(eth_dev);
 
-   return roc_nix_npc_mcast_config(&dev->nix, true, false);
+   return roc_nix_npc_mcast_config(&dev->nix, true,
+   eth_dev->data->promiscuous);
 }
 
 int
-- 
2.25.1

Re: [PATCH] ip_frag: add IPv4 options fragment and unit test data

[Dariusz Sosnowski]

On Wed, 24 Nov 2021 16:47:06 +0800, Huichao Cai wrote:
> +/*
> + * Options "fragmenting", just fill options not
> + * allowed in fragments with NOOPs.
> + * Simple and stupid 8), but the most efficient way.
> + */
> +static inline void ip_options_fragment(struct rte_ipv4_hdr *iph)
> +{
> +   unsigned char *optptr = (unsigned char *)iph +
> +   sizeof(struct rte_ipv4_hdr);
> +   int l = (iph->version_ihl & RTE_IPV4_HDR_IHL_MASK) *
> +   RTE_IPV4_IHL_MULTIPLIER - sizeof(struct rte_ipv4_hdr);
> +   int optlen;
> +
> +   while (l > 0) {
> +   switch (*optptr) {
> +   case RTE_IPOPT_END:
> +   return;
> +   case RTE_IPOPT_NOOP:
> +   l--;
> +   optptr++;
> +   continue;
> +   }
> +   optlen = optptr[1];
> +   if (optlen < 2 || optlen > l)
> +   return;
> +   if (!RTE_IPOPT_COPIED(*optptr))
> +   memset(optptr, RTE_IPOPT_NOOP, optlen);
> +   l -= optlen;
> +   optptr += optlen;
> +   }
> +}
> +
I have a few concerns regarding this implementation:
- Any IPv4 option longer than 2 bytes with copied flag unset, will not be 
substituted by NOOP option. In effect it will be copied to all fragments.
- Substituting options with NOOP might cause rte_ipv4_fragment_packet to 
produce more fragments than necessary, since options with copied flag unset 
will still occupy space in IPv4 header.
It would require some benchmarking, but maybe a better solution would be to 
prepare a separate IPv4 header for fragments without unnecessary options. 

PRM release mailing list

[Asaf Penso]

If you wish to get the PRM release mail, you can subscribe to nbu-adapter-prm 
mailing list via https://dlrequest/

Regards,
Asaf Penso

[PATCH] pipeline: add check against loops

[Cristian Dumitrescu]

Detect when a jump instruction, either conditional or unconditional,
is jumping to itself, thus creating a loop, which is not allowed in
data plane code.

Signed-off-by: Cristian Dumitrescu 
Signed-off-by: Harshad Narayane 
---
 lib/pipeline/rte_swx_pipeline.c | 13 +
 1 file changed, 13 insertions(+)

diff --git a/lib/pipeline/rte_swx_pipeline.c b/lib/pipeline/rte_swx_pipeline.c
index c332d44bd1..1a50c4bb72 100644
--- a/lib/pipeline/rte_swx_pipeline.c
+++ b/lib/pipeline/rte_swx_pipeline.c
@@ -6000,6 +6000,19 @@ instr_label_check(struct instruction_data 
*instruction_data,
CHECK(strcmp(label, instruction_data[j].label), EINVAL);
}
 
+   /* Check that no jump instruction (either conditional or not) can jump 
to itself (loop). */
+   for (i = 0; i < n_instructions; i++) {
+   struct instruction_data *data = &instruction_data[i];
+   char *label = data->label;
+   char *jmp_label = data->jmp_label;
+
+   /* Continue if this instruction does not have a label or it is 
not a jump. */
+   if (!label[0] || !jmp_label[0])
+   continue;
+
+   CHECK(strcmp(label, jmp_label), EINVAL);
+   }
+
/* Get users for each instruction label. */
for (i = 0; i < n_instructions; i++) {
struct instruction_data *data = &instruction_data[i];
-- 
2.17.1

RE: PRM release mailing list

[Asaf Penso]

Sent by mistake to this ML. Sorry for this.

Regards,
Asaf Penso

From: Asaf Penso 
Sent: Wednesday, December 1, 2021 1:55 PM
To: dpdk-dev 
Subject: PRM release mailing list

If you wish to get the PRM release mail, you can subscribe to nbu-adapter-prm 
mailing list via https://dlrequest/

Regards,
Asaf Penso

[PATCH v3] devtools: spell check

[Henry Nadeau]

 A spell check script to check for errors in patches. An
 example of usage being cat PATCH_FILE | spell_check.sh.
 Errors will be printed to console, following the file
 they are located in. Word exclusions can be made by
 making additions to the dictionary, and problematic
 patterns can be added to the regex filter.

---
 devtools/spell_check.sh |  115 ++
 devtools/spell_check_dictionary.txt | 2160 +++
 devtools/spell_check_regex.txt  |2 +
 3 files changed, 2277 insertions(+)
 create mode 100755 devtools/spell_check.sh
 create mode 100644 devtools/spell_check_dictionary.txt
 create mode 100644 devtools/spell_check_regex.txt

diff --git a/devtools/spell_check.sh b/devtools/spell_check.sh
new file mode 100755
index 00..a8f8331fd2
--- /dev/null
+++ b/devtools/spell_check.sh
@@ -0,0 +1,115 @@
+#!/bin/bash
+
+file_count=0
+error_count=0
+non_doc=0
+simple_out=''
+output_dir=''
+regex_pattern=''
+output=''
+dir=$(git rev-parse --show-toplevel)
+if [ -z "$dir" ]; then
+  echo "Please execute this script from within a git repo"
+  exit 1
+fi
+
+# function to spell check a single file
+function spellcheck() {
+   echo "$3" | sed "$2" | aspell --lang=en --encoding=utf-8 --ignore-case 
--ignore=3 --ignore-repl list 
--personal="$1""/devtools/spell_check_dictionary.txt"
+}
+
+function read_input {
+  while read -r data; do
+echo "$data"
+  done
+}
+
+while test $# -gt 0; do
+   case "$1" in
+   -h | --help)
+   echo "Spell Check"
+   echo " "
+   echo "spell_check [options] [arguments]"
+   echo " "
+   echo "options:"
+   echo "-h, --helpShow shows list of flags and usages."
+   echo "-eExcludes file (and dir) from being printed."
+   echo "-output-dir=Output file."
+   exit 0
+   ;;
+   -e)
+   shift
+   export simple_out='true'
+   ;;
+   --output-dir*)
+   export output_dir=$(echo $1 | sed -e 's/^[^=]*=//g')
+   shift
+   ;;
+   *)
+   break
+   ;;
+   esac
+done
+
+# requires patch file be piped into script
+PATCH_FILE=$(read_input)
+PATCH_FILE=$(echo "$PATCH_FILE" | sed 's/``.*``//' | grep ^+ | tr -d '*')
+
+# Build regex pattern from files
+while IFS= read -r line; do
+  if [[ ! $line =~ "#" ]]; then
+regex_pattern+="s/$line/ /; "
+  fi
+done < "$dir/devtools/spell_check_regex.txt"
+
+if [ -n "$regex_pattern" ]; then
+  regex_pattern="${regex_pattern::-1}"
+fi
+
+if [ 's// /;' = "$regex_pattern" ]; then
+  regex_pattern=''
+fi
+
+
+while IFS= read -r line; do
+  if [[ ($line =~ ^\+\+\+) && ($line =~ .rst$)]]; then # if line is a file and 
rst echo and reset flag
+output=$output"${line//+++ b\//}"$'\n'
+((file_count=file_count+1))
+non_doc=0
+continue
+  elif [[ ($line =~ ^\+\+\+) && (! $line =~ .rst$)]]; then # if line is a file 
and not rst set flag
+non_doc=1
+continue;
+  fi
+
+  if [[ ($non_doc = 0 ) && (! $line =~ ^\+\+\+)]]; then # only when non_doc 
isn't set & line is not a file
+line=${line/+  /}
+line=${line/+/}
+for word in $line;
+do
+   error=$(spellcheck "$dir" "$regex_pattern" "$(echo "$word" |
+   sed 's/>/ /;
+s/> "$output_dir"
+fi
+
+exit 0
diff --git a/devtools/spell_check_dictionary.txt 
b/devtools/spell_check_dictionary.txt
new file mode 100644
index 00..8a10d942d3
--- /dev/null
+++ b/devtools/spell_check_dictionary.txt
@@ -0,0 +1,2160 @@
+personal_ws-1.1 en 1
+AArch
+ABI's
+ABIs
+ACAT
+ACKs
+ACLs
+ADAT
+AEEOT
+AENQ
+AESN
+AFUs
+AIOP
+ALEF
+ALTQ
+AMDA
+APCI
+API's
+APIs
+ARKA
+ARKP
+ARMv
+ASCAT
+ASIC
+ASLR
+ASan
+AUPE
+AddressSanitizer
+AdminQ
+Agilex
+Agilio
+AltArch
+AltiVec
+Alveo
+Aquantia
+Aquantia's
+Arkville
+Arkville's
+Arria
+BBDev
+BBdev
+BCAT
+BDFs
+BMan
+BPSK
+BSPs
+BUFS
+BXNT
+Backports
+Blocklisting
+BlueField
+Broadcom
+Broadwell
+Broder
+BusA
+CAVP
+CAVS
+CBAR
+CBDMA
+CBFC
+CCAT
+CCITT
+CCed
+CDAT
+CEIL
+CFLAG
+CIDR
+CIMC
+CLIs
+CNTRS
+COMPDEV
+CPUFLAG
+CPUFreq
+CPUIdle
+CPUs
+CamelCase
+Cavium
+Cavium's
+CentOS
+Centos
+Cesnet
+Chelsio
+CollectD
+CommitID
+CompressDev
+ConnectX
+Coverity
+CryptoDev
+DALLOW
+DCBX
+DDEBUG
+DDIO
+DEINTERLEAVE
+DEINTERLEAVER
+DESTDIR
+DEVMEM
+DEVS
+DIMM
+DIMMs
+DLActive
+DLB's
+DMAdev
+DMFS
+DOCSISBPI
+DPBP
+DPBPs
+DPDK's
+DPDKVMS
+DPIO
+DPIOs
+DPMAC
+DPMCP
+DPRCs
+DRHD
+DRTE
+DSPs
+DTLB
+DVPT
+DWRR
+Dbuildtype
+Ddisable
+Deinitialising
+Denable
+DevA
+DevB
+DevC
+DevD
+DevF
+DevX
+DevxEnabled
+DevxFsRules
+Dexamples
+Dflexran
+DiffServ
+Diffie
+Dmachine
+Dmax
+Dmitry
+Dockerfile
+DomBDF
+Doption
+Dplatform
+DrvA
+DrvB
+DstMAC
+Dwerror
+EBUSY
+ECAT
+ECDSA
+ECPFs
+ECPM
+EDAT
+EEARBC
+EEXIST
+EFD's
+EINVAL
+ENAv
+ENOSPC
+ENOSYS
+ENOTSUP
+EPOLLIN
+ERANGE
+ESXi
+ETAG
+ETHPORTS
+ETQF
+ETrackId
+EVLAN
+EZchip
+Enablement
+EncryptExtIV
+EqualizationComplete

RE: [PATCH v3 1/2] doc: note KNI alternatives

[Morten Brørup]

> From: Ferruh Yigit [mailto:ferruh.yi...@intel.com]
> Sent: Wednesday, 24 November 2021 18.16
> 
> Add more information on alternatives of KNI and the cons of KNI against
> these alternatives.
> 
> Signed-off-by: Ferruh Yigit 
> ---
> Cc: Olivier Matz Olivier Matz 
> Cc: David Marchand David Marchand 
> Cc: Stephen Hemminger Stephen Hemminger 
> Cc: Elad Nachman 
> Cc: Igor Ryzhov 
> Cc: Dan Gora 
> 
> v3:
> * reference tap document directly instead of adding label to it.
> ---
>  .../prog_guide/kernel_nic_interface.rst   | 34 +--
>  1 file changed, 31 insertions(+), 3 deletions(-)
> 
> diff --git a/doc/guides/prog_guide/kernel_nic_interface.rst
> b/doc/guides/prog_guide/kernel_nic_interface.rst
> index 1ce03ec1a374..f5a8b7c0782c 100644
> --- a/doc/guides/prog_guide/kernel_nic_interface.rst
> +++ b/doc/guides/prog_guide/kernel_nic_interface.rst
> @@ -6,16 +6,44 @@
>  Kernel NIC Interface
>  
> 
> +.. Note::
> +
> +   :ref:`virtio_user_as_exceptional_path` alternative is the preferred
> way for
> +   interfacing with the Linux network stack as it is an in-kernel
> solution and
> +   has similar performance expectations.
> +
>  The DPDK Kernel NIC Interface (KNI) allows userspace applications
> access to the Linux* control plane.
> 
> -The benefits of using the DPDK KNI are:
> +KNI provides an interface with the kernel network stack and allows
> management of
> +DPDK ports using standard Linux net tools such as ``ethtool``,
> ``ifconfig`` and
> +``tcpdump``.
> +
> +The main use case of KNI is to get/receive exception packets from/to
> Linux network
> +stack while main datapath IO is done bypassing the networking stack.
> +
> +There are other alternatives to KNI, all are available in the upstream
> Linux:
> +
> +#. :ref:`virtio_user_as_exceptional_path`
> +
> +#. :doc:`../nics/tap` as wrapper to `Linux tun/tap
> +   `_
> +
> +The benefits of using the KNI against alternatives are:
> 
>  *   Faster than existing Linux TUN/TAP interfaces
>  (by eliminating system calls and copy_to_user()/copy_from_user()
> operations.
> 
> -*   Allows management of DPDK ports using standard Linux net tools
> such as ethtool, ifconfig and tcpdump.
> +The cons of the KNI are:

Consider if "disadvantages" is more appropriate than "cons". (I'm not a native 
English speaker.)

> +
> +* It is out-of-tree Linux kernel module and it can't be distributed as
> binary as
> +  part of operating system vendor DPDK packages. This makes it harder
> to
> +  consume, although it is always possible to compile it from the
> source code.
> +
> +* As it shares memory between userspace and kernelspace, and kernel
> part
> +  directly uses input provided by userspace, it is not safe. This
> makes hard to
> +  upstream the module.
> 
> -*   Allows an interface with the kernel network stack.
> +* Only a subset of net devices control commands are supported by KNI.

If it is still relevant, add something along the lines of:
* Requires dedicated kernel cores.

> 
>  The components of an application using the DPDK Kernel NIC Interface
> are shown in :numref:`figure_kernel_nic_intf`.
> 
> --
> 2.31.1
> 

Could you perhaps also promote the virtio-user documentation by moving it from 
the HowTo Guides to the Programmer's Guide?

-Morten

[PATCH v4] devtools: spell check

[Henry Nadeau]

 A spell check script to check for errors in patches. An
 example of usage being cat PATCH_FILE | spell_check.sh. Errors will be
 printed to console, following the file they are located in. Word exclusions
 can be made by making additions to the dictionary, and problematic patterns
 can be added to the regex filter.

Signed-off-by: Henry Nadeau 
---
 devtools/spell_check.sh |  121 ++
 devtools/spell_check_dictionary.txt | 2160 +++
 devtools/spell_check_regex.txt  |2 +
 3 files changed, 2283 insertions(+)
 create mode 100755 devtools/spell_check.sh
 create mode 100644 devtools/spell_check_dictionary.txt
 create mode 100644 devtools/spell_check_regex.txt

diff --git a/devtools/spell_check.sh b/devtools/spell_check.sh
new file mode 100755
index 00..16ebe02406
--- /dev/null
+++ b/devtools/spell_check.sh
@@ -0,0 +1,121 @@
+#!/bin/bash
+
+file_count=0
+error_count=0
+non_doc=0
+simple_out=''
+output_dir=''
+regex_pattern=''
+output=''
+dir=$(git rev-parse --show-toplevel)
+if [ -z "$dir" ]; then
+  echo "Please execute this script from within a git repo"
+  exit 1
+fi
+
+# Function to spell check a single file
+function spellcheck() {
+   echo "$3" | sed "$2" | aspell --lang=en \
+ --encoding=utf-8 \
+ --ignore-case \
+ --ignore=3 \
+ --ignore-repl \
+ list \
+ 
--personal="$1""/devtools/spell_check_dictionary.txt"
+}
+
+function read_input {
+  while read -r data; do
+echo "$data"
+  done
+}
+
+while test $# -gt 0; do
+   case "$1" in
+   -h | --help)
+   echo "Spell Check"
+   echo " "
+   echo "spell_check [options] [arguments]"
+   echo " "
+   echo "options:"
+   echo "-h, --helpShow shows list of flags and usages."
+   echo "-eExcludes file (and dir) from being printed."
+   echo "-output-dir=Output file."
+   exit 0
+   ;;
+   -e)
+   shift
+   export simple_out='true'
+   ;;
+   --output-dir*)
+   export output_dir=$(echo $1 | sed -e 's/^[^=]*=//g')
+   shift
+   ;;
+   *)
+   break
+   ;;
+   esac
+done
+
+# Requires patch file be piped into script
+PATCH_FILE=$(read_input)
+PATCH_FILE=$(echo "$PATCH_FILE" | sed 's/``.*``//' | grep ^+ | tr -d '*')
+
+# Build regex pattern from files
+while IFS= read -r line; do
+  if [[ ! $line =~ "#" ]]; then
+regex_pattern+="s/$line/ /; "
+  fi
+done < "$dir/devtools/spell_check_regex.txt"
+
+if [ -n "$regex_pattern" ]; then
+  regex_pattern="${regex_pattern::-1}"
+fi
+
+if [ 's// /;' = "$regex_pattern" ]; then
+  regex_pattern=''
+fi
+
+
+while IFS= read -r line; do
+  if [[ ($line =~ ^\+\+\+) && ($line =~ .rst$)]]; then
+output=$output"${line//+++ b\//}"$'\n'
+((file_count=file_count+1))
+non_doc=0
+continue
+  elif [[ ($line =~ ^\+\+\+) && (! $line =~ .rst$)]]; then
+non_doc=1
+continue;
+  fi
+
+  if [[ ($non_doc = 0 ) && (! $line =~ ^\+\+\+)]]; then
+line=${line/+  /}
+line=${line/+/}
+for word in $line;
+do
+   error=$(spellcheck "$dir" "$regex_pattern" "$(echo "$word" |
+   sed 's/>/ /;
+s/> "$output_dir"
+fi
+
+exit 0
diff --git a/devtools/spell_check_dictionary.txt 
b/devtools/spell_check_dictionary.txt
new file mode 100644
index 00..8a10d942d3
--- /dev/null
+++ b/devtools/spell_check_dictionary.txt
@@ -0,0 +1,2160 @@
+personal_ws-1.1 en 1
+AArch
+ABI's
+ABIs
+ACAT
+ACKs
+ACLs
+ADAT
+AEEOT
+AENQ
+AESN
+AFUs
+AIOP
+ALEF
+ALTQ
+AMDA
+APCI
+API's
+APIs
+ARKA
+ARKP
+ARMv
+ASCAT
+ASIC
+ASLR
+ASan
+AUPE
+AddressSanitizer
+AdminQ
+Agilex
+Agilio
+AltArch
+AltiVec
+Alveo
+Aquantia
+Aquantia's
+Arkville
+Arkville's
+Arria
+BBDev
+BBdev
+BCAT
+BDFs
+BMan
+BPSK
+BSPs
+BUFS
+BXNT
+Backports
+Blocklisting
+BlueField
+Broadcom
+Broadwell
+Broder
+BusA
+CAVP
+CAVS
+CBAR
+CBDMA
+CBFC
+CCAT
+CCITT
+CCed
+CDAT
+CEIL
+CFLAG
+CIDR
+CIMC
+CLIs
+CNTRS
+COMPDEV
+CPUFLAG
+CPUFreq
+CPUIdle
+CPUs
+CamelCase
+Cavium
+Cavium's
+CentOS
+Centos
+Cesnet
+Chelsio
+CollectD
+CommitID
+CompressDev
+ConnectX
+Coverity
+CryptoDev
+DALLOW
+DCBX
+DDEBUG
+DDIO
+DEINTERLEAVE
+DEINTERLEAVER
+DESTDIR
+DEVMEM
+DEVS
+DIMM
+DIMMs
+DLActive
+DLB's
+DMAdev
+DMFS
+DOCSISBPI
+DPBP
+DPBPs
+DPDK's
+DPDKVMS
+DPIO
+DPIOs
+DPMAC
+DPMCP
+DPRCs
+DRHD
+DRTE
+DSPs
+DTLB
+DVPT
+DWRR
+Dbuildtype
+Ddisable
+Deinitialising
+Denable
+DevA
+DevB
+DevC
+DevD
+DevF
+DevX
+DevxEnabled
+DevxFsRules
+Dexamples
+Dflexran
+DiffServ
+Diffie
+Dmachine
+Dmax
+Dmitry
+Dockerfile
+DomBDF
+Doption
+Dplatform
+DrvA
+DrvB
+DstMAC
+Dwerror
+EBUSY
+ECAT
+ECDSA
+ECPFs
+ECPM
+EDAT
+EEARBC
+EEXIST
+EFD's
+EINVAL
+ENAv
+ENOSPC
+ENOSYS
+EN

[PATCH v9 1/9] eal/windows: return ENOTSUP for not supported API

[Jie Zhou]

UT memory_autotest on Windows has 2 failed cases on eal APIs
eal_memalloc_get_seg_fd and eal_memalloc_get_seg_fd_offset. These 2
APIs are not supported on Windows yet. Should return ENOTSUP such that
in test_memory.c these 2 ENOTSUP cases will not be marked as failures,
same as other ENOTSUP cases.

Signed-off-by: Jie Zhou 
Acked-by: Dmitry Kozlyuk 

---
 lib/eal/windows/eal_memalloc.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/lib/eal/windows/eal_memalloc.c b/lib/eal/windows/eal_memalloc.c
index 55d6dcc71c..aa7589b81d 100644
--- a/lib/eal/windows/eal_memalloc.c
+++ b/lib/eal/windows/eal_memalloc.c
@@ -17,7 +17,7 @@ eal_memalloc_get_seg_fd(int list_idx, int seg_idx)
RTE_SET_USED(list_idx);
RTE_SET_USED(seg_idx);
EAL_LOG_NOT_IMPLEMENTED();
-   return -1;
+   return -ENOTSUP;
 }
 
 int
@@ -28,7 +28,7 @@ eal_memalloc_get_seg_fd_offset(int list_idx, int seg_idx, 
size_t *offset)
RTE_SET_USED(seg_idx);
RTE_SET_USED(offset);
EAL_LOG_NOT_IMPLEMENTED();
-   return -1;
+   return -ENOTSUP;
 }
 
 static int
@@ -428,7 +428,7 @@ eal_memalloc_sync_with_primary(void)
 {
/* No multi-process support. */
EAL_LOG_NOT_IMPLEMENTED();
-   return -1;
+   return -ENOTSUP;
 }
 
 int
-- 
2.31.0.vfs.0.1

[PATCH v9 3/9] app/test: fix incorrect errno variable

[Jie Zhou]

Fix incorrect errno variable used in memory autotest.
Use rte_errno instead.

Fixes: 086d426406bd ("app/test: fix memory autotests on FreeBSD")
Cc: bruce.richard...@intel.com

Signed-off-by: Jie Zhou 

---
 app/test/test_memory.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/app/test/test_memory.c b/app/test/test_memory.c
index dbf6871e71..140ac3f3cf 100644
--- a/app/test/test_memory.c
+++ b/app/test/test_memory.c
@@ -63,7 +63,7 @@ check_seg_fds(const struct rte_memseg_list *msl, const struct 
rte_memseg *ms,
/* we're able to get memseg fd - try getting its offset */
ret = rte_memseg_get_fd_offset_thread_unsafe(ms, &offset);
if (ret < 0) {
-   if (errno == ENOTSUP)
+   if (rte_errno == ENOTSUP)
return 1;
return -1;
}
-- 
2.31.0.vfs.0.1

[PATCH v9 0/9] app/test: enable subset of tests on Windows

[Jie Zhou]

From: Jie Zhou 

This patchset is to enable a subset of unit tests on windows. It mainly 
includes:
- Replace POSIX specific codes
- Add test stubs for not supported ones on Windows
- Fix some lib and tests per failures investigation
- Replace .sh script with .py script for meson.build
- Enable build and run subset of unit tests on Windows

Future work:
- Work with CI/CD lab to onboard unit tests on Windows to catch regression
- Enable more tests

---
V2 changes:
- Fix compilation error on FreeBSD
- Fix email mismatch issue
- Add a missing space around "*"

---
V3 changes:
- Fix a misc c coding style issue
- Revise some commit title and message body
- Fix violations of PEP8 in new added Python scripts
- Add error handling in get_coremask.py
- Fix has_hugepage.py to check system support of hugepages
  instead of checking privileges
- Fix test meson.build to run Python scripts using py3
- Consolidate lists of source files, test dep, etc. across all 
  platforms, with conditional extending on some platform(s)

---
V4 changes:
- Remove building of ip_frag, rib, and reorder libraries on Windows.
  These three libs usually can be built on Windows without change.
  However, in between the time of V3 and V4, there is regression in
  upstream caused build failures of these three libs. Will separately
  investigate and enable these libraries.

- Remove previous patch#2 (Enable mempool/stack on Windows) from this
  patchset as it was separated out and merged as patch-19314.

- Consolidate the source files, deps, tests lists across platforms as
  much as possible.

---
V5 changes:
- Remove a space between function name and open parenthesis '('
- Add back a header mistakenly deleted

---
V6 changes:
   - Fix inconsistent static vs. non-static declarations

---
V7 changes:
   - Remove get_coremask.py as it is not needed any more in meson.build
   - Remove enablement of efd and lpm and their corresponding unit tests.
 The enablement of these two libs and their UTs will be in separate
 patches after this patch set.

V8 changes:
   - Fix coding style issue of using C99 // comments

V9 changes:
   - Fix has_hugepage.py with adding failure handling on Linux, using
 proper variable name to follow Python convention, and removing
 unnecessary comment.
   - Enable previously skipped test_cmdline_socket_fns test cases
   - Revise title and message, and add Fixes info for current Patch#3 
   - Combine 2 patches (previous #2 and #3 in V8) into one and with 
 more detailed message

Jie Zhou (9):
  eal/windows: return ENOTSUP for not supported API
  app/test: remove POSIX-specific code
  app/test: fix incorrect errno variable
  app/test: skip interrupt tests on Windows
  app/test: skip two logs_autotest cases on Windows
  app/test: differentiate a strerror on different OS
  app/test: remove two alarm_autotest cases
  app/test: replace .sh script with .py script
  app/test: enable subset of unit tests on Windows

 app/test/commands.c  |   2 -
 app/test/has-hugepage.sh |  11 ---
 app/test/has_hugepage.py |  26 +
 app/test/meson.build | 119 ---
 app/test/packet_burst_generator.c|   1 +
 app/test/process.h   |   4 +-
 app/test/test.c  |   5 +-
 app/test/test_acl.c  |  12 +++
 app/test/test_alarm.c|   4 +
 app/test/test_bpf.c  |  15 ++-
 app/test/test_byteorder.c|   2 +-
 app/test/test_cksum.c|  12 +++
 app/test/test_cmdline.c  |   2 +
 app/test/test_cmdline_ipaddr.c   |   5 +
 app/test/test_cmdline_lib.c  |  13 ++-
 app/test/test_crc.c  |   1 -
 app/test/test_cryptodev.c|   4 +
 app/test/test_cryptodev_asym.c   |   4 +
 app/test/test_cryptodev_blockcipher.c|   4 +
 app/test/test_cryptodev_security_ipsec.c |   4 +
 app/test/test_cryptodev_security_pdcp.c  |   4 +
 app/test/test_debug.c|  17 +++-
 app/test/test_distributor.c  |  13 +++
 app/test/test_distributor_perf.c |  13 +++
 app/test/test_dmadev.c   |  14 ++-
 app/test/test_dmadev_api.c   |   4 +
 app/test/test_eal_flags.c|  90 +
 app/test/test_eal_fs.c   |  12 +++
 app/test/test_efd.c  |  15 ++-
 app/test/test_efd_perf.c |  16 ++-
 app/test/test_errno.c|  12 ++-
 app/test/test_event_crypto_adapter.c |  15 ++-
 app/test/test_event_eth_rx_adapter.c |  25 -
 app/test/test_event_eth_tx_adapter.c |  12 +++
 app/test/test_event_ring.c   |  16 ++-
 app/test/test_event_timer_adapter.c  |  16 ++-
 app/test/test_eventdev.c |  20 +++-
 app/test/te

[PATCH v9 2/9] app/test: remove POSIX-specific code

[Jie Zhou]

- Remove header inclusion of netinet/in.h and terminos.h
- Include rte_os_shim.h
- Replace sleep and usleep with rte_delay_us_sleep
- Use NUL on Windows as /dev/null for Linux
- Exclude tests not yet supported on Windows, e.g. multi-process, and
  IP address parsing (the test cases use linux netinet/in.h u6_addr
  and better have a separate patch to add such tests on Windows)

Signed-off-by: Jie Zhou 

---
 app/test/commands.c   |  2 --
 app/test/packet_burst_generator.c |  1 +
 app/test/process.h|  4 +++-
 app/test/test.c   |  5 -
 app/test/test_byteorder.c |  2 +-
 app/test/test_cmdline.c   |  2 ++
 app/test/test_cmdline_lib.c   | 13 +
 app/test/test_crc.c   |  1 -
 app/test/test_pmd_perf.c  |  6 +-
 app/test/test_ring_stress_impl.h  |  2 +-
 app/test/test_telemetry_data.c|  2 ++
 11 files changed, 28 insertions(+), 12 deletions(-)

diff --git a/app/test/commands.c b/app/test/commands.c
index 2dced3bc44..887cabad64 100644
--- a/app/test/commands.c
+++ b/app/test/commands.c
@@ -8,8 +8,6 @@
 #include 
 #include 
 #include 
-#include 
-#include 
 #include 
 #include 
 #include 
diff --git a/app/test/packet_burst_generator.c 
b/app/test/packet_burst_generator.c
index 8ac24577ba..6b42b9b83b 100644
--- a/app/test/packet_burst_generator.c
+++ b/app/test/packet_burst_generator.c
@@ -5,6 +5,7 @@
 #include 
 #include 
 #include 
+#include 
 
 #include "packet_burst_generator.h"
 
diff --git a/app/test/process.h b/app/test/process.h
index 5b10cf64df..1f073b9c5c 100644
--- a/app/test/process.h
+++ b/app/test/process.h
@@ -7,12 +7,14 @@
 
 #include   /* errno */
 #include  /* PATH_MAX */
+#ifndef RTE_EXEC_ENV_WINDOWS
 #include  /* basename et al */
+#include 
+#endif
 #include  /* NULL */
 #include  /* strerror */
 #include  /* readlink */
 #include 
-#include 
 
 #include  /* strlcpy */
 
diff --git a/app/test/test.c b/app/test/test.c
index 5194131026..e69cae3eea 100644
--- a/app/test/test.c
+++ b/app/test/test.c
@@ -8,7 +8,6 @@
 #include 
 #include 
 #include 
-#include 
 #include 
 #include 
 
@@ -63,7 +62,9 @@ do_recursive_call(void)
const char *env_var;
int (*action_fn)(void);
} actions[] =  {
+#ifndef RTE_EXEC_ENV_WINDOWS
{ "run_secondary_instances", test_mp_secondary },
+#endif
 #ifdef RTE_LIB_PDUMP
 #ifdef RTE_NET_RING
{ "run_pdump_server_tests", test_pdump },
@@ -82,7 +83,9 @@ do_recursive_call(void)
{ "test_file_prefix", no_action },
{ "test_no_huge_flag", no_action },
 #ifdef RTE_LIB_TIMER
+#ifndef RTE_EXEC_ENV_WINDOWS
{ "timer_secondary_spawn_wait", test_timer_secondary },
+#endif
 #endif
};
 
diff --git a/app/test/test_byteorder.c b/app/test/test_byteorder.c
index 03c08d9abf..de14ed539e 100644
--- a/app/test/test_byteorder.c
+++ b/app/test/test_byteorder.c
@@ -46,7 +46,7 @@ test_byteorder(void)
return -1;
 
res_u16 = rte_bswap16(0x1337);
-   printf("const %"PRIx16" -> %"PRIx16"\n", 0x1337, res_u16);
+   printf("const %"PRIx16" -> %"PRIx16"\n", (uint16_t)0x1337, res_u16);
if (res_u16 != 0x3713)
return -1;
 
diff --git a/app/test/test_cmdline.c b/app/test/test_cmdline.c
index 115bee966d..9a76bd299f 100644
--- a/app/test/test_cmdline.c
+++ b/app/test/test_cmdline.c
@@ -31,6 +31,7 @@ test_cmdline(void)
return -1;
if (test_parse_num_invalid_param() < 0)
return -1;
+#ifndef RTE_EXEC_ENV_WINDOWS
printf("Testing parsing IP addresses...\n");
if (test_parse_ipaddr_valid() < 0)
return -1;
@@ -38,6 +39,7 @@ test_cmdline(void)
return -1;
if (test_parse_ipaddr_invalid_param() < 0)
return -1;
+#endif
printf("Testing parsing strings...\n");
if (test_parse_string_valid() < 0)
return -1;
diff --git a/app/test/test_cmdline_lib.c b/app/test/test_cmdline_lib.c
index f50ccdb599..fcd58cb76a 100644
--- a/app/test/test_cmdline_lib.c
+++ b/app/test/test_cmdline_lib.c
@@ -8,7 +8,6 @@
 #include 
 #include 
 #include 
-#include 
 #include 
 #include 
 
@@ -22,6 +21,12 @@
 
 #include "test_cmdline.h"
 
+#ifndef RTE_EXEC_ENV_WINDOWS
+#define NULL_INPUT "/dev/null"
+#else
+#define NULL_INPUT "NUL"
+#endif
+
 //
 /* static functions required for some tests */
 static void
@@ -156,10 +161,10 @@ test_cmdline_socket_fns(void)
cl = cmdline_stdin_new(&ctx, NULL);
if (cl != NULL)
goto error;
-   cl = cmdline_file_new(NULL, "prompt", "/dev/null");
+   cl = cmdline_file_new(NULL, "prompt", NULL_INPUT);
if (cl != NULL)
goto error;
-   cl = cmdline_file_new(&ctx, NULL, "/dev/null");
+   cl = cmdline_file_new(&ctx, NULL, NULL_INPUT);
if (c

[PATCH v9 4/9] app/test: skip interrupt tests on Windows

[Jie Zhou]

Even though test_interrupts.c can compile on Windows, skip interrupt
tests for now since majority of eal_interrupt on Windows are stubs.
Will remove the skip after interrupt being fully enabled on Windows.

Signed-off-by: Jie Zhou 

---
 app/test/test_interrupts.c | 10 ++
 1 file changed, 10 insertions(+)

diff --git a/app/test/test_interrupts.c b/app/test/test_interrupts.c
index 2a05399f96..eec9b2805b 100644
--- a/app/test/test_interrupts.c
+++ b/app/test/test_interrupts.c
@@ -12,6 +12,15 @@
 
 #include "test.h"
 
+#ifdef RTE_EXEC_ENV_WINDOWS
+static int
+test_interrupt(void)
+{
+   printf("Interrupt on Windows is not fully supported yet, skipping 
test\n");
+   return TEST_SKIPPED;
+}
+#else
+
 #define TEST_INTERRUPT_CHECK_INTERVAL 100 /* ms */
 
 /* predefined interrupt handle types */
@@ -590,5 +599,6 @@ test_interrupt(void)
 
return ret;
 }
+#endif /*ifdef RTE_EXEC_ENV_WINDOWS*/
 
 REGISTER_TEST_COMMAND(interrupt_autotest, test_interrupt);
-- 
2.31.0.vfs.0.1

[PATCH v9 5/9] app/test: skip two logs_autotest cases on Windows

[Jie Zhou]

DPDK logs_autotest on Windows failed at "dynamic log types" tests.
The failures are on 2 test cases for rte_log_set_level_regexp API,
due to regular expression is not supported on Windows in DPDK yet
and regcomp/regexec are just stubs on Windows (in regex.h).

In app\test\test_logs.c, ifndef these two test cases, and for the
rte_log_set_level_pattern validation case following these two cases,
differentiate the expected log level passed into macro CHECK_LEVELS

Now logs_autotest completes for all dynamic log types and static log types.

Signed-off-by: Jie Zhou 

---
 app/test/test_logs.c | 6 +-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/app/test/test_logs.c b/app/test/test_logs.c
index 7abb6eeca2..7c001c1ab3 100644
--- a/app/test/test_logs.c
+++ b/app/test/test_logs.c
@@ -113,6 +113,7 @@ test_logs(void)
rte_log_set_level(logtype1, RTE_LOG_ERR);
CHECK_LEVELS(RTE_LOG_ERR, RTE_LOG_INFO, RTE_LOG_ERR);
 
+#ifndef RTE_EXEC_ENV_WINDOWS
rte_log_set_level_regexp("type$", RTE_LOG_EMERG);
CHECK_LEVELS(RTE_LOG_ERR, RTE_LOG_INFO, RTE_LOG_ERR);
 
@@ -121,7 +122,10 @@ test_logs(void)
 
rte_log_set_level_pattern("logtype", RTE_LOG_DEBUG);
CHECK_LEVELS(RTE_LOG_ERR, RTE_LOG_EMERG, RTE_LOG_EMERG);
-
+#else
+   ret = rte_log_set_level_pattern("logtype", RTE_LOG_DEBUG);
+   CHECK_LEVELS(RTE_LOG_ERR, RTE_LOG_INFO, RTE_LOG_ERR);
+#endif
/* set logtype level low to so we can test global level */
rte_log_set_level_pattern("logtype*", RTE_LOG_DEBUG);
CHECK_LEVELS(RTE_LOG_DEBUG, RTE_LOG_DEBUG, RTE_LOG_DEBUG);
-- 
2.31.0.vfs.0.1

[PATCH v9 8/9] app/test: replace .sh script with .py script

[Jie Zhou]

- Add python script to check if system supports hugepages
- Remove corresponding .sh script
- Replace calling of .sh with corresponding .py in meson.build

Signed-off-by: Jie Zhou 

---
 app/test/has-hugepage.sh | 11 ---
 app/test/has_hugepage.py | 26 ++
 app/test/meson.build |  2 +-
 3 files changed, 27 insertions(+), 12 deletions(-)
 delete mode 100755 app/test/has-hugepage.sh
 create mode 100644 app/test/has_hugepage.py

diff --git a/app/test/has-hugepage.sh b/app/test/has-hugepage.sh
deleted file mode 100755
index d600fad319..00
--- a/app/test/has-hugepage.sh
+++ /dev/null
@@ -1,11 +0,0 @@
-#! /bin/sh
-# SPDX-License-Identifier: BSD-3-Clause
-# Copyright 2020 Mellanox Technologies, Ltd
-
-if [ "$(uname)" = "Linux" ] ; then
-   cat /proc/sys/vm/nr_hugepages || echo 0
-elif [ "$(uname)" = "FreeBSD" ] ; then
-   echo 1 # assume FreeBSD always has hugepages
-else
-   echo 0
-fi
diff --git a/app/test/has_hugepage.py b/app/test/has_hugepage.py
new file mode 100644
index 00..f8e7087d1c
--- /dev/null
+++ b/app/test/has_hugepage.py
@@ -0,0 +1,26 @@
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright (c) 2021 Microsoft Corporation
+"""This script checks if the system supports huge pages"""
+
+import platform
+import ctypes
+
+os_name = platform.system()
+if os_name == "Linux":
+try:
+with open("/proc/sys/vm/nr_hugepages") as file_o:
+content = file_o.read()
+print(content)
+except:
+print("0")
+
+elif os_name == "FreeBSD":
+# Assume FreeBSD always has hugepages enabled
+print("1")
+elif os_name == "Windows":
+if ctypes.windll.kernel32.GetLargePageMinimum() > 0:
+print("1")
+else:
+print("0")
+else:
+print("0")
diff --git a/app/test/meson.build b/app/test/meson.build
index 2b480adfba..97ee83029e 100644
--- a/app/test/meson.build
+++ b/app/test/meson.build
@@ -492,7 +492,7 @@ dpdk_test = executable('dpdk-test',
  driver_install_path),
 install: true)
 
-has_hugepage = run_command('has-hugepage.sh').stdout().strip() != '0'
+has_hugepage = run_command(py3, 'has_hugepage.py').stdout().strip() != '0'
 message('hugepage availability: @0@'.format(has_hugepage))
 
 # some perf tests (eg: memcpy perf autotest)take very long
-- 
2.31.0.vfs.0.1

[PATCH v9 7/9] app/test: remove two alarm_autotest cases

[Jie Zhou]

Remove two alarm_autotest test cases which do bogus range check
on Windows.

Signed-off-by: Jie Zhou 

---
 app/test/test_alarm.c | 4 
 1 file changed, 4 insertions(+)

diff --git a/app/test/test_alarm.c b/app/test/test_alarm.c
index b4034339b8..70e97a3109 100644
--- a/app/test/test_alarm.c
+++ b/app/test/test_alarm.c
@@ -10,6 +10,7 @@
 
 #include "test.h"
 
+#ifndef RTE_EXEC_ENV_WINDOWS
 static volatile int flag;
 
 static void
@@ -18,6 +19,7 @@ test_alarm_callback(void *cb_arg)
flag = 1;
printf("Callback setting flag - OK. [cb_arg = %p]\n", cb_arg);
 }
+#endif
 
 static int
 test_alarm(void)
@@ -27,6 +29,7 @@ test_alarm(void)
return 0;
 #endif
 
+#ifndef RTE_EXEC_ENV_WINDOWS
/* check if it will fail to set alarm with wrong us value */
printf("check if it will fail to set alarm with wrong ms values\n");
if (rte_eal_alarm_set(0, test_alarm_callback,
@@ -39,6 +42,7 @@ test_alarm(void)
printf("should not be successful with (UINT64_MAX-1) us 
value\n");
return -1;
}
+#endif
 
/* check if it will fail to set alarm with null callback parameter */
printf("check if it will fail to set alarm with null callback 
parameter\n");
-- 
2.31.0.vfs.0.1

[PATCH v9 6/9] app/test: differentiate a strerror on different OS

[Jie Zhou]

On Windows, strerror returns just "Unknown error" for errnum greater
than MAX_ERRNO, while linux and freebsd returns "Unknown error ",
which is the current expectation for errno_autotest. Differentiate
the error string on Windows to remove a "duplicate error code" failure.

Signed-off-by: Jie Zhou 

---
 app/test/test_errno.c | 12 +++-
 lib/eal/common/eal_common_errno.c |  4 
 2 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/app/test/test_errno.c b/app/test/test_errno.c
index 3ff0456a58..0db4fbc8b3 100644
--- a/app/test/test_errno.c
+++ b/app/test/test_errno.c
@@ -18,13 +18,19 @@ test_errno(void)
 {
const char *rte_retval;
const char *libc_retval;
+
+#ifndef RTE_EXEC_ENV_WINDOWS
 #ifdef RTE_EXEC_ENV_FREEBSD
/* BSD has a colon in the string, unlike linux */
const char unknown_code_result[] = "Unknown error: %d";
 #else
const char unknown_code_result[] = "Unknown error %d";
 #endif
-   char expected_libc_retval[sizeof(unknown_code_result)+3];
+   char expected_libc_retval[sizeof(unknown_code_result) + 3];
+#else
+   /* Windows doesn't return error number for error greater than 
MAX_errno*/
+   static const char expected_libc_retval[] = "Unknown error";
+#endif
 
/* use a small selection of standard errors for testing */
int std_errs[] = {EAGAIN, EBADF, EACCES, EINTR, EINVAL};
@@ -54,11 +60,13 @@ test_errno(void)
rte_retval, libc_retval);
if (strcmp(rte_retval, libc_retval) == 0)
return -1;
+#ifndef RTE_EXEC_ENV_WINDOWS
/* generate appropriate error string for unknown error number
 * and then check that this is what we got back. If not, we have
 * a duplicate error number that conflicts with errno.h */
snprintf(expected_libc_retval, sizeof(expected_libc_retval),
unknown_code_result, rte_errs[i]);
+#endif
if ((strcmp(expected_libc_retval, libc_retval) != 0) &&
(strcmp("", libc_retval) != 0)){
printf("Error, duplicate error code %d\n", rte_errs[i]);
@@ -69,8 +77,10 @@ test_errno(void)
/* ensure that beyond RTE_MAX_ERRNO, we always get an unknown code */
rte_retval = rte_strerror(RTE_MAX_ERRNO + 1);
libc_retval = strerror(RTE_MAX_ERRNO + 1);
+#ifndef RTE_EXEC_ENV_WINDOWS
snprintf(expected_libc_retval, sizeof(expected_libc_retval),
unknown_code_result, RTE_MAX_ERRNO + 1);
+#endif
printf("rte_strerror: '%s', strerror: '%s'\n",
rte_retval, libc_retval);
if ((strcmp(rte_retval, libc_retval) != 0) ||
diff --git a/lib/eal/common/eal_common_errno.c 
b/lib/eal/common/eal_common_errno.c
index f86802705a..4c4abb802e 100644
--- a/lib/eal/common/eal_common_errno.c
+++ b/lib/eal/common/eal_common_errno.c
@@ -37,7 +37,11 @@ rte_strerror(int errnum)
/* since some implementations of strerror_r throw an error
 * themselves if errnum is too big, we handle that case here */
if (errnum >= RTE_MAX_ERRNO)
+#ifdef RTE_EXEC_ENV_WINDOWS
+   snprintf(ret, RETVAL_SZ, "Unknown error%s", sep);
+#else
snprintf(ret, RETVAL_SZ, "Unknown error%s %d", sep, errnum);
+#endif
else
switch (errnum){
case E_RTE_SECONDARY:
-- 
2.31.0.vfs.0.1

[PATCH v9 9/9] app/test: enable subset of unit tests on Windows

[Jie Zhou]

- For fast tests and perf tests, add test stubs to skip not supported
  ones.
- For driver tests, for now skip on Windows totally to avoid
  unnecessary amount of test stubs. For example, there are about 30
  cryptodev related tests (even though in the meson for CI it only
  listed about half) which will be enabled by "patch-18949: app/test:
  enable crypto unit tests on Windows".
- For dump tests, currently the tests hang on Windows which require
  further investigation. Keep the dump test list just for non-Windows
  for easier tracking.

Signed-off-by: Jie Zhou 

---
 app/test/meson.build | 117 ---
 app/test/test_acl.c  |  12 +++
 app/test/test_bpf.c  |  15 ++-
 app/test/test_cksum.c|  12 +++
 app/test/test_cmdline_ipaddr.c   |   5 +
 app/test/test_cryptodev.c|   4 +
 app/test/test_cryptodev_asym.c   |   4 +
 app/test/test_cryptodev_blockcipher.c|   4 +
 app/test/test_cryptodev_security_ipsec.c |   4 +
 app/test/test_cryptodev_security_pdcp.c  |   4 +
 app/test/test_debug.c|  17 +++-
 app/test/test_distributor.c  |  13 +++
 app/test/test_distributor_perf.c |  13 +++
 app/test/test_dmadev.c   |  14 ++-
 app/test/test_dmadev_api.c   |   4 +
 app/test/test_eal_flags.c|  90 +
 app/test/test_eal_fs.c   |  12 +++
 app/test/test_efd.c  |  15 ++-
 app/test/test_efd_perf.c |  16 +++-
 app/test/test_event_crypto_adapter.c |  15 ++-
 app/test/test_event_eth_rx_adapter.c |  25 -
 app/test/test_event_eth_tx_adapter.c |  12 +++
 app/test/test_event_ring.c   |  16 +++-
 app/test/test_event_timer_adapter.c  |  16 +++-
 app/test/test_eventdev.c |  20 +++-
 app/test/test_external_mem.c |  18 +++-
 app/test/test_fib.c  |  22 -
 app/test/test_fib6.c |  24 -
 app/test/test_fib6_perf.c|  16 +++-
 app/test/test_fib_perf.c |  15 ++-
 app/test/test_flow_classify.c|  13 +++
 app/test/test_func_reentrancy.c  |  12 +++
 app/test/test_graph.c|  18 +++-
 app/test/test_graph_perf.c   |  16 +++-
 app/test/test_hash_perf.c|  12 +++
 app/test/test_ipfrag.c   |  16 +++-
 app/test/test_ipsec.c|  15 ++-
 app/test/test_ipsec_perf.c   |  15 ++-
 app/test/test_ipsec_sad.c|  14 ++-
 app/test/test_kni.c  |  10 +-
 app/test/test_lcores.c   |  12 +++
 app/test/test_lpm.c  |  14 ++-
 app/test/test_lpm6.c |  14 ++-
 app/test/test_lpm6_perf.c|  14 ++-
 app/test/test_lpm_perf.c |  13 ++-
 app/test/test_malloc.c   |  17 +++-
 app/test/test_mbuf.c |  13 ++-
 app/test/test_member.c   |  16 +++-
 app/test/test_member_perf.c  |  16 +++-
 app/test/test_memcpy_perf.c  |  18 +++-
 app/test/test_mempool_perf.c |  12 +++
 app/test/test_mp_secondary.c |  12 +++
 app/test/test_pie.c  |  30 +-
 app/test/test_rawdev.c   |  17 +++-
 app/test/test_rcu_qsbr_perf.c|  12 +++
 app/test/test_reciprocal_division.c  |  12 +++
 app/test/test_reciprocal_division_perf.c |  12 +++
 app/test/test_red.c  |  29 +-
 app/test/test_reorder.c  |  15 ++-
 app/test/test_rib.c  |  22 -
 app/test/test_rib6.c |  22 -
 app/test/test_sched.c|  14 ++-
 app/test/test_security.c |  16 +++-
 app/test/test_table.c|  13 +++
 app/test/test_table_acl.c|   3 +
 app/test/test_table_combined.c   |   4 +
 app/test/test_table_pipeline.c   |   4 +
 app/test/test_table_ports.c  |   4 +
 app/test/test_table_tables.c |   4 +
 app/test/test_timer_secondary.c  |  13 +++
 app/test/test_trace.c|  32 ++-
 71 files changed, 1038 insertions(+), 126 deletions(-)

diff --git a/app/test/meson.build b/app/test/meson.build
index 97ee83029e..52eddab36d 100644
--- a/app/test/meson.build
+++ b/app/test/meson.build
@@ -1,12 +1,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 # Copyright(c) 2017 Intel Corporation
 
-if is_windows
-build = false
-reason = 'not supported on Windows'
-subdir_done()
-endif
-
 if not get_option('tests')
 subdir_done()
 endif
@@ -158,34 +152,17 @@ test_sources = files(
 )
 
 test_deps = [
-'acl',
 'bus_pci',
 'bus_vdev',
-'bpf',
+'bitratestats',
 'cfgfile',
 'cmdline',
-'cryptodev',
- 

[PATCH v10 1/9] eal/windows: return ENOTSUP for not supported API

[Jie Zhou]

UT memory_autotest on Windows has 2 failed cases on eal APIs
eal_memalloc_get_seg_fd and eal_memalloc_get_seg_fd_offset. These 2
APIs are not supported on Windows yet. Should return ENOTSUP such that
in test_memory.c these 2 ENOTSUP cases will not be marked as failures,
same as other ENOTSUP cases.

Signed-off-by: Jie Zhou 
Acked-by: Dmitry Kozlyuk 

---
 lib/eal/windows/eal_memalloc.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/lib/eal/windows/eal_memalloc.c b/lib/eal/windows/eal_memalloc.c
index 55d6dcc71c..aa7589b81d 100644
--- a/lib/eal/windows/eal_memalloc.c
+++ b/lib/eal/windows/eal_memalloc.c
@@ -17,7 +17,7 @@ eal_memalloc_get_seg_fd(int list_idx, int seg_idx)
RTE_SET_USED(list_idx);
RTE_SET_USED(seg_idx);
EAL_LOG_NOT_IMPLEMENTED();
-   return -1;
+   return -ENOTSUP;
 }
 
 int
@@ -28,7 +28,7 @@ eal_memalloc_get_seg_fd_offset(int list_idx, int seg_idx, 
size_t *offset)
RTE_SET_USED(seg_idx);
RTE_SET_USED(offset);
EAL_LOG_NOT_IMPLEMENTED();
-   return -1;
+   return -ENOTSUP;
 }
 
 static int
@@ -428,7 +428,7 @@ eal_memalloc_sync_with_primary(void)
 {
/* No multi-process support. */
EAL_LOG_NOT_IMPLEMENTED();
-   return -1;
+   return -ENOTSUP;
 }
 
 int
-- 
2.31.0.vfs.0.1

[PATCH v10 3/9] app/test: fix incorrect errno variable

[Jie Zhou]

Fix incorrect errno variable used in memory autotest.
Use rte_errno instead.

Fixes: 086d426406bd ("app/test: fix memory autotests on FreeBSD")

Signed-off-by: Jie Zhou 

---
 app/test/test_memory.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/app/test/test_memory.c b/app/test/test_memory.c
index dbf6871e71..140ac3f3cf 100644
--- a/app/test/test_memory.c
+++ b/app/test/test_memory.c
@@ -63,7 +63,7 @@ check_seg_fds(const struct rte_memseg_list *msl, const struct 
rte_memseg *ms,
/* we're able to get memseg fd - try getting its offset */
ret = rte_memseg_get_fd_offset_thread_unsafe(ms, &offset);
if (ret < 0) {
-   if (errno == ENOTSUP)
+   if (rte_errno == ENOTSUP)
return 1;
return -1;
}
-- 
2.31.0.vfs.0.1

[PATCH v10 4/9] app/test: skip interrupt tests on Windows

[Jie Zhou]

Even though test_interrupts.c can compile on Windows, skip interrupt
tests for now since majority of eal_interrupt on Windows are stubs.
Will remove the skip after interrupt being fully enabled on Windows.

Signed-off-by: Jie Zhou 

---
 app/test/test_interrupts.c | 10 ++
 1 file changed, 10 insertions(+)

diff --git a/app/test/test_interrupts.c b/app/test/test_interrupts.c
index 2a05399f96..eec9b2805b 100644
--- a/app/test/test_interrupts.c
+++ b/app/test/test_interrupts.c
@@ -12,6 +12,15 @@
 
 #include "test.h"
 
+#ifdef RTE_EXEC_ENV_WINDOWS
+static int
+test_interrupt(void)
+{
+   printf("Interrupt on Windows is not fully supported yet, skipping 
test\n");
+   return TEST_SKIPPED;
+}
+#else
+
 #define TEST_INTERRUPT_CHECK_INTERVAL 100 /* ms */
 
 /* predefined interrupt handle types */
@@ -590,5 +599,6 @@ test_interrupt(void)
 
return ret;
 }
+#endif /*ifdef RTE_EXEC_ENV_WINDOWS*/
 
 REGISTER_TEST_COMMAND(interrupt_autotest, test_interrupt);
-- 
2.31.0.vfs.0.1

[PATCH v10 2/9] app/test: remove POSIX-specific code

[Jie Zhou]

- Remove header inclusion of netinet/in.h and terminos.h
- Include rte_os_shim.h
- Replace sleep and usleep with rte_delay_us_sleep
- Use NUL on Windows as /dev/null for Linux
- Exclude tests not supported on Windows yet, e.g. multi-process, and
  IP address parsing (the test cases use linux netinet/in.h u6_addr
  and better have a sperate patch to add such tests on Windows)

Signed-off-by: Jie Zhou 

---
 app/test/commands.c   |  2 --
 app/test/packet_burst_generator.c |  1 +
 app/test/process.h|  4 +++-
 app/test/test.c   |  5 -
 app/test/test_byteorder.c |  2 +-
 app/test/test_cmdline.c   |  2 ++
 app/test/test_cmdline_lib.c   | 13 +
 app/test/test_crc.c   |  1 -
 app/test/test_pmd_perf.c  |  6 +-
 app/test/test_ring_stress_impl.h  |  2 +-
 app/test/test_telemetry_data.c|  2 ++
 11 files changed, 28 insertions(+), 12 deletions(-)

diff --git a/app/test/commands.c b/app/test/commands.c
index 2dced3bc44..887cabad64 100644
--- a/app/test/commands.c
+++ b/app/test/commands.c
@@ -8,8 +8,6 @@
 #include 
 #include 
 #include 
-#include 
-#include 
 #include 
 #include 
 #include 
diff --git a/app/test/packet_burst_generator.c 
b/app/test/packet_burst_generator.c
index 8ac24577ba..6b42b9b83b 100644
--- a/app/test/packet_burst_generator.c
+++ b/app/test/packet_burst_generator.c
@@ -5,6 +5,7 @@
 #include 
 #include 
 #include 
+#include 
 
 #include "packet_burst_generator.h"
 
diff --git a/app/test/process.h b/app/test/process.h
index 5b10cf64df..1f073b9c5c 100644
--- a/app/test/process.h
+++ b/app/test/process.h
@@ -7,12 +7,14 @@
 
 #include   /* errno */
 #include  /* PATH_MAX */
+#ifndef RTE_EXEC_ENV_WINDOWS
 #include  /* basename et al */
+#include 
+#endif
 #include  /* NULL */
 #include  /* strerror */
 #include  /* readlink */
 #include 
-#include 
 
 #include  /* strlcpy */
 
diff --git a/app/test/test.c b/app/test/test.c
index 5194131026..e69cae3eea 100644
--- a/app/test/test.c
+++ b/app/test/test.c
@@ -8,7 +8,6 @@
 #include 
 #include 
 #include 
-#include 
 #include 
 #include 
 
@@ -63,7 +62,9 @@ do_recursive_call(void)
const char *env_var;
int (*action_fn)(void);
} actions[] =  {
+#ifndef RTE_EXEC_ENV_WINDOWS
{ "run_secondary_instances", test_mp_secondary },
+#endif
 #ifdef RTE_LIB_PDUMP
 #ifdef RTE_NET_RING
{ "run_pdump_server_tests", test_pdump },
@@ -82,7 +83,9 @@ do_recursive_call(void)
{ "test_file_prefix", no_action },
{ "test_no_huge_flag", no_action },
 #ifdef RTE_LIB_TIMER
+#ifndef RTE_EXEC_ENV_WINDOWS
{ "timer_secondary_spawn_wait", test_timer_secondary },
+#endif
 #endif
};
 
diff --git a/app/test/test_byteorder.c b/app/test/test_byteorder.c
index 03c08d9abf..de14ed539e 100644
--- a/app/test/test_byteorder.c
+++ b/app/test/test_byteorder.c
@@ -46,7 +46,7 @@ test_byteorder(void)
return -1;
 
res_u16 = rte_bswap16(0x1337);
-   printf("const %"PRIx16" -> %"PRIx16"\n", 0x1337, res_u16);
+   printf("const %"PRIx16" -> %"PRIx16"\n", (uint16_t)0x1337, res_u16);
if (res_u16 != 0x3713)
return -1;
 
diff --git a/app/test/test_cmdline.c b/app/test/test_cmdline.c
index 115bee966d..9a76bd299f 100644
--- a/app/test/test_cmdline.c
+++ b/app/test/test_cmdline.c
@@ -31,6 +31,7 @@ test_cmdline(void)
return -1;
if (test_parse_num_invalid_param() < 0)
return -1;
+#ifndef RTE_EXEC_ENV_WINDOWS
printf("Testing parsing IP addresses...\n");
if (test_parse_ipaddr_valid() < 0)
return -1;
@@ -38,6 +39,7 @@ test_cmdline(void)
return -1;
if (test_parse_ipaddr_invalid_param() < 0)
return -1;
+#endif
printf("Testing parsing strings...\n");
if (test_parse_string_valid() < 0)
return -1;
diff --git a/app/test/test_cmdline_lib.c b/app/test/test_cmdline_lib.c
index f50ccdb599..fcd58cb76a 100644
--- a/app/test/test_cmdline_lib.c
+++ b/app/test/test_cmdline_lib.c
@@ -8,7 +8,6 @@
 #include 
 #include 
 #include 
-#include 
 #include 
 #include 
 
@@ -22,6 +21,12 @@
 
 #include "test_cmdline.h"
 
+#ifndef RTE_EXEC_ENV_WINDOWS
+#define NULL_INPUT "/dev/null"
+#else
+#define NULL_INPUT "NUL"
+#endif
+
 //
 /* static functions required for some tests */
 static void
@@ -156,10 +161,10 @@ test_cmdline_socket_fns(void)
cl = cmdline_stdin_new(&ctx, NULL);
if (cl != NULL)
goto error;
-   cl = cmdline_file_new(NULL, "prompt", "/dev/null");
+   cl = cmdline_file_new(NULL, "prompt", NULL_INPUT);
if (cl != NULL)
goto error;
-   cl = cmdline_file_new(&ctx, NULL, "/dev/null");
+   cl = cmdline_file_new(&ctx, NULL, NULL_INPUT);
if (cl

[PATCH v10 0/9] app/test: enable subset of tests on Windows

[Jie Zhou]

This patchset is to enable a subset of unit tests on windows. It mainly 
includes:
- Replace POSIX specific codes
- Add test stubs for not supported ones on Windows
- Fix some lib and tests per failures investigation
- Replace .sh script with .py script for meson.build
- Enable build and run subset of unit tests on Windows

Future work:
- Work with CI/CD lab to onboard unit tests on Windows to catch regression
- Enable more tests

---
V2 changes:
- Fix compilation error on FreeBSD
- Fix email mismatch issue
- Add a missing space around "*"

---
V3 changes:
- Fix a misc c coding style issue
- Revise some commit title and message body
- Fix violations of PEP8 in new added Python scripts
- Add error handling in get_coremask.py
- Fix has_hugepage.py to check system support of hugepages
  instead of checking privileges
- Fix test meson.build to run Python scripts using py3
- Consolidate lists of source files, test dep, etc. across all 
  platforms, with conditional extending on some platform(s)

---
V4 changes:
- Remove building of ip_frag, rib, and reorder libraries on Windows.
  These three libs usually can be built on Windows without change.
  However, in between the time of V3 and V4, there is regression in
  upstream caused build failures of these three libs. Will separately
  investigate and enable these libraries.

- Remove previous patch#2 (Enable mempool/stack on Windows) from this
  patchset as it was separated out and merged as patch-19314.

- Consolidate the source files, deps, tests lists across platforms as
  much as possible.

---
V5 changes:
- Remove a space between function name and open parenthesis '('
- Add back a header mistakenly deleted

---
V6 changes:
   - Fix inconsistent static vs. non-static declarations

---
V7 changes:
   - Remove get_coremask.py as it is not needed any more in meson.build
   - Remove enablement of efd and lpm and their corresponding unit tests.
 The enablement of these two libs and their UTs will be in separate
 patches after this patch set.

V8 changes:
   - Fix coding style issue of using C99 // comments

V9 changes:
   - Fix has_hugepage.py with adding failure handling on Linux, using
 proper variable name to follow Python convention, and removing
 unnecessary comment.
   - Enable previously skipped test_cmdline_socket_fns test cases
   - Revise title and message, and add Fixes info for current Patch#3 
   - Combine 2 patches (previous #2 and #3 in V8) into one and with 
 more detailed message

V10 changes:
   - Fix indentation

Jie Zhou (9):
  eal/windows: return ENOTSUP for not supported API
  app/test: remove POSIX-specific code
  app/test: fix incorrect errno variable
  app/test: skip interrupt tests on Windows
  app/test: skip two logs_autotest cases on Windows
  app/test: differentiate a strerror on different OS
  app/test: remove two alarm_autotest cases
  app/test: replace .sh script with .py script
  app/test: enable subset of unit tests on Windows

 app/test/commands.c  |   2 -
 app/test/has-hugepage.sh |  11 ---
 app/test/has_hugepage.py |  26 +
 app/test/meson.build | 119 ---
 app/test/packet_burst_generator.c|   1 +
 app/test/process.h   |   4 +-
 app/test/test.c  |   5 +-
 app/test/test_acl.c  |  12 +++
 app/test/test_alarm.c|   4 +
 app/test/test_bpf.c  |  15 ++-
 app/test/test_byteorder.c|   2 +-
 app/test/test_cksum.c|  12 +++
 app/test/test_cmdline.c  |   2 +
 app/test/test_cmdline_ipaddr.c   |   5 +
 app/test/test_cmdline_lib.c  |  13 ++-
 app/test/test_crc.c  |   1 -
 app/test/test_cryptodev.c|   4 +
 app/test/test_cryptodev_asym.c   |   4 +
 app/test/test_cryptodev_blockcipher.c|   4 +
 app/test/test_cryptodev_security_ipsec.c |   4 +
 app/test/test_cryptodev_security_pdcp.c  |   4 +
 app/test/test_debug.c|  17 +++-
 app/test/test_distributor.c  |  13 +++
 app/test/test_distributor_perf.c |  13 +++
 app/test/test_dmadev.c   |  14 ++-
 app/test/test_dmadev_api.c   |   4 +
 app/test/test_eal_flags.c|  90 +
 app/test/test_eal_fs.c   |  12 +++
 app/test/test_efd.c  |  15 ++-
 app/test/test_efd_perf.c |  16 ++-
 app/test/test_errno.c|  12 ++-
 app/test/test_event_crypto_adapter.c |  15 ++-
 app/test/test_event_eth_rx_adapter.c |  25 -
 app/test/test_event_eth_tx_adapter.c |  12 +++
 app/test/test_event_ring.c   |  16 ++-
 app/test/test_event_timer_adapter.c  |  16 ++-
 app/test/test_eventdev.c |  20

[PATCH v10 7/9] app/test: remove two alarm_autotest cases

[Jie Zhou]

Remove two alarm_autotest test cases which do bogus range check
on Windows.

Signed-off-by: Jie Zhou 

---
 app/test/test_alarm.c | 4 
 1 file changed, 4 insertions(+)

diff --git a/app/test/test_alarm.c b/app/test/test_alarm.c
index b4034339b8..70e97a3109 100644
--- a/app/test/test_alarm.c
+++ b/app/test/test_alarm.c
@@ -10,6 +10,7 @@
 
 #include "test.h"
 
+#ifndef RTE_EXEC_ENV_WINDOWS
 static volatile int flag;
 
 static void
@@ -18,6 +19,7 @@ test_alarm_callback(void *cb_arg)
flag = 1;
printf("Callback setting flag - OK. [cb_arg = %p]\n", cb_arg);
 }
+#endif
 
 static int
 test_alarm(void)
@@ -27,6 +29,7 @@ test_alarm(void)
return 0;
 #endif
 
+#ifndef RTE_EXEC_ENV_WINDOWS
/* check if it will fail to set alarm with wrong us value */
printf("check if it will fail to set alarm with wrong ms values\n");
if (rte_eal_alarm_set(0, test_alarm_callback,
@@ -39,6 +42,7 @@ test_alarm(void)
printf("should not be successful with (UINT64_MAX-1) us 
value\n");
return -1;
}
+#endif
 
/* check if it will fail to set alarm with null callback parameter */
printf("check if it will fail to set alarm with null callback 
parameter\n");
-- 
2.31.0.vfs.0.1

[PATCH v10 8/9] app/test: replace .sh script with .py script

[Jie Zhou]

- Add python script to check if system supports hugepages
- Remove corresponding .sh script
- Replace calling of .sh with corresponding .py in meson.build

Signed-off-by: Jie Zhou 

---
 app/test/has-hugepage.sh | 11 ---
 app/test/has_hugepage.py | 26 ++
 app/test/meson.build |  2 +-
 3 files changed, 27 insertions(+), 12 deletions(-)
 delete mode 100755 app/test/has-hugepage.sh
 create mode 100644 app/test/has_hugepage.py

diff --git a/app/test/has-hugepage.sh b/app/test/has-hugepage.sh
deleted file mode 100755
index d600fad319..00
--- a/app/test/has-hugepage.sh
+++ /dev/null
@@ -1,11 +0,0 @@
-#! /bin/sh
-# SPDX-License-Identifier: BSD-3-Clause
-# Copyright 2020 Mellanox Technologies, Ltd
-
-if [ "$(uname)" = "Linux" ] ; then
-   cat /proc/sys/vm/nr_hugepages || echo 0
-elif [ "$(uname)" = "FreeBSD" ] ; then
-   echo 1 # assume FreeBSD always has hugepages
-else
-   echo 0
-fi
diff --git a/app/test/has_hugepage.py b/app/test/has_hugepage.py
new file mode 100644
index 00..f8e7087d1c
--- /dev/null
+++ b/app/test/has_hugepage.py
@@ -0,0 +1,26 @@
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright (c) 2021 Microsoft Corporation
+"""This script checks if the system supports huge pages"""
+
+import platform
+import ctypes
+
+os_name = platform.system()
+if os_name == "Linux":
+try:
+with open("/proc/sys/vm/nr_hugepages") as file_o:
+content = file_o.read()
+print(content)
+except:
+print("0")
+
+elif os_name == "FreeBSD":
+# Assume FreeBSD always has hugepages enabled
+print("1")
+elif os_name == "Windows":
+if ctypes.windll.kernel32.GetLargePageMinimum() > 0:
+print("1")
+else:
+print("0")
+else:
+print("0")
diff --git a/app/test/meson.build b/app/test/meson.build
index 2b480adfba..97ee83029e 100644
--- a/app/test/meson.build
+++ b/app/test/meson.build
@@ -492,7 +492,7 @@ dpdk_test = executable('dpdk-test',
  driver_install_path),
 install: true)
 
-has_hugepage = run_command('has-hugepage.sh').stdout().strip() != '0'
+has_hugepage = run_command(py3, 'has_hugepage.py').stdout().strip() != '0'
 message('hugepage availability: @0@'.format(has_hugepage))
 
 # some perf tests (eg: memcpy perf autotest)take very long
-- 
2.31.0.vfs.0.1

[PATCH v10 5/9] app/test: skip two logs_autotest cases on Windows

[Jie Zhou]

DPDK logs_autotest on Windows failed at "dynamic log types" tests.
The failures are on 2 test cases for rte_log_set_level_regexp API,
due to regular expression is not supported on Windows in DPDK yet
and regcomp/regexec are just stubs on Windows (in regex.h).

In app\test\test_logs.c, ifndef these two test cases, and for the
rte_log_set_level_pattern validation case following these two cases,
differentiate the expected log level passed into macro CHECK_LEVELS

Now logs_autotest completes for all dynamic log types and static log types.

Signed-off-by: Jie Zhou 

---
 app/test/test_logs.c | 6 +-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/app/test/test_logs.c b/app/test/test_logs.c
index 7abb6eeca2..7c001c1ab3 100644
--- a/app/test/test_logs.c
+++ b/app/test/test_logs.c
@@ -113,6 +113,7 @@ test_logs(void)
rte_log_set_level(logtype1, RTE_LOG_ERR);
CHECK_LEVELS(RTE_LOG_ERR, RTE_LOG_INFO, RTE_LOG_ERR);
 
+#ifndef RTE_EXEC_ENV_WINDOWS
rte_log_set_level_regexp("type$", RTE_LOG_EMERG);
CHECK_LEVELS(RTE_LOG_ERR, RTE_LOG_INFO, RTE_LOG_ERR);
 
@@ -121,7 +122,10 @@ test_logs(void)
 
rte_log_set_level_pattern("logtype", RTE_LOG_DEBUG);
CHECK_LEVELS(RTE_LOG_ERR, RTE_LOG_EMERG, RTE_LOG_EMERG);
-
+#else
+   ret = rte_log_set_level_pattern("logtype", RTE_LOG_DEBUG);
+   CHECK_LEVELS(RTE_LOG_ERR, RTE_LOG_INFO, RTE_LOG_ERR);
+#endif
/* set logtype level low to so we can test global level */
rte_log_set_level_pattern("logtype*", RTE_LOG_DEBUG);
CHECK_LEVELS(RTE_LOG_DEBUG, RTE_LOG_DEBUG, RTE_LOG_DEBUG);
-- 
2.31.0.vfs.0.1

[PATCH v10 6/9] app/test: differentiate a strerror on different OS

[Jie Zhou]

On Windows, strerror returns just "Unknown error" for errnum greater
than MAX_ERRNO, while linux and freebsd returns "Unknown error ",
which is the current expectation for errno_autotest. Differentiate
the error string on Windows to remove a "duplicate error code" failure.

Signed-off-by: Jie Zhou 

---
 app/test/test_errno.c | 12 +++-
 lib/eal/common/eal_common_errno.c |  4 
 2 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/app/test/test_errno.c b/app/test/test_errno.c
index 3ff0456a58..0db4fbc8b3 100644
--- a/app/test/test_errno.c
+++ b/app/test/test_errno.c
@@ -18,13 +18,19 @@ test_errno(void)
 {
const char *rte_retval;
const char *libc_retval;
+
+#ifndef RTE_EXEC_ENV_WINDOWS
 #ifdef RTE_EXEC_ENV_FREEBSD
/* BSD has a colon in the string, unlike linux */
const char unknown_code_result[] = "Unknown error: %d";
 #else
const char unknown_code_result[] = "Unknown error %d";
 #endif
-   char expected_libc_retval[sizeof(unknown_code_result)+3];
+   char expected_libc_retval[sizeof(unknown_code_result) + 3];
+#else
+   /* Windows doesn't return error number for error greater than 
MAX_errno*/
+   static const char expected_libc_retval[] = "Unknown error";
+#endif
 
/* use a small selection of standard errors for testing */
int std_errs[] = {EAGAIN, EBADF, EACCES, EINTR, EINVAL};
@@ -54,11 +60,13 @@ test_errno(void)
rte_retval, libc_retval);
if (strcmp(rte_retval, libc_retval) == 0)
return -1;
+#ifndef RTE_EXEC_ENV_WINDOWS
/* generate appropriate error string for unknown error number
 * and then check that this is what we got back. If not, we have
 * a duplicate error number that conflicts with errno.h */
snprintf(expected_libc_retval, sizeof(expected_libc_retval),
unknown_code_result, rte_errs[i]);
+#endif
if ((strcmp(expected_libc_retval, libc_retval) != 0) &&
(strcmp("", libc_retval) != 0)){
printf("Error, duplicate error code %d\n", rte_errs[i]);
@@ -69,8 +77,10 @@ test_errno(void)
/* ensure that beyond RTE_MAX_ERRNO, we always get an unknown code */
rte_retval = rte_strerror(RTE_MAX_ERRNO + 1);
libc_retval = strerror(RTE_MAX_ERRNO + 1);
+#ifndef RTE_EXEC_ENV_WINDOWS
snprintf(expected_libc_retval, sizeof(expected_libc_retval),
unknown_code_result, RTE_MAX_ERRNO + 1);
+#endif
printf("rte_strerror: '%s', strerror: '%s'\n",
rte_retval, libc_retval);
if ((strcmp(rte_retval, libc_retval) != 0) ||
diff --git a/lib/eal/common/eal_common_errno.c 
b/lib/eal/common/eal_common_errno.c
index f86802705a..4c4abb802e 100644
--- a/lib/eal/common/eal_common_errno.c
+++ b/lib/eal/common/eal_common_errno.c
@@ -37,7 +37,11 @@ rte_strerror(int errnum)
/* since some implementations of strerror_r throw an error
 * themselves if errnum is too big, we handle that case here */
if (errnum >= RTE_MAX_ERRNO)
+#ifdef RTE_EXEC_ENV_WINDOWS
+   snprintf(ret, RETVAL_SZ, "Unknown error%s", sep);
+#else
snprintf(ret, RETVAL_SZ, "Unknown error%s %d", sep, errnum);
+#endif
else
switch (errnum){
case E_RTE_SECONDARY:
-- 
2.31.0.vfs.0.1

[PATCH v10 9/9] app/test: enable subset of unit tests on Windows

[Jie Zhou]

- For fast tests and perf tests, add test stubs to skip not supported
  ones.
- For driver tests, for now skip on Windows totally to avoid
  unnecessary amount of test stubs. For example, there are about 30
  cryptodev related tests (even though in the meson for CI it only
  listed about half) which will be enabled by "patch-18949: app/test:
  enable crypto unit tests on Windows".
- For dump tests, currently the tests hang on Windows which require
  further investigation. Keep the dump test list just for non-Windows
  for easier tracking.

Signed-off-by: Jie Zhou 

---
 app/test/meson.build | 117 ---
 app/test/test_acl.c  |  12 +++
 app/test/test_bpf.c  |  15 ++-
 app/test/test_cksum.c|  12 +++
 app/test/test_cmdline_ipaddr.c   |   5 +
 app/test/test_cryptodev.c|   4 +
 app/test/test_cryptodev_asym.c   |   4 +
 app/test/test_cryptodev_blockcipher.c|   4 +
 app/test/test_cryptodev_security_ipsec.c |   4 +
 app/test/test_cryptodev_security_pdcp.c  |   4 +
 app/test/test_debug.c|  17 +++-
 app/test/test_distributor.c  |  13 +++
 app/test/test_distributor_perf.c |  13 +++
 app/test/test_dmadev.c   |  14 ++-
 app/test/test_dmadev_api.c   |   4 +
 app/test/test_eal_flags.c|  90 +
 app/test/test_eal_fs.c   |  12 +++
 app/test/test_efd.c  |  15 ++-
 app/test/test_efd_perf.c |  16 +++-
 app/test/test_event_crypto_adapter.c |  15 ++-
 app/test/test_event_eth_rx_adapter.c |  25 -
 app/test/test_event_eth_tx_adapter.c |  12 +++
 app/test/test_event_ring.c   |  16 +++-
 app/test/test_event_timer_adapter.c  |  16 +++-
 app/test/test_eventdev.c |  20 +++-
 app/test/test_external_mem.c |  18 +++-
 app/test/test_fib.c  |  22 -
 app/test/test_fib6.c |  24 -
 app/test/test_fib6_perf.c|  16 +++-
 app/test/test_fib_perf.c |  15 ++-
 app/test/test_flow_classify.c|  13 +++
 app/test/test_func_reentrancy.c  |  12 +++
 app/test/test_graph.c|  18 +++-
 app/test/test_graph_perf.c   |  16 +++-
 app/test/test_hash_perf.c|  12 +++
 app/test/test_ipfrag.c   |  16 +++-
 app/test/test_ipsec.c|  15 ++-
 app/test/test_ipsec_perf.c   |  15 ++-
 app/test/test_ipsec_sad.c|  14 ++-
 app/test/test_kni.c  |  10 +-
 app/test/test_lcores.c   |  12 +++
 app/test/test_lpm.c  |  14 ++-
 app/test/test_lpm6.c |  14 ++-
 app/test/test_lpm6_perf.c|  14 ++-
 app/test/test_lpm_perf.c |  13 ++-
 app/test/test_malloc.c   |  17 +++-
 app/test/test_mbuf.c |  13 ++-
 app/test/test_member.c   |  16 +++-
 app/test/test_member_perf.c  |  16 +++-
 app/test/test_memcpy_perf.c  |  18 +++-
 app/test/test_mempool_perf.c |  12 +++
 app/test/test_mp_secondary.c |  12 +++
 app/test/test_pie.c  |  30 +-
 app/test/test_rawdev.c   |  17 +++-
 app/test/test_rcu_qsbr_perf.c|  12 +++
 app/test/test_reciprocal_division.c  |  12 +++
 app/test/test_reciprocal_division_perf.c |  12 +++
 app/test/test_red.c  |  29 +-
 app/test/test_reorder.c  |  15 ++-
 app/test/test_rib.c  |  22 -
 app/test/test_rib6.c |  22 -
 app/test/test_sched.c|  14 ++-
 app/test/test_security.c |  16 +++-
 app/test/test_table.c|  13 +++
 app/test/test_table_acl.c|   3 +
 app/test/test_table_combined.c   |   4 +
 app/test/test_table_pipeline.c   |   4 +
 app/test/test_table_ports.c  |   4 +
 app/test/test_table_tables.c |   4 +
 app/test/test_timer_secondary.c  |  13 +++
 app/test/test_trace.c|  32 ++-
 71 files changed, 1038 insertions(+), 126 deletions(-)

diff --git a/app/test/meson.build b/app/test/meson.build
index 97ee83029e..52eddab36d 100644
--- a/app/test/meson.build
+++ b/app/test/meson.build
@@ -1,12 +1,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 # Copyright(c) 2017 Intel Corporation
 
-if is_windows
-build = false
-reason = 'not supported on Windows'
-subdir_done()
-endif
-
 if not get_option('tests')
 subdir_done()
 endif
@@ -158,34 +152,17 @@ test_sources = files(
 )
 
 test_deps = [
-'acl',
 'bus_pci',
 'bus_vdev',
-'bpf',
+'bitratestats',
 'cfgfile',
 'cmdline',
-'cryptodev',
- 

Re: [PATCH v9 9/9] app/test: enable subset of unit tests on Windows

[Stephen Hemminger]

On Wed,  1 Dec 2021 10:05:38 -0800
Jie Zhou  wrote:

> diff --git a/app/test/test_dmadev.c b/app/test/test_dmadev.c
> index b206db27ae..c2ea9090cf 100644
> --- a/app/test/test_dmadev.c
> +++ b/app/test/test_dmadev.c
> @@ -2,6 +2,17 @@
>   * Copyright(c) 2021 HiSilicon Limited
>   * Copyright(c) 2021 Intel Corporation
>   */
> +#include "test.h"
> +
> +#ifdef RTE_EXEC_ENV_WINDOWS
> +static int
> +test_dma(void)
> +{
> + printf("dma not supported on Windows, skipping test\n");
> + return TEST_SKIPPED;
> +}
> +
> +#else

Rather than littering code with ifdefs is it not possible to
just not build these tests on Windows?

Re: [PATCH v9 9/9] app/test: enable subset of unit tests on Windows

[Jie Zhou]

On Wed, Dec 01, 2021 at 10:45:35AM -0800, Stephen Hemminger wrote:
> On Wed,  1 Dec 2021 10:05:38 -0800
> Jie Zhou  wrote:
> 
> > diff --git a/app/test/test_dmadev.c b/app/test/test_dmadev.c
> > index b206db27ae..c2ea9090cf 100644
> > --- a/app/test/test_dmadev.c
> > +++ b/app/test/test_dmadev.c
> > @@ -2,6 +2,17 @@
> >   * Copyright(c) 2021 HiSilicon Limited
> >   * Copyright(c) 2021 Intel Corporation
> >   */
> > +#include "test.h"
> > +
> > +#ifdef RTE_EXEC_ENV_WINDOWS
> > +static int
> > +test_dma(void)
> > +{
> > +   printf("dma not supported on Windows, skipping test\n");
> > +   return TEST_SKIPPED;
> > +}
> > +
> > +#else
> 
> Rather than littering code with ifdefs is it not possible to
> just not build these tests on Windows?

Originally I took the approach not building on Windows, but then the community 
reached to a point that instead of maintaining two lists of source files to be 
build on Windows and on non-Windows, we use the same list, and add test stubs 
for not-yet-supported-on-windows ones.

Re: [PATCH v1] gpudev: return EINVAL if invalid input pointer for free and unregister

[Tyler Retzlaff]

On Wed, Nov 24, 2021 at 06:04:56PM +, Bruce Richardson wrote:
> On Wed, Nov 24, 2021 at 09:24:42AM -0800, Tyler Retzlaff wrote:
> > On Fri, Nov 19, 2021 at 10:56:36AM +0100, Thomas Monjalon wrote:
> > > 19/11/2021 10:34, Ferruh Yigit:
> > > > >> +if (ptr == NULL) {
> > > > >> +rte_errno = EINVAL;
> > > > >> +return -rte_errno;
> > > > >> +}
> > > > > 
> > > > > in general dpdk has real problems with how it indicates that an error
> > > > > occurred and what error occurred consistently.
> > > > > 
> > > > > some api's return 0 on success
> > > > >and maybe return -errno if ! 0
> > > > >and maybe return errno if ! 0
> > > 
> > > Which function returns a positive errno?
> > 
> > i may have mispoke about this variant, it may be something i recall
> > seeing in a posted patch that was resolved before integration.
> > 
> > > 
> > > > >and maybe set rte_errno if ! 0
> > > > > 
> > > > > some api's return -1 on failure
> > > > >and set rte_errno if -1
> > > > > 
> > > > > some api's return < 0 on failure
> > > > >and maybe set rte_errno
> > > > >and maybe return -errno
> > > > >and maybe set rte_errno and return -rte_errno
> > > > 
> > > > This is a generic comment, cc'ed a few more folks to make the comment 
> > > > more
> > > > visible.
> > > > 
> > > > > this isn't isiolated to only this change but since additions and 
> > > > > context
> > > > > in this patch highlight it maybe it's a good time to bring it up.
> > > > > 
> > > > > it's frustrating to have to carefully read the implementation every 
> > > > > time
> > > > > you want to make a function call to make sure you're handling the 
> > > > > flavor
> > > > > of error reporting for a particular function.
> > > > > 
> > > > > if this is new code could we please clearly identify the current best
> > > > > practice and follow it as a standard going forward for all new public
> > > > > apis.
> > > 
> > > I think this patch is following the best practice.
> > > 1/ Return negative value in case of error
> > > 2/ Set rte_errno
> > > 3/ Set same absolute value in rte_errno and return code
> > 
> > with the approach proposed as best practice above it results in at least 
> > the 
> > applicaiton code variations as follows.
> > 
> > int rv = rte_func_call();
> > 
> > 1. if (rv < 0 && rte_errno == EAGAIN)
> > 
> > 2. if (rv == -1 && rte_errno == EAGAIN)
> > 
> > 3. if (rv < 0 && -rv == EAGAIN)
> > 
> > 4. if (rv < 0 && rv == -EAGAIN)
> > 
> > (and incorrectly)
> > 
> > 5. // ignore rv
> >   if (rte_errno == EAGAIN)
> > 
> > it might be better practice if indication that an error occurs is
> > signaled distinctly from the error that occurred. otherwise why use
> > rte_errno at all instead returning -rte_errno always?
> > 
> > this philosophy would align better with modern posix / unix platform
> > apis. often documented in the RETURN VALUE section of the manpage as:
> > 
> > ``Upon successful completion, somefunction() shall return 0;
> >   otherwise, -1 shall be returned and errno set to indicate the
> >   error.''
> > 
> > therefore returning a value outside of the set {0, -1} is an abi break.
>  
> I like using this standard, because it also allows consistent behaviour for
> non-integer returning functions, e.g. object creation functions returning
> pointers.
> 
>   if (ret < 0 && rte_errno == EAGAIN)

i only urge that this be explicit as opposed to a range i.e. ret == -1
preferred over ret < 0

> 
> becomes for a pointer:
> 
>   if (ret == NULL && rte_errno == EAGAIN)
> 
> Regards,
> /Bruce

but otherwise i agree, ret indicates an error happened and rte_errno
provides the detail.

Re:Re: [PATCH] ip_frag: add IPv4 options fragment and unit test data

[Huichao Cai]

Hi Dariusz


Substituting options with NOOP might cause rte_ipv4_fragment_packet to produce 
more fragments than necessary, since options with copied flag unset will still 
occupy space in IPv4 header.


--The "ip_options_fragment" just make a replacement and doesn't change the 
length of the IPv4 header.So I don't quite understand why it leads to produce 
more fragments.


but maybe a better solution would be to prepare a separate IPv4 header for 
fragments without unnecessary options.
--Yes, we can do this, but it adds some extra work, such as generating a new 
IPv4 header and reassembling the data,which has some performance implications.

Huichao Cai

[PATCH] net/ixgbe: add vector Rx parameter check

[Bin Zheng]

Under the circumstance that `rx_tail` wrap back to zero
and the advance speed of `rx_tail` is greater than `rxrearm_start`,
`rx_tail` will catch up with `rxrearm_start` and surpass it.
This may cause some mbufs be reused by applicaion.

So we need to make some restrictions to ensure that
 `rx_tail` will not exceed `rxrearm_start`.

e.g.

RDH: 972 RDT: 991 rxrearm_nb: 991 rxrearm_start: 992 rx_tail: 959
RDH: 1004 RDT: 1023 rxrearm_nb: 991 rxrearm_start: 0 rx_tail: 991
RDH: 12 RDT: 31 rxrearm_nb: 991 rxrearm_start: 32 rx_tail: 1023
RDH: 31 RDT: 63 rxrearm_nb: 960 rxrearm_start: 64 rx_tail: 0
RDH: 95 RDT: 95 rxrearm_nb: 1016 rxrearm_start: 96 rx_tail: 88
RDH: 95 RDT: 127 rxrearm_nb: 991 rxrearm_start: 128 rx_tail: 95
...
RDH: 908 RDT: 927 rxrearm_nb: 991 rxrearm_start: 928 rx_tail: 895
RDH: 940 RDT: 959 rxrearm_nb: 991 rxrearm_start: 960 rx_tail: 927
RDH: 980 RDT: 991 rxrearm_nb: 991 rxrearm_start: 992 rx_tail: 959
RDH: 991 RDT: 991 rxrearm_nb: 1026 rxrearm_start: 992 rx_tail: 994

when `rx_tail` catches up with `rxrearm_start`,
2(994 - 992) mbufs be reused by applicaion !

Bugzilla ID: 882
Fixes: 5a3cca342417 ("net/ixgbe: fix vector Rx")
Cc: jia@intel.com
Cc: sta...@dpdk.org

Signed-off-by: Bin Zheng 
---
 drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c | 18 +++---
 1 file changed, 15 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c 
b/drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c
index 1eed949495..33d2e96266 100644
--- a/drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c
+++ b/drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c
@@ -364,9 +364,6 @@ _recv_raw_pkts_vec(struct ixgbe_rx_queue *rxq, struct 
rte_mbuf **rx_pkts,
uint8_t vlan_flags;
uint16_t udp_p_flag = 0; /* Rx Descriptor UDP header present */
 
-   /* nb_pkts has to be floor-aligned to RTE_IXGBE_DESCS_PER_LOOP */
-   nb_pkts = RTE_ALIGN_FLOOR(nb_pkts, RTE_IXGBE_DESCS_PER_LOOP);
-
/* Just the act of getting into the function from the application is
 * going to cost about 7 cycles
 */
@@ -380,6 +377,21 @@ _recv_raw_pkts_vec(struct ixgbe_rx_queue *rxq, struct 
rte_mbuf **rx_pkts,
if (rxq->rxrearm_nb > RTE_IXGBE_RXQ_REARM_THRESH)
ixgbe_rxq_rearm(rxq);
 
+   /*
+* Under the circumstance that `rx_tail` wrap back to zero
+* and the advance speed of `rx_tail` is greater than `rxrearm_start`,
+* `rx_tail` will catch up with `rxrearm_start` and surpass it.
+* This may cause some mbufs be reused by applicaion.
+*
+* So we need to make some restrictions to ensure that
+* `rx_tail` will not exceed `rxrearm_start`.
+*/
+   if (rxq->rx_tail < rxq->rxrearm_start)
+   nb_pkts = RTE_MIN(nb_pkts, rxq->rxrearm_start - rxq->rx_tail - 
1);
+
+   /* nb_pkts has to be floor-aligned to RTE_IXGBE_DESCS_PER_LOOP */
+   nb_pkts = RTE_ALIGN_FLOOR(nb_pkts, RTE_IXGBE_DESCS_PER_LOOP);
+
/* Before we start moving massive data around, check to see if
 * there is actually a packet available
 */
-- 
2.25.1

[PATCH] net/ixgbe: add vector Rx parameter check

[Bin Zheng]

Under the circumstance that `rx_tail` wrap back to zero
and the advance speed of `rx_tail` is greater than `rxrearm_start`,
`rx_tail` will catch up with `rxrearm_start` and surpass it.
This may cause some mbufs be reused by applicaion.

So we need to make some restrictions to ensure that
 `rx_tail` will not exceed `rxrearm_start`.

e.g.

RDH: 972 RDT: 991 rxrearm_nb: 991 rxrearm_start: 992 rx_tail: 959
RDH: 1004 RDT: 1023 rxrearm_nb: 991 rxrearm_start: 0 rx_tail: 991
RDH: 12 RDT: 31 rxrearm_nb: 991 rxrearm_start: 32 rx_tail: 1023
RDH: 31 RDT: 63 rxrearm_nb: 960 rxrearm_start: 64 rx_tail: 0
RDH: 95 RDT: 95 rxrearm_nb: 1016 rxrearm_start: 96 rx_tail: 88
RDH: 95 RDT: 127 rxrearm_nb: 991 rxrearm_start: 128 rx_tail: 95
...
RDH: 908 RDT: 927 rxrearm_nb: 991 rxrearm_start: 928 rx_tail: 895
RDH: 940 RDT: 959 rxrearm_nb: 991 rxrearm_start: 960 rx_tail: 927
RDH: 980 RDT: 991 rxrearm_nb: 991 rxrearm_start: 992 rx_tail: 959
RDH: 991 RDT: 991 rxrearm_nb: 1026 rxrearm_start: 992 rx_tail: 994

when `rx_tail` catches up with `rxrearm_start`,
2(994 - 992) mbufs be reused by applicaion !

Bugzilla ID: 882
Fixes: 5a3cca342417 ("net/ixgbe: fix vector Rx")
Cc: jia@intel.com
Cc: sta...@dpdk.org

Signed-off-by: Bin Zheng 
---
 drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c | 11 +++
 1 file changed, 11 insertions(+)

diff --git a/drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c 
b/drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c
index 1eed949495..5811749b95 100644
--- a/drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c
+++ b/drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c
@@ -364,6 +364,17 @@ _recv_raw_pkts_vec(struct ixgbe_rx_queue *rxq, struct 
rte_mbuf **rx_pkts,
uint8_t vlan_flags;
uint16_t udp_p_flag = 0; /* Rx Descriptor UDP header present */
 
+   /*
+* Under the circumstance that `rx_tail` wrap back to zero
+* and the advance speed of `rx_tail` is greater than `rxrearm_start`,
+* `rx_tail` will catch up with `rxrearm_start` and surpass it.
+* This may cause some mbufs be reused by applicaion.
+*
+* So we need to make some restrictions to ensure that
+* `rx_tail` will not exceed `rxrearm_start`.
+*/
+   nb_pkts = RTE_MIN(nb_pkts, RTE_IXGBE_RXQ_REARM_THRESH);
+
/* nb_pkts has to be floor-aligned to RTE_IXGBE_DESCS_PER_LOOP */
nb_pkts = RTE_ALIGN_FLOOR(nb_pkts, RTE_IXGBE_DESCS_PER_LOOP);
 
-- 
2.25.1

RE: [PATCH] net/ixgbe: add vector Rx parameter check

[Wang, Haiyue]

> -Original Message-
> From: Bin Zheng 
> Sent: Thursday, December 2, 2021 11:19
> To: dev@dpdk.org
> Cc: Wang, Haiyue ; lian...@liangbit.com; Bin Zheng
> ; jia@intel.com; sta...@dpdk.org
> Subject: [PATCH] net/ixgbe: add vector Rx parameter check
> 
> Under the circumstance that `rx_tail` wrap back to zero
> and the advance speed of `rx_tail` is greater than `rxrearm_start`,
> `rx_tail` will catch up with `rxrearm_start` and surpass it.
> This may cause some mbufs be reused by applicaion.
> 
> So we need to make some restrictions to ensure that
>  `rx_tail` will not exceed `rxrearm_start`.
> 
> e.g.
> 
> RDH: 972 RDT: 991 rxrearm_nb: 991 rxrearm_start: 992 rx_tail: 959
> RDH: 1004 RDT: 1023 rxrearm_nb: 991 rxrearm_start: 0 rx_tail: 991
> RDH: 12 RDT: 31 rxrearm_nb: 991 rxrearm_start: 32 rx_tail: 1023
> RDH: 31 RDT: 63 rxrearm_nb: 960 rxrearm_start: 64 rx_tail: 0
> RDH: 95 RDT: 95 rxrearm_nb: 1016 rxrearm_start: 96 rx_tail: 88
> RDH: 95 RDT: 127 rxrearm_nb: 991 rxrearm_start: 128 rx_tail: 95
> ...
> RDH: 908 RDT: 927 rxrearm_nb: 991 rxrearm_start: 928 rx_tail: 895
> RDH: 940 RDT: 959 rxrearm_nb: 991 rxrearm_start: 960 rx_tail: 927
> RDH: 980 RDT: 991 rxrearm_nb: 991 rxrearm_start: 992 rx_tail: 959
> RDH: 991 RDT: 991 rxrearm_nb: 1026 rxrearm_start: 992 rx_tail: 994
> 
> when `rx_tail` catches up with `rxrearm_start`,
> 2(994 - 992) mbufs be reused by applicaion !
> 
> Bugzilla ID: 882
> Fixes: 5a3cca342417 ("net/ixgbe: fix vector Rx")
> Cc: jia@intel.com
> Cc: sta...@dpdk.org
> 
> Signed-off-by: Bin Zheng 
> ---
>  drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c | 11 +++
>  1 file changed, 11 insertions(+)
> 

Both ? Or ?

https://patchwork.dpdk.org/project/dpdk/patch/20211202031530.1808112-1-zhengbin.89...@bytedance.com/
https://patchwork.dpdk.org/project/dpdk/patch/20211202031857.1808705-1-zhengbin.89...@bytedance.com/

> diff --git a/drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c 
> b/drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c
> index 1eed949495..5811749b95 100644
> --- a/drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c
> +++ b/drivers/net/ixgbe/ixgbe_rxtx_vec_sse.c
> @@ -364,6 +364,17 @@ _recv_raw_pkts_vec(struct ixgbe_rx_queue *rxq, struct 
> rte_mbuf **rx_pkts,
>   uint8_t vlan_flags;
>   uint16_t udp_p_flag = 0; /* Rx Descriptor UDP header present */
> 
> + /*
> +  * Under the circumstance that `rx_tail` wrap back to zero
> +  * and the advance speed of `rx_tail` is greater than `rxrearm_start`,
> +  * `rx_tail` will catch up with `rxrearm_start` and surpass it.
> +  * This may cause some mbufs be reused by applicaion.
> +  *
> +  * So we need to make some restrictions to ensure that
> +  * `rx_tail` will not exceed `rxrearm_start`.
> +  */
> + nb_pkts = RTE_MIN(nb_pkts, RTE_IXGBE_RXQ_REARM_THRESH);
> +
>   /* nb_pkts has to be floor-aligned to RTE_IXGBE_DESCS_PER_LOOP */
>   nb_pkts = RTE_ALIGN_FLOOR(nb_pkts, RTE_IXGBE_DESCS_PER_LOOP);
> 
> --
> 2.25.1

RE: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

[Namburu, Chandu-babu]

[Public]

Hi Arsalan,

Thank you for your comments. 

Regards,
Chandu
-Original Message-
From: Awan, Arsalan  
Sent: Wednesday, December 1, 2021 2:19 PM
To: Namburu, Chandu-babu ; dev@dpdk.org
Cc: Somalapuram, Amaranath ; Sebastian, Selwin 
; Giriyapura, Maheshwaramurthy 

Subject: Re: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

Hi Chandu,

I have provided my comments on your patch acknowledging your solution as a 
better approach than this one.

Thanks and regards,
Arsalan

From: Namburu, Chandu-babu 
Sent: Friday, November 26, 2021 3:35 PM
To: Awan, Arsalan; dev@dpdk.org
Cc: Somalapuram, Amaranath; Sebastian, Selwin; Giriyapura, Maheshwaramurthy
Subject: RE: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

[Public]

Hi Arsalan,

I have submitted new patch to community review and looped you in. Please share 
your comments.

https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fpatches.dpdk.org%2Fproject%2Fdpdk%2Flist%2F%3Fseries%3D20788&data=04%7C01%7Cchandu%40amd.com%7C6355ab7462c44e390bcc08d9b4a75e16%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637739453199074315%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=I9m0teM9abhf21%2BHn90dJCLNKS28D%2BY%2BCya%2FADdS5nI%3D&reserved=0

Regards,
Chandu

-Original Message-
From: Awan, Arsalan 
Sent: Thursday, November 25, 2021 2:37 PM
To: Namburu, Chandu-babu ; dev@dpdk.org
Cc: Somalapuram, Amaranath ; Sebastian, Selwin 
; Giriyapura, Maheshwaramurthy 

Subject: Re: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

Thanks, Chandu!

This will help a lot!

Regards,
Arsalan

From: Namburu, Chandu-babu 
Sent: Thursday, November 25, 2021 1:23 PM
To: Awan, Arsalan; dev@dpdk.org
Cc: Somalapuram, Amaranath; Sebastian, Selwin; Giriyapura, Maheshwaramurthy
Subject: RE: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

[Public]

Hi Arsalan,

We have identified fix for this issue and will submit the patch for community 
review asap.

Thanks,
Chandu

-Original Message-
From: Awan, Arsalan 
Sent: Thursday, November 25, 2021 1:42 PM
To: Namburu, Chandu-babu ; dev@dpdk.org
Cc: Somalapuram, Amaranath ; Sebastian, Selwin 
; Giriyapura, Maheshwaramurthy 

Subject: Re: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

Hi Chandu,

Do we have an update on this?

Can we go ahead with this patch for now as it does not break anything else, but 
fixes the problem.

DPDK will be working fine as is on the e3000.

This patch only adjusts the code for the v1000 and it works fine and there 
isn't a difference in the approach of fixing the problem compared to that being 
done currently to distinguish between the two platforms i.e. e3000 & v1000.

Hoping to hear from you soon.

Thanks,
Arsalan

From: Awan, Arsalan
Sent: Monday, November 15, 2021 3:15 PM
To: Namburu, Chandu-babu; dev@dpdk.org
Cc: Somalapuram, Amaranath; Sebastian, Selwin; Giriyapura, Maheshwaramurthy
Subject: Re: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

Thanks, Chandu!

Appreciate you guys looking into this!

Looking forward to hear from you soon.

Thanks and regards,
Arsalan

From: Namburu, Chandu-babu 
Sent: Friday, November 12, 2021 7:48 PM
To: Awan, Arsalan; dev@dpdk.org
Cc: Somalapuram, Amaranath; Sebastian, Selwin; Giriyapura, Maheshwaramurthy
Subject: RE: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

[AMD Official Use Only]

Hi Arsalan Awan,

Thank you for sharing the patch. DPDK not working on V1000 is regression due to 
"bus/pci: optimize bus scan" patch. As RV Root Complex device does not have any 
Linux kernel driver assigned, this device is removed from PCI scan list and not 
found using pcie_search_device() routine.

Patch submitted is potential fix for the issue, but we are checking if this is 
the best way to fix regression. Will update you soon on the fix shared.

Regards,
Chandu

-Original Message-
From: Arsalan H. Awan 
Sent: Wednesday, November 10, 2021 5:15 PM
To: dev@dpdk.org
Cc: Namburu, Chandu-babu ; Somalapuram, Amaranath 
; Sebastian, Selwin ; 
Giriyapura, Maheshwaramurthy ; Arsalan H. 
Awan 
Subject: [PATCH] net/axgbe: fix PCI ID to distinguish v1000 vs e3000

The RV PCI Root Complex Device is not visible to DPDK on the PCI bus via 
pci_search_device on some OSs including Ubuntu and Yocto. This makes it 
impossible to determine which machine DPDK/axgbe is running on. As a result, 
DPDK/axgbe does not work on v1000.

Let's use the Raven Internal PCIe GPP Bridge ID that DPDK can see on the PCI 
bus that exists on the v1000 to differentiate between RV AXGBE and SNOWY AXGBE, 
and set the registers accordingly. This fixes DPDK not working on v1000 
platform.

Signed-off-by: Arsalan H. Awan 
---
 drivers/net/axgbe/axgbe_ethdev.c | 6 +++---
 1 file changed, 3 ins

RE: [PATCH v1] net/axgbe: use PCI root complex device to distinguish AMD hardware

[Namburu, Chandu-babu]

[Public]

Hi David Marchand,

Thank you for your comments. Will change patch and submit v2 for review

Regards,
Chandu

-Original Message-
From: David Marchand  
Sent: Wednesday, December 1, 2021 2:31 PM
To: Namburu, Chandu-babu 
Cc: dev ; Sebastian, Selwin ; 
arsalan_a...@mentor.com; Yigit, Ferruh 
Subject: Re: [PATCH v1] net/axgbe: use PCI root complex device to distinguish 
AMD hardware

On Fri, Nov 26, 2021 at 11:24 AM Chandubabu Namburu  wrote:
>
> "bus/pci: optimize bus scan" broke axgbe on V1000/R1000.
> RV root complex pci device does not have any kernel driver assigned so 
> it is removed from pci scan list which is used in
> "net/axgbe: add a HW quirk for register definitions"
>
> Get root complex device id directly from pci sysfs instead of pci scan 
> list

Please end the sentences with a '.'.

As you described, this change is a fix.
Please add a Fixes: tag and Cc: sta...@dpdk.org in the commitlog if you think 
it should be backported to LTS releases.
https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdoc.dpdk.org%2Fguides%2Fcontributing%2Fpatches.html%23patch-for-stable-releases&data=04%7C01%7Cchandu%40amd.com%7C6618c303ecb146be0b5908d9b4a91c01%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637739460691977618%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=piC1U%2BPcPP2A0xwM3%2FA5qDUFXNLAj2SrLBywTFc3DFM%3D&reserved=0


>
> Signed-off-by: Chandubabu Namburu 
> ---
>  drivers/net/axgbe/axgbe_ethdev.c | 39 
> 
>  1 file changed, 20 insertions(+), 19 deletions(-)
>
> diff --git a/drivers/net/axgbe/axgbe_ethdev.c 
> b/drivers/net/axgbe/axgbe_ethdev.c
> index 9cd056d04a..26babde354 100644
> --- a/drivers/net/axgbe/axgbe_ethdev.c
> +++ b/drivers/net/axgbe/axgbe_ethdev.c
> @@ -10,6 +10,8 @@
>  #include "axgbe_regs.h"
>  #include "rte_time.h"
>
> +#include "eal_filesystem.h"
> +
>  static int eth_axgbe_dev_init(struct rte_eth_dev *eth_dev);  static 
> int  axgbe_dev_configure(struct rte_eth_dev *dev);  static int  
> axgbe_dev_start(struct rte_eth_dev *dev); @@ -1923,28 +1925,27 @@ 
> static void axgbe_default_config(struct axgbe_port *pdata)
> pdata->power_down = 0;
>  }
>
> -static int
> -pci_device_cmp(const struct rte_device *dev, const void *_pci_id)
> +/*
> + * Return PCI root complex device id on success else 0  */ static 
> +uint16_t
> +get_pci_rc_devid(void)
>  {
> -   const struct rte_pci_device *pdev = RTE_DEV_TO_PCI_CONST(dev);
> -   const struct rte_pci_id *pcid = _pci_id;
> +   char pci_sysfs[PATH_MAX];
> +   const struct rte_pci_addr pci_rc_addr = {0, 0, 0, 0};
> +   unsigned long device_id;
>
> -   if (pdev->id.vendor_id == AMD_PCI_VENDOR_ID &&
> -   pdev->id.device_id == pcid->device_id)
> -   return 0;
> -   return 1;
> -}
> +   snprintf(pci_sysfs, sizeof(pci_sysfs), "%s/" PCI_PRI_FMT "/device",
> +rte_pci_get_sysfs_path(), pci_rc_addr.domain,
> +pci_rc_addr.bus, pci_rc_addr.devid, 
> + pci_rc_addr.function);

The use of rte_pci_get_sysfs_path() and in general is ugly because it only 
works for Linux.
We could come up with a better API in the pci bus, but I don't think it is 
worth it atm.

We can live with your implementation since this driver is only compiled on 
Linux.


--
David Marchand

Re: [PATCH v1] gpudev: return EINVAL if invalid input pointer for free and unregister

[Thomas Monjalon]

01/12/2021 22:37, Tyler Retzlaff:
> On Wed, Nov 24, 2021 at 06:04:56PM +, Bruce Richardson wrote:
> >   if (ret < 0 && rte_errno == EAGAIN)
> 
> i only urge that this be explicit as opposed to a range i.e. ret == -1
> preferred over ret < 0

I don't understand why you think it is important to limit return value to -1.
Why "if (ret == -1)" is better than "if (ret < 0)" ?