bus/pci: fix numa node return value
PATCH 1/1 bus/pci: fix numa node return value
On older CPUs, currently numa_node returns value only for socket 0. Instead, application should be able to make correct decision and also to keep consistent with the Linux code,
replace the return value to -1.
--- drivers/bus/pci/windows/pci.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/bus/pci/windows/pci.c +2 -1
@@ -5,6 +5,7 @@#include <rte_errno.h> #include <rte_log.h> #include <rte_eal.h> +#include <rte_memory.h> #include "private.h" #include "pci_netuio.h"
@@ -243,7 +244,7 @@ get_device_resource_info(HDEVINFO dev_info,DWORD error =GetLastError ();if (error ==ERROR_NOT_FOUND ) {/* On older CPUs, NUMA is not bound to PCIe locality. */ - dev->device.numa_node =0 ; + dev->device.numa_node =SOCKET_ID_ANY ;return ERROR_SUCCESS ; } RTE_LOG_WIN32_ERR("SetupDiGetDevicePropertyW"
Signature · 1 line
2.31.1.windows.1
Quoted 41 lines
Subject: [PATCH] bus/pci: fix numa node return value
External email: Use caution opening links or attachments
On older CPUs, currently
numa_nodereturns value only for socket 0. Instead, application should be able to make correct decision and also to keep consistent with the Linux code, replace the return value to -1.Fixes ac7c98d04f2c ("bus/pci: ignore missing NUMA node on Windows")Reported-by Vipin Varghese <[email protected]>Signed-off-by Pallavi Kadam <[email protected]>--- drivers/bus/pci/windows/pci.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)drivers/bus/pci/windows/pci.c +2
-2@@ -5,6 +5,7 @@#include <rte_errno.h> #include <rte_log.h> #include <rte_eal.h> +#include <rte_memory.h> #include "private.h" #include "pci_netuio.h" @@ -243,7 +244,7 @@ get_device_resource_info(HDEVINFO dev_info,DWORD error =GetLastError ();if (error ==ERROR_NOT_FOUND ) {/* On older CPUs, NUMA is not bound to PCIe locality. */ - dev->device.numa_node =0 ; + dev->device.numa_node =SOCKET_ID_ANY ;return ERROR_SUCCESS ; } RTE_LOG_WIN32_ERR("SetupDiGetDevicePropertyW" --
Quoted 1 line
2.31.1.windows.1
You should add "on Windows" to the end of the patch name to clarity it's a Windows issue.
Aside from that LGTM.
Thomas Monjalon
Applied23/09/2021 17:44, Tal Shnaiderman:
Quoted 6 lines
Quoted 6 lines
Subject: [PATCH] bus/pci: fix numa node return value
On older CPUs, currently
numa_nodereturns value only for socket 0. Instead, application should be able to make correct decision and also to keep consistent with the Linux code, replace the return value to -1.
Quoted 8 lines
Quoted 6 lines
Fixes ac7c98d04f2c ("bus/pci: ignore missing NUMA node on Windows")Reported-by Vipin Varghese <[email protected]>Signed-off-by Pallavi Kadam <[email protected]>Acked-by Tal Shnaiderman <[email protected]>
Applied, thanks
Hi Pallavi,
Assuming the intention is set SOCKET_ID_ANY for faulty cases, couple of observation. In Linux
```
if (access(filename, F_OK) != -1) {
if (eal_parse_sysfs_value(filename, &tmp) == 0)
dev->device.numa_node = tmp;
else
dev->device.numa_node = -1;
} else {
dev->device.numa_node = 0;
}
```[snipped]
drivers/bus/pci/windows/pci.c get_device_resource_info(HDEVINFO dev_info, · line 244
Quoted 2 lines
drivers/bus/pci/windows/pci.c +0
-0@@ -243,7 +244,7 @@ get_device_resource_info(HDEVINFO dev_info,DWORD error =GetLastError ();
Set the default value as dev->device.numa_node = SOCKET_ID_ANY;
drivers/bus/pci/windows/pci.c get_device_resource_info(HDEVINFO dev_info, · line 247
Quoted 5 lines
(quoted diff) +0
-0if (error == ERROR_NOT_FOUND) { /* On older CPUs, NUMA is not bound to PCIelocality. */
drivers/bus/pci/windows/pci.c +1
-1- dev->device.numa_node =0 ; + dev->device.numa_node =SOCKET_ID_ANY ;
If the API is unsuccessful for ERROR_NOT_FOUND then you can set it to 0 (as recommened in documentation)
drivers/bus/pci/windows/pci.c get_device_resource_info(HDEVINFO dev_info, · line 250
Quoted 3 lines
drivers/bus/pci/windows/pci.c +0
-0return ERROR_SUCCESS ; } RTE_LOG_WIN32_ERR("SetupDiGetDevicePropertyW"
Else if if it other errors it will SOCKET_ID_ANY.
Quoted 2 lines
--
2.31.1.windows.1