On Thursday 10 November 2016 08:36 AM, Jianbo Liu wrote:
> On 28 October 2016 at 20:26, Shreyansh Jain <shreyansh.jain at nxp.com> wrote:
>> From: Jan Viktorin <viktorin at rehivetech.com>
>>
>> SoC devices would be linked in a separate list (from PCI). This is used for
>> probe function.
>> A helper for dumping the device list is added.
>>
>> Signed-off-by: Jan Viktorin <viktorin at rehivetech.com>
>> Signed-off-by: Shreyansh Jain <shreyansh.jain at nxp.com>
>> Signed-off-by: Hemant Agrawal <hemant.agrawal at nxp.com>
>> ---
>> lib/librte_eal/bsdapp/eal/rte_eal_version.map | 2 ++
>> lib/librte_eal/common/eal_common_soc.c | 34
>> +++++++++++++++++++++++++
>> lib/librte_eal/common/include/rte_soc.h | 9 +++++++
>> lib/librte_eal/linuxapp/eal/rte_eal_version.map | 2 ++
>> 4 files changed, 47 insertions(+)
>>
>> diff --git a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
>> b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
>> index cf6fb8e..86e3cfd 100644
>> --- a/lib/librte_eal/bsdapp/eal/rte_eal_version.map
>> +++ b/lib/librte_eal/bsdapp/eal/rte_eal_version.map
>> @@ -171,11 +171,13 @@ DPDK_16.11 {
>> rte_eal_dev_attach;
>> rte_eal_dev_detach;
>> rte_eal_map_resource;
>> + rte_eal_soc_dump;
>> rte_eal_soc_register;
>> rte_eal_soc_unregister;
>> rte_eal_unmap_resource;
>> rte_eal_vdrv_register;
>> rte_eal_vdrv_unregister;
>> + soc_device_list;
>> soc_driver_list;
>>
>> } DPDK_16.07;
>> diff --git a/lib/librte_eal/common/eal_common_soc.c
>> b/lib/librte_eal/common/eal_common_soc.c
>> index 56135ed..5dcddc5 100644
>> --- a/lib/librte_eal/common/eal_common_soc.c
>> +++ b/lib/librte_eal/common/eal_common_soc.c
>> @@ -31,6 +31,8 @@
>> * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
>> */
>>
>> +#include <stddef.h>
>> +#include <stdio.h>
>> #include <sys/queue.h>
>>
>> #include <rte_log.h>
>> @@ -40,6 +42,38 @@
>> /* Global SoC driver list */
>> struct soc_driver_list soc_driver_list =
>> TAILQ_HEAD_INITIALIZER(soc_driver_list);
>> +struct soc_device_list soc_device_list =
>> + TAILQ_HEAD_INITIALIZER(soc_device_list);
>> +
>> +/* dump one device */
>> +static int
>> +soc_dump_one_device(FILE *f, struct rte_soc_device *dev)
>> +{
>> + int i;
>> +
>> + fprintf(f, "%s", dev->addr.name);
>> + fprintf(f, " - fdt_path: %s\n",
>> + dev->addr.fdt_path ? dev->addr.fdt_path : "(none)");
>> +
>> + for (i = 0; dev->id && dev->id[i].compatible; ++i)
>> + fprintf(f, " %s\n", dev->id[i].compatible);
>> +
>> + return 0;
>> +}
>> +
>> +/* dump devices on the bus to an output stream */
>> +void
>> +rte_eal_soc_dump(FILE *f)
>> +{
>> + struct rte_soc_device *dev = NULL;
>> +
>> + if (!f)
>> + return;
>> +
>> + TAILQ_FOREACH(dev, &soc_device_list, next) {
>> + soc_dump_one_device(f, dev);
>> + }
>> +}
>>
>> /* register a driver */
>> void
>> diff --git a/lib/librte_eal/common/include/rte_soc.h
>> b/lib/librte_eal/common/include/rte_soc.h
>> index 23b06a9..347e611 100644
>> --- a/lib/librte_eal/common/include/rte_soc.h
>> +++ b/lib/librte_eal/common/include/rte_soc.h
>> @@ -56,8 +56,12 @@ extern "C" {
>>
>> extern struct soc_driver_list soc_driver_list;
>> /**< Global list of SoC Drivers */
>> +extern struct soc_device_list soc_device_list;
>> +/**< Global list of SoC Devices */
>>
>> TAILQ_HEAD(soc_driver_list, rte_soc_driver); /**< SoC drivers in D-linked
>> Q. */
>> +TAILQ_HEAD(soc_device_list, rte_soc_device); /**< SoC devices in D-linked
>> Q. */
>> +
>>
>> struct rte_soc_id {
>> const char *compatible; /**< OF compatible specification */
>> @@ -142,6 +146,11 @@ rte_eal_compare_soc_addr(const struct rte_soc_addr *a0,
>> }
>>
>> /**
>> + * Dump discovered SoC devices.
>> + */
>> +void rte_eal_soc_dump(FILE *f);
>
> If it is to dump device information (not driver), is it proper to
> rename it rte_eal_soc_device_dump()?
Well, 'SoC'=='device' in this context. Isn't it?
In which case, 'soc_device' is just redundant/reuse of a descriptive word.
Or, on a second thought, 'SoC' represents a bus type here. In which
case, this name sounds better. This would also warrant
rte_eal_soc_register->rte_eal_soc_device_register.
But, if I need to keep it in sync with PCI (rte_eal_pci_register), I
think original naming is apt.
I prefer continuing with existing (rte_eal_soc_dump) for this reason.
>
>> +
>> +/**
>> * Register a SoC driver.
>> */
>> void rte_eal_soc_register(struct rte_soc_driver *driver);
>> diff --git a/lib/librte_eal/linuxapp/eal/rte_eal_version.map
>> b/lib/librte_eal/linuxapp/eal/rte_eal_version.map
>> index ab6b985..0155025 100644
>> --- a/lib/librte_eal/linuxapp/eal/rte_eal_version.map
>> +++ b/lib/librte_eal/linuxapp/eal/rte_eal_version.map
>> @@ -175,11 +175,13 @@ DPDK_16.11 {
>> rte_eal_dev_attach;
>> rte_eal_dev_detach;
>> rte_eal_map_resource;
>> + rte_eal_soc_dump;
>> rte_eal_soc_register;
>> rte_eal_soc_unregister;
>> rte_eal_unmap_resource;
>> rte_eal_vdrv_register;
>> rte_eal_vdrv_unregister;
>> + soc_device_list;
>> soc_driver_list;
>>
>> } DPDK_16.07;
>> --
>> 2.7.4
>>
>
-
Shreyansh
