[dpdk-dev] [PATCH v3 4/4] eal: fix end character check in --lcores argument

Dai, Wei wei.dai at intel.com
Tue Aug 2 10:22:29 CEST 2016


Hi, Thomas, Yigit

> -----Original Message-----
> From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com]
> Sent: Friday, July 29, 2016 12:06 AM
> To: Dai, Wei <wei.dai at intel.com>
> Cc: dev at dpdk.org; Yigit, Ferruh <ferruh.yigit at intel.com>
> Subject: Re: [dpdk-dev] [PATCH v3 4/4] eal: fix end character check in --lcores
> argument
> 
> 2016-07-28 16:26, Ferruh Yigit:
> > On 7/27/2016 12:27 PM, Wei Dai wrote:
> > > With --lcores 'a-b at c-d', eal_parse_cores() fails because
> > > eal_parse_set() fails due to the next character after lcore set a-b,
> > > which is '@'and not ',' or '\0'.
> > > There is also a right check immediately after this incorrect check.
> > >
> > > Fixes: 53e54bf81700 ("eal: new option --lcores for cpu assignment")
> > >
> > > Signed-off-by: Wei Dai <wei.dai at intel.com>
> >
> > I am not sure if a-b at c-d syntax should be supported. (a-b)@(c-d) is
> > supported and already working.
> 
> Agreed.
> 
> Series applied, except this last patch 4, thanks

Sorry, I can't find any document to clarify whether a-b at c-d shoule be supported.
So I design following case to verify all kinds of combination of - and @. 
I tested my patch with --lcores '0-3 at 12-15, 4-7@(8-11),  (8-11)@4-7, (12-15)@(0-3), 16-19, (20-23) ',
After calling eal_parse_args(argc, argv),  I observed lcore_config[i].cpuset to check cpuset of each lcore.
I added following codes immediately after calling fctret = eal_parse_args(argc, argv); in function rte_eal_init( )
in eal.c to check this patch.

	{
		struct rte_config *p;
		const char *rte_role_str[2] = {"ROLE_RTE", "ROLE_OFF"};
		p = rte_eal_get_configuration();
		for (i=0; i<28; i++) {
			printf("lcore %2d status = %s,  runs on cpuset = 0x%08lX\n", 
					i, rte_role_str[p->lcore_role[i]], lcore_config[i].cpuset.__bits[0]);
		}
	}

The result is as follow:
lcore  0 status = ROLE_RTE,  runs on cpuset = 0x0000F000
lcore  1 status = ROLE_RTE,  runs on cpuset = 0x0000F000
lcore  2 status = ROLE_RTE,  runs on cpuset = 0x0000F000
lcore  3 status = ROLE_RTE,  runs on cpuset = 0x0000F000
lcore  4 status = ROLE_RTE,  runs on cpuset = 0x00000F00
lcore  5 status = ROLE_RTE,  runs on cpuset = 0x00000F00
lcore  6 status = ROLE_RTE,  runs on cpuset = 0x00000F00
lcore  7 status = ROLE_RTE,  runs on cpuset = 0x00000F00
lcore  8 status = ROLE_RTE,  runs on cpuset = 0x000000F0
lcore  9 status = ROLE_RTE,  runs on cpuset = 0x000000F0
lcore 10 status = ROLE_RTE,  runs on cpuset = 0x000000F0
lcore 11 status = ROLE_RTE,  runs on cpuset = 0x000000F0
lcore 12 status = ROLE_RTE,  runs on cpuset = 0x0000000F
lcore 13 status = ROLE_RTE,  runs on cpuset = 0x0000000F
lcore 14 status = ROLE_RTE,  runs on cpuset = 0x0000000F
lcore 15 status = ROLE_RTE,  runs on cpuset = 0x0000000F
lcore 16 status = ROLE_RTE,  runs on cpuset = 0x00010000
lcore 17 status = ROLE_RTE,  runs on cpuset = 0x00020000
lcore 18 status = ROLE_RTE,  runs on cpuset = 0x00040000
lcore 19 status = ROLE_RTE,  runs on cpuset = 0x00080000
lcore 20 status = ROLE_RTE,  runs on cpuset = 0x00F00000
lcore 21 status = ROLE_RTE,  runs on cpuset = 0x00F00000
lcore 22 status = ROLE_RTE,  runs on cpuset = 0x00F00000
lcore 23 status = ROLE_RTE,  runs on cpuset = 0x00F00000
lcore 24 status = ROLE_OFF,  runs on cpuset = 0x00000000
lcore 25 status = ROLE_OFF,  runs on cpuset = 0x00000000
lcore 26 status = ROLE_OFF,  runs on cpuset = 0x00000000
lcore 27 status = ROLE_OFF,  runs on cpuset = 0x00000000

	


More information about the dev mailing list