2011/10/25, Giuseppe CAVALLARO <peppe.cavallaro@xxxxxx>: > On 10/25/2011 9:09 AM, Giuseppe CAVALLARO wrote: >> On 10/25/2011 4:12 AM, Kelvin Cheung wrote: >>> 2011/10/24, Giuseppe CAVALLARO <peppe.cavallaro@xxxxxx>: >>>> On 10/24/2011 4:05 PM, Kelvin Cheung wrote: >>>>> 2011/10/24, Giuseppe CAVALLARO <peppe.cavallaro@xxxxxx>: >>>>>> Hello Kelvin. >>>>>> >>>>>> On 10/24/2011 12:36 PM, Kelvin Cheung wrote: >>>>>> >>>>>> [snip] >>>>>> >>>>>>> According to datasheet of Loongson 1B, the buffer size in RX/TX >>>>>>> descriptor is only 2KB. So the Loongson1B's GMAC could not handle >>>>>>> jumbo frames. And the second buffer is useless in this case. Am I >>>>>>> right? Is there a better way than ifdef CONFIG_MACH_LOONGSON1 to >>>>>>> avoid duplicate code? >>>>>> >>>>>> Sorry for my misunderstanding. >>>>>> >>>>>> I think you have to use the normal descriptor and remove the enh_desc >>>>>> from the platform w/o modifying the driver at all. >>>>>> >>>>>> The driver will be able to select/configure all automatically (also >>>>>> jumbo). >>>>>> >>>>>> Let me know. >>>>> >>>>> That's the problem. >>>>> The bitfield definition of Loongson1B is also different from normal >>>>> descriptor. >>>> >>>> The problem is not in the Loongson1B gmac. >>> >>> I found that the bit checksum_insertion is not existed in normal >>> descriptor. >>> >>>> The normal descriptor fields in the stmmac refer to an old synopsys >>>> databook. >>> >>> Could you send me the new databook of Synopsys GMAC? >>> >>>> New chips have the same structure you have added; so we should fix this >>>> in the driver w/o breaking the compatibility for old chips. >>> >>> Agree. >>> >>>> I kindly ask you to confirm if the currently normal descriptor structure >>>> (w/o your changes) doesn't work on your platform. >>>> Did you test it? >>> >>> Well, the normal descriptor works on my platform except TX checksum >>> offload. >> >> ok! I suspected that. >> >> >>>>> Moreover, I want to enable the TX checksum offload function which is >>>>> not supported in normal descriptor. >>>>> Any suggestions? >>>> >>>> It is supported but you have to pass from the platform: tx_coe = 1. >>> >>> I noticed that the flag csum_insertion is passed to >>> ndesc_prepare_tx_desc() in stmmac_xmit(). But ndesc_prepare_tx_desc() >>> just ignores it. >>> In other words, the TX checksum offload function is disabled in normal >>> descriptor currently. >>> >>> Should we fix this problem for normal descriptor? >> >> Yes, we should. If you agree, I'll update the normal descriptor >> structure to yours. This is the normal descriptor used in newer GMAC. >> Tx csum will be done for normal descriptors in case of these GMAC >> devices and not for old MAC10/100. For the MAC10/100 some bits for >> normal descriptors are reserved and won't be used at all. >> >> I'll also verify that the patch doesn't break the back-compatibility >> with old MAC10/100. I have the HW where doing the tests. >> >> After that, I'll prepare the patch for net-next and for your kernel. > > Hello Kelvin > > attached the patch tested on my development kernel. > It runs fine on old and new mac devices. > > Can you try it on your side? Hmm, it is likely it won't apply fine on > your tree but you know the changes ;-). Sure. > If ok, I'll rework it for net-next and send it to the mailing list. > > Thanks > Peppe > >> >>> >>>> Peppe >>>>> >>>>>> Note: >>>>>> IIRC, there is a bit difference in case of normal descriptors for >>>>>> Synopsys databook newer than the 1.91 (I used for testing this mode). >>>>>> In any case, I remember that, on some platforms, the normal >>>>>> descriptors >>>>>> have been used w/o problems also on these new chip generations. >>>>>> >>>>>> Peppe >>>>>> >>>>>> >>>>> >>>>> >>>> >>>> >>> >>> >> >> -- >> To unsubscribe from this list: send the line "unsubscribe netdev" in >> the body of a message to majordomo@xxxxxxxxxxxxxxx >> More majordomo info at http://vger.kernel.org/majordomo-info.html >> > > -- Best Regards! Kelvin
