Re: [PATCH] parisc: use per-pagetable spinlock

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 2019-04-13 12:23 p.m., Helge Deller wrote:
> On 10.04.19 21:27, John David Anglin wrote:
>> On 2019-04-10 3:10 p.m., John David Anglin wrote:
>>> Was merced bus used on PA 1.x?  If not, then the locks could be
>>> removed.
>> Openpa states that merced bus was used in L1500, L3000 and N4000.  All are Pa 2.0,
>> so locks are not needed.
> And according to https://www.openpa.net/pa-risc_chipsets_stretch.html
> it seems that a IKE I/O controller is used on Stretch only.
> So, detecting an IKE should be sufficient to check if we need
> to avoid multiple TLB flushed on the bus...
>
In hardware.c, we have:

        {HPHW_BCPORT, 0x800, 0x0000C, 0x10, "DEW BC Merced Port"},
        {HPHW_BCPORT, 0x801, 0x0000C, 0x10, "SMC Bus Interface Merced Bus0"},
        {HPHW_BCPORT, 0x802, 0x0000C, 0x10, "SMC Bus INterface Merced Bus1"},
        {HPHW_BCPORT, 0x803, 0x0000C, 0x10, "IKE I/O BC Merced Port"},
        {HPHW_BCPORT, 0x781, 0x0000C, 0x00, "IKE I/O BC Ropes Port"},
        {HPHW_BCPORT, 0x804, 0x0000C, 0x10, "REO I/O BC Merced Port"},
        {HPHW_BCPORT, 0x782, 0x0000C, 0x00, "REO I/O BC Ropes Port"},

Dave

-- 
John David Anglin  dave.anglin@xxxxxxxx




[Index of Archives]     [Linux SoC]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux