Rune,
I don't think that there is a bug.
I have PCI working on an 8250 target.
The new structure agrees with the rev 1 of the MPC8xx manual.
The old structure agreed with the rev 0 manual (without PCI).
Conor
Rune Torgersen wrote:
> I think I found a bug in immap_cmp2.h
>
> (in structue sit_cpm2_t)
>
> I know that res7 was changed to make room for the PCI structure
> But res6 is 2 bytes shorter in the new one.
>
> I compared to the same structure in u-boot (that got the PCI change at
> about the same time) and it has it at 94 bytes long also.
>
>
> Old structure:
> /* System Integration Timers.
> */
> typedef struct sys_int_timers {
> char res1[32];
> ushort sit_tmcntsc;
> char res2[2];
> uint sit_tmcnt;
> char res3[4];
> uint sit_tmcntal;
> char res4[16];
> ushort sit_piscr;
> char res5[2];
> uint sit_pitc;
> uint sit_pitr;
> char res6[94];
> char res7[2390];
> } sit8260_t;
>
> New structue:
> typedef struct sys_int_timers {
> u8 res1[32];
> u16 sit_tmcntsc;
> u8 res2[2];
> u32 sit_tmcnt;
> u8 res3[4];
> u32 sit_tmcntal;
> u8 res4[16];
> u16 sit_piscr;
> u8 res5[2];
> u32 sit_pitc;
> u32 sit_pitr;
> u8 res6[92];
> u8 res7[390];
> } sit_cpm2_t;
>
> Rune Torgersen
> System Developer
> Innovative Systems LLC
> 1000 Innovative Drive
> Mitchell, SD 57301
> Ph: 605-995-6120
> www.innovsys.com
> _______________________________________________
> Linuxppc-embedded mailing list
> Linuxppc-embedded at ozlabs.org
> https://ozlabs.org/mailman/listinfo/linuxppc-embedded
>
>
