Re: [PATCH v2 6/7] s390x/mmu: DAT table definition overhaul

[Thomas Huth <thuth@redhat.com>]

On 26/09/2019 09.38, David Hildenbrand wrote:
> On 26.09.19 09:35, Thomas Huth wrote:
>> On 25/09/2019 14.52, David Hildenbrand wrote:
>>> Let's use consitent names for the region/section/page table entries and
>>> for the macros to extract relevant parts from virtual address. Make them
>>> match the definitions in the PoP - e.g., how the televant bits are actually
>>
>> s/televant/relevant/
>>
>>> called.
>>>
>>> Introduce defines for all bits declared in the PoP. This will come in
>>> handy in follow-up patches.
>>>
>>> Signed-off-by: David Hildenbrand <david@redhat.com>
>>> ---
>>>  target/s390x/cpu.h        | 77 +++++++++++++++++++++++++++++----------
>>>  target/s390x/mem_helper.c | 12 +++---
>>>  target/s390x/mmu_helper.c | 37 ++++++++++---------
>>>  3 files changed, 83 insertions(+), 43 deletions(-)
>>>
>>> diff --git a/target/s390x/cpu.h b/target/s390x/cpu.h
>>> index 163dae13d7..e74a809257 100644
>>> --- a/target/s390x/cpu.h
>>> +++ b/target/s390x/cpu.h
>>> @@ -558,26 +558,63 @@ QEMU_BUILD_BUG_ON(sizeof(SysIB) != 4096);
>>>  #define ASCE_TYPE_SEGMENT     0x00        /* segment table type               */
>>>  #define ASCE_TABLE_LENGTH     0x03        /* region table length              */
>>>  
>>> -#define REGION_ENTRY_ORIGIN   (~0xfffULL) /* region/segment table origin    */
>>> -#define REGION_ENTRY_RO       0x200       /* region/segment protection bit  */
>>> -#define REGION_ENTRY_TF       0xc0        /* region/segment table offset    */
>>> -#define REGION_ENTRY_INV      0x20        /* invalid region table entry     */
>>> -#define REGION_ENTRY_TYPE_MASK 0x0c       /* region/segment table type mask */
>>> -#define REGION_ENTRY_TYPE_R1  0x0c        /* region first table type        */
>>> -#define REGION_ENTRY_TYPE_R2  0x08        /* region second table type       */
>>> -#define REGION_ENTRY_TYPE_R3  0x04        /* region third table type        */
>>> -#define REGION_ENTRY_LENGTH   0x03        /* region third length            */
>>> -
>>> -#define SEGMENT_ENTRY_ORIGIN  (~0x7ffULL) /* segment table origin        */
>>> -#define SEGMENT_ENTRY_FC      0x400       /* format control              */
>>> -#define SEGMENT_ENTRY_RO      0x200       /* page protection bit         */
>>> -#define SEGMENT_ENTRY_INV     0x20        /* invalid segment table entry */
>>> -
>>> -#define VADDR_PX              0xff000     /* page index bits   */
>>> -
>>> -#define PAGE_RO               0x200       /* HW read-only bit  */
>>> -#define PAGE_INVALID          0x400       /* HW invalid bit    */
>>> -#define PAGE_RES0             0x800       /* bit must be zero  */
>>> +#define REGION_ENTRY_ORIGIN         0xfffffffffffff000ULL
>>> +#define REGION_ENTRY_P              0x0000000000000200ULL
>>> +#define REGION_ENTRY_TF             0x00000000000000c0ULL
>>> +#define REGION_ENTRY_I              0x0000000000000020ULL
>>> +#define REGION_ENTRY_TT             0x000000000000000cULL
>>> +#define REGION_ENTRY_TL             0x0000000000000003ULL
>>
>> Any chance that you could keep the comments after the definitions? I
>> think they are useful for people who are not 100% familiar with the DAT
>> on s390x.
> 
> I thought about that, but do we expect people that don't have a clue
> about s390x DAT and don't compare the code against the PoP to understand
> our DAT translation just by comments on defines?

I'm not sure that everybody is aware of the PoP ... maybe you could just
put a comment in front of the block a la:

/*
 * For details on the following definitions, see the "Dynamic Address
 * Translation" section in chapter 3 of the "z/Architecture Principles
 * of Operations - SA22-7832-11"
 */

?

 Thomas