[v3,18/18] infiniband: cxgb4: Eliminate duplicate barriers on weakly-ordered archs

Commit Message

Sinan Kaya March 17, 2018, 6:30 p.m. UTC

[e-mail]

On 3/17/2018 11:05 AM, Jason Gunthorpe wrote:
> On Sat, Mar 17, 2018 at 12:25:14AM -0400, Sinan Kaya wrote:
>> On 3/17/2018 12:03 AM, Sinan Kaya wrote:
>>> On 3/16/2018 11:40 PM, Sinan Kaya wrote:
>>>> I'll change writel_relaxed() with __raw_writel() in the series like you suggested
>>>> and also look at your other comments.
>>>
>>> I spoke too soon.
>>>
>>> Now that I realized, code needs to follow one of the following patterns for correctness
>>>
>>> 1)
>>> wmb()
>>> writel()/writel_relaxed()
>>>
>>> or 
>>>
>>> 2)
>>> wmb()
>>> __raw_wrltel()
>>> mmiowb()
>>>
>>> but definitely not
>>>
>>> wmb()
>>> __raw_wrltel()
>>>
>>> Since #1 == #2, I'll stick to my current implementation of writel_relaxed()
>>>
>>> Changing writel() to writel_relaxed() or __raw_writel() isn't enough. PowerPC needs mmiowb()
>>> for correctness. ARM's mmiowb() implementation is empty.
>>>
>>> So, there is no one size fits all solution with the current state of affairs.
>>>
>>>
>>
>> I think I finally got what you mean.
>>
>> Code seems to have
>>
>> wmb()
>> writel()/writeq()
>> wmb()
>>
>> this can be safely replaced with
>>
>> wmb()
>> __raw_writel()/__raw_writeq()
>> wmb()
>>
>> This will work on all arches. Below is the new version. Let me know if this is OK.
>>
>> +++ b/drivers/infiniband/hw/cxgb4/t4.h
>> @@ -457,7 +457,7 @@ static inline void pio_copy(u64 __iomem *dst, u64 *src)
>>         int count = 8;
>>
>>         while (count) {
>> -               writeq(*src, dst);
>> +               __raw_writeq(*src, dst);
>>                 src++;
>>                 dst++;
>>                 count--;
>> @@ -477,15 +477,16 @@ static inline void t4_ring_sq_db(struct t4_wq *wq, u16 inc, union t4_wr *wqe)
>>                                  (u64 *)wqe);
>>                 } else {
>>                         pr_debug("DB wq->sq.pidx = %d\n", wq->sq.pidx);
>> -                       writel(PIDX_T5_V(inc) | QID_V(wq->sq.bar2_qid),
>> -                              wq->sq.bar2_va + SGE_UDB_KDOORBELL);
>> +                       __raw_writel(PIDX_T5_V(inc) | QID_V(wq->sq.bar2_qid),
>> +                                    wq->sq.bar2_va + SGE_UDB_KDOORBELL);
>>                 }
>>
>>                 /* Flush user doorbell area writes. */
>>                 wmb();
>>                 return;
>>         }
>> -       writel(QID_V(wq->sq.qid) | PIDX_V(inc), wq->db);
>> +       __raw_writel(QID_V(wq->sq.qid) | PIDX_V(inc), wq->db);
>> +       mmiowmb()
>>  }
>>
>>  static inline void t4_ring_rq_db(struct t4_wq *wq, u16 inc,
>> @@ -502,15 +503,16 @@ static inline void t4_ring_rq_db(struct t4_wq *wq, u16 inc,
>>                                  (void *)wqe);
>>                 } else {
>>                         pr_debug("DB wq->rq.pidx = %d\n", wq->rq.pidx);
>> -                       writel(PIDX_T5_V(inc) | QID_V(wq->rq.bar2_qid),
>> -                              wq->rq.bar2_va + SGE_UDB_KDOORBELL);
>> +                       __raw_writel(PIDX_T5_V(inc) | QID_V(wq->rq.bar2_qid),
>> +                                    wq->rq.bar2_va + SGE_UDB_KDOORBELL);
>>                 }
>>
>>                 /* Flush user doorbell area writes. */
>>                 wmb();
>>                 return;
>>         }
>> -       writel(QID_V(wq->rq.qid) | PIDX_V(inc), wq->db);
>> +       __raw_writel(QID_V(wq->rq.qid) | PIDX_V(inc), wq->db);
>> +       mmiowmb();
> 
> No! NAK! Adding *new* barriers to use the *wrong* accessor is crazy!
> 
> Your first patch was right, replacing
>   wmb()
>   writel()
> 
> With wmb(); writel_relaxed() is fine, and helps ARM.
> 
> The problem with PPC has nothing to do with the writel, it is with the
> spinlock, and we can't improve it without adding some new
> infrastructure. Certainly don't hack around the problem by using
> __raw_writel in multi-arch drivers!
> 
> In userspace we settled on something like this as a pattern:
> 
>  mmio_wc_spin_lock()
>  writel_wc_mmio()
>  mmio_wc_spin_unlock()

> 
> Where mmio_wc_spin_unlock incorporates the mmiowmb and is defined to
> fully contain all MMIO access to WC and UC memory.
> 
> Due to our userspace the pattern is specifically used with MMIO writes
> to WC BAR memory, but it could be generalized..
> 
> This allows all known arches to use the best code in all cases. x86
> can even optimize a lot and combine the mmiowmb() into the
> spin_unlock, apparently.

Given both Dave [1] and Jason's feedback, we have to ask PowerPC developers
to fix writel_relaxed().

Otherwise, PowerPC won't be able to take advantage of the optimizations
introduced in this series.

I don't think we need writel_really_relaxed() API.

Somebody also has to take a task and work very hard to get rid of __raw_writeX()
APIs in drivers/net directory. It looked like a very common practice though
it clearly violates multiarch portability concerns Jason and Deve highlighted.

This will be the next version:

iff --git a/drivers/infiniband/hw/cxgb4/t4.h b/drivers/infiniband/hw/cxgb4/t4.h
index 8369c7c..6e5658a 100644
[1] https://lkml.org/lkml/2018/3/17/100

> 
> Jason
>

Comments

Jason Gunthorpe March 19, 2018, 1:48 a.m. UTC | #1

On Sat, Mar 17, 2018 at 02:30:10PM -0400, Sinan Kaya wrote:

> Somebody also has to take a task and work very hard to get rid of __raw_writeX()
> APIs in drivers/net directory. It looked like a very common practice though
> it clearly violates multiarch portability concerns Jason and Deve highlighted.

When you posted your list I thought most of the hits were in what I'd
think of 'one-arch drivers', eg an IRQ controller or clock driver or
something.. Some might have a reason for it (eg avoiding the swap, for
instance), maybe it is a hold over from before writel_relaxed, or
maybe it is just a cargo-cult behavior..

It is the obviously multi-arch drivers that probably need some
attention..

Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch

--- a/drivers/infiniband/hw/cxgb4/t4.h
+++ b/drivers/infiniband/hw/cxgb4/t4.h
@@ -457,7 +457,7 @@  static inline void pio_copy(u64 __iomem *dst, u64 *src)
        int count = 8;
 
        while (count) {
-               writeq(*src, dst);
+               writeq_relaxed(*src, dst);
                src++;
                dst++;
                count--;
@@ -477,15 +477,15 @@  static inline void t4_ring_sq_db(struct t4_wq *wq, u16 inc, union t4_wr *wqe)
                                 (u64 *)wqe);
                } else {
                        pr_debug("DB wq->sq.pidx = %d\n", wq->sq.pidx);
-                       writel(PIDX_T5_V(inc) | QID_V(wq->sq.bar2_qid),
-                              wq->sq.bar2_va + SGE_UDB_KDOORBELL);
+                       writel_relaxed(PIDX_T5_V(inc) | QID_V(wq->sq.bar2_qid),
+                                      wq->sq.bar2_va + SGE_UDB_KDOORBELL);
                }
 
                /* Flush user doorbell area writes. */
                wmb();
                return;
        }
-       writel(QID_V(wq->sq.qid) | PIDX_V(inc), wq->db);
+       writel_relaxed(QID_V(wq->sq.qid) | PIDX_V(inc), wq->db);
 }
 
 static inline void t4_ring_rq_db(struct t4_wq *wq, u16 inc,
@@ -502,15 +502,15 @@  static inline void t4_ring_rq_db(struct t4_wq *wq, u16 inc,
                                 (void *)wqe);
                } else {
                        pr_debug("DB wq->rq.pidx = %d\n", wq->rq.pidx);
-                       writel(PIDX_T5_V(inc) | QID_V(wq->rq.bar2_qid),
-                              wq->rq.bar2_va + SGE_UDB_KDOORBELL);
+                       writel_relaxed(PIDX_T5_V(inc) | QID_V(wq->rq.bar2_qid),
+                                      wq->rq.bar2_va + SGE_UDB_KDOORBELL);
                }
 
                /* Flush user doorbell area writes. */
                wmb();
                return;
        }
-       writel(QID_V(wq->rq.qid) | PIDX_V(inc), wq->db);
+       writel_relaxed(QID_V(wq->rq.qid) | PIDX_V(inc), wq->db);
 }