Re: [PATCH 1/5] mm/hmm: Update HMM documentation

[Ralph Campbell <rcampbell@nvidia.com>]

On 6/6/19 7:02 AM, Jason Gunthorpe wrote:
> On Mon, May 06, 2019 at 04:29:38PM -0700, rcampbell@nvidia.com wrote:
>> From: Ralph Campbell <rcampbell@nvidia.com>
>>
>> Update the HMM documentation to reflect the latest API and make a few minor
>> wording changes.
>>
>> Signed-off-by: Ralph Campbell <rcampbell@nvidia.com>
>> Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
>> Cc: John Hubbard <jhubbard@nvidia.com>
>> Cc: Ira Weiny <ira.weiny@intel.com>
>> Cc: Dan Williams <dan.j.williams@intel.com>
>> Cc: Arnd Bergmann <arnd@arndb.de>
>> Cc: Balbir Singh <bsingharora@gmail.com>
>> Cc: Dan Carpenter <dan.carpenter@oracle.com>
>> Cc: Matthew Wilcox <willy@infradead.org>
>> Cc: Souptick Joarder <jrdr.linux@gmail.com>
>> Cc: Andrew Morton <akpm@linux-foundation.org>
>>   Documentation/vm/hmm.rst | 139 ++++++++++++++++++++-------------------
>>   1 file changed, 73 insertions(+), 66 deletions(-)
> 
> Okay, lets start picking up hmm patches in to the new shared hmm.git,
> as promised I will take responsibility to send these to Linus. The
> tree is here:
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma.git/log/?h=hmm
> 
> This looks fine to me with one minor comment:
> 
>> diff --git a/Documentation/vm/hmm.rst b/Documentation/vm/hmm.rst
>> index ec1efa32af3c..7c1e929931a0 100644
>> +++ b/Documentation/vm/hmm.rst
>>   
>> @@ -151,21 +151,27 @@ registration of an hmm_mirror struct::
>>   
>>    int hmm_mirror_register(struct hmm_mirror *mirror,
>>                            struct mm_struct *mm);
>> - int hmm_mirror_register_locked(struct hmm_mirror *mirror,
>> -                                struct mm_struct *mm);
>>   
>> -
>> -The locked variant is to be used when the driver is already holding mmap_sem
>> -of the mm in write mode. The mirror struct has a set of callbacks that are used
>> +The mirror struct has a set of callbacks that are used
>>   to propagate CPU page tables::
>>   
>>    struct hmm_mirror_ops {
>> +     /* release() - release hmm_mirror
>> +      *
>> +      * @mirror: pointer to struct hmm_mirror
>> +      *
>> +      * This is called when the mm_struct is being released.
>> +      * The callback should make sure no references to the mirror occur
>> +      * after the callback returns.
>> +      */
> 
> This is not quite accurate (at least, as the other series I sent
> intends), the struct hmm_mirror is valid up until
> hmm_mirror_unregister() is called - specifically it remains valid
> after the release() callback.
> 
> I will revise it (and the hmm.h comment it came from) to read the
> below. Please let me know if you'd like something else:
> 
> 	/* release() - release hmm_mirror
> 	 *
> 	 * @mirror: pointer to struct hmm_mirror
> 	 *
> 	 * This is called when the mm_struct is being released.  The callback
> 	 * must ensure that all access to any pages obtained from this mirror
> 	 * is halted before the callback returns. All future access should
> 	 * fault.
> 	 */
> 
> The key task for release is to fence off all device access to any
> related pages as the mm is about to recycle them and the device must
> not cause a use-after-free.
> 
> I applied it to hmm.git
> 
> Thanks,
> Jason
> 

Yes, I agree this is better.

Also, I noticed the sample code for hmm_range_register() is wrong.
If you could merge this minor change into this patch, that
would be appreciated.

diff --git a/Documentation/vm/hmm.rst b/Documentation/vm/hmm.rst
index dc8fe4241a18..b5fb9bc02aa2 100644
--- a/Documentation/vm/hmm.rst
+++ b/Documentation/vm/hmm.rst
@@ -245,7 +245,7 @@ The usage pattern is::
              hmm_range_wait_until_valid(&range, TIMEOUT_IN_MSEC);
              goto again;
            }
-          hmm_mirror_unregister(&range);
+          hmm_range_unregister(&range);
            return ret;
        }
        take_lock(driver->update);
@@ -257,7 +257,7 @@ The usage pattern is::

        // Use pfns array content to update device page table

-      hmm_mirror_unregister(&range);
+      hmm_range_unregister(&range);
        release_lock(driver->update);
        up_read(&mm->mmap_sem);
        return 0;