Re: [PATCH RESEND v3 1/2] mm/tlb: skip redundant IPI when TLB flush already synchronized

[Lance Yang <lance.yang@linux.dev>]

On 2026/1/6 23:19, David Hildenbrand (Red Hat) wrote:
>>   static void tlb_table_flush(struct mmu_gather *tlb)
>> @@ -367,7 +378,7 @@ void tlb_remove_table(struct mmu_gather *tlb, void 
>> *table)
>>           *batch = (struct mmu_table_batch *)__get_free_page(GFP_NOWAIT);
>>           if (*batch == NULL) {
>>               tlb_table_invalidate(tlb);
>> -            tlb_remove_table_one(table);
>> +            tlb_remove_table_one(table, tlb);
>>               return;
>>           }
>>           (*batch)->nr = 0;
>> @@ -427,6 +438,7 @@ static void __tlb_gather_mmu(struct mmu_gather 
>> *tlb, struct mm_struct *mm,
>>       tlb->vma_pfn = 0;
>>       tlb->fully_unshared_tables = 0;
>> +    tlb->tlb_flush_sent_ipi = 0;
>>       __tlb_reset_range(tlb);
>>       inc_tlb_flush_pending(tlb->mm);
>>   }
> 
> But when would we have to reset tlb->tlb_flush_sent_ipi = 0 later? 
> That's where it gets tricky. Just imagine the MMU gather gets reused later.
> 
> Also,
> 
>      +    if (info->freed_tables && info->tlb)
>      +        info->tlb->tlb_flush_sent_ipi = true;
> 
> in native_flush_tlb_multi() misses the fact that we have different 
> flushing types for removed/unshared tables vs. other flush.
> 
> So this approach more here certainly gets more complicated and error prone.

Agreed. Tracking the flag through mmu_gather lifecycle does get
more complicated and error-prone ...

> 
> tlb_table_flush_implies_ipi_broadcast() was clearer in that regard: if 
> you flushed the TLB after removing /unsharing tables, the IPI for 
> handling page tables can be skipped. It's on the code flow to assure that.

v2 was definitely simpler.

> 
> What could work is tracking "tlb_table_flush_sent_ipi" really when we 
> are flushing the TLB for removed/unshared tables, and maybe resetting 
> it ... I don't know when from the top of my head.

Not sure what's the best way forward here :(

> 
> v2 was simpler IMHO.

The main concern Dave raised was that with PV hypercalls or when
INVLPGB is available, we can't tell from a static check whether IPIs
were actually sent.

Maybe that's acceptable, or we could find a simpler way to track that ...

Open to suggestions!