search for: xa_load

...c Look for the word 'implicit' mlx5_ib_invalidate_range() releases the interval_notifier when there are no populated shadow PTEs in its leaf pagefault_implicit_mr() creates an interval_notifier that covers the level in the page table that needs population. Notice it just uses an unlocked xa_load to find the page table level. The locking is pretty tricky as it relies on RCU, but the fault flow is fairly lightweight. Jason

...n(range, pfn) | cpu_flags; > } > - if (hmm_vma_walk->pgmap) { > - put_dev_pagemap(hmm_vma_walk->pgmap); > - hmm_vma_walk->pgmap = NULL; Putting the value in the hmm_vma_walk would have made some sense to me if the pgmap was not set to NULL all over the place. Then the most xa_loads would be eliminated, as I would expect the pgmap tends to be mostly uniform for these use cases. Is there some reason the pgmap ref can't be held across faulting/sleeping? ie like below. Anyhow, I looked over this pretty carefully and the change looks functionally OK, I just don't know w...

...vma_walk->pgmap) { > > - put_dev_pagemap(hmm_vma_walk->pgmap); > > - hmm_vma_walk->pgmap = NULL; > > Putting the value in the hmm_vma_walk would have made some sense to me > if the pgmap was not set to NULL all over the place. Then the most > xa_loads would be eliminated, as I would expect the pgmap tends to be > mostly uniform for these use cases. > > Is there some reason the pgmap ref can't be held across > faulting/sleeping? ie like below. No restriction on holding refs over faulting / sleeping. > > Anyhow, I looked o...

...gt;cx << 32; + break; + default: + hi = 0; + } + + return hi | lo; +} + +static bool sev_es_check_cpuid_cache(struct es_em_ctxt *ctxt, + unsigned long cache_index) +{ + struct sev_es_cpuid_cache_entry *cache_entry; + + if (cache_index == ULONG_MAX) + return false; + + cache_entry = xa_load(&sev_es_cpuid_cache, cache_index); + if (!cache_entry) + return false; + + ctxt->regs->ax = cache_entry->eax; + ctxt->regs->bx = cache_entry->ebx; + ctxt->regs->cx = cache_entry->ecx; + ctxt->regs->dx = cache_entry->edx; + + return true; +} + +static void se...

On Wed, Jan 15, 2020 at 02:09:47PM -0800, Ralph Campbell wrote: > I don't understand the lifetime/membership issue. The driver is the only thing > that allocates, inserts, or removes struct mmu_interval_notifier and thus > completely controls the lifetime. If the returned value is on the defered list it could be freed at any moment. The existing locks do not prevent it. > >

...ed long pfn; > + void *ptr; > + > + ptr = bounce->ptr + ((start - bounce->addr) & PAGE_MASK); > + > + for (pfn = start >> PAGE_SHIFT; pfn < (end >> PAGE_SHIFT); pfn++) { > + void *entry; > + struct page *page; > + void *tmp; > + > + entry = xa_load(&dmirror->pt, pfn); > + page = xa_untag_pointer(entry); > + if (!page) > + return -ENOENT; > + > + tmp = kmap(page); > + memcpy(ptr, tmp, PAGE_SIZE); > + kunmap(page); > + > + ptr += PAGE_SIZE; > + bounce->cpages++; > + } > + > + return 0; &...

On 3/17/20 5:59 AM, Christoph Hellwig wrote: > On Tue, Mar 17, 2020 at 09:47:55AM -0300, Jason Gunthorpe wrote: >> I've been using v7 of Ralph's tester and it is working well - it has >> DEVICE_PRIVATE support so I think it can test this flow too. Ralph are >> you able? >> >> This hunk seems trivial enough to me, can we include it now? > > I can send

...gt; + >> + ptr = bounce->ptr + ((start - bounce->addr) & PAGE_MASK); >> + >> + for (pfn = start >> PAGE_SHIFT; pfn < (end >> PAGE_SHIFT); pfn++) { >> + void *entry; >> + struct page *page; >> + void *tmp; >> + >> + entry = xa_load(&dmirror->pt, pfn); >> + page = xa_untag_pointer(entry); >> + if (!page) >> + return -ENOENT; >> + >> + tmp = kmap(page); >> + memcpy(ptr, tmp, PAGE_SIZE); >> + kunmap(page); >> + >> + ptr += PAGE_SIZE; >> + bounce->cpag...

This series adds new functions to the mmu interval notifier API to allow device drivers with MMUs to dynamically mirror a process' page tables based on device faults and invalidation callbacks. The Nouveau driver is updated to use the extended API and a set of stand alone self tests is added to help validate and maintain correctness. The patches are based on linux-5.5.0-rc6 and are for

Hi Jérôme, Ben, Felix and Jason, below is a series against the hmm tree which cleans up various minor bits and allows HMM_MIRROR to be built on all architectures. Diffstat: 11 files changed, 94 insertions(+), 210 deletions(-) A git tree is also available at: git://git.infradead.org/users/hch/misc.git hmm-cleanups.2 Gitweb:

Hi, here is the next version of changes to enable Linux to run as an SEV-ES guest. The code was rebased to v5.7-rc3 and got a fair number of changes since the last version. What is SEV-ES ============== SEV-ES is an acronym for 'Secure Encrypted Virtualization - Encrypted State' and means a hardware feature of AMD processors which hides the register state of VCPUs to the hypervisor by

Hi, here is the next version of changes to enable Linux to run as an SEV-ES guest. The code was rebased to v5.7-rc3 and got a fair number of changes since the last version. What is SEV-ES ============== SEV-ES is an acronym for 'Secure Encrypted Virtualization - Encrypted State' and means a hardware feature of AMD processors which hides the register state of VCPUs to the hypervisor by