| Message ID | 20230706075612.67404-2-david@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | virtio-mem: Support "x-ignore-shared" migration | expand |
David Hildenbrand <david@redhat.com> wrote: > ram_block_discard_range() cannot possibly do the right thing in > MAP_PRIVATE file mappings in the general case. > > To achieve the documented semantics, we also have to punch a hole into > the file, possibly messing with other MAP_PRIVATE/MAP_SHARED mappings > of such a file. > > For example, using VM templating -- see commit b17fbbe55cba ("migration: > allow private destination ram with x-ignore-shared") -- in combination with > any mechanism that relies on discarding of RAM is problematic. This > includes: > * Postcopy live migration > * virtio-balloon inflation/deflation or free-page-reporting > * virtio-mem > > So at least warn that there is something possibly dangerous is going on > when using ram_block_discard_range() in these cases. > > Acked-by: Peter Xu <peterx@redhat.com> > Tested-by: Mario Casquero <mcasquer@redhat.com> > Signed-off-by: David Hildenbrand <david@redhat.com> Reviewed-by: Juan Quintela <quintela@redhat.com> (at least we give a warning) But I wonder if we can do better and test that: * Postcopy live migration We can check if we are on postcopy, or put a marker so we know that postcopy can have problems when started. * virtio-balloon inflation/deflation or free-page-reporting We can check if we have ever used virtio-balloon. * virtio-mem We can check if we have used virtio-men I am just wondering if that is even possible? Thanks, Juan.
On 06.07.23 10:10, Juan Quintela wrote: > David Hildenbrand <david@redhat.com> wrote: >> ram_block_discard_range() cannot possibly do the right thing in >> MAP_PRIVATE file mappings in the general case. >> >> To achieve the documented semantics, we also have to punch a hole into >> the file, possibly messing with other MAP_PRIVATE/MAP_SHARED mappings >> of such a file. >> >> For example, using VM templating -- see commit b17fbbe55cba ("migration: >> allow private destination ram with x-ignore-shared") -- in combination with >> any mechanism that relies on discarding of RAM is problematic. This >> includes: >> * Postcopy live migration >> * virtio-balloon inflation/deflation or free-page-reporting >> * virtio-mem >> >> So at least warn that there is something possibly dangerous is going on >> when using ram_block_discard_range() in these cases. >> >> Acked-by: Peter Xu <peterx@redhat.com> >> Tested-by: Mario Casquero <mcasquer@redhat.com> >> Signed-off-by: David Hildenbrand <david@redhat.com> > > Reviewed-by: Juan Quintela <quintela@redhat.com> > > (at least we give a warning) > > But I wonder if we can do better and test that: > * Postcopy live migration > > We can check if we are on postcopy, or put a marker so we know that > postcopy can have problems when started. > > * virtio-balloon inflation/deflation or free-page-reporting > > We can check if we have ever used virtio-balloon. > > * virtio-mem > > We can check if we have used virtio-men > > > I am just wondering if that is even possible? Now we warn when any of these features actually tries discarding RAM (calling ram_block_discard_range()). As these features trigger discarding of RAM, once we reach this point we know that they are getting used. (in comparison to default libvirt attaching a virtio-balloon device without anybody ever using it) The alternative would be checking the RAM for compatibility when configuring each features. I decided to warn at a central place for now, which covers any users. Thanks!
> #ifdef CONFIG_FALLOCATE_PUNCH_HOLE > + /* > + * We'll discard data from the actual file, even though we only > + * have a MAP_PRIVATE mapping, possibly messing with other > + * MAP_PRIVATE/MAP_SHARED mappings. There is no easy way to > + * change that behavior whithout violating the promised > + * semantics of ram_block_discard_range(). > + * > + * Only warn, because it work as long as nobody else uses that > I'll fixup s/work/works/
David Hildenbrand <david@redhat.com> wrote: > On 06.07.23 10:10, Juan Quintela wrote: >> David Hildenbrand <david@redhat.com> wrote: >>> ram_block_discard_range() cannot possibly do the right thing in >>> MAP_PRIVATE file mappings in the general case. >>> >>> To achieve the documented semantics, we also have to punch a hole into >>> the file, possibly messing with other MAP_PRIVATE/MAP_SHARED mappings >>> of such a file. >>> >>> For example, using VM templating -- see commit b17fbbe55cba ("migration: >>> allow private destination ram with x-ignore-shared") -- in combination with >>> any mechanism that relies on discarding of RAM is problematic. This >>> includes: >>> * Postcopy live migration >>> * virtio-balloon inflation/deflation or free-page-reporting >>> * virtio-mem >>> >>> So at least warn that there is something possibly dangerous is going on >>> when using ram_block_discard_range() in these cases. >>> >>> Acked-by: Peter Xu <peterx@redhat.com> >>> Tested-by: Mario Casquero <mcasquer@redhat.com> >>> Signed-off-by: David Hildenbrand <david@redhat.com> >> Reviewed-by: Juan Quintela <quintela@redhat.com> >> (at least we give a warning) >> But I wonder if we can do better and test that: >> * Postcopy live migration >> We can check if we are on postcopy, or put a marker so we know >> that >> postcopy can have problems when started. >> * virtio-balloon inflation/deflation or free-page-reporting >> We can check if we have ever used virtio-balloon. >> * virtio-mem >> We can check if we have used virtio-men >> I am just wondering if that is even possible? > > Now we warn when any of these features actually tries discarding RAM > (calling ram_block_discard_range()). > > As these features trigger discarding of RAM, once we reach this point > we know that they are getting used. (in comparison to default libvirt > attaching a virtio-balloon device without anybody ever using it) > > > The alternative would be checking the RAM for compatibility when > configuring each features. I decided to warn at a central place for > now, which covers any users. I think this is the right thing to do. Patient: It hurts when I do this. Doctor: Don't do that. O:-) Now more seriously, at this point we are very late to do anything sensible. I think that the normal thing when we are configuring incompatible things just flag it. We are following that approach with migration for some time now, and everybody is happier. Later, Juan.
On 06.07.23 15:20, Juan Quintela wrote: > David Hildenbrand <david@redhat.com> wrote: >> On 06.07.23 10:10, Juan Quintela wrote: >>> David Hildenbrand <david@redhat.com> wrote: >>>> ram_block_discard_range() cannot possibly do the right thing in >>>> MAP_PRIVATE file mappings in the general case. >>>> >>>> To achieve the documented semantics, we also have to punch a hole into >>>> the file, possibly messing with other MAP_PRIVATE/MAP_SHARED mappings >>>> of such a file. >>>> >>>> For example, using VM templating -- see commit b17fbbe55cba ("migration: >>>> allow private destination ram with x-ignore-shared") -- in combination with >>>> any mechanism that relies on discarding of RAM is problematic. This >>>> includes: >>>> * Postcopy live migration >>>> * virtio-balloon inflation/deflation or free-page-reporting >>>> * virtio-mem >>>> >>>> So at least warn that there is something possibly dangerous is going on >>>> when using ram_block_discard_range() in these cases. >>>> >>>> Acked-by: Peter Xu <peterx@redhat.com> >>>> Tested-by: Mario Casquero <mcasquer@redhat.com> >>>> Signed-off-by: David Hildenbrand <david@redhat.com> >>> Reviewed-by: Juan Quintela <quintela@redhat.com> >>> (at least we give a warning) >>> But I wonder if we can do better and test that: >>> * Postcopy live migration >>> We can check if we are on postcopy, or put a marker so we know >>> that >>> postcopy can have problems when started. >>> * virtio-balloon inflation/deflation or free-page-reporting >>> We can check if we have ever used virtio-balloon. >>> * virtio-mem >>> We can check if we have used virtio-men >>> I am just wondering if that is even possible? >> >> Now we warn when any of these features actually tries discarding RAM >> (calling ram_block_discard_range()). >> >> As these features trigger discarding of RAM, once we reach this point >> we know that they are getting used. (in comparison to default libvirt >> attaching a virtio-balloon device without anybody ever using it) >> >> >> The alternative would be checking the RAM for compatibility when >> configuring each features. I decided to warn at a central place for >> now, which covers any users. > > I think this is the right thing to do. > > Patient: It hurts when I do this. > Doctor: Don't do that. > > O:-) :) > > Now more seriously, at this point we are very late to do anything > sensible. I think that the normal thing when we are configuring > incompatible things just flag it. > > We are following that approach with migration for some time now, and > everybody is happier. For the time being I'll move forward with this patch. I agree that warning early is nicer (but warning for example for virtio-balloon early doesn't make too much sense: libvirt adds it blindly to each VM just to query guest statistics and never inflate the balloon). In any case we'll want to warn here as well, because we know that new callers will easily ignore that limitation / checks.
David, On 7/6/2023 3:56 PM, David Hildenbrand wrote: > ram_block_discard_range() cannot possibly do the right thing in > MAP_PRIVATE file mappings in the general case. > > To achieve the documented semantics, we also have to punch a hole into > the file, possibly messing with other MAP_PRIVATE/MAP_SHARED mappings > of such a file. > > For example, using VM templating -- see commit b17fbbe55cba ("migration: > allow private destination ram with x-ignore-shared") -- in combination with > any mechanism that relies on discarding of RAM is problematic. This > includes: > * Postcopy live migration > * virtio-balloon inflation/deflation or free-page-reporting > * virtio-mem > > So at least warn that there is something possibly dangerous is going on > when using ram_block_discard_range() in these cases. > > Acked-by: Peter Xu <peterx@redhat.com> > Tested-by: Mario Casquero <mcasquer@redhat.com> > Signed-off-by: David Hildenbrand <david@redhat.com> > --- > softmmu/physmem.c | 18 ++++++++++++++++++ > 1 file changed, 18 insertions(+) > > diff --git a/softmmu/physmem.c b/softmmu/physmem.c > index bda475a719..4ee157bda4 100644 > --- a/softmmu/physmem.c > +++ b/softmmu/physmem.c > @@ -3456,6 +3456,24 @@ int ram_block_discard_range(RAMBlock *rb, uint64_t start, size_t length) > * so a userfault will trigger. > */ > #ifdef CONFIG_FALLOCATE_PUNCH_HOLE > + /* > + * We'll discard data from the actual file, even though we only > + * have a MAP_PRIVATE mapping, possibly messing with other > + * MAP_PRIVATE/MAP_SHARED mappings. There is no easy way to > + * change that behavior whithout violating the promised > + * semantics of ram_block_discard_range(). > + * > + * Only warn, because it work as long as nobody else uses that > + * file. > + */ > + if (!qemu_ram_is_shared(rb)) { > + warn_report_once("ram_block_discard_range: Discarding RAM" > + " in private file mappings is possibly" > + " dangerous, because it will modify the" > + " underlying file and will affect other" > + " users of the file"); > + } > + TDX has two types of memory backend for each RAM, shared memory and private memory. Private memory is serviced by guest memfd and shared memory can also be backed with a fd. At any time, only one type needs to be valid, which means the opposite can be discarded. We do implement the memory discard when TDX converts the memory[1]. It will trigger this warning 100% because by default the guest memfd is not mapped as shared (MAP_SHARED). Simply remove the warning will fail the purpose of this patch. The other option is to skip the warning for TDX case, which looks vary hacky. Do you have any idea? [1] https://lore.kernel.org/all/20230914035117.3285885-18-xiaoyao.li@intel.com/ > ret = fallocate(rb->fd, FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE, > start, length); > if (ret) {
On 18.10.23 05:02, Xiaoyao Li wrote: > David, > > On 7/6/2023 3:56 PM, David Hildenbrand wrote: >> ram_block_discard_range() cannot possibly do the right thing in >> MAP_PRIVATE file mappings in the general case. >> >> To achieve the documented semantics, we also have to punch a hole into >> the file, possibly messing with other MAP_PRIVATE/MAP_SHARED mappings >> of such a file. >> >> For example, using VM templating -- see commit b17fbbe55cba ("migration: >> allow private destination ram with x-ignore-shared") -- in combination with >> any mechanism that relies on discarding of RAM is problematic. This >> includes: >> * Postcopy live migration >> * virtio-balloon inflation/deflation or free-page-reporting >> * virtio-mem >> >> So at least warn that there is something possibly dangerous is going on >> when using ram_block_discard_range() in these cases. >> >> Acked-by: Peter Xu <peterx@redhat.com> >> Tested-by: Mario Casquero <mcasquer@redhat.com> >> Signed-off-by: David Hildenbrand <david@redhat.com> >> --- >> softmmu/physmem.c | 18 ++++++++++++++++++ >> 1 file changed, 18 insertions(+) >> >> diff --git a/softmmu/physmem.c b/softmmu/physmem.c >> index bda475a719..4ee157bda4 100644 >> --- a/softmmu/physmem.c >> +++ b/softmmu/physmem.c >> @@ -3456,6 +3456,24 @@ int ram_block_discard_range(RAMBlock *rb, uint64_t start, size_t length) >> * so a userfault will trigger. >> */ >> #ifdef CONFIG_FALLOCATE_PUNCH_HOLE >> + /* >> + * We'll discard data from the actual file, even though we only >> + * have a MAP_PRIVATE mapping, possibly messing with other >> + * MAP_PRIVATE/MAP_SHARED mappings. There is no easy way to >> + * change that behavior whithout violating the promised >> + * semantics of ram_block_discard_range(). >> + * >> + * Only warn, because it work as long as nobody else uses that >> + * file. >> + */ >> + if (!qemu_ram_is_shared(rb)) { >> + warn_report_once("ram_block_discard_range: Discarding RAM" >> + " in private file mappings is possibly" >> + " dangerous, because it will modify the" >> + " underlying file and will affect other" >> + " users of the file"); >> + } >> + > > TDX has two types of memory backend for each RAM, shared memory and > private memory. Private memory is serviced by guest memfd and shared > memory can also be backed with a fd. > > At any time, only one type needs to be valid, which means the opposite > can be discarded. We do implement the memory discard when TDX converts > the memory[1]. It will trigger this warning 100% because by default the > guest memfd is not mapped as shared (MAP_SHARED). If MAP_PRIVATE is not involved and you are taking the pages directly out of the memfd, you should mark that thing as shared. Anonymous memory is never involved. "Private memory" is only private from the guest POV, not from a mmap() point of view. Two different concepts of "private". > > Simply remove the warning will fail the purpose of this patch. The other > option is to skip the warning for TDX case, which looks vary hacky. Do > you have any idea? For TDX, all memory backends / RAMBlocks should be marked as "shared", and you should fail if that is not provided by the user.
On 10/18/2023 3:42 PM, David Hildenbrand wrote: > On 18.10.23 05:02, Xiaoyao Li wrote: >> David, >> >> On 7/6/2023 3:56 PM, David Hildenbrand wrote: >>> ram_block_discard_range() cannot possibly do the right thing in >>> MAP_PRIVATE file mappings in the general case. >>> >>> To achieve the documented semantics, we also have to punch a hole into >>> the file, possibly messing with other MAP_PRIVATE/MAP_SHARED mappings >>> of such a file. >>> >>> For example, using VM templating -- see commit b17fbbe55cba ("migration: >>> allow private destination ram with x-ignore-shared") -- in >>> combination with >>> any mechanism that relies on discarding of RAM is problematic. This >>> includes: >>> * Postcopy live migration >>> * virtio-balloon inflation/deflation or free-page-reporting >>> * virtio-mem >>> >>> So at least warn that there is something possibly dangerous is going on >>> when using ram_block_discard_range() in these cases. >>> >>> Acked-by: Peter Xu <peterx@redhat.com> >>> Tested-by: Mario Casquero <mcasquer@redhat.com> >>> Signed-off-by: David Hildenbrand <david@redhat.com> >>> --- >>> softmmu/physmem.c | 18 ++++++++++++++++++ >>> 1 file changed, 18 insertions(+) >>> >>> diff --git a/softmmu/physmem.c b/softmmu/physmem.c >>> index bda475a719..4ee157bda4 100644 >>> --- a/softmmu/physmem.c >>> +++ b/softmmu/physmem.c >>> @@ -3456,6 +3456,24 @@ int ram_block_discard_range(RAMBlock *rb, >>> uint64_t start, size_t length) >>> * so a userfault will trigger. >>> */ >>> #ifdef CONFIG_FALLOCATE_PUNCH_HOLE >>> + /* >>> + * We'll discard data from the actual file, even though >>> we only >>> + * have a MAP_PRIVATE mapping, possibly messing with other >>> + * MAP_PRIVATE/MAP_SHARED mappings. There is no easy way to >>> + * change that behavior whithout violating the promised >>> + * semantics of ram_block_discard_range(). >>> + * >>> + * Only warn, because it work as long as nobody else >>> uses that >>> + * file. >>> + */ >>> + if (!qemu_ram_is_shared(rb)) { >>> + warn_report_once("ram_block_discard_range: >>> Discarding RAM" >>> + " in private file mappings is >>> possibly" >>> + " dangerous, because it will modify >>> the" >>> + " underlying file and will affect >>> other" >>> + " users of the file"); >>> + } >>> + >> >> TDX has two types of memory backend for each RAM, shared memory and >> private memory. Private memory is serviced by guest memfd and shared >> memory can also be backed with a fd. >> >> At any time, only one type needs to be valid, which means the opposite >> can be discarded. We do implement the memory discard when TDX converts >> the memory[1]. It will trigger this warning 100% because by default the >> guest memfd is not mapped as shared (MAP_SHARED). > > If MAP_PRIVATE is not involved and you are taking the pages directly out > of the memfd, you should mark that thing as shared. Is it the general rule of Linux? Of just the rule of QEMU memory discard? > Anonymous memory is never involved. Could you please elaborate more on this? What do you want to express here regrading anonymous memory? (Sorry that I'm newbie for mmap stuff) > > "Private memory" is only private from the guest POV, not from a mmap() > point of view. > > Two different concepts of "private". > >> >> Simply remove the warning will fail the purpose of this patch. The other >> option is to skip the warning for TDX case, which looks vary hacky. Do >> you have any idea? > > For TDX, all memory backends / RAMBlocks should be marked as "shared", > and you should fail if that is not provided by the user. As I asked above, I want to understand the logic clearly. Is mapped as shared is a must to support the memory discard? i.e., if we want to support memory discard after memory type change, then the memory must be mapped with MAP_SHARED?
On 18.10.23 11:02, Xiaoyao Li wrote: > On 10/18/2023 3:42 PM, David Hildenbrand wrote: >> On 18.10.23 05:02, Xiaoyao Li wrote: >>> David, >>> >>> On 7/6/2023 3:56 PM, David Hildenbrand wrote: >>>> ram_block_discard_range() cannot possibly do the right thing in >>>> MAP_PRIVATE file mappings in the general case. >>>> >>>> To achieve the documented semantics, we also have to punch a hole into >>>> the file, possibly messing with other MAP_PRIVATE/MAP_SHARED mappings >>>> of such a file. >>>> >>>> For example, using VM templating -- see commit b17fbbe55cba ("migration: >>>> allow private destination ram with x-ignore-shared") -- in >>>> combination with >>>> any mechanism that relies on discarding of RAM is problematic. This >>>> includes: >>>> * Postcopy live migration >>>> * virtio-balloon inflation/deflation or free-page-reporting >>>> * virtio-mem >>>> >>>> So at least warn that there is something possibly dangerous is going on >>>> when using ram_block_discard_range() in these cases. >>>> >>>> Acked-by: Peter Xu <peterx@redhat.com> >>>> Tested-by: Mario Casquero <mcasquer@redhat.com> >>>> Signed-off-by: David Hildenbrand <david@redhat.com> >>>> --- >>>> softmmu/physmem.c | 18 ++++++++++++++++++ >>>> 1 file changed, 18 insertions(+) >>>> >>>> diff --git a/softmmu/physmem.c b/softmmu/physmem.c >>>> index bda475a719..4ee157bda4 100644 >>>> --- a/softmmu/physmem.c >>>> +++ b/softmmu/physmem.c >>>> @@ -3456,6 +3456,24 @@ int ram_block_discard_range(RAMBlock *rb, >>>> uint64_t start, size_t length) >>>> * so a userfault will trigger. >>>> */ >>>> #ifdef CONFIG_FALLOCATE_PUNCH_HOLE >>>> + /* >>>> + * We'll discard data from the actual file, even though >>>> we only >>>> + * have a MAP_PRIVATE mapping, possibly messing with other >>>> + * MAP_PRIVATE/MAP_SHARED mappings. There is no easy way to >>>> + * change that behavior whithout violating the promised >>>> + * semantics of ram_block_discard_range(). >>>> + * >>>> + * Only warn, because it work as long as nobody else >>>> uses that >>>> + * file. >>>> + */ >>>> + if (!qemu_ram_is_shared(rb)) { >>>> + warn_report_once("ram_block_discard_range: >>>> Discarding RAM" >>>> + " in private file mappings is >>>> possibly" >>>> + " dangerous, because it will modify >>>> the" >>>> + " underlying file and will affect >>>> other" >>>> + " users of the file"); >>>> + } >>>> + >>> >>> TDX has two types of memory backend for each RAM, shared memory and >>> private memory. Private memory is serviced by guest memfd and shared >>> memory can also be backed with a fd. >>> >>> At any time, only one type needs to be valid, which means the opposite >>> can be discarded. We do implement the memory discard when TDX converts >>> the memory[1]. It will trigger this warning 100% because by default the >>> guest memfd is not mapped as shared (MAP_SHARED). >> >> If MAP_PRIVATE is not involved and you are taking the pages directly out >> of the memfd, you should mark that thing as shared. > > Is it the general rule of Linux? Of just the rule of QEMU memory discard? > MAP_SHARED vs. MAP_PRIVATE is a common UNIX principle, and that's what this flag and the check is about. From mmap(2) MAP_SHARED: Share this mapping. Updates to the mapping are visible to other processes mapping the same region, and (in the case of file-backed mappings) are carried through to the underlying file. MAP_PRIVATE: Create a private copy-on-write mapping. Updates to the mapping are not visible to other processes mapping the same file, and are not carried through to the underlying file. It is unspecified whether changes made to the file after the mmap() call are visible in the mapped region. For your purpose (no mmap() at all), we behave like MAP_SHARED -- as if nothing special is done. No Copy-on-write, no anonymous memory. >> Anonymous memory is never involved. > > Could you please elaborate more on this? What do you want to express > here regrading anonymous memory? (Sorry that I'm newbie for mmap stuff) Anonymous memory is memory that is private to a specific process, and (see MAP_PRIVATE) modifications remain private to the process and are not reflected to the file. If you have a MAP_PRIVATE file mapping and write to a virtual memory location, you'll get a process-private copy of the underlying pagecache page. that's what we call anonymous memory, because it does not belong to a specific file. fallocate(punch) would not free up that anonymous memory. > >> >> "Private memory" is only private from the guest POV, not from a mmap() >> point of view. >> >> Two different concepts of "private". >> >>> >>> Simply remove the warning will fail the purpose of this patch. The other >>> option is to skip the warning for TDX case, which looks vary hacky. Do >>> you have any idea? >> >> For TDX, all memory backends / RAMBlocks should be marked as "shared", >> and you should fail if that is not provided by the user. > > As I asked above, I want to understand the logic clearly. Is mapped as > shared is a must to support the memory discard? i.e., if we want to > support memory discard after memory type change, then the memory must be > mapped with MAP_SHARED? MAP_PIRVATE means that it's not sufficient to only fallocate(punch) the fd to free up all memory for a virtual address, because there might be anonymous memory in a private mapping that has to be freed up using MADV_DONTNEED. That's why this functions handles both cases differently, and warns if something possibly nasty is being done.
On 10/18/2023 5:26 PM, David Hildenbrand wrote: > On 18.10.23 11:02, Xiaoyao Li wrote: >> On 10/18/2023 3:42 PM, David Hildenbrand wrote: >>> On 18.10.23 05:02, Xiaoyao Li wrote: >>>> David, >>>> >>>> On 7/6/2023 3:56 PM, David Hildenbrand wrote: >>>>> ram_block_discard_range() cannot possibly do the right thing in >>>>> MAP_PRIVATE file mappings in the general case. >>>>> >>>>> To achieve the documented semantics, we also have to punch a hole into >>>>> the file, possibly messing with other MAP_PRIVATE/MAP_SHARED mappings >>>>> of such a file. >>>>> >>>>> For example, using VM templating -- see commit b17fbbe55cba >>>>> ("migration: >>>>> allow private destination ram with x-ignore-shared") -- in >>>>> combination with >>>>> any mechanism that relies on discarding of RAM is problematic. This >>>>> includes: >>>>> * Postcopy live migration >>>>> * virtio-balloon inflation/deflation or free-page-reporting >>>>> * virtio-mem >>>>> >>>>> So at least warn that there is something possibly dangerous is >>>>> going on >>>>> when using ram_block_discard_range() in these cases. >>>>> >>>>> Acked-by: Peter Xu <peterx@redhat.com> >>>>> Tested-by: Mario Casquero <mcasquer@redhat.com> >>>>> Signed-off-by: David Hildenbrand <david@redhat.com> >>>>> --- >>>>> softmmu/physmem.c | 18 ++++++++++++++++++ >>>>> 1 file changed, 18 insertions(+) >>>>> >>>>> diff --git a/softmmu/physmem.c b/softmmu/physmem.c >>>>> index bda475a719..4ee157bda4 100644 >>>>> --- a/softmmu/physmem.c >>>>> +++ b/softmmu/physmem.c >>>>> @@ -3456,6 +3456,24 @@ int ram_block_discard_range(RAMBlock *rb, >>>>> uint64_t start, size_t length) >>>>> * so a userfault will trigger. >>>>> */ >>>>> #ifdef CONFIG_FALLOCATE_PUNCH_HOLE >>>>> + /* >>>>> + * We'll discard data from the actual file, even though >>>>> we only >>>>> + * have a MAP_PRIVATE mapping, possibly messing with >>>>> other >>>>> + * MAP_PRIVATE/MAP_SHARED mappings. There is no easy >>>>> way to >>>>> + * change that behavior whithout violating the promised >>>>> + * semantics of ram_block_discard_range(). >>>>> + * >>>>> + * Only warn, because it work as long as nobody else >>>>> uses that >>>>> + * file. >>>>> + */ >>>>> + if (!qemu_ram_is_shared(rb)) { >>>>> + warn_report_once("ram_block_discard_range: >>>>> Discarding RAM" >>>>> + " in private file mappings is >>>>> possibly" >>>>> + " dangerous, because it will modify >>>>> the" >>>>> + " underlying file and will affect >>>>> other" >>>>> + " users of the file"); >>>>> + } >>>>> + >>>> >>>> TDX has two types of memory backend for each RAM, shared memory and >>>> private memory. Private memory is serviced by guest memfd and shared >>>> memory can also be backed with a fd. >>>> >>>> At any time, only one type needs to be valid, which means the opposite >>>> can be discarded. We do implement the memory discard when TDX converts >>>> the memory[1]. It will trigger this warning 100% because by default the >>>> guest memfd is not mapped as shared (MAP_SHARED). >>> >>> If MAP_PRIVATE is not involved and you are taking the pages directly out >>> of the memfd, you should mark that thing as shared. >> >> Is it the general rule of Linux? Of just the rule of QEMU memory discard? >> > > MAP_SHARED vs. MAP_PRIVATE is a common UNIX principle, and that's what > this flag and the check is about. > > From mmap(2) > > MAP_SHARED: Share this mapping. Updates to the mapping are visible to > other processes mapping the same region, and (in the case of file-backed > mappings) are carried through to the underlying file. > > MAP_PRIVATE: Create a private copy-on-write mapping. Updates to the > mapping are not visible to other processes mapping the same file, and > are not carried through to the underlying file. It is unspecified > whether changes made to the file after the mmap() call are visible in > the mapped region. > > For your purpose (no mmap() at all), we behave like MAP_SHARED -- as if > nothing special is done. No Copy-on-write, no anonymous memory. > >>> Anonymous memory is never involved. >> >> Could you please elaborate more on this? What do you want to express >> here regrading anonymous memory? (Sorry that I'm newbie for mmap stuff) > > Anonymous memory is memory that is private to a specific process, and > (see MAP_PRIVATE) modifications remain private to the process and are > not reflected to the file. > > If you have a MAP_PRIVATE file mapping and write to a virtual memory > location, you'll get a process-private copy of the underlying pagecache > page. that's what we call anonymous memory, because it does not belong > to a specific file. fallocate(punch) would not free up that anonymous > memory. For guest memfd, it does implement kvm_gmem_fallocate as .fallocate() callback, which calls truncate_inode_pages_range() [*]. I'm not sure if it frees up the memory. I need to learn it. [*] https://github.com/kvm-x86/linux/blob/911b515af3ec5f53992b9cc162cf7d3893c2fbe2/virt/kvm/guest_memfd.c#L147C73-L147C73 >> >>> >>> "Private memory" is only private from the guest POV, not from a mmap() >>> point of view. >>> >>> Two different concepts of "private". >>> >>>> >>>> Simply remove the warning will fail the purpose of this patch. The >>>> other >>>> option is to skip the warning for TDX case, which looks vary hacky. Do >>>> you have any idea? >>> >>> For TDX, all memory backends / RAMBlocks should be marked as "shared", >>> and you should fail if that is not provided by the user. >> >> As I asked above, I want to understand the logic clearly. Is mapped as >> shared is a must to support the memory discard? i.e., if we want to >> support memory discard after memory type change, then the memory must be >> mapped with MAP_SHARED? > > MAP_PIRVATE means that it's not sufficient to only fallocate(punch) the > fd to free up all memory for a virtual address, because there might be > anonymous memory in a private mapping that has to be freed up using > MADV_DONTNEED. I can understand this. But it seems unrelated to my question: Is mapped as shared is a must to support the memory discard? e.g., if use below parameters to specify the RAM for a VM -object memory-backend-memfd,id=mem0,size=2G \ -machine memory-backend=mem0 since not specifying "share" property, the ram_block doesn't have RAM_SHARED set. If want to discard some range of this memfd, it triggers the warning. Is this warning expected? I know it is not a possible case for current QEMU, but it's true for future TDX VM. TDX VM can have memory-backend-memfd as the backend for shared memory and kvm gmem memfd for private memory. At any time, for any memory range, only one type is valid, thus the range in opposite memfd can be fallocated. Here I get your message as "the ramblock needs to have RAM_SHARED flag to allow the fallocate of the memfd". This is what I don't understand. That's why this functions handles both cases differently, > and warns if something possibly nasty is being done. >
On 18.10.23 18:27, Xiaoyao Li wrote: > On 10/18/2023 5:26 PM, David Hildenbrand wrote: >> On 18.10.23 11:02, Xiaoyao Li wrote: >>> On 10/18/2023 3:42 PM, David Hildenbrand wrote: >>>> On 18.10.23 05:02, Xiaoyao Li wrote: >>>>> David, >>>>> >>>>> On 7/6/2023 3:56 PM, David Hildenbrand wrote: >>>>>> ram_block_discard_range() cannot possibly do the right thing in >>>>>> MAP_PRIVATE file mappings in the general case. >>>>>> >>>>>> To achieve the documented semantics, we also have to punch a hole into >>>>>> the file, possibly messing with other MAP_PRIVATE/MAP_SHARED mappings >>>>>> of such a file. >>>>>> >>>>>> For example, using VM templating -- see commit b17fbbe55cba >>>>>> ("migration: >>>>>> allow private destination ram with x-ignore-shared") -- in >>>>>> combination with >>>>>> any mechanism that relies on discarding of RAM is problematic. This >>>>>> includes: >>>>>> * Postcopy live migration >>>>>> * virtio-balloon inflation/deflation or free-page-reporting >>>>>> * virtio-mem >>>>>> >>>>>> So at least warn that there is something possibly dangerous is >>>>>> going on >>>>>> when using ram_block_discard_range() in these cases. >>>>>> >>>>>> Acked-by: Peter Xu <peterx@redhat.com> >>>>>> Tested-by: Mario Casquero <mcasquer@redhat.com> >>>>>> Signed-off-by: David Hildenbrand <david@redhat.com> >>>>>> --- >>>>>> softmmu/physmem.c | 18 ++++++++++++++++++ >>>>>> 1 file changed, 18 insertions(+) >>>>>> >>>>>> diff --git a/softmmu/physmem.c b/softmmu/physmem.c >>>>>> index bda475a719..4ee157bda4 100644 >>>>>> --- a/softmmu/physmem.c >>>>>> +++ b/softmmu/physmem.c >>>>>> @@ -3456,6 +3456,24 @@ int ram_block_discard_range(RAMBlock *rb, >>>>>> uint64_t start, size_t length) >>>>>> * so a userfault will trigger. >>>>>> */ >>>>>> #ifdef CONFIG_FALLOCATE_PUNCH_HOLE >>>>>> + /* >>>>>> + * We'll discard data from the actual file, even though >>>>>> we only >>>>>> + * have a MAP_PRIVATE mapping, possibly messing with >>>>>> other >>>>>> + * MAP_PRIVATE/MAP_SHARED mappings. There is no easy >>>>>> way to >>>>>> + * change that behavior whithout violating the promised >>>>>> + * semantics of ram_block_discard_range(). >>>>>> + * >>>>>> + * Only warn, because it work as long as nobody else >>>>>> uses that >>>>>> + * file. >>>>>> + */ >>>>>> + if (!qemu_ram_is_shared(rb)) { >>>>>> + warn_report_once("ram_block_discard_range: >>>>>> Discarding RAM" >>>>>> + " in private file mappings is >>>>>> possibly" >>>>>> + " dangerous, because it will modify >>>>>> the" >>>>>> + " underlying file and will affect >>>>>> other" >>>>>> + " users of the file"); >>>>>> + } >>>>>> + >>>>> >>>>> TDX has two types of memory backend for each RAM, shared memory and >>>>> private memory. Private memory is serviced by guest memfd and shared >>>>> memory can also be backed with a fd. >>>>> >>>>> At any time, only one type needs to be valid, which means the opposite >>>>> can be discarded. We do implement the memory discard when TDX converts >>>>> the memory[1]. It will trigger this warning 100% because by default the >>>>> guest memfd is not mapped as shared (MAP_SHARED). >>>> >>>> If MAP_PRIVATE is not involved and you are taking the pages directly out >>>> of the memfd, you should mark that thing as shared. >>> >>> Is it the general rule of Linux? Of just the rule of QEMU memory discard? >>> >> >> MAP_SHARED vs. MAP_PRIVATE is a common UNIX principle, and that's what >> this flag and the check is about. >> >> From mmap(2) >> >> MAP_SHARED: Share this mapping. Updates to the mapping are visible to >> other processes mapping the same region, and (in the case of file-backed >> mappings) are carried through to the underlying file. >> >> MAP_PRIVATE: Create a private copy-on-write mapping. Updates to the >> mapping are not visible to other processes mapping the same file, and >> are not carried through to the underlying file. It is unspecified >> whether changes made to the file after the mmap() call are visible in >> the mapped region. >> >> For your purpose (no mmap() at all), we behave like MAP_SHARED -- as if >> nothing special is done. No Copy-on-write, no anonymous memory. >> >>>> Anonymous memory is never involved. >>> >>> Could you please elaborate more on this? What do you want to express >>> here regrading anonymous memory? (Sorry that I'm newbie for mmap stuff) >> >> Anonymous memory is memory that is private to a specific process, and >> (see MAP_PRIVATE) modifications remain private to the process and are >> not reflected to the file. >> >> If you have a MAP_PRIVATE file mapping and write to a virtual memory >> location, you'll get a process-private copy of the underlying pagecache >> page. that's what we call anonymous memory, because it does not belong >> to a specific file. fallocate(punch) would not free up that anonymous >> memory. > > For guest memfd, it does implement kvm_gmem_fallocate as .fallocate() > callback, which calls truncate_inode_pages_range() [*]. > > I'm not sure if it frees up the memory. I need to learn it. > > [*] > https://github.com/kvm-x86/linux/blob/911b515af3ec5f53992b9cc162cf7d3893c2fbe2/virt/kvm/guest_memfd.c#L147C73-L147C73 > >>> >>>> >>>> "Private memory" is only private from the guest POV, not from a mmap() >>>> point of view. >>>> >>>> Two different concepts of "private". >>>> >>>>> >>>>> Simply remove the warning will fail the purpose of this patch. The >>>>> other >>>>> option is to skip the warning for TDX case, which looks vary hacky. Do >>>>> you have any idea? >>>> >>>> For TDX, all memory backends / RAMBlocks should be marked as "shared", >>>> and you should fail if that is not provided by the user. >>> >>> As I asked above, I want to understand the logic clearly. Is mapped as >>> shared is a must to support the memory discard? i.e., if we want to >>> support memory discard after memory type change, then the memory must be >>> mapped with MAP_SHARED? >> >> MAP_PIRVATE means that it's not sufficient to only fallocate(punch) the >> fd to free up all memory for a virtual address, because there might be >> anonymous memory in a private mapping that has to be freed up using >> MADV_DONTNEED. > > I can understand this. But it seems unrelated to my question: Is mapped > as shared is a must to support the memory discard? Sorry, I don't quite get what you are asking that I haven't answered yet. Let's talk about the issue you are seeing below. > > e.g., if use below parameters to specify the RAM for a VM > > -object memory-backend-memfd,id=mem0,size=2G \ > -machine memory-backend=mem0 > > since not specifying "share" property, the ram_block doesn't have > RAM_SHARED set. If want to discard some range of this memfd, it triggers > the warning. Is this warning expected? That should not be the case. See "memfd_backend_instance_init" where we set share=true. In memfd_backend_memory_alloc(), we set RAM_SHARED. We only default to share=off for memory-backend-file (well, and memory-backend-ram). So are you sure you get this error message in the configuration you are describing here? > > I know it is not a possible case for current QEMU, but it's true for > future TDX VM. TDX VM can have memory-backend-memfd as the backend for > shared memory and kvm gmem memfd for private memory. At any time, for > any memory range, only one type is valid, thus the range in opposite > memfd can be fallocated. Right. > > Here I get your message as "the ramblock needs to have RAM_SHARED flag > to allow the fallocate of the memfd". This is what I don't understand. The problem I am seeing is that either (a) Someone explicitly sets share=off for some reason for memory-backend-memfd, triggering the warning. (b) We somehow lose RAM_SHARED in above configuration, which would be bad and trigger the warning. Can you make sure that (a) is not the case?
On 10/19/2023 4:26 PM, David Hildenbrand wrote: > On 18.10.23 18:27, Xiaoyao Li wrote: >> On 10/18/2023 5:26 PM, David Hildenbrand wrote: >>> On 18.10.23 11:02, Xiaoyao Li wrote: >>>> On 10/18/2023 3:42 PM, David Hildenbrand wrote: >>>>> On 18.10.23 05:02, Xiaoyao Li wrote: >>>>>> David, >>>>>> >>>>>> On 7/6/2023 3:56 PM, David Hildenbrand wrote: >>>>>>> ram_block_discard_range() cannot possibly do the right thing in >>>>>>> MAP_PRIVATE file mappings in the general case. >>>>>>> >>>>>>> To achieve the documented semantics, we also have to punch a hole >>>>>>> into >>>>>>> the file, possibly messing with other MAP_PRIVATE/MAP_SHARED >>>>>>> mappings >>>>>>> of such a file. >>>>>>> >>>>>>> For example, using VM templating -- see commit b17fbbe55cba >>>>>>> ("migration: >>>>>>> allow private destination ram with x-ignore-shared") -- in >>>>>>> combination with >>>>>>> any mechanism that relies on discarding of RAM is problematic. This >>>>>>> includes: >>>>>>> * Postcopy live migration >>>>>>> * virtio-balloon inflation/deflation or free-page-reporting >>>>>>> * virtio-mem >>>>>>> >>>>>>> So at least warn that there is something possibly dangerous is >>>>>>> going on >>>>>>> when using ram_block_discard_range() in these cases. >>>>>>> >>>>>>> Acked-by: Peter Xu <peterx@redhat.com> >>>>>>> Tested-by: Mario Casquero <mcasquer@redhat.com> >>>>>>> Signed-off-by: David Hildenbrand <david@redhat.com> >>>>>>> --- >>>>>>> softmmu/physmem.c | 18 ++++++++++++++++++ >>>>>>> 1 file changed, 18 insertions(+) >>>>>>> >>>>>>> diff --git a/softmmu/physmem.c b/softmmu/physmem.c >>>>>>> index bda475a719..4ee157bda4 100644 >>>>>>> --- a/softmmu/physmem.c >>>>>>> +++ b/softmmu/physmem.c >>>>>>> @@ -3456,6 +3456,24 @@ int ram_block_discard_range(RAMBlock *rb, >>>>>>> uint64_t start, size_t length) >>>>>>> * so a userfault will trigger. >>>>>>> */ >>>>>>> #ifdef CONFIG_FALLOCATE_PUNCH_HOLE >>>>>>> + /* >>>>>>> + * We'll discard data from the actual file, even though >>>>>>> we only >>>>>>> + * have a MAP_PRIVATE mapping, possibly messing with >>>>>>> other >>>>>>> + * MAP_PRIVATE/MAP_SHARED mappings. There is no easy >>>>>>> way to >>>>>>> + * change that behavior whithout violating the promised >>>>>>> + * semantics of ram_block_discard_range(). >>>>>>> + * >>>>>>> + * Only warn, because it work as long as nobody else >>>>>>> uses that >>>>>>> + * file. >>>>>>> + */ >>>>>>> + if (!qemu_ram_is_shared(rb)) { >>>>>>> + warn_report_once("ram_block_discard_range: >>>>>>> Discarding RAM" >>>>>>> + " in private file mappings is >>>>>>> possibly" >>>>>>> + " dangerous, because it will >>>>>>> modify >>>>>>> the" >>>>>>> + " underlying file and will affect >>>>>>> other" >>>>>>> + " users of the file"); >>>>>>> + } >>>>>>> + >>>>>> >>>>>> TDX has two types of memory backend for each RAM, shared memory and >>>>>> private memory. Private memory is serviced by guest memfd and shared >>>>>> memory can also be backed with a fd. >>>>>> >>>>>> At any time, only one type needs to be valid, which means the >>>>>> opposite >>>>>> can be discarded. We do implement the memory discard when TDX >>>>>> converts >>>>>> the memory[1]. It will trigger this warning 100% because by >>>>>> default the >>>>>> guest memfd is not mapped as shared (MAP_SHARED). >>>>> >>>>> If MAP_PRIVATE is not involved and you are taking the pages >>>>> directly out >>>>> of the memfd, you should mark that thing as shared. >>>> >>>> Is it the general rule of Linux? Of just the rule of QEMU memory >>>> discard? >>>> >>> >>> MAP_SHARED vs. MAP_PRIVATE is a common UNIX principle, and that's what >>> this flag and the check is about. >>> >>> From mmap(2) >>> >>> MAP_SHARED: Share this mapping. Updates to the mapping are visible to >>> other processes mapping the same region, and (in the case of file-backed >>> mappings) are carried through to the underlying file. >>> >>> MAP_PRIVATE: Create a private copy-on-write mapping. Updates to the >>> mapping are not visible to other processes mapping the same file, and >>> are not carried through to the underlying file. It is unspecified >>> whether changes made to the file after the mmap() call are visible in >>> the mapped region. >>> >>> For your purpose (no mmap() at all), we behave like MAP_SHARED -- as if >>> nothing special is done. No Copy-on-write, no anonymous memory. >>> >>>>> Anonymous memory is never involved. >>>> >>>> Could you please elaborate more on this? What do you want to express >>>> here regrading anonymous memory? (Sorry that I'm newbie for mmap stuff) >>> >>> Anonymous memory is memory that is private to a specific process, and >>> (see MAP_PRIVATE) modifications remain private to the process and are >>> not reflected to the file. >>> >>> If you have a MAP_PRIVATE file mapping and write to a virtual memory >>> location, you'll get a process-private copy of the underlying pagecache >>> page. that's what we call anonymous memory, because it does not belong >>> to a specific file. fallocate(punch) would not free up that anonymous >>> memory. >> >> For guest memfd, it does implement kvm_gmem_fallocate as .fallocate() >> callback, which calls truncate_inode_pages_range() [*]. >> >> I'm not sure if it frees up the memory. I need to learn it. >> >> [*] >> https://github.com/kvm-x86/linux/blob/911b515af3ec5f53992b9cc162cf7d3893c2fbe2/virt/kvm/guest_memfd.c#L147C73-L147C73 >> >>>> >>>>> >>>>> "Private memory" is only private from the guest POV, not from a mmap() >>>>> point of view. >>>>> >>>>> Two different concepts of "private". >>>>> >>>>>> >>>>>> Simply remove the warning will fail the purpose of this patch. The >>>>>> other >>>>>> option is to skip the warning for TDX case, which looks vary >>>>>> hacky. Do >>>>>> you have any idea? >>>>> >>>>> For TDX, all memory backends / RAMBlocks should be marked as "shared", >>>>> and you should fail if that is not provided by the user. >>>> >>>> As I asked above, I want to understand the logic clearly. Is mapped as >>>> shared is a must to support the memory discard? i.e., if we want to >>>> support memory discard after memory type change, then the memory >>>> must be >>>> mapped with MAP_SHARED? >>> >>> MAP_PIRVATE means that it's not sufficient to only fallocate(punch) the >>> fd to free up all memory for a virtual address, because there might be >>> anonymous memory in a private mapping that has to be freed up using >>> MADV_DONTNEED. >> >> I can understand this. But it seems unrelated to my question: Is mapped >> as shared is a must to support the memory discard? > > Sorry, I don't quite get what you are asking that I haven't answered > yet. Let's talk about the issue you are seeing below. > >> >> e.g., if use below parameters to specify the RAM for a VM >> >> -object memory-backend-memfd,id=mem0,size=2G \ >> -machine memory-backend=mem0 >> >> since not specifying "share" property, the ram_block doesn't have >> RAM_SHARED set. If want to discard some range of this memfd, it triggers >> the warning. Is this warning expected? > > That should not be the case. See "memfd_backend_instance_init" where we > set share=true. In memfd_backend_memory_alloc(), we set RAM_SHARED. > > We only default to share=off for memory-backend-file (well, and > memory-backend-ram). > > So are you sure you get this error message in the configuration you are > describing here? Sorry, I made an mistake. I was using "-object memory-backend-ram,id=mem0,size=2G" instead of "memory-backend-memfd". yes, when using "memory-backend-memfd" as the backend for TDX shared memory, it doesn't trigger the warning because memfd_backend_instance_init() set "share" to true. When using "memory-backend-ram" as the backend for TDX shared memory, the warning is triggered converting memory from private (kvm gmem) to shared (backend-ram). In this case, there is a valid fd (kvm gmem fd), so it goes to the path of need_fallocate. However, qemu_ram_is_shared(rb) returns false because "memory-backend-ram" doesn't have "share" default on. Thus, the warning is triggered. It seems I need figure out a more proper solution to refactor the ram_block_discard_range(), to make it applicable for kvm gmem discard case. >> >> I know it is not a possible case for current QEMU, but it's true for >> future TDX VM. TDX VM can have memory-backend-memfd as the backend for >> shared memory and kvm gmem memfd for private memory. At any time, for >> any memory range, only one type is valid, thus the range in opposite >> memfd can be fallocated. > > Right. > >> >> Here I get your message as "the ramblock needs to have RAM_SHARED flag >> to allow the fallocate of the memfd". This is what I don't understand. > > The problem I am seeing is that either > > (a) Someone explicitly sets share=off for some reason for > memory-backend-memfd, triggering the warning. > > (b) We somehow lose RAM_SHARED in above configuration, which would be > bad and trigger the warning. > > Can you make sure that (a) is not the case? Above reply answers it. Sorry for it.
On 19.10.23 11:26, Xiaoyao Li wrote: > On 10/19/2023 4:26 PM, David Hildenbrand wrote: >> On 18.10.23 18:27, Xiaoyao Li wrote: >>> On 10/18/2023 5:26 PM, David Hildenbrand wrote: >>>> On 18.10.23 11:02, Xiaoyao Li wrote: >>>>> On 10/18/2023 3:42 PM, David Hildenbrand wrote: >>>>>> On 18.10.23 05:02, Xiaoyao Li wrote: >>>>>>> David, >>>>>>> >>>>>>> On 7/6/2023 3:56 PM, David Hildenbrand wrote: >>>>>>>> ram_block_discard_range() cannot possibly do the right thing in >>>>>>>> MAP_PRIVATE file mappings in the general case. >>>>>>>> >>>>>>>> To achieve the documented semantics, we also have to punch a hole >>>>>>>> into >>>>>>>> the file, possibly messing with other MAP_PRIVATE/MAP_SHARED >>>>>>>> mappings >>>>>>>> of such a file. >>>>>>>> >>>>>>>> For example, using VM templating -- see commit b17fbbe55cba >>>>>>>> ("migration: >>>>>>>> allow private destination ram with x-ignore-shared") -- in >>>>>>>> combination with >>>>>>>> any mechanism that relies on discarding of RAM is problematic. This >>>>>>>> includes: >>>>>>>> * Postcopy live migration >>>>>>>> * virtio-balloon inflation/deflation or free-page-reporting >>>>>>>> * virtio-mem >>>>>>>> >>>>>>>> So at least warn that there is something possibly dangerous is >>>>>>>> going on >>>>>>>> when using ram_block_discard_range() in these cases. >>>>>>>> >>>>>>>> Acked-by: Peter Xu <peterx@redhat.com> >>>>>>>> Tested-by: Mario Casquero <mcasquer@redhat.com> >>>>>>>> Signed-off-by: David Hildenbrand <david@redhat.com> >>>>>>>> --- >>>>>>>> softmmu/physmem.c | 18 ++++++++++++++++++ >>>>>>>> 1 file changed, 18 insertions(+) >>>>>>>> >>>>>>>> diff --git a/softmmu/physmem.c b/softmmu/physmem.c >>>>>>>> index bda475a719..4ee157bda4 100644 >>>>>>>> --- a/softmmu/physmem.c >>>>>>>> +++ b/softmmu/physmem.c >>>>>>>> @@ -3456,6 +3456,24 @@ int ram_block_discard_range(RAMBlock *rb, >>>>>>>> uint64_t start, size_t length) >>>>>>>> * so a userfault will trigger. >>>>>>>> */ >>>>>>>> #ifdef CONFIG_FALLOCATE_PUNCH_HOLE >>>>>>>> + /* >>>>>>>> + * We'll discard data from the actual file, even though >>>>>>>> we only >>>>>>>> + * have a MAP_PRIVATE mapping, possibly messing with >>>>>>>> other >>>>>>>> + * MAP_PRIVATE/MAP_SHARED mappings. There is no easy >>>>>>>> way to >>>>>>>> + * change that behavior whithout violating the promised >>>>>>>> + * semantics of ram_block_discard_range(). >>>>>>>> + * >>>>>>>> + * Only warn, because it work as long as nobody else >>>>>>>> uses that >>>>>>>> + * file. >>>>>>>> + */ >>>>>>>> + if (!qemu_ram_is_shared(rb)) { >>>>>>>> + warn_report_once("ram_block_discard_range: >>>>>>>> Discarding RAM" >>>>>>>> + " in private file mappings is >>>>>>>> possibly" >>>>>>>> + " dangerous, because it will >>>>>>>> modify >>>>>>>> the" >>>>>>>> + " underlying file and will affect >>>>>>>> other" >>>>>>>> + " users of the file"); >>>>>>>> + } >>>>>>>> + >>>>>>> >>>>>>> TDX has two types of memory backend for each RAM, shared memory and >>>>>>> private memory. Private memory is serviced by guest memfd and shared >>>>>>> memory can also be backed with a fd. >>>>>>> >>>>>>> At any time, only one type needs to be valid, which means the >>>>>>> opposite >>>>>>> can be discarded. We do implement the memory discard when TDX >>>>>>> converts >>>>>>> the memory[1]. It will trigger this warning 100% because by >>>>>>> default the >>>>>>> guest memfd is not mapped as shared (MAP_SHARED). >>>>>> >>>>>> If MAP_PRIVATE is not involved and you are taking the pages >>>>>> directly out >>>>>> of the memfd, you should mark that thing as shared. >>>>> >>>>> Is it the general rule of Linux? Of just the rule of QEMU memory >>>>> discard? >>>>> >>>> >>>> MAP_SHARED vs. MAP_PRIVATE is a common UNIX principle, and that's what >>>> this flag and the check is about. >>>> >>>> From mmap(2) >>>> >>>> MAP_SHARED: Share this mapping. Updates to the mapping are visible to >>>> other processes mapping the same region, and (in the case of file-backed >>>> mappings) are carried through to the underlying file. >>>> >>>> MAP_PRIVATE: Create a private copy-on-write mapping. Updates to the >>>> mapping are not visible to other processes mapping the same file, and >>>> are not carried through to the underlying file. It is unspecified >>>> whether changes made to the file after the mmap() call are visible in >>>> the mapped region. >>>> >>>> For your purpose (no mmap() at all), we behave like MAP_SHARED -- as if >>>> nothing special is done. No Copy-on-write, no anonymous memory. >>>> >>>>>> Anonymous memory is never involved. >>>>> >>>>> Could you please elaborate more on this? What do you want to express >>>>> here regrading anonymous memory? (Sorry that I'm newbie for mmap stuff) >>>> >>>> Anonymous memory is memory that is private to a specific process, and >>>> (see MAP_PRIVATE) modifications remain private to the process and are >>>> not reflected to the file. >>>> >>>> If you have a MAP_PRIVATE file mapping and write to a virtual memory >>>> location, you'll get a process-private copy of the underlying pagecache >>>> page. that's what we call anonymous memory, because it does not belong >>>> to a specific file. fallocate(punch) would not free up that anonymous >>>> memory. >>> >>> For guest memfd, it does implement kvm_gmem_fallocate as .fallocate() >>> callback, which calls truncate_inode_pages_range() [*]. >>> >>> I'm not sure if it frees up the memory. I need to learn it. >>> >>> [*] >>> https://github.com/kvm-x86/linux/blob/911b515af3ec5f53992b9cc162cf7d3893c2fbe2/virt/kvm/guest_memfd.c#L147C73-L147C73 >>> >>>>> >>>>>> >>>>>> "Private memory" is only private from the guest POV, not from a mmap() >>>>>> point of view. >>>>>> >>>>>> Two different concepts of "private". >>>>>> >>>>>>> >>>>>>> Simply remove the warning will fail the purpose of this patch. The >>>>>>> other >>>>>>> option is to skip the warning for TDX case, which looks vary >>>>>>> hacky. Do >>>>>>> you have any idea? >>>>>> >>>>>> For TDX, all memory backends / RAMBlocks should be marked as "shared", >>>>>> and you should fail if that is not provided by the user. >>>>> >>>>> As I asked above, I want to understand the logic clearly. Is mapped as >>>>> shared is a must to support the memory discard? i.e., if we want to >>>>> support memory discard after memory type change, then the memory >>>>> must be >>>>> mapped with MAP_SHARED? >>>> >>>> MAP_PIRVATE means that it's not sufficient to only fallocate(punch) the >>>> fd to free up all memory for a virtual address, because there might be >>>> anonymous memory in a private mapping that has to be freed up using >>>> MADV_DONTNEED. >>> >>> I can understand this. But it seems unrelated to my question: Is mapped >>> as shared is a must to support the memory discard? >> >> Sorry, I don't quite get what you are asking that I haven't answered >> yet. Let's talk about the issue you are seeing below. >> >>> >>> e.g., if use below parameters to specify the RAM for a VM >>> >>> -object memory-backend-memfd,id=mem0,size=2G \ >>> -machine memory-backend=mem0 >>> >>> since not specifying "share" property, the ram_block doesn't have >>> RAM_SHARED set. If want to discard some range of this memfd, it triggers >>> the warning. Is this warning expected? >> >> That should not be the case. See "memfd_backend_instance_init" where we >> set share=true. In memfd_backend_memory_alloc(), we set RAM_SHARED. >> >> We only default to share=off for memory-backend-file (well, and >> memory-backend-ram). >> >> So are you sure you get this error message in the configuration you are >> describing here? > > Sorry, I made an mistake. I was using "-object > memory-backend-ram,id=mem0,size=2G" instead of "memory-backend-memfd". > > yes, when using "memory-backend-memfd" as the backend for TDX shared > memory, it doesn't trigger the warning because > memfd_backend_instance_init() set "share" to true. > > When using "memory-backend-ram" as the backend for TDX shared memory, > the warning is triggered converting memory from private (kvm gmem) to > shared (backend-ram). In this case, there is a valid fd (kvm gmem fd), > so it goes to the path of need_fallocate. However, > qemu_ram_is_shared(rb) returns false because "memory-backend-ram" > doesn't have "share" default on. Thus, the warning is triggered. > > It seems I need figure out a more proper solution to refactor the > ram_block_discard_range(), to make it applicable for kvm gmem discard case. You should probably completely ignore any ramblock flags when fallocate(punch) the kvm_gmem_fd. kvm_gmem_fd is a rather special "secondary storage that's never mapped", independent of the ordinary RAMBlock memory.
diff --git a/softmmu/physmem.c b/softmmu/physmem.c index bda475a719..4ee157bda4 100644 --- a/softmmu/physmem.c +++ b/softmmu/physmem.c @@ -3456,6 +3456,24 @@ int ram_block_discard_range(RAMBlock *rb, uint64_t start, size_t length) * so a userfault will trigger. */ #ifdef CONFIG_FALLOCATE_PUNCH_HOLE + /* + * We'll discard data from the actual file, even though we only + * have a MAP_PRIVATE mapping, possibly messing with other + * MAP_PRIVATE/MAP_SHARED mappings. There is no easy way to + * change that behavior whithout violating the promised + * semantics of ram_block_discard_range(). + * + * Only warn, because it work as long as nobody else uses that + * file. + */ + if (!qemu_ram_is_shared(rb)) { + warn_report_once("ram_block_discard_range: Discarding RAM" + " in private file mappings is possibly" + " dangerous, because it will modify the" + " underlying file and will affect other" + " users of the file"); + } + ret = fallocate(rb->fd, FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE, start, length); if (ret) {
ram_block_discard_range() cannot possibly do the right thing in MAP_PRIVATE file mappings in the general case. To achieve the documented semantics, we also have to punch a hole into the file, possibly messing with other MAP_PRIVATE/MAP_SHARED mappings of such a file. For example, using VM templating -- see commit b17fbbe55cba ("migration: allow private destination ram with x-ignore-shared") -- in combination with any mechanism that relies on discarding of RAM is problematic. This includes: * Postcopy live migration * virtio-balloon inflation/deflation or free-page-reporting * virtio-mem So at least warn that there is something possibly dangerous is going on when using ram_block_discard_range() in these cases. Acked-by: Peter Xu <peterx@redhat.com> Tested-by: Mario Casquero <mcasquer@redhat.com> Signed-off-by: David Hildenbrand <david@redhat.com> --- softmmu/physmem.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+)