On Thu, 18 Jul 2024 05:07:53 -0400
Yao Xingtao <yaoxt.f...@fujitsu.com> wrote:
> When injecting a new poisoned region through qmp_cxl_inject_poison(),
> the newly injected region should not overlap with existing poisoned
> regions.
>
> The current validation method does not consider the following
> overlapping region:
> ┌───┬───────┬───┐
> │a │ b(a) │a │
> └───┴───────┴───┘
> (a is a newly added region, b is an existing region, and b is a
> subregion of a)
>
> Signed-off-by: Yao Xingtao <yaoxt.f...@fujitsu.com>
Looks correct to me.
Reviewed-by: Jonathan Cameron <jonathan.came...@huwei.com>
I've queued it on my local branch.
I need to put together an updated public one.
No huge rush to queue this up though I think as the effects
are minor.
Jonathan
> ---
> hw/mem/cxl_type3.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c
> index 35ac59883a5b..8e32de327908 100644
> --- a/hw/mem/cxl_type3.c
> +++ b/hw/mem/cxl_type3.c
> @@ -1331,9 +1331,7 @@ void qmp_cxl_inject_poison(const char *path, uint64_t
> start, uint64_t length,
> ct3d = CXL_TYPE3(obj);
>
> QLIST_FOREACH(p, &ct3d->poison_list, node) {
> - if (((start >= p->start) && (start < p->start + p->length)) ||
> - ((start + length > p->start) &&
> - (start + length <= p->start + p->length))) {
> + if ((start < p->start + p->length) && (start + length > p->start)) {
> error_setg(errp,
> "Overlap with existing poisoned region not
> supported");
> return;
