Thread (21 messages) flat view 21 messages, 6 authors, 2016-11-10

Re: [PATCH kernel v4 3/4] vfio/spapr: Reference mm in tce_container

From: Nicholas Piggin <npiggin@gmail.com>
Date: 2016-10-24 07:14:35

On Mon, 24 Oct 2016 17:53:09 +1100
Alexey Kardashevskiy [off-list ref] wrote:
In some situations the userspace memory context may live longer than
the userspace process itself so if we need to do proper memory context
cleanup, we better have tce_container take a reference to mm_struct and
use it later when the process is gone (@current or @current->mm is NULL).

This references mm and stores the pointer in the container; this is done
when a container is just created so checking for !current->mm in other
places becomes pointless.

This replaces current->mm with container->mm everywhere except debug
prints.

This adds a check that current->mm is the same as the one stored in
the container to prevent userspace from making changes to a memory
context of other processes; in order to add this check,
VFIO_CHECK_EXTENSION is moved out of the switch(cmd) as it is
quite special anyway - it is the only ioctl() called when neither
container nor container->mm is initialized.

Signed-off-by: Alexey Kardashevskiy <redacted>
---
Changes:
v4:
* added check for container->mm!=current->mm in tce_iommu_ioctl()
for all ioctls and removed other redundand checks
[...]
quoted hunk ↗ jump to hunk
@@ -326,13 +314,18 @@ static void *tce_iommu_open(unsigned long arg)
 
 	container->v2 = arg == VFIO_SPAPR_TCE_v2_IOMMU;
 
+	/* current->mm cannot be NULL in this context */
+	container->mm = current->mm;
+	atomic_inc(&container->mm->mm_count);
[...]
quoted hunk ↗ jump to hunk
@@ -733,7 +728,13 @@ static long tce_iommu_ioctl(void *iommu_data,
 		}
 
 		return (ret < 0) ? 0 : ret;
+	}
 
+	/* tce_iommu_open() initializes container->mm so it can't be NULL here */
+	if (container->mm != current->mm)
+		return -ESRCH;
+
+	switch (cmd) {
 	case VFIO_IOMMU_SPAPR_TCE_GET_INFO: {
 		struct vfio_iommu_spapr_tce_info info;
 		struct tce_iommu_group *tcegrp;
I think doing the mm checks like this is a great improvement.

Reviewed-by: Nicholas Piggin <npiggin@gmail.com>

Thanks,
Nick
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help