[v2] tap: fix write-after-free and double free of intr_handle

  rte_pmd_tun/tap_probe() allocates pmd->intr_handle in eth_dev_tap_create()
and it should not be freed until rte_pmd_tap_remove() is called.

Inspection of tap_rx_intr_vec_set() shows that the call to
tap_tx_intr_vec_uninstall() was calling rte_intr_instance_free() but
tap_tx_intr_vec_install() can then be immediately called, and this then
uses pmd->intr_handle without it being reallocated.

This commit moves the call of rte_intr_instance_free() from
tap_tx_intr_vec_uninstall() to rte_pmd_tap_remove().

Fixes: d61138d4f0e2 ("drivers: remove direct access to interrupt handle")

Changes in v2:
  Move rte_intr_instance_free() from tap_rx_intr_vec_uninstall()
  to tap_dev_close().

Signed-off-by: Quentin Armitage <quentin@armitage.org.uk>
---
 drivers/net/tap/rte_eth_tap.c | 2 ++
 drivers/net/tap/tap_intr.c    | 2 --
 2 files changed, 2 insertions(+), 2 deletions(-)
  

On Tue, May 3, 2022 at 8:23 PM Quentin Armitage <quentin@armitage.org.uk> wrote:
>
> rte_pmd_tun/tap_probe() allocates pmd->intr_handle in eth_dev_tap_create()
> and it should not be freed until rte_pmd_tap_remove() is called.
>
> Inspection of tap_rx_intr_vec_set() shows that the call to
> tap_tx_intr_vec_uninstall() was calling rte_intr_instance_free() but
> tap_tx_intr_vec_install() can then be immediately called, and this then
> uses pmd->intr_handle without it being reallocated.
>
> This commit moves the call of rte_intr_instance_free() from
> tap_tx_intr_vec_uninstall() to rte_pmd_tap_remove().
>
> Fixes: d61138d4f0e2 ("drivers: remove direct access to interrupt handle")

Cc: stable@dpdk.org

https://doc.dpdk.org/guides/contributing/patches.html#patch-for-stable-releases
The reason is that backport scripts look for a "Cc: stable@dpdk.org"
in the commitlog itself.

(no need for a v3 just for this, it can be fixed when applying)

>
> Changes in v2:
>   Move rte_intr_instance_free() from tap_rx_intr_vec_uninstall()
>   to tap_dev_close().

Nit: revisions changelog should be added as annotations (i.e. put
after the --- after the commitlog).


>
> Signed-off-by: Quentin Armitage <quentin@armitage.org.uk>

I did not test the change, but the fix lgtm.
The CI failure from UNH is a false positive.

Reviewed-by: David Marchand <david.marchand@redhat.com>
  

On 5/4/22 14:17, David Marchand wrote:
> On Tue, May 3, 2022 at 8:23 PM Quentin Armitage <quentin@armitage.org.uk> wrote:
>>
>> rte_pmd_tun/tap_probe() allocates pmd->intr_handle in eth_dev_tap_create()
>> and it should not be freed until rte_pmd_tap_remove() is called.
>>
>> Inspection of tap_rx_intr_vec_set() shows that the call to
>> tap_tx_intr_vec_uninstall() was calling rte_intr_instance_free() but
>> tap_tx_intr_vec_install() can then be immediately called, and this then
>> uses pmd->intr_handle without it being reallocated.
>>
>> This commit moves the call of rte_intr_instance_free() from
>> tap_tx_intr_vec_uninstall() to rte_pmd_tap_remove().
>>
>> Fixes: d61138d4f0e2 ("drivers: remove direct access to interrupt handle")
> 
> Cc: stable@dpdk.org
> 
> https://doc.dpdk.org/guides/contributing/patches.html#patch-for-stable-releases
> The reason is that backport scripts look for a "Cc: stable@dpdk.org"
> in the commitlog itself.
> 
> (no need for a v3 just for this, it can be fixed when applying)
> 
>>
>> Changes in v2:
>>    Move rte_intr_instance_free() from tap_rx_intr_vec_uninstall()
>>    to tap_dev_close().
> 
> Nit: revisions changelog should be added as annotations (i.e. put
> after the --- after the commitlog).
> 
> 
>>
>> Signed-off-by: Quentin Armitage <quentin@armitage.org.uk>
> 
> I did not test the change, but the fix lgtm.
> The CI failure from UNH is a false positive.
> 
> Reviewed-by: David Marchand <david.marchand@redhat.com>
> 
> 

Applied with minor fixes in summary and description, thanks.