Re: [PATCH v2] dmaengine: idxd: fix use-after-free in idxd_free() and idxd_alloc() error paths

From: Frank Li

Date: Fri Oct 16 2026 - 11:32:15 EST


On Thu, Jun 18, 2026 at 02:32:54PM +0300, Bogdan Codres (Wind River) wrote:
> From: Bogdan Codres <bogdan.codres@xxxxxxxxxxxxx>
>
> [ 18.628791] idxd 0000:00:01.0: Device is HALTED!
> [ 18.631447] idxd 0000:00:01.0: Intel(R) IDXD DMA Engine init failed
> [ 18.631450] ------------[ cut here ]------------
> [ 18.631451] ida_free called for id=0 which is not allocated.
> [ 18.631462] WARNING: CPU: 0 PID: 11 at lib/idr.c:525 ida_free+0xd3/0x130
> [ 18.631474] CPU: 0 UID: 0 PID: 11 Comm: kworker/0:1 Not tainted 6.12.0-1-rt-amd64 #1
> [ 18.631477] Hardware name: Dell Inc. PowerEdge XR8720t/0J91KV, BIOS 1.1.3 02/03/2026
> [ 18.631478] Workqueue: events work_for_cpu_fn
> [ 18.631480] RIP: 0010:ida_free+0xd3/0x130
> [ 18.631492] Call Trace:
> [ 18.631494] <TASK>
> [ 18.631495] idxd_pci_probe+0x1b0/0x1860 [idxd]
> [ 18.631506] local_pci_probe+0x43/0xa0
> [ 18.631508] work_for_cpu_fn+0x13/0x20
> [ 18.631510] process_one_work+0x179/0x390
> [ 18.631512] worker_thread+0x237/0x340
> [ 18.631517] kthread+0xc6/0x100
> [ 18.631520] ret_from_fork+0x2d/0x50
> [ 18.631524] ret_from_fork_asm+0x1a/0x30
> [ 18.631526] </TASK>
>
> idxd_free() calls put_device(idxd_confdev(idxd)) which drops the last
> reference and synchronously invokes idxd_conf_device_release(). That
> release callback already frees idxd->opcap_bmap, idxd->id (via
> ida_free), and the idxd structure itself. The subsequent bitmap_free(),
> ida_free(), and kfree() in idxd_free() therefore operate on freed
> memory - a double-free that corrupts the slab allocator.
>
> The same pattern exists in idxd_alloc() at the err_name label where
> put_device() is followed by bitmap_free() fall-through.
>
> Fix both by letting put_device() handle all resource cleanup via the
> release callback, removing the duplicate frees.
>
> Fixes: 90022b3a6981 ("dmaengine: idxd: fix memory leak in error handling path of idxd_pci_probe")
> Fixes: 46a5cca76c76 ("dmaengine: idxd: fix memory leak in error handling path of idxd_alloc")
> Cc: stable@xxxxxxxxxxxxxxx
> Signed-off-by: Bogdan Codres <bogdan.codres@xxxxxxxxxxxxx>
> ---

Avoid post new patch at old thread.

There are more fixes at
https://patchwork.kernel.org/project/linux-dmaengine/list/?series=1184207

similar fix at
https://patchwork.kernel.org/project/linux-dmaengine/patch/20261012-wt-idxd-fixes-v1-2-30a35a445077@xxxxxxxxx/

Anyways, you post early.

Reviewed-by: Frank Li <Frank.Li@xxxxxxx>

> drivers/dma/idxd/init.c | 15 +++++++++++----
> 1 file changed, 11 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/dma/idxd/init.c b/drivers/dma/idxd/init.c
> index 18997f80bdc9..ce4d58740c88 100644
> --- a/drivers/dma/idxd/init.c
> +++ b/drivers/dma/idxd/init.c
> @@ -585,15 +585,18 @@ static void idxd_read_caps(struct idxd_device *idxd)
> idxd->hw.iaa_cap.bits = ioread64(idxd->reg_base + IDXD_IAACAP_OFFSET);
> }
>
> +/*
> + * Release an idxd device that was allocated (device_initialize() was called)
> + * but never successfully registered. put_device() drops the last reference and
> + * triggers idxd_conf_device_release() which frees all resources including the
> + * ida, opcap_bmap, and the idxd structure itself.
> + */
> static void idxd_free(struct idxd_device *idxd)
> {
> if (!idxd)
> return;
>
> put_device(idxd_confdev(idxd));
> - bitmap_free(idxd->opcap_bmap);
> - ida_free(&idxd_ida, idxd->id);
> - kfree(idxd);
> }
>
> static struct idxd_device *idxd_alloc(struct pci_dev *pdev, struct idxd_driver_data *data)
> @@ -633,8 +636,12 @@ static struct idxd_device *idxd_alloc(struct pci_dev *pdev, struct idxd_driver_d
> return idxd;
>
> err_name:
> + /* device_initialize() was called, so put_device() will trigger
> + * idxd_conf_device_release() which frees ida, opcap_bmap, and idxd.
> + * Do not fall through to err_opcap/err_ida.
> + */
> put_device(conf_dev);
> - bitmap_free(idxd->opcap_bmap);
> + return NULL;
> err_opcap:
> ida_free(&idxd_ida, idxd->id);
> err_ida:
> --
> 2.51.0
>