[PATCH net 0/3] There are some bugfix for the HNS3 ethernet driver

[PATCH net 0/3] There are some bugfix for the HNS3 ethernet driver

[Jijie Shao]

There are some bugfix for the HNS3 ethernet driver

Jian Shen (3):
  net: hns3: using the num_tqps in the vf driver to apply for resources
  net: hns3: using the num_tqps to check whether tqp_index is out of
    range when vf get ring info from mbx
  net: hns3: add VLAN id validation before using

 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c   | 3 +++
 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c    | 4 ++--
 drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c | 4 ++--
 3 files changed, 7 insertions(+), 4 deletions(-)

-- 
2.33.0

[PATCH net 1/3] net: hns3: using the num_tqps in the vf driver to apply for resources

[Jijie Shao]

From: Jian Shen <shenjian15@huawei.com>

Currently, hdev->htqp is allocated using hdev->num_tqps, and kinfo->tqp
is allocated using kinfo->num_tqps. However, kinfo->num_tqps is set to
min(new_tqps, hdev->num_tqps);  Therefore, kinfo->num_tqps may be smaller
than hdev->num_tqps, which causes some hdev->htqp[i] to remain
uninitialized in hclgevf_knic_setup().

Thus, this patch allocates hdev->htqp and kinfo->tqp using hdev->num_tqps,
ensuring that the lengths of hdev->htqp and kinfo->tqp are consistent
and that all elements are properly initialized.

Fixes: e2cb1dec9779 ("net: hns3: Add HNS3 VF HCL(Hardware Compatibility Layer) Support")
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
index 8fcf220a120d..70327a73dee3 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
@@ -368,12 +368,12 @@ static int hclgevf_knic_setup(struct hclgevf_dev *hdev)
 	new_tqps = kinfo->rss_size * num_tc;
 	kinfo->num_tqps = min(new_tqps, hdev->num_tqps);
 
-	kinfo->tqp = devm_kcalloc(&hdev->pdev->dev, kinfo->num_tqps,
+	kinfo->tqp = devm_kcalloc(&hdev->pdev->dev, hdev->num_tqps,
 				  sizeof(struct hnae3_queue *), GFP_KERNEL);
 	if (!kinfo->tqp)
 		return -ENOMEM;
 
-	for (i = 0; i < kinfo->num_tqps; i++) {
+	for (i = 0; i < hdev->num_tqps; i++) {
 		hdev->htqp[i].q.handle = &hdev->nic;
 		hdev->htqp[i].q.tqp_index = i;
 		kinfo->tqp[i] = &hdev->htqp[i].q;
-- 
2.33.0

[PATCH net 2/3] net: hns3: using the num_tqps to check whether tqp_index is out of range when vf get ring info from mbx

[Jijie Shao]

From: Jian Shen <shenjian15@huawei.com>

Currently, rss_size = num_tqps / tc_num. If tc_num is 1, then num_tqps
equals rss_size. However, if the tc_num is greater than 1, then rss_size
will be less than num_tqps, causing the tqp_index check for subsequent TCs
using rss_size to always fail.

This patch uses the num_tqps to check whether tqp_index is out of range,
instead of rss_size.

Fixes: 326334aad024 ("net: hns3: add a check for tqp_index in hclge_get_ring_chain_from_mbx()")
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
index c7ff12a6c076..b7d4e06a55d4 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
@@ -193,10 +193,10 @@ static int hclge_get_ring_chain_from_mbx(
 		return -EINVAL;
 
 	for (i = 0; i < ring_num; i++) {
-		if (req->msg.param[i].tqp_index >= vport->nic.kinfo.rss_size) {
+		if (req->msg.param[i].tqp_index >= vport->nic.kinfo.num_tqps) {
 			dev_err(&hdev->pdev->dev, "tqp index(%u) is out of range(0-%u)\n",
 				req->msg.param[i].tqp_index,
-				vport->nic.kinfo.rss_size - 1U);
+				vport->nic.kinfo.num_tqps - 1U);
 			return -EINVAL;
 		}
 	}
-- 
2.33.0

[PATCH net 3/3] net: hns3: add VLAN id validation before using

[Jijie Shao]

From: Jian Shen <shenjian15@huawei.com>

Currently, the VLAN id may be used without validation when
receive a VLAN configuration mailbox from VF. It may cause
out-of-bounds memory access once the VLAN id is bigger than
4095.

Fixes: fe4144d47eef ("net: hns3: sync VLAN filter entries when kill VLAN ID failed")
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index cf8abbe01840..c589baea7c77 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -10555,6 +10555,9 @@ int hclge_set_vlan_filter(struct hnae3_handle *handle, __be16 proto,
 	bool writen_to_tbl = false;
 	int ret = 0;
 
+	if (vlan_id >= VLAN_N_VID)
+		return -EINVAL;
+
 	/* When device is resetting or reset failed, firmware is unable to
 	 * handle mailbox. Just record the vlan id, and remove it after
 	 * reset finished.
-- 
2.33.0

Re: [PATCH net 3/3] net: hns3: add VLAN id validation before using

[Simon Horman]

On Tue, Dec 09, 2025 at 09:38:25PM +0800, Jijie Shao wrote:
> From: Jian Shen <shenjian15@huawei.com>
> 
> Currently, the VLAN id may be used without validation when
> receive a VLAN configuration mailbox from VF. It may cause
> out-of-bounds memory access once the VLAN id is bigger than
> 4095.
> 
> Fixes: fe4144d47eef ("net: hns3: sync VLAN filter entries when kill VLAN ID failed")
> Signed-off-by: Jian Shen <shenjian15@huawei.com>
> Signed-off-by: Jijie Shao <shaojijie@huawei.com>

Hi Jijie,

Can you clarify that the (only) oob access is to vlan_del_fail_bmap?

If so, I agree with this change and that the problem was introduced in
the cited commit. But I think it would be worth mentioning vlan_del_fail_bmap
in the commit message.


...

Re: [PATCH net 1/3] net: hns3: using the num_tqps in the vf driver to apply for resources

[Simon Horman]

On Tue, Dec 09, 2025 at 09:38:23PM +0800, Jijie Shao wrote:
> From: Jian Shen <shenjian15@huawei.com>
> 
> Currently, hdev->htqp is allocated using hdev->num_tqps, and kinfo->tqp
> is allocated using kinfo->num_tqps. However, kinfo->num_tqps is set to
> min(new_tqps, hdev->num_tqps);  Therefore, kinfo->num_tqps may be smaller
> than hdev->num_tqps, which causes some hdev->htqp[i] to remain
> uninitialized in hclgevf_knic_setup().
> 
> Thus, this patch allocates hdev->htqp and kinfo->tqp using hdev->num_tqps,
> ensuring that the lengths of hdev->htqp and kinfo->tqp are consistent
> and that all elements are properly initialized.
> 
> Fixes: e2cb1dec9779 ("net: hns3: Add HNS3 VF HCL(Hardware Compatibility Layer) Support")
> Signed-off-by: Jian Shen <shenjian15@huawei.com>
> Signed-off-by: Jijie Shao <shaojijie@huawei.com>

Reviewed-by: Simon Horman <horms@kernel.org>

Re: [PATCH net 2/3] net: hns3: using the num_tqps to check whether tqp_index is out of range when vf get ring info from mbx

[Simon Horman]

On Tue, Dec 09, 2025 at 09:38:24PM +0800, Jijie Shao wrote:
> From: Jian Shen <shenjian15@huawei.com>
> 
> Currently, rss_size = num_tqps / tc_num. If tc_num is 1, then num_tqps
> equals rss_size. However, if the tc_num is greater than 1, then rss_size
> will be less than num_tqps, causing the tqp_index check for subsequent TCs
> using rss_size to always fail.
> 
> This patch uses the num_tqps to check whether tqp_index is out of range,
> instead of rss_size.
> 
> Fixes: 326334aad024 ("net: hns3: add a check for tqp_index in hclge_get_ring_chain_from_mbx()")
> Signed-off-by: Jian Shen <shenjian15@huawei.com>
> Signed-off-by: Jijie Shao <shaojijie@huawei.com>

Reviewed-by: Simon Horman <horms@kernel.org>

Re: [PATCH net 3/3] net: hns3: add VLAN id validation before using

[Jijie Shao]

on 2025/12/10 0:50, Simon Horman wrote:
> On Tue, Dec 09, 2025 at 09:38:25PM +0800, Jijie Shao wrote:
>> From: Jian Shen <shenjian15@huawei.com>
>>
>> Currently, the VLAN id may be used without validation when
>> receive a VLAN configuration mailbox from VF. It may cause
>> out-of-bounds memory access once the VLAN id is bigger than
>> 4095.
>>
>> Fixes: fe4144d47eef ("net: hns3: sync VLAN filter entries when kill VLAN ID failed")
>> Signed-off-by: Jian Shen <shenjian15@huawei.com>
>> Signed-off-by: Jijie Shao <shaojijie@huawei.com>
> Hi Jijie,
>
> Can you clarify that the (only) oob access is to vlan_del_fail_bmap?
>
> If so, I agree with this change and that the problem was introduced in
> the cited commit. But I think it would be worth mentioning vlan_del_fail_bmap
> in the commit message.
>
Yes, the length of vlan_del_fail_bmap is BITS_TO_LONGS(VLAN_N_VID).
Therefore, vlan_id needs to be checked to ensure it is within the range of VLAN_N_VID.

I will add this in V2

Thanks,
Jijie Shao


> ...
>

Re: [PATCH net 3/3] net: hns3: add VLAN id validation before using

[Simon Horman]

On Wed, Dec 10, 2025 at 02:39:11PM +0800, Jijie Shao wrote:
> 
> on 2025/12/10 0:50, Simon Horman wrote:
> > On Tue, Dec 09, 2025 at 09:38:25PM +0800, Jijie Shao wrote:
> > > From: Jian Shen <shenjian15@huawei.com>
> > > 
> > > Currently, the VLAN id may be used without validation when
> > > receive a VLAN configuration mailbox from VF. It may cause
> > > out-of-bounds memory access once the VLAN id is bigger than
> > > 4095.
> > > 
> > > Fixes: fe4144d47eef ("net: hns3: sync VLAN filter entries when kill VLAN ID failed")
> > > Signed-off-by: Jian Shen <shenjian15@huawei.com>
> > > Signed-off-by: Jijie Shao <shaojijie@huawei.com>
> > Hi Jijie,
> > 
> > Can you clarify that the (only) oob access is to vlan_del_fail_bmap?
> > 
> > If so, I agree with this change and that the problem was introduced in
> > the cited commit. But I think it would be worth mentioning vlan_del_fail_bmap
> > in the commit message.
> > 
> Yes, the length of vlan_del_fail_bmap is BITS_TO_LONGS(VLAN_N_VID).
> Therefore, vlan_id needs to be checked to ensure it is within the range of VLAN_N_VID.
> 
> I will add this in V2

Thanks.

Feel free to also add:

Reviewed-by: Simon Horman <horms@kernel.org>