From: sashiko-bot@kernel.org
To: "Mukesh Ojha" <mukesh.ojha@oss.qualcomm.com>
Cc: dri-devel@lists.freedesktop.org, media-ci@linuxtv.org
Subject: Re: [PATCH v2 7/9] wifi: ath12k: migrate firmware loading to qcom_mdt_pas_load() and qcom_pas_prepare_and_auth_reset()
Date: Sat, 14 Nov 2026 18:54:58 +0000 [thread overview]
Message-ID: <sashiko-outbox-214918@kernel.org> (raw)
In-Reply-To: <20261113185009.157893-8-mukesh.ojha@oss.qualcomm.com>
Thank you for your contribution! Sashiko AI review found 1 potential
issue to consider:
- [High] wifi: ath12k: PAS metadata leak in ath12k_ahb_power_up() and
shared FW load
--
commit 7ea070a02821a21185ed124f14f936a8b5576395
Author: Mukesh Ojha <mukesh.ojha@oss.qualcomm.com>
wifi: ath12k: migrate firmware loading to qcom_mdt_pas_load() and qcom_pas_prepare_and_auth_reset()
This commit replaces qcom_mdt_load() and qcom_pas_auth_and_reset() in
ath12k_ahb_power_up() and ath12k_ahb_load_auth_shared_fw() with
qcom_pas_context_init(), qcom_mdt_pas_load(), and
qcom_pas_prepare_and_auth_reset().
> diff --git a/drivers/net/wireless/ath/ath12k/ahb.c b/drivers/net/wireless/ath/ath12k/ahb.c
> index 2c55aae06dd06..355f1f5509975 100644
> --- a/drivers/net/wireless/ath/ath12k/ahb.c
> +++ b/drivers/net/wireless/ath/ath12k/ahb.c
> @@ -411,6 +411,7 @@ static int ath12k_ahb_power_up(struct ath12k_base *ab)
> struct ath12k_ahb_rproc_info *rproc_info = ab_ahb->rproc_info;
> char fw_name[ATH12K_USERPD_FW_NAME_LEN];
> char fw2_name[ATH12K_USERPD_FW_NAME_LEN];
> + struct qcom_pas_context ctx;
> struct device *dev = ab->dev;
> const struct firmware *fw, *fw2;
> unsigned long time_left;
> @@ -444,14 +445,15 @@ static int ath12k_ahb_power_up(struct ath12k_base *ab)
> ATH12K_AHB_UPD_SWID;
>
> /* Load FW image to a reserved memory location */
> - if (ab_ahb->scm_auth_enabled)
> - ret = qcom_mdt_load(dev, fw, fw_name, pasid, rproc_info->mem_region,
> - rproc_info->mem_phys, rproc_info->mem_size,
> - NULL);
> - else
> + if (ab_ahb->scm_auth_enabled) {
> + qcom_pas_context_init(&ctx, dev, pasid,
> + rproc_info->mem_phys, rproc_info->mem_size);
> + ret = qcom_mdt_pas_load(&ctx, fw, fw_name, NULL);
> + } else {
> ret = qcom_mdt_load_no_init(dev, fw, fw_name, rproc_info->mem_region,
> rproc_info->mem_phys, rproc_info->mem_size,
> NULL);
> + }
> if (ret) {
> ath12k_err(ab, "Failed to load MDT segments: %d\n", ret);
> goto err_fw;
> @@ -485,7 +487,7 @@ static int ath12k_ahb_power_up(struct ath12k_base *ab)
>
> if (ab_ahb->scm_auth_enabled) {
> /* Authenticate FW image using peripheral ID */
> - ret = qcom_pas_auth_and_reset(pasid);
> + ret = qcom_pas_prepare_and_auth_reset(&ctx);
> if (ret) {
> ath12k_err(ab, "failed to boot the remote processor %d\n", ret);
> goto err_fw2;
> @@ -1023,6 +1025,7 @@ static int ath12k_ahb_load_auth_shared_fw(struct ath12k_base *ab,
> struct ath12k_ahb_rproc_info *rproc_info,
> const char *fw_name, u32 pasid)
> {
> + struct qcom_pas_context ctx;
> int ret;
>
> const struct firmware *fw __free(firmware) = NULL;
> @@ -1041,14 +1044,16 @@ static int ath12k_ahb_load_auth_shared_fw(struct ath12k_base *ab,
> ath12k_dbg(ab, ATH12K_DBG_AHB, "loading firmware %s, size %zd\n", fw_name,
> fw->size);
>
> - ret = qcom_mdt_load(ab->dev, fw, fw_name, pasid, rproc_info->mem_region,
> - rproc_info->mem_phys, rproc_info->mem_size, NULL);
> + qcom_pas_context_init(&ctx, ab->dev, pasid,
> + rproc_info->mem_phys, rproc_info->mem_size);
> +
> + ret = qcom_mdt_pas_load(&ctx, fw, fw_name, NULL);
> if (ret) {
> ath12k_err(ab, "failed to load RO firmware %s: %d\n", fw_name, ret);
> return ret;
> }
>
> - ret = qcom_pas_auth_and_reset(pasid);
> + ret = qcom_pas_prepare_and_auth_reset(&ctx);
> if (ret)
> ath12k_err(ab, "failed to authenticate and boot shared firmware: %d\n",
> ret);
[Severity: High]
Does this leak the PAS metadata buffer stored in ctx.ptr in both
ath12k_ahb_power_up() and ath12k_ahb_load_auth_shared_fw()?
Previously, qcom_mdt_load() passed a NULL context pointer to
__qcom_mdt_pas_init() -> qcom_pas_init_image(), so the underlying backend
freed the metadata allocation immediately. qcom_mdt_pas_load() instead
passes &ctx, which retains the allocated metadata buffer in ctx.ptr until
the caller releases it with qcom_pas_metadata_release(&ctx):
drivers/firmware/qcom/qcom_scm.c:qcom_scm_pas_init_image() {
...
if (ret < 0 || !ctx) {
dma_free_coherent(dev, size, mdata_buf, mdata_phys);
} else if (ctx) {
ctx->ptr = mdata_buf;
ctx->phys = mdata_phys;
ctx->size = size;
}
...
}
Because neither function calls qcom_pas_metadata_release(&ctx) after
qcom_pas_prepare_and_auth_reset(&ctx) or on the error paths after
qcom_mdt_pas_load() (including if qcom_mdt_pas_load() fails after
qcom_pas_init_image() succeeds), the stack-allocated ctx goes out of scope
and leaks the metadata buffer on every firmware load.
--
Sashiko AI review · https://sashiko.dev/#/patchset/20261113185009.157893-1-mukesh.ojha@oss.qualcomm.com?part=7
next prev parent reply other threads:[~2026-11-14 18:55 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-13 18:50 [PATCH v2 0/9] Migrate PAS clients to qcom_mdt_pas_load() and qcom_pas_prepare_and_auth_reset() Mukesh Ojha
2026-11-13 18:50 ` [PATCH v2 1/9] firmware: qcom: pas: add qcom_pas_context_init() for pre-allocated context Mukesh Ojha
2026-11-13 19:03 ` Alex Elder
2026-11-13 18:50 ` [PATCH v2 2/9] remoteproc: qcom_wcnss: migrate to qcom_mdt_pas_load() and qcom_pas_prepare_and_auth_reset() Mukesh Ojha
2026-11-14 7:29 ` kernel test robot
2026-11-14 14:49 ` kernel test robot
2026-11-14 18:54 ` sashiko-bot
2026-11-13 18:50 ` [PATCH v2 3/9] drm/msm/adreno: migrate zap shader loading " Mukesh Ojha
2026-11-14 18:54 ` sashiko-bot
2026-11-13 18:50 ` [PATCH v2 4/9] media: iris: migrate firmware " Mukesh Ojha
2026-11-14 18:54 ` sashiko-bot
2026-11-13 18:50 ` [PATCH v2 5/9] media: venus: " Mukesh Ojha
2026-11-14 18:54 ` sashiko-bot
2026-11-13 18:50 ` [PATCH v2 6/9] net: ipa: " Mukesh Ojha
2026-11-13 19:03 ` Alex Elder
2026-11-14 18:54 ` sashiko-bot
2026-11-13 18:50 ` [PATCH v2 7/9] wifi: ath12k: " Mukesh Ojha
2026-11-13 21:54 ` Jeff Johnson
2026-11-14 18:54 ` sashiko-bot [this message]
2026-11-13 18:50 ` [PATCH v2 8/9] soc: qcom: mdt_loader: remove qcom_mdt_load() Mukesh Ojha
2026-11-14 18:54 ` sashiko-bot
2026-11-13 18:50 ` [PATCH v2 9/9] firmware: qcom: pas: remove qcom_pas_auth_and_reset() Mukesh Ojha
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=sashiko-outbox-214918@kernel.org \
--to=sashiko-bot@kernel.org \
--cc=dri-devel@lists.freedesktop.org \
--cc=media-ci@linuxtv.org \
--cc=mukesh.ojha@oss.qualcomm.com \
--cc=sashiko-reviews@lists.linux.dev \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox