linux-kernel - [PATCH] drm/msm: Replace unsafe snprintf usage with scnprintf

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <20251130012834.142585-2-veyga@veygax.dev>
Date: Sun, 30 Nov 2025 01:28:54 +0000
From: veygax <veyga@...gax.dev>
To: robin.clark@....qualcomm.com, lumag@...nel.org
Cc: abhinav.kumar@...ux.dev, jessica.zhang@....qualcomm.com, sean@...rly.run, marijn.suijten@...ainline.org, airlied@...il.com, simona@...ll.ch, linux-arm-msm@...r.kernel.org, dri-devel@...ts.freedesktop.org, freedreno@...ts.freedesktop.org, linux-kernel@...r.kernel.org, veygax <veyga@...gax.dev>
Subject: [PATCH] drm/msm: Replace unsafe snprintf usage with scnprintf

The refill_buf function uses snprintf to append to a fixed-size buffer.
snprintf returns the length that would have been written, which can
exceed the remaining buffer size. If this happens, ptr advances beyond
the buffer and rem becomes negative. In the 2nd iteration, rem is
treated as a large unsigned integer, causing snprintf to write oob.

While this behavior is technically mitigated by num_perfcntrs being
locked at 5, it's still unsafe if num_perfcntrs were ever to change/a
second source was added.

Signed-off-by: veygax <veyga@...gax.dev>
---
 drivers/gpu/drm/msm/msm_perf.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/msm/msm_perf.c b/drivers/gpu/drm/msm/msm_perf.c
index d3c7889aaf26..c369d4acc378 100644
--- a/drivers/gpu/drm/msm/msm_perf.c
+++ b/drivers/gpu/drm/msm/msm_perf.c
@@ -65,13 +65,13 @@ static int refill_buf(struct msm_perf_state *perf)
 
 	if ((perf->cnt++ % 32) == 0) {
 		/* Header line: */
-		n = snprintf(ptr, rem, "%%BUSY");
+		n = scnprintf(ptr, rem, "%%BUSY");
 		ptr += n;
 		rem -= n;
 
 		for (i = 0; i < gpu->num_perfcntrs; i++) {
 			const struct msm_gpu_perfcntr *perfcntr = &gpu->perfcntrs[i];
-			n = snprintf(ptr, rem, "\t%s", perfcntr->name);
+			n = scnprintf(ptr, rem, "\t%s", perfcntr->name);
 			ptr += n;
 			rem -= n;
 		}
@@ -93,21 +93,21 @@ static int refill_buf(struct msm_perf_state *perf)
 			return ret;
 
 		val = totaltime ? 1000 * activetime / totaltime : 0;
-		n = snprintf(ptr, rem, "%3d.%d%%", val / 10, val % 10);
+		n = scnprintf(ptr, rem, "%3d.%d%%", val / 10, val % 10);
 		ptr += n;
 		rem -= n;
 
 		for (i = 0; i < ret; i++) {
 			/* cycle counters (I think).. convert to MHz.. */
 			val = cntrs[i] / 10000;
-			n = snprintf(ptr, rem, "\t%5d.%02d",
+			n = scnprintf(ptr, rem, "\t%5d.%02d",
 					val / 100, val % 100);
 			ptr += n;
 			rem -= n;
 		}
 	}
 
-	n = snprintf(ptr, rem, "\n");
+	n = scnprintf(ptr, rem, "\n");
 	ptr += n;
 	rem -= n;
 
-- 
2.52.0