Skip to contentSkip to conversation

media: uvcvideo: Silence memcpy() run-time false positive warnings

PATCH In review Kees Cook 4 participants 7 messages Jan 6, 2023 – Feb 22, 2023

Subsystems (2)
lore
Share & tools
lore

Kees Cook

author
#0 lore raw

PATCH 1/1 media: uvcvideo: Silence memcpy() run-time false positive warnings

The memcpy() in uvc_video_decode_meta() intentionally copies across the length and flags members and into the trailing buf flexible array. Split the copy so that the compiler can better reason about (the lack of) buffer overflows here. Avoid the run-time false positive warning:

memcpy: detected field-spanning write (size 12) of single field "&meta->length" at drivers/media/usb/uvc/uvc_video.c:1355 (size 1)

Additionally fix a typo in the documentation for struct uvc_meta_buf.

---
 drivers/media/usb/uvc/uvc_video.c | 4 +++-
 include/uapi/linux/uvcvideo.h     | 2 +-
 2 files changed, 4 insertions(+), 2 deletions(-)
drivers/media/usb/uvc/uvc_video.c +3 -1
index d2eb9066e4dc..b67347ab4181 100644
@@ -1352,7 +1352,9 @@ static void uvc_video_decode_meta(struct uvc_streaming *stream,
 	if (has_scr)
 		memcpy(stream->clock.last_scr, scr, 6);
 
-	memcpy(&meta->length, mem, length);
+	meta->length = mem[0];
+	meta->flags  = mem[1];
+	memcpy(meta->buf, &mem[2], length - 2);
 	meta_buf->bytesused += length + sizeof(meta->ns) + sizeof(meta->sof);
 
 	uvc_dbg(stream->dev, FRAME,
include/uapi/linux/uvcvideo.h +1 -1
index 8288137387c0..a9d0a64007ba 100644
@@ -86,7 +86,7 @@ struct uvc_xu_control_query {
  * struct. The first two fields are added by the driver, they can be used for
  * clock synchronisation. The rest is an exact copy of a UVC payload header.
  * Only complete objects with complete buffers are included. Therefore it's
- * always sizeof(meta->ts) + sizeof(meta->sof) + meta->length bytes large.
+ * always sizeof(meta->ns) + sizeof(meta->sof) + meta->length bytes large.
  */
 struct uvc_meta_buf {
 	__u64 ns;
Signature · 1 line
2.34.1
#1 lore raw

Hi Kees

Thanks for the patch

Would it make more sense to replace *mem with a structure/union. Something like:
https://patchwork.linuxtv.org/project/linux-media/patch/[email protected]/

Regards!

Quoted 52 lines · Kees Cook

The memcpy() in uvc_video_decode_meta() intentionally copies across the length and flags members and into the trailing buf flexible array. Split the copy so that the compiler can better reason about (the lack of) buffer overflows here. Avoid the run-time false positive warning:

memcpy: detected field-spanning write (size 12) of single field "&meta->length" at drivers/media/usb/uvc/uvc_video.c:1355 (size 1)

Additionally fix a typo in the documentation for struct uvc_meta_buf.

---
 drivers/media/usb/uvc/uvc_video.c | 4 +++-
 include/uapi/linux/uvcvideo.h     | 2 +-
 2 files changed, 4 insertions(+), 2 deletions(-)
drivers/media/usb/uvc/uvc_video.c +3 -1
index d2eb9066e4dc..b67347ab4181 100644
@@ -1352,7 +1352,9 @@ static void uvc_video_decode_meta(struct uvc_streaming *stream,
        if (has_scr)
                memcpy(stream->clock.last_scr, scr, 6);
 
-       memcpy(&meta->length, mem, length);
+       meta->length = mem[0];
+       meta->flags  = mem[1];
+       memcpy(meta->buf, &mem[2], length - 2);
        meta_buf->bytesused += length + sizeof(meta->ns) + sizeof(meta->sof);
 
        uvc_dbg(stream->dev, FRAME,
include/uapi/linux/uvcvideo.h +1 -2
index 8288137387c0..a9d0a64007ba 100644
@@ -86,7 +86,7 @@ struct uvc_xu_control_query {
  * struct. The first two fields are added by the driver, they can be used for
  * clock synchronisation. The rest is an exact copy of a UVC payload header.
  * Only complete objects with complete buffers are included. Therefore it's
- * always sizeof(meta->ts) + sizeof(meta->sof) + meta->length bytes large.
+ * always sizeof(meta->ns) + sizeof(meta->sof) + meta->length bytes large.
  */
 struct uvc_meta_buf {
        __u64 ns;
--

2.34.1

Signature · 1 line
Ricardo Ribalda
author
#2 lore raw
Quoted 52 lines · Ricardo Ribalda
Quoted 49 lines · Kees Cook

The memcpy() in uvc_video_decode_meta() intentionally copies across the length and flags members and into the trailing buf flexible array. Split the copy so that the compiler can better reason about (the lack of) buffer overflows here. Avoid the run-time false positive warning:

memcpy: detected field-spanning write (size 12) of single field "&meta->length" at drivers/media/usb/uvc/uvc_video.c:1355 (size 1)

Additionally fix a typo in the documentation for struct uvc_meta_buf.

---
 drivers/media/usb/uvc/uvc_video.c | 4 +++-
 include/uapi/linux/uvcvideo.h     | 2 +-
 2 files changed, 4 insertions(+), 2 deletions(-)
drivers/media/usb/uvc/uvc_video.c +3 -1
index d2eb9066e4dc..b67347ab4181 100644
@@ -1352,7 +1352,9 @@ static void uvc_video_decode_meta(struct uvc_streaming *stream,
        if (has_scr)
                memcpy(stream->clock.last_scr, scr, 6);
 
-       memcpy(&meta->length, mem, length);
+       meta->length = mem[0];
+       meta->flags  = mem[1];
+       memcpy(meta->buf, &mem[2], length - 2);
        meta_buf->bytesused += length + sizeof(meta->ns) + sizeof(meta->sof);
 
        uvc_dbg(stream->dev, FRAME,
include/uapi/linux/uvcvideo.h +1 -1
index 8288137387c0..a9d0a64007ba 100644
@@ -86,7 +86,7 @@ struct uvc_xu_control_query {
  * struct. The first two fields are added by the driver, they can be used for
  * clock synchronisation. The rest is an exact copy of a UVC payload header.
  * Only complete objects with complete buffers are included. Therefore it's
- * always sizeof(meta->ts) + sizeof(meta->sof) + meta->length bytes large.
+ * always sizeof(meta->ns) + sizeof(meta->sof) + meta->length bytes large.
  */
 struct uvc_meta_buf {
        __u64 ns;

[...]

Quoted 2 lines

Would it make more sense to replace *mem with a structure/union. Something like:
https://patchwork.linuxtv.org/project/linux-media/patch/[email protected]/

I wasn't sure -- it seemed like this routine was doing the serializing into a struct already and an additional struct overlay wasn't going to improve readability. But I can certainly do that if it's preferred!

-Kees

Signature · 1 line
Kees Cook
maintainer Reviewed-by
#3 lore raw

Hello,

Quoted 53 lines · Kees Cook
Quoted 52 lines · Ricardo Ribalda
Quoted 49 lines · Kees Cook

The memcpy() in uvc_video_decode_meta() intentionally copies across the length and flags members and into the trailing buf flexible array. Split the copy so that the compiler can better reason about (the lack of) buffer overflows here. Avoid the run-time false positive warning:

memcpy: detected field-spanning write (size 12) of single field "&meta->length" at drivers/media/usb/uvc/uvc_video.c:1355 (size 1)

Additionally fix a typo in the documentation for struct uvc_meta_buf.

---
 drivers/media/usb/uvc/uvc_video.c | 4 +++-
 include/uapi/linux/uvcvideo.h     | 2 +-
 2 files changed, 4 insertions(+), 2 deletions(-)
drivers/media/usb/uvc/uvc_video.c +3 -1
index d2eb9066e4dc..b67347ab4181 100644
@@ -1352,7 +1352,9 @@ static void uvc_video_decode_meta(struct uvc_streaming *stream,
        if (has_scr)
                memcpy(stream->clock.last_scr, scr, 6);
 
-       memcpy(&meta->length, mem, length);
+       meta->length = mem[0];
+       meta->flags  = mem[1];
+       memcpy(meta->buf, &mem[2], length - 2);
        meta_buf->bytesused += length + sizeof(meta->ns) + sizeof(meta->sof);
 
        uvc_dbg(stream->dev, FRAME,
include/uapi/linux/uvcvideo.h +1 -1
index 8288137387c0..a9d0a64007ba 100644
@@ -86,7 +86,7 @@ struct uvc_xu_control_query {
  * struct. The first two fields are added by the driver, they can be used for
  * clock synchronisation. The rest is an exact copy of a UVC payload header.
  * Only complete objects with complete buffers are included. Therefore it's
- * always sizeof(meta->ts) + sizeof(meta->sof) + meta->length bytes large.
+ * always sizeof(meta->ns) + sizeof(meta->sof) + meta->length bytes large.
  */
 struct uvc_meta_buf {
        __u64 ns;

[...]

Quoted 6 lines
Quoted 2 lines

Would it make more sense to replace *mem with a structure/union. Something like:
https://patchwork.linuxtv.org/project/linux-media/patch/[email protected]/

I wasn't sure -- it seemed like this routine was doing the serializing into a struct already and an additional struct overlay wasn't going to improve readability. But I can certainly do that if it's preferred!

I'm not sure to see how using an additional struct or union would help. We can't use struct assignment as the data may be unaligned, so memcpy() is required. The issue isn't with the source operand of the memcpy() but with the destination operand. Ricardo, if I'm missing something, please submit an alternative patch to explain what you meant.

Signature · 3 lines
Regards,

Laurent Pinchart
#4 lore raw

Hi Laurent

I was thinking about something on the line of the attached patch,

uvc_frame_header->data could also be replaced with a union.

Warning, not tested ;)

Quoted 76 lines · Laurent Pinchart

Hello,

Quoted 53 lines · Kees Cook
Quoted 52 lines · Ricardo Ribalda
Quoted 49 lines · Kees Cook

The memcpy() in uvc_video_decode_meta() intentionally copies across the length and flags members and into the trailing buf flexible array. Split the copy so that the compiler can better reason about (the lack of) buffer overflows here. Avoid the run-time false positive warning:

memcpy: detected field-spanning write (size 12) of single field "&meta->length" at drivers/media/usb/uvc/uvc_video.c:1355 (size 1)

Additionally fix a typo in the documentation for struct uvc_meta_buf.

---
 drivers/media/usb/uvc/uvc_video.c | 4 +++-
 include/uapi/linux/uvcvideo.h     | 2 +-
 2 files changed, 4 insertions(+), 2 deletions(-)
drivers/media/usb/uvc/uvc_video.c +3 -1
index d2eb9066e4dc..b67347ab4181 100644
@@ -1352,7 +1352,9 @@ static void uvc_video_decode_meta(struct uvc_streaming *stream,
        if (has_scr)
                memcpy(stream->clock.last_scr, scr, 6);
 
-       memcpy(&meta->length, mem, length);
+       meta->length = mem[0];
+       meta->flags  = mem[1];
+       memcpy(meta->buf, &mem[2], length - 2);
        meta_buf->bytesused += length + sizeof(meta->ns) + sizeof(meta->sof);
 
        uvc_dbg(stream->dev, FRAME,
include/uapi/linux/uvcvideo.h +1 -1
index 8288137387c0..a9d0a64007ba 100644
@@ -86,7 +86,7 @@ struct uvc_xu_control_query {
  * struct. The first two fields are added by the driver, they can be used for
  * clock synchronisation. The rest is an exact copy of a UVC payload header.
  * Only complete objects with complete buffers are included. Therefore it's
- * always sizeof(meta->ts) + sizeof(meta->sof) + meta->length bytes large.
+ * always sizeof(meta->ns) + sizeof(meta->sof) + meta->length bytes large.
  */
 struct uvc_meta_buf {
        __u64 ns;

[...]

Quoted 6 lines
Quoted 2 lines

Would it make more sense to replace *mem with a structure/union. Something like:
https://patchwork.linuxtv.org/project/linux-media/patch/[email protected]/

I wasn't sure -- it seemed like this routine was doing the serializing into a struct already and an additional struct overlay wasn't going to improve readability. But I can certainly do that if it's preferred!

I'm not sure to see how using an additional struct or union would help. We can't use struct assignment as the data may be unaligned, so memcpy() is required. The issue isn't with the source operand of the memcpy() but with the destination operand. Ricardo, if I'm missing something, please submit an alternative patch to explain what you meant.

--
Regards,

Laurent Pinchart

Signature · 1 line
Ricardo Ribalda

From dce72fe7f003ba40ba2d534e5f84bafc73b37314 Mon Sep 17 00:00:00 2001
From: Ricardo Ribalda <[email protected]>
Date: Mon, 9 Jan 2023 11:42:21 +0100
Subject: [PATCH] media: uvcvideo: Refactor uvc_video_decode_meta

NOT TESTED!

---
 drivers/media/usb/uvc/uvc_video.c | 58 ++++++++++++++++++++-----------
 1 file changed, 37 insertions(+), 21 deletions(-)
drivers/media/usb/uvc/uvc_video.c +37 -21
index 04f452d95fd6..f6c091626f3c 100644
@@ -1330,6 +1330,17 @@ static int uvc_video_encode_data(struct uvc_streaming *stream,
  * Metadata
  */
 
+struct uvc_frame_header {
+	u8 length;
+	u8 flags;
+	u8 data[];
+} __packed;
+
+struct uvc_scr {
+	u32 scr;
+	u16 frame_sof;
+} __packed;
+
 /*
  * Additionally to the payload headers we also want to provide the user with USB
  * Frame Numbers and system time values. The resulting buffer is thus composed
@@ -1343,7 +1354,8 @@ static int uvc_video_encode_data(struct uvc_streaming *stream,
  */
 static void uvc_video_decode_meta(struct uvc_streaming *stream,
 				  struct uvc_buffer *meta_buf,
-				  const u8 *mem, unsigned int length)
+				  const struct uvc_frame_header *header,
+				  unsigned int length)
 {
 	struct uvc_meta_buf *meta;
 	size_t len_std = 2;
@@ -1351,7 +1363,8 @@ static void uvc_video_decode_meta(struct uvc_streaming *stream,
 	unsigned long flags;
 	unsigned int sof;
 	ktime_t time;
-	const u8 *scr;
+	const struct uvc_scr *scr;
+	const u32 *pts;
 
 	if (!meta_buf || length == 2)
 		return;
@@ -1362,28 +1375,31 @@ static void uvc_video_decode_meta(struct uvc_streaming *stream,
 		return;
 	}
 
-	has_pts = mem[1] & UVC_STREAM_PTS;
-	has_scr = mem[1] & UVC_STREAM_SCR;
+	has_pts = header->flags & UVC_STREAM_PTS;
+	has_scr = header->flags & UVC_STREAM_SCR;
 
 	if (has_pts) {
-		len_std += 4;
-		scr = mem + 6;
-	} else {
-		scr = mem + 2;
+		pts = (u32 *) header->data;
+		len_std += sizeof(*pts);
 	}
 
-	if (has_scr)
-		len_std += 6;
+	if (has_scr) {
+		u8 offset;
+
+		offset = has_pts ? sizeof(*pts) : 0;
+		scr = (struct uvc_scr *) header->data + offset;
+		len_std += sizeof(struct uvc_scr);
+	}
 
 	if (stream->meta.format == V4L2_META_FMT_UVC)
 		length = len_std;
 
 	if (length == len_std && (!has_scr ||
-				  !memcmp(scr, stream->clock.last_scr, 6)))
+				  !memcmp(scr, stream->clock.last_scr, sizeof(struct uvc_scr))))
 		return;
 
 	meta = (struct uvc_meta_buf *)((u8 *)meta_buf->mem + meta_buf->bytesused);
-	local_irq_save(flags);
+	local_irq_save(flags); //do we need this?
 	time = uvc_video_get_time();
 	sof = usb_get_current_frame_number(stream->dev->udev);
 	local_irq_restore(flags);
@@ -1391,20 +1407,20 @@ static void uvc_video_decode_meta(struct uvc_streaming *stream,
 	put_unaligned(sof, &meta->sof);
 
 	if (has_scr)
-		memcpy(stream->clock.last_scr, scr, 6);
+		memcpy(stream->clock.last_scr, scr, sizeof(struct uvc_scr));
 
-	meta->length = mem[0];
-	meta->flags  = mem[1];
-	memcpy(meta->buf, &mem[2], length - 2);
+	meta->length = header->length;
+	meta->flags  = header->flags;
+	memcpy(meta->buf, header->data, length - offsetof(struct uvc_frame_header, data));
 	meta_buf->bytesused += length + sizeof(meta->ns) + sizeof(meta->sof);
 
 	uvc_dbg(stream->dev, FRAME,
 		"%s(): t-sys %lluns, SOF %u, len %u, flags 0x%x, PTS %u, STC %u frame SOF %u\n",
 		__func__, ktime_to_ns(time), meta->sof, meta->length,
 		meta->flags,
-		has_pts ? *(u32 *)meta->buf : 0,
-		has_scr ? *(u32 *)scr : 0,
-		has_scr ? *(u32 *)(scr + 4) & 0x7ff : 0);
+		has_pts ? *pts : 0,
+		has_scr ? scr->scr : 0,
+		has_scr ? scr->frame_sof & 0x7ff : 0);
 }
 
 /* ------------------------------------------------------------------------
@@ -1479,7 +1495,7 @@ static void uvc_video_decode_isoc(struct uvc_urb *uvc_urb,
 		if (ret < 0)
 			continue;
 
-		uvc_video_decode_meta(stream, meta_buf, mem, ret);
+		uvc_video_decode_meta(stream, meta_buf, (struct uvc_frame_header *) mem, ret);
 
 		/* Decode the payload data. */
 		uvc_video_decode_data(uvc_urb, buf, mem + ret,
@@ -1531,7 +1547,7 @@ static void uvc_video_decode_bulk(struct uvc_urb *uvc_urb,
 			memcpy(stream->bulk.header, mem, ret);
 			stream->bulk.header_size = ret;
 
-			uvc_video_decode_meta(stream, meta_buf, mem, ret);
+			uvc_video_decode_meta(stream, meta_buf, (struct uvc_frame_header *)mem, ret);
 
 			mem += ret;
 			len -= ret;
Signature · 1 line
2.39.0.314.g84b9a713c41-goog
#5 lore raw
Quoted 7 lines · Ricardo Ribalda

Hi Laurent

I was thinking about something on the line of the attached patch,

uvc_frame_header->data could also be replaced with a union.

Warning, not tested ;)

...

drivers/media/usb/uvc/uvc_video.c static int uvc_video_encode_data(struct uvc_streaming *stream, · line 1337

Quoted 5 lines
drivers/media/usb/uvc/uvc_video.c +5 -0
+struct uvc_frame_header {
+	u8 length;
+	u8 flags;
+	u8 data[];
+} __packed;

__packed! (See below)

...

drivers/media/usb/uvc/uvc_video.c static void uvc_video_decode_meta(struct uvc_streaming *stream, · line 1382

Quoted 1 line
drivers/media/usb/uvc/uvc_video.c +1 -0
+		pts = (u32 *) header->data;

Ai-ai-ai.
Here is just a yellow flag...

...

drivers/media/usb/uvc/uvc_video.c static void uvc_video_decode_meta(struct uvc_streaming *stream, · line 1421

Quoted 5 lines
drivers/media/usb/uvc/uvc_video.c +1 -0
 	uvc_dbg(stream->dev, FRAME,
 		"%s(): t-sys %lluns, SOF %u, len %u, flags 0x%x, PTS %u, STC %u frame SOF %u\n",
 		__func__, ktime_to_ns(time), meta->sof, meta->length,
 		meta->flags,
+		has_pts ? *pts : 0,

...and here is a red flag. What you need to have is

	void *pts;
	u32 pts_val;

	pts_val = get_unaligned_be32(); // or le32

	...use pts_val...
Quoted 2 lines
drivers/media/usb/uvc/uvc_video.c +2 -0
+		has_scr ? scr->scr : 0,
+		has_scr ? scr->frame_sof & 0x7ff : 0);
Signature · 2 lines
With Best Regards,
Andy Shevchenko
#6 lore raw

Hi Andy

Quoted 29 lines · Andy Shevchenko
Quoted 7 lines · Ricardo Ribalda

Hi Laurent

I was thinking about something on the line of the attached patch,

uvc_frame_header->data could also be replaced with a union.

Warning, not tested ;)

...

drivers/media/usb/uvc/uvc_video.c static int uvc_video_encode_data(struct uvc_streaming *stream, · line 1337

Quoted 5 lines
drivers/media/usb/uvc/uvc_video.c +5 -0
+struct uvc_frame_header {
+     u8 length;
+     u8 flags;
+     u8 data[];
+} __packed;

__packed! (See below)

...

Quoted 1 line

+ pts = (u32 *) header->data;

Ai-ai-ai.
Here is just a yellow flag...

...

Quoted 7 lines
Quoted 5 lines
      uvc_dbg(stream->dev, FRAME,
              "%s(): t-sys %lluns, SOF %u, len %u, flags 0x%x, PTS %u, STC %u frame SOF %u\n",
              __func__, ktime_to_ns(time), meta->sof, meta->length,
              meta->flags,

+ has_pts ? *pts : 0,

...and here is a red flag. What you need to have is

Thanks! you are absolutely right :)

Luckily Laurent merged the original patch not mine.

Regards!

Quoted 17 lines
        void *pts;
        u32 pts_val;

        pts_val = get_unaligned_be32(); // or le32

        ...use pts_val...

drivers/media/usb/uvc/uvc_video.c static void uvc_video_decode_meta(struct uvc_streaming *stream, · line 1423

Quoted 2 lines
drivers/media/usb/uvc/uvc_video.c +2 -0
+             has_scr ? scr->scr : 0,
+             has_scr ? scr->frame_sof & 0x7ff : 0);

--
With Best Regards,
Andy Shevchenko

Signature · 1 line
Ricardo Ribalda