Skip to contentSkip to conversation

uas: add stream number sanity checks (maybe 6.1)

PATCH In review Gerd Hoffmann 5 participants 9 messages Aug 18, 2021 – Dec 9, 2021

lore
Share & tools
lore
#0 lore raw

cover uas: add stream number sanity checks (maybe 6.1)

Security fix. Sorry for the last-minute patch, I had completely forgotten this one until the CVE number for it arrived today.

Given that the classic usb storage device is way more popular than the uas (usb attached scsi) device the impact should be pretty low and we might consider to not screw up our release schedule for this.

take care,
Gerd

Gerd Hoffmann (1):
uas: add stream number sanity checks.

 hw/usb/dev-uas.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
Signature · 1 line
2.31.1
#1 lore raw

PATCH 1/1 uas: add stream number sanity checks.

The device uses the guest-supplied stream number unchecked, which can lead to guest-triggered out-of-band access to the UASDevice->data3 and UASDevice->status3 fields. Add the missing checks.

Fixes CVE-2021-3713
---
 hw/usb/dev-uas.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
hw/usb/dev-uas.c +11 -0
index 263056231c79..f6309a5ebfdc 100644
@@ -840,6 +840,9 @@ static void usb_uas_handle_data(USBDevice *dev, USBPacket *p)
         }
         break;
     case UAS_PIPE_ID_STATUS:
+        if (p->stream > UAS_MAX_STREAMS) {
+            goto err_stream;
+        }
         if (p->stream) {
             QTAILQ_FOREACH(st, &uas->results, next) {
                 if (st->stream == p->stream) {
@@ -867,6 +870,9 @@ static void usb_uas_handle_data(USBDevice *dev, USBPacket *p)
         break;
     case UAS_PIPE_ID_DATA_IN:
     case UAS_PIPE_ID_DATA_OUT:
+        if (p->stream > UAS_MAX_STREAMS) {
+            goto err_stream;
+        }
         if (p->stream) {
             req = usb_uas_find_request(uas, p->stream);
         } else {
@@ -902,6 +908,11 @@ static void usb_uas_handle_data(USBDevice *dev, USBPacket *p)
         p->status = USB_RET_STALL;
         break;
     }
+
+err_stream:
+    error_report("%s: invalid stream %d", __func__, p->stream);
+    p->status = USB_RET_STALL;
+    return;
 }
 
 static void usb_uas_unrealize(USBDevice *dev)
Signature · 1 line
2.31.1
Reviewed-by on v1 1/1
#2 lore raw

Re: [PATCH 1/1] uas: add stream number sanity checks.

Quoted 5 lines · Gerd Hoffmann

The device uses the guest-supplied stream number unchecked, which can lead to guest-triggered out-of-band access to the UASDevice->data3 and UASDevice->status3 fields. Add the missing checks.

Fixes CVE-2021-3713
Quoted 30 lines
---
 hw/usb/dev-uas.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
hw/usb/dev-uas.c +6 -0
index 263056231c79..f6309a5ebfdc 100644
@@ -840,6 +840,9 @@ static void usb_uas_handle_data(USBDevice *dev, USBPacket *p)
         }
         break;
     case UAS_PIPE_ID_STATUS:
+        if (p->stream > UAS_MAX_STREAMS) {
+            goto err_stream;
+        }
         if (p->stream) {
             QTAILQ_FOREACH(st, &uas->results, next) {
                 if (st->stream == p->stream) {
@@ -867,6 +870,9 @@ static void usb_uas_handle_data(USBDevice *dev, USBPacket *p)
         break;
     case UAS_PIPE_ID_DATA_IN:
     case UAS_PIPE_ID_DATA_OUT:
+        if (p->stream > UAS_MAX_STREAMS) {
+            goto err_stream;
+        }
         if (p->stream) {
             req = usb_uas_find_request(uas, p->stream);
         } else {

hw/usb/dev-uas.c static void usb_uas_handle_data(USBDevice *dev, USBPacket *p) · line 916

Quoted 9 lines
hw/usb/dev-uas.c +5 -0
@@ -902,6 +908,11 @@ static void usb_uas_handle_data(USBDevice *dev, USBPacket *p)
         p->status = USB_RET_STALL;
         break;
     }
+
+err_stream:
+    error_report("%s: invalid stream %d", __func__, p->stream);
+    p->status = USB_RET_STALL;
+    return;
 }

Thanks for tackling this, I was looking but thought it was too late for 6.1. Beside your obvious checks, I'd also modify usb_uas_alloc_status() to feel safer, by checking 'tag'. Eventually returning NULL and adapt the call sites? Anyway this can wait 6.2, so for this patch:

#3 lore raw

Re: [PATCH 1/1] uas: add stream number sanity checks.

In reply to Gerd Hoffmann
Quoted 38 lines · Gerd Hoffmann

The device uses the guest-supplied stream number unchecked, which can lead to guest-triggered out-of-band access to the UASDevice->data3 and UASDevice->status3 fields. Add the missing checks.

---
 hw/usb/dev-uas.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
hw/usb/dev-uas.c +6 -0
index 263056231c79..f6309a5ebfdc 100644
@@ -840,6 +840,9 @@ static void usb_uas_handle_data(USBDevice *dev, USBPacket *p)
         }
         break;
     case UAS_PIPE_ID_STATUS:
+        if (p->stream > UAS_MAX_STREAMS) {
+            goto err_stream;
+        }
         if (p->stream) {
             QTAILQ_FOREACH(st, &uas->results, next) {
                 if (st->stream == p->stream) {
@@ -867,6 +870,9 @@ static void usb_uas_handle_data(USBDevice *dev, USBPacket *p)
         break;
     case UAS_PIPE_ID_DATA_IN:
     case UAS_PIPE_ID_DATA_OUT:
+        if (p->stream > UAS_MAX_STREAMS) {
+            goto err_stream;
+        }
         if (p->stream) {
             req = usb_uas_find_request(uas, p->stream);
         } else {

hw/usb/dev-uas.c static void usb_uas_handle_data(USBDevice *dev, USBPacket *p) · line 915

Quoted 8 lines
hw/usb/dev-uas.c +5 -0
@@ -902,6 +908,11 @@ static void usb_uas_handle_data(USBDevice *dev, USBPacket *p)
         p->status = USB_RET_STALL;
         break;
     }
+
+err_stream:
+    error_report("%s: invalid stream %d", __func__, p->stream);
+    p->status = USB_RET_STALL;
+    return;

How is this supposed to work ? It results in messages such as the following.

qemu-system-sparc64: usb_uas_handle_data: invalid stream 1
qemu-system-sparc64: usb_uas_handle_data: invalid stream 1

It also sets the status unconditionally to USB_RET_STALL, and UAS is simply broken after this patch is applied because the error handling code is executed literally for each call of usb_uas_handle_data().

Guenter

#4 lore raw
In reply to Gerd Hoffmann
Quoted 6 lines · Gerd Hoffmann

Security fix. Sorry for the last-minute patch, I had completely forgotten this one until the CVE number for it arrived today.

Given that the classic usb storage device is way more popular than the uas (usb attached scsi) device the impact should be pretty low and we might consider to not screw up our release schedule for this.

So 6.1-rc5 or stable?

#5 lore raw
In reply to Gerd Hoffmann
Quoted 7 lines · Gerd Hoffmann

Security fix. Sorry for the last-minute patch, I had completely forgotten this one until the CVE number for it arrived today.

Given that the classic usb storage device is way more popular than the uas (usb attached scsi) device the impact should be pretty low and we might consider to not screw up our release schedule for this.

What's the impact if the bug is exploited ?

-- PMM

#6 lore raw

Cc'ing Mauro to double-check.

Quoted 10 lines · Peter Maydell
Quoted 7 lines · Gerd Hoffmann

Security fix. Sorry for the last-minute patch, I had completely forgotten this one until the CVE number for it arrived today.

Given that the classic usb storage device is way more popular than the uas (usb attached scsi) device the impact should be pretty low and we might consider to not screw up our release schedule for this.

What's the impact if the bug is exploited ?

Bug class: "guest-triggered user-after-free".

Being privileged (root) in the guest, you can leak some data from the host process then DoS the host or potentially exploit the use-after-free to execute code on the host.

#7 lore raw

Hi,

Quoted 13 lines · Philippe Mathieu-Daudé

Cc'ing Mauro to double-check.

Quoted 9 lines · Peter Maydell
Quoted 7 lines · Gerd Hoffmann

Security fix. Sorry for the last-minute patch, I had completely forgotten this one until the CVE number for it arrived today.

Given that the classic usb storage device is way more popular than the uas (usb attached scsi) device the impact should be pretty low and we might consider to not screw up our release schedule for this.

Quoted 8 lines
Quoted 1 line

What's the impact if the bug is exploited ?

Bug class: "guest-triggered user-after-free".

Being privileged (root) in the guest, you can leak some data from the host process then DoS the host or potentially exploit the use-after-free to execute code on the host.

This is actually an out-of-bounds access issue (not UAF). It's still potentially bad, but I agree with Gerd the impact is low. Plus there's an assert right before [1] that makes it a DoS if the accessed memory is not NULL.

[1] https://gitlab.com/qemu-project/qemu/-/blob/master/hw/usb/dev-uas.c#L850

Regards.
--
Mauro Matteo Cascella
Red Hat Product Security
PGP-Key ID: BB3410B0

#8 lore raw
Quoted 27 lines · Mauro Matteo Cascella

Hi,

Quoted 13 lines · Philippe Mathieu-Daudé

Cc'ing Mauro to double-check.

Quoted 9 lines · Peter Maydell
Quoted 7 lines · Gerd Hoffmann

Security fix. Sorry for the last-minute patch, I had completely forgotten this one until the CVE number for it arrived today.

Given that the classic usb storage device is way more popular than the uas (usb attached scsi) device the impact should be pretty low and we might consider to not screw up our release schedule for this.

Quoted 8 lines
Quoted 1 line

What's the impact if the bug is exploited ?

Bug class: "guest-triggered user-after-free".

Being privileged (root) in the guest, you can leak some data from the host process then DoS the host or potentially exploit the use-after-free to execute code on the host.

Quoted 4 lines

This is actually an out-of-bounds access issue (not UAF). It's still potentially bad, but I agree with Gerd the impact is low. Plus there's an assert right before [1] that makes it a DoS if the accessed memory is not NULL.

Thanks. OK, (and following discussion of this on irc on Friday) we won't put this fix into 6.1. (That is, we treat it the same way we would if the CVE patch had arrived the day after we tagged 6.1, ie distros and other interested parties pick up the patch as they would any other security fix.)

-- PMM