| Message ID | 20200323164318.26567-1-peter.maydell@linaro.org |
|---|---|
| State | Superseded |
| Headers | show |
| Series | [RFC] hw/usb/xen-usb.c: Pass struct usbback_req* to usbback_packet_complete() | expand |
On Mon, Mar 23, 2020 at 04:43:18PM +0000, Peter Maydell wrote: > The function usbback_packet_complete() currently takes a USBPacket*, > which must be a pointer to the packet field within a struct > usbback_req; the function uses container_of() to get the struct > usbback_req* given the USBPacket*. > > This is unnecessarily confusing (and in particular it confuses the > Coverity Scan analysis, resulting in the false positive CID 1421919 > where it thinks that we write off the end of the structure). Since > both callsites already have the pointer to the struct usbback_req, > just pass that in directly. Looks sane from usb point of view. Acked-by: Gerd Hoffmann <kraxel@redhat.com> cheers, Gerd
On Mon, Mar 23, 2020 at 04:43:18PM +0000, Peter Maydell wrote: > The function usbback_packet_complete() currently takes a USBPacket*, > which must be a pointer to the packet field within a struct > usbback_req; the function uses container_of() to get the struct > usbback_req* given the USBPacket*. > > This is unnecessarily confusing (and in particular it confuses the > Coverity Scan analysis, resulting in the false positive CID 1421919 > where it thinks that we write off the end of the structure). Since > both callsites already have the pointer to the struct usbback_req, > just pass that in directly. > > Signed-off-by: Peter Maydell <peter.maydell@linaro.org> > --- > This is an RFC because: > * I'm not very familiar with the Xen bits of QEMU > * the main rationale here is to change something that's > confusing Coverity -- the code as it stands isn't wrong > * the only testing I've done is "make check" > Still, the change seems like a good thing to me as a human reader... > > PS: QEMU's MAINTAINERS file stanza for Xen doesn't pick up > that this file is Xen related, so it could use an extra F: line. Looks good, Reviewed-by: Anthony PERARD <anthony.perard@citrix.com> Thanks, -- Anthony PERARD
diff --git a/hw/usb/xen-usb.c b/hw/usb/xen-usb.c index 1fc2f32ce93..961190d0f78 100644 --- a/hw/usb/xen-usb.c +++ b/hw/usb/xen-usb.c @@ -347,13 +347,11 @@ static int32_t usbback_xlat_status(int status) return -ESHUTDOWN; } -static void usbback_packet_complete(USBPacket *packet) +static void usbback_packet_complete(struct usbback_req *usbback_req) { - struct usbback_req *usbback_req; + USBPacket *packet = &usbback_req->packet; int32_t status; - usbback_req = container_of(packet, struct usbback_req, packet); - QTAILQ_REMOVE(&usbback_req->stub->submit_q, usbback_req, q); status = usbback_xlat_status(packet->status); @@ -566,7 +564,7 @@ static void usbback_dispatch(struct usbback_req *usbback_req) usb_handle_packet(usbback_req->stub->dev, &usbback_req->packet); if (usbback_req->packet.status != USB_RET_ASYNC) { - usbback_packet_complete(&usbback_req->packet); + usbback_packet_complete(usbback_req); } return; @@ -993,7 +991,7 @@ static void xen_bus_complete(USBPort *port, USBPacket *packet) usbif = usbback_req->usbif; TR_REQ(&usbif->xendev, "\n"); - usbback_packet_complete(packet); + usbback_packet_complete(usbback_req); } static USBPortOps xen_usb_port_ops = {
The function usbback_packet_complete() currently takes a USBPacket*, which must be a pointer to the packet field within a struct usbback_req; the function uses container_of() to get the struct usbback_req* given the USBPacket*. This is unnecessarily confusing (and in particular it confuses the Coverity Scan analysis, resulting in the false positive CID 1421919 where it thinks that we write off the end of the structure). Since both callsites already have the pointer to the struct usbback_req, just pass that in directly. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- This is an RFC because: * I'm not very familiar with the Xen bits of QEMU * the main rationale here is to change something that's confusing Coverity -- the code as it stands isn't wrong * the only testing I've done is "make check" Still, the change seems like a good thing to me as a human reader... PS: QEMU's MAINTAINERS file stanza for Xen doesn't pick up that this file is Xen related, so it could use an extra F: line. hw/usb/xen-usb.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) -- 2.20.1