[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH 02/11] xen: Fix and improve handling of device_add usb-host e
From: |
Markus Armbruster |
Subject: |
Re: [PATCH 02/11] xen: Fix and improve handling of device_add usb-host errors |
Date: |
Wed, 29 Apr 2020 07:48:55 +0200 |
User-agent: |
Gnus/5.13 (Gnus v5.13) Emacs/26.3 (gnu/linux) |
Paul Durrant <address@hidden> writes:
>> -----Original Message-----
>> From: Markus Armbruster <address@hidden>
>> Sent: 24 April 2020 20:20
>> To: address@hidden
>> Cc: Stefano Stabellini <address@hidden>; Anthony Perard <address@hidden>;
>> Paul
>> Durrant <address@hidden>; Gerd Hoffmann <address@hidden>; address@hidden
>> Subject: [PATCH 02/11] xen: Fix and improve handling of device_add usb-host
>> errors
>>
>> usbback_portid_add() leaks the error when qdev_device_add() fails.
>> Fix that. While there, use the error to improve the error message.
>>
>> The qemu_opts_from_qdict() similarly leaks on failure. But any
>> failure there is a programming error. Pass &error_abort.
>>
>> Fixes: 816ac92ef769f9ffc534e49a1bb6177bddce7aa2
>> Cc: Stefano Stabellini <address@hidden>
>> Cc: Anthony Perard <address@hidden>
>> Cc: Paul Durrant <address@hidden>
>> Cc: Gerd Hoffmann <address@hidden>
>> Cc: address@hidden
>> Signed-off-by: Markus Armbruster <address@hidden>
>> ---
>> hw/usb/xen-usb.c | 18 ++++++++----------
>> 1 file changed, 8 insertions(+), 10 deletions(-)
>>
>> diff --git a/hw/usb/xen-usb.c b/hw/usb/xen-usb.c
>> index 961190d0f7..42643c3390 100644
>> --- a/hw/usb/xen-usb.c
>> +++ b/hw/usb/xen-usb.c
>> @@ -30,6 +30,7 @@
>> #include "hw/usb.h"
>> #include "hw/xen/xen-legacy-backend.h"
>> #include "monitor/qdev.h"
>> +#include "qapi/error.h"
>> #include "qapi/qmp/qdict.h"
>> #include "qapi/qmp/qstring.h"
>>
>> @@ -755,13 +756,15 @@ static void usbback_portid_add(struct usbback_info
>> *usbif, unsigned port,
>> qdict_put_int(qdict, "port", port);
>> qdict_put_int(qdict, "hostbus", atoi(busid));
>> qdict_put_str(qdict, "hostport", portname);
>> - opts = qemu_opts_from_qdict(qemu_find_opts("device"), qdict,
>> &local_err);
>> - if (local_err) {
>> - goto err;
>> - }
>> + opts = qemu_opts_from_qdict(qemu_find_opts("device"), qdict,
>> + &error_abort);
>> usbif->ports[port - 1].dev = USB_DEVICE(qdev_device_add(opts,
>> &local_err));
>> if (!usbif->ports[port - 1].dev) {
>> - goto err;
>> + qobject_unref(qdict);
>> + xen_pv_printf(&usbif->xendev, 0,
>> + "device %s could not be opened: %s\n",
>> + busid, error_get_pretty(local_err));
>> + error_free(local_err);
>
> Previously the goto caused the function to bail out. Should there not be a
> 'return' here?
Owww, of course. Thanks!
>
>> }
>> qobject_unref(qdict);
>> speed = usbif->ports[port - 1].dev->speed;
>> @@ -793,11 +796,6 @@ static void usbback_portid_add(struct usbback_info
>> *usbif, unsigned port,
>> usbback_hotplug_enq(usbif, port);
>>
>> TR_BUS(&usbif->xendev, "port %d attached\n", port);
>> - return;
>> -
>> -err:
>> - qobject_unref(qdict);
>> - xen_pv_printf(&usbif->xendev, 0, "device %s could not be opened\n",
>> busid);
>> }
>>
>> static void usbback_process_port(struct usbback_info *usbif, unsigned port)
>> --
>> 2.21.1
- [PATCH 03/11] s390x/cpumodel: Fix harmless misuse of visit_check_struct(), (continued)
- [PATCH 03/11] s390x/cpumodel: Fix harmless misuse of visit_check_struct(), Markus Armbruster, 2020/04/24
- [PATCH 10/11] arm/sabrelite: Consistently use &error_fatal in sabrelite_init(), Markus Armbruster, 2020/04/24
- [PATCH 01/11] nvdimm: Plug memory leak in uuid property setter, Markus Armbruster, 2020/04/24
- [PATCH 08/11] mips/boston: Fix boston_mach_init() error handling, Markus Armbruster, 2020/04/24
- [PATCH 11/11] i386: Fix x86_cpu_load_model() error API violation, Markus Armbruster, 2020/04/24
- [PATCH 02/11] xen: Fix and improve handling of device_add usb-host errors, Markus Armbruster, 2020/04/24
- [PATCH 09/11] mips/boston: Plug memory leak in boston_mach_init(), Markus Armbruster, 2020/04/24