Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1328084
| From | Dmitry Vyukov <dvyukov@google.com> |
|---|---|
| Newsgroups | linux.kernel |
| Subject | Re: gigaset: memory leak in gigaset_initcshw |
| Date | 2016-02-05 22:30 +0100 |
| Message-ID | <qYW6Z-5D5-1@gated-at.bofh.it> (permalink) |
| References | (7 earlier) <qYty2-3iA-11@gated-at.bofh.it> <qYtHI-3B3-27@gated-at.bofh.it> <qYOMa-P6-15@gated-at.bofh.it> <qYR7j-2vG-3@gated-at.bofh.it> <qYTst-3Q7-1@gated-at.bofh.it> |
| Organization | linux.* mail to news gateway |
On Fri, Feb 5, 2016 at 7:36 PM, Paul Bolle <pebolle@tiscali.nl> wrote:
> On vr, 2016-02-05 at 17:06 +0100, Paul Bolle wrote:
>> If that would happen, then cs can be reused while the previous
>> > cs->hw.ser is not freed yet. Just a guess.
>>
>> I'll have to ponder on that a bit, sorry.
>
> This is from the hit-the-code-until-it-confesses department:
> --- a/drivers/isdn/gigaset/ser-gigaset.c
> +++ b/drivers/isdn/gigaset/ser-gigaset.c
> @@ -373,13 +373,9 @@ static void gigaset_freecshw(struct cardstate *cs)
>
> static void gigaset_device_release(struct device *dev)
> {
> - struct cardstate *cs = dev_get_drvdata(dev);
> -
> - if (!cs)
> - return;
> + struct ser_cardstate *scs = dev_get_drvdata(dev);
> dev_set_drvdata(dev, NULL);
> - kfree(cs->hw.ser);
> - cs->hw.ser = NULL;
> + kfree(scs);
> }
>
> /*
> @@ -408,7 +404,7 @@ static int gigaset_initcshw(struct cardstate *cs)
> cs->hw.ser = NULL;
> return rc;
> }
> - dev_set_drvdata(&cs->hw.ser->dev.dev, cs);
> + dev_set_drvdata(&cs->hw.ser->dev.dev, scs);
>
> tasklet_init(&cs->write_tasklet,
> gigaset_modem_fill, (unsigned long) cs);
>
> Does that make any difference?
Nope.
Almost 500 objects leaked in less than 10 seconds:
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 1992 2015 2520 13 8 : tunables 0 0
0 : slabdata 155 155 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2024 2041 2520 13 8 : tunables 0 0
0 : slabdata 157 157 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2061 2080 2520 13 8 : tunables 0 0
0 : slabdata 160 160 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2091 2119 2520 13 8 : tunables 0 0
0 : slabdata 163 163 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2147 2171 2520 13 8 : tunables 0 0
0 : slabdata 167 167 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2228 2236 2520 13 8 : tunables 0 0
0 : slabdata 172 172 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2261 2288 2520 13 8 : tunables 0 0
0 : slabdata 176 176 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2289 2301 2520 13 8 : tunables 0 0
0 : slabdata 177 177 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2316 2340 2520 13 8 : tunables 0 0
0 : slabdata 180 180 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2324 2366 2520 13 8 : tunables 0 0
0 : slabdata 182 182 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2356 2379 2520 13 8 : tunables 0 0
0 : slabdata 183 183 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2450 2509 2520 13 8 : tunables 0 0
0 : slabdata 193 193 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2450 2509 2520 13 8 : tunables 0 0
0 : slabdata 193 193 0
# cat /proc/slabinfo | egrep "^kmalloc-2048"
kmalloc-2048 2450 2509 2520 13 8 : tunables 0 0
0 : slabdata 193 193 0
Back to linux.kernel | Previous | Next — Previous in thread | Find similar | Unroll thread
gigaset: memory leak in gigaset_initcshw Dmitry Vyukov <dvyukov@google.com> - 2016-02-03 16:40 +0100
Re: gigaset: memory leak in gigaset_initcshw Paul Bolle <pebolle@tiscali.nl> - 2016-02-03 17:20 +0100
Re: gigaset: memory leak in gigaset_initcshw Paul Bolle <pebolle@tiscali.nl> - 2016-02-03 20:20 +0100
Re: gigaset: memory leak in gigaset_initcshw Dmitry Vyukov <dvyukov@google.com> - 2016-02-04 11:50 +0100
Re: gigaset: memory leak in gigaset_initcshw Paul Bolle <pebolle@tiscali.nl> - 2016-02-04 14:20 +0100
Re: gigaset: memory leak in gigaset_initcshw Paul Bolle <pebolle@tiscali.nl> - 2016-02-04 15:00 +0100
Re: gigaset: memory leak in gigaset_initcshw Dmitry Vyukov <dvyukov@google.com> - 2016-02-04 16:00 +0100
Re: gigaset: memory leak in gigaset_initcshw Paul Bolle <pebolle@tiscali.nl> - 2016-02-04 16:10 +0100
Re: gigaset: memory leak in gigaset_initcshw Dmitry Vyukov <dvyukov@google.com> - 2016-02-05 14:40 +0100
Re: gigaset: memory leak in gigaset_initcshw Paul Bolle <pebolle@tiscali.nl> - 2016-02-05 17:10 +0100
Re: gigaset: memory leak in gigaset_initcshw Paul Bolle <pebolle@tiscali.nl> - 2016-02-05 19:40 +0100
Re: gigaset: memory leak in gigaset_initcshw Dmitry Vyukov <dvyukov@google.com> - 2016-02-05 22:30 +0100
csiph-web