With 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec) converting
ps3disk to new bvec helpers, incrementing the offset was accidently
lost, corrupting consecutive buffers. Restore index for non-corrupted
data transfers.
Fixes: 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec)
Signed-off-by: René Rebe <rene@exactco.de>
--- a/drivers/block/ps3disk.c 2023-08-07 17:40:55.200957746 +0200
+++ b/drivers/block/ps3disk.c 2023-08-07 17:46:39.702964129 +0200
@@ -85,10 +93,14 @@
struct bio_vec bvec;
rq_for_each_segment(bvec, req, iter) {
+ dev_dbg(&dev->sbd.core, "%s:%u: %u sectors from %llu\n",
+ __func__, __LINE__, bio_sectors(iter.bio),
+ iter.bio->bi_iter.bi_sector);
if (gather)
memcpy_from_bvec(dev->bounce_buf + offset, &bvec);
else
memcpy_to_bvec(&bvec, dev->bounce_buf + offset);
+ offset += bvec.bv_len;
}
}
--
René Rebe, ExactCODE GmbH, Lietzenburger Str. 42, DE-10789 Berlin
https://exactcode.com | https://t2linux.com | https://rene.rebe.deOn Fri, Nov 14, 2025 at 03:30:33PM +0100, Rene Rebe wrote:
> With 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec) converting
> ps3disk to new bvec helpers, incrementing the offset was accidently
> lost, corrupting consecutive buffers. Restore index for non-corrupted
> data transfers.
>
> Fixes: 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec)
>
> Signed-off-by: René Rebe <rene@exactco.de>
>
> --- a/drivers/block/ps3disk.c 2023-08-07 17:40:55.200957746 +0200
> +++ b/drivers/block/ps3disk.c 2023-08-07 17:46:39.702964129 +0200
> @@ -85,10 +93,14 @@
> struct bio_vec bvec;
>
> rq_for_each_segment(bvec, req, iter) {
> + dev_dbg(&dev->sbd.core, "%s:%u: %u sectors from %llu\n",
> + __func__, __LINE__, bio_sectors(iter.bio),
In the future, please note that all dev_dbg() calls have __func__ built
into them, and I think __LINE__ as well, so there's no need to ever add
that to the output. Take a look at the documentation for more details
on how to see that in the log.
thanks,
greg k-h
Hi Greg,
On Sun, 16 Nov 2025 at 13:14, Greg Kroah-Hartman
<gregkh@linuxfoundation.org> wrote:
> On Fri, Nov 14, 2025 at 03:30:33PM +0100, Rene Rebe wrote:
> > With 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec) converting
> > ps3disk to new bvec helpers, incrementing the offset was accidently
> > lost, corrupting consecutive buffers. Restore index for non-corrupted
> > data transfers.
> >
> > Fixes: 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec)
> >
> > Signed-off-by: René Rebe <rene@exactco.de>
> >
> > --- a/drivers/block/ps3disk.c 2023-08-07 17:40:55.200957746 +0200
> > +++ b/drivers/block/ps3disk.c 2023-08-07 17:46:39.702964129 +0200
> > @@ -85,10 +93,14 @@
> > struct bio_vec bvec;
> >
> > rq_for_each_segment(bvec, req, iter) {
> > + dev_dbg(&dev->sbd.core, "%s:%u: %u sectors from %llu\n",
> > + __func__, __LINE__, bio_sectors(iter.bio),
>
> In the future, please note that all dev_dbg() calls have __func__ built
> into them, and I think __LINE__ as well, so there's no need to ever add
> that to the output. Take a look at the documentation for more details
> on how to see that in the log.
That does not seem to be true in v6.18-rc6: the function name and line
number are not included.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
On Mon, Nov 17, 2025 at 10:47:49AM +0100, Geert Uytterhoeven wrote:
> Hi Greg,
>
> On Sun, 16 Nov 2025 at 13:14, Greg Kroah-Hartman
> <gregkh@linuxfoundation.org> wrote:
> > On Fri, Nov 14, 2025 at 03:30:33PM +0100, Rene Rebe wrote:
> > > With 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec) converting
> > > ps3disk to new bvec helpers, incrementing the offset was accidently
> > > lost, corrupting consecutive buffers. Restore index for non-corrupted
> > > data transfers.
> > >
> > > Fixes: 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec)
> > >
> > > Signed-off-by: René Rebe <rene@exactco.de>
> > >
> > > --- a/drivers/block/ps3disk.c 2023-08-07 17:40:55.200957746 +0200
> > > +++ b/drivers/block/ps3disk.c 2023-08-07 17:46:39.702964129 +0200
> > > @@ -85,10 +93,14 @@
> > > struct bio_vec bvec;
> > >
> > > rq_for_each_segment(bvec, req, iter) {
> > > + dev_dbg(&dev->sbd.core, "%s:%u: %u sectors from %llu\n",
> > > + __func__, __LINE__, bio_sectors(iter.bio),
> >
> > In the future, please note that all dev_dbg() calls have __func__ built
> > into them, and I think __LINE__ as well, so there's no need to ever add
> > that to the output. Take a look at the documentation for more details
> > on how to see that in the log.
>
> That does not seem to be true in v6.18-rc6: the function name and line
> number are not included.
What changed to remove them? Look at
Documentation/admin-guide/dynamic-debug-howto.rst for how to enable both
of these for any dynamic debug message in the kernel.
Does that not work properly here?
thanks,
greg k-h
Hi Greg,
On Mon, 17 Nov 2025 at 13:39, Greg Kroah-Hartman
<gregkh@linuxfoundation.org> wrote:
> On Mon, Nov 17, 2025 at 10:47:49AM +0100, Geert Uytterhoeven wrote:
> > On Sun, 16 Nov 2025 at 13:14, Greg Kroah-Hartman
> > <gregkh@linuxfoundation.org> wrote:
> > > On Fri, Nov 14, 2025 at 03:30:33PM +0100, Rene Rebe wrote:
> > > > With 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec) converting
> > > > ps3disk to new bvec helpers, incrementing the offset was accidently
> > > > lost, corrupting consecutive buffers. Restore index for non-corrupted
> > > > data transfers.
> > > >
> > > > Fixes: 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec)
> > > >
> > > > Signed-off-by: René Rebe <rene@exactco.de>
> > > >
> > > > --- a/drivers/block/ps3disk.c 2023-08-07 17:40:55.200957746 +0200
> > > > +++ b/drivers/block/ps3disk.c 2023-08-07 17:46:39.702964129 +0200
> > > > @@ -85,10 +93,14 @@
> > > > struct bio_vec bvec;
> > > >
> > > > rq_for_each_segment(bvec, req, iter) {
> > > > + dev_dbg(&dev->sbd.core, "%s:%u: %u sectors from %llu\n",
> > > > + __func__, __LINE__, bio_sectors(iter.bio),
> > >
> > > In the future, please note that all dev_dbg() calls have __func__ built
> > > into them, and I think __LINE__ as well, so there's no need to ever add
> > > that to the output. Take a look at the documentation for more details
> > > on how to see that in the log.
> >
> > That does not seem to be true in v6.18-rc6: the function name and line
> > number are not included.
>
> What changed to remove them? Look at
> Documentation/admin-guide/dynamic-debug-howto.rst for how to enable both
> of these for any dynamic debug message in the kernel.
>
> Does that not work properly here?
Oh, dynamic debug, dunno ;-)
It does not work with #define DEBUG at the top of a source file.
So perhaps lots of cleanup work to do?
$ git grep -Ew "(pr_debug|dev_dbg).*__func__" v6.18-rc6 | wc -l
4512
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
On Mon, Nov 17, 2025 at 02:05:33PM +0100, Geert Uytterhoeven wrote:
> Hi Greg,
>
> On Mon, 17 Nov 2025 at 13:39, Greg Kroah-Hartman
> <gregkh@linuxfoundation.org> wrote:
> > On Mon, Nov 17, 2025 at 10:47:49AM +0100, Geert Uytterhoeven wrote:
> > > On Sun, 16 Nov 2025 at 13:14, Greg Kroah-Hartman
> > > <gregkh@linuxfoundation.org> wrote:
> > > > On Fri, Nov 14, 2025 at 03:30:33PM +0100, Rene Rebe wrote:
> > > > > With 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec) converting
> > > > > ps3disk to new bvec helpers, incrementing the offset was accidently
> > > > > lost, corrupting consecutive buffers. Restore index for non-corrupted
> > > > > data transfers.
> > > > >
> > > > > Fixes: 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec)
> > > > >
> > > > > Signed-off-by: René Rebe <rene@exactco.de>
> > > > >
> > > > > --- a/drivers/block/ps3disk.c 2023-08-07 17:40:55.200957746 +0200
> > > > > +++ b/drivers/block/ps3disk.c 2023-08-07 17:46:39.702964129 +0200
> > > > > @@ -85,10 +93,14 @@
> > > > > struct bio_vec bvec;
> > > > >
> > > > > rq_for_each_segment(bvec, req, iter) {
> > > > > + dev_dbg(&dev->sbd.core, "%s:%u: %u sectors from %llu\n",
> > > > > + __func__, __LINE__, bio_sectors(iter.bio),
> > > >
> > > > In the future, please note that all dev_dbg() calls have __func__ built
> > > > into them, and I think __LINE__ as well, so there's no need to ever add
> > > > that to the output. Take a look at the documentation for more details
> > > > on how to see that in the log.
> > >
> > > That does not seem to be true in v6.18-rc6: the function name and line
> > > number are not included.
> >
> > What changed to remove them? Look at
> > Documentation/admin-guide/dynamic-debug-howto.rst for how to enable both
> > of these for any dynamic debug message in the kernel.
> >
> > Does that not work properly here?
>
> Oh, dynamic debug, dunno ;-)
> It does not work with #define DEBUG at the top of a source file.
Nor should it, that's a really old scheme :)
> So perhaps lots of cleanup work to do?
>
> $ git grep -Ew "(pr_debug|dev_dbg).*__func__" v6.18-rc6 | wc -l
> 4512
Huh? We want those to be used.
confused,
greg k-h
Hi Greg,
On Mon, 17 Nov 2025 at 14:13, Greg Kroah-Hartman
<gregkh@linuxfoundation.org> wrote:
> On Mon, Nov 17, 2025 at 02:05:33PM +0100, Geert Uytterhoeven wrote:
> > On Mon, 17 Nov 2025 at 13:39, Greg Kroah-Hartman
> > <gregkh@linuxfoundation.org> wrote:
> > > On Mon, Nov 17, 2025 at 10:47:49AM +0100, Geert Uytterhoeven wrote:
> > > > On Sun, 16 Nov 2025 at 13:14, Greg Kroah-Hartman
> > > > <gregkh@linuxfoundation.org> wrote:
> > > > > On Fri, Nov 14, 2025 at 03:30:33PM +0100, Rene Rebe wrote:
> > > > > > With 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec) converting
> > > > > > ps3disk to new bvec helpers, incrementing the offset was accidently
> > > > > > lost, corrupting consecutive buffers. Restore index for non-corrupted
> > > > > > data transfers.
> > > > > >
> > > > > > Fixes: 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec)
> > > > > >
> > > > > > Signed-off-by: René Rebe <rene@exactco.de>
> > > > > >
> > > > > > --- a/drivers/block/ps3disk.c 2023-08-07 17:40:55.200957746 +0200
> > > > > > +++ b/drivers/block/ps3disk.c 2023-08-07 17:46:39.702964129 +0200
> > > > > > @@ -85,10 +93,14 @@
> > > > > > struct bio_vec bvec;
> > > > > >
> > > > > > rq_for_each_segment(bvec, req, iter) {
> > > > > > + dev_dbg(&dev->sbd.core, "%s:%u: %u sectors from %llu\n",
> > > > > > + __func__, __LINE__, bio_sectors(iter.bio),
> > > > >
> > > > > In the future, please note that all dev_dbg() calls have __func__ built
> > > > > into them, and I think __LINE__ as well, so there's no need to ever add
> > > > > that to the output. Take a look at the documentation for more details
> > > > > on how to see that in the log.
> > > >
> > > > That does not seem to be true in v6.18-rc6: the function name and line
> > > > number are not included.
> > >
> > > What changed to remove them? Look at
> > > Documentation/admin-guide/dynamic-debug-howto.rst for how to enable both
> > > of these for any dynamic debug message in the kernel.
> > >
> > > Does that not work properly here?
> >
> > Oh, dynamic debug, dunno ;-)
> > It does not work with #define DEBUG at the top of a source file.
>
> Nor should it, that's a really old scheme :)
That method is still documented in
Documentation/admin-guide/dynamic-debug-howto.rst, and I use it a lot.
It does mean "all dev_dbg() calls have __func__ built in to them" is false...
> > So perhaps lots of cleanup work to do?
> >
> > $ git grep -Ew "(pr_debug|dev_dbg).*__func__" v6.18-rc6 | wc -l
> > 4512
>
> Huh? We want those to be used.
You just said they should not print __func__ explicitly??
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
On Mon, Nov 17, 2025 at 02:31:49PM +0100, Geert Uytterhoeven wrote:
> Hi Greg,
>
> On Mon, 17 Nov 2025 at 14:13, Greg Kroah-Hartman
> <gregkh@linuxfoundation.org> wrote:
> > On Mon, Nov 17, 2025 at 02:05:33PM +0100, Geert Uytterhoeven wrote:
> > > On Mon, 17 Nov 2025 at 13:39, Greg Kroah-Hartman
> > > <gregkh@linuxfoundation.org> wrote:
> > > > On Mon, Nov 17, 2025 at 10:47:49AM +0100, Geert Uytterhoeven wrote:
> > > > > On Sun, 16 Nov 2025 at 13:14, Greg Kroah-Hartman
> > > > > <gregkh@linuxfoundation.org> wrote:
> > > > > > On Fri, Nov 14, 2025 at 03:30:33PM +0100, Rene Rebe wrote:
> > > > > > > With 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec) converting
> > > > > > > ps3disk to new bvec helpers, incrementing the offset was accidently
> > > > > > > lost, corrupting consecutive buffers. Restore index for non-corrupted
> > > > > > > data transfers.
> > > > > > >
> > > > > > > Fixes: 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec)
> > > > > > >
> > > > > > > Signed-off-by: René Rebe <rene@exactco.de>
> > > > > > >
> > > > > > > --- a/drivers/block/ps3disk.c 2023-08-07 17:40:55.200957746 +0200
> > > > > > > +++ b/drivers/block/ps3disk.c 2023-08-07 17:46:39.702964129 +0200
> > > > > > > @@ -85,10 +93,14 @@
> > > > > > > struct bio_vec bvec;
> > > > > > >
> > > > > > > rq_for_each_segment(bvec, req, iter) {
> > > > > > > + dev_dbg(&dev->sbd.core, "%s:%u: %u sectors from %llu\n",
> > > > > > > + __func__, __LINE__, bio_sectors(iter.bio),
> > > > > >
> > > > > > In the future, please note that all dev_dbg() calls have __func__ built
> > > > > > into them, and I think __LINE__ as well, so there's no need to ever add
> > > > > > that to the output. Take a look at the documentation for more details
> > > > > > on how to see that in the log.
> > > > >
> > > > > That does not seem to be true in v6.18-rc6: the function name and line
> > > > > number are not included.
> > > >
> > > > What changed to remove them? Look at
> > > > Documentation/admin-guide/dynamic-debug-howto.rst for how to enable both
> > > > of these for any dynamic debug message in the kernel.
> > > >
> > > > Does that not work properly here?
> > >
> > > Oh, dynamic debug, dunno ;-)
> > > It does not work with #define DEBUG at the top of a source file.
> >
> > Nor should it, that's a really old scheme :)
>
> That method is still documented in
> Documentation/admin-guide/dynamic-debug-howto.rst, and I use it a lot.
>
> It does mean "all dev_dbg() calls have __func__ built in to them" is false...
But it's there, just set it from the debugfs api, right?
> > > So perhaps lots of cleanup work to do?
> > >
> > > $ git grep -Ew "(pr_debug|dev_dbg).*__func__" v6.18-rc6 | wc -l
> > > 4512
> >
> > Huh? We want those to be used.
>
> You just said they should not print __func__ explicitly??
Yes, they should not.
thanks,
greg k-h
Hi Greg,
On Tue, 18 Nov 2025 at 03:36, Greg Kroah-Hartman
<gregkh@linuxfoundation.org> wrote:
> On Mon, Nov 17, 2025 at 02:31:49PM +0100, Geert Uytterhoeven wrote:
> > On Mon, 17 Nov 2025 at 14:13, Greg Kroah-Hartman
> > <gregkh@linuxfoundation.org> wrote:
> > > On Mon, Nov 17, 2025 at 02:05:33PM +0100, Geert Uytterhoeven wrote:
> > > > On Mon, 17 Nov 2025 at 13:39, Greg Kroah-Hartman
> > > > <gregkh@linuxfoundation.org> wrote:
> > > > > On Mon, Nov 17, 2025 at 10:47:49AM +0100, Geert Uytterhoeven wrote:
> > > > > > On Sun, 16 Nov 2025 at 13:14, Greg Kroah-Hartman
> > > > > > <gregkh@linuxfoundation.org> wrote:
> > > > > > > On Fri, Nov 14, 2025 at 03:30:33PM +0100, Rene Rebe wrote:
> > > > > > > > With 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec) converting
> > > > > > > > ps3disk to new bvec helpers, incrementing the offset was accidently
> > > > > > > > lost, corrupting consecutive buffers. Restore index for non-corrupted
> > > > > > > > data transfers.
> > > > > > > >
> > > > > > > > Fixes: 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec)
> > > > > > > >
> > > > > > > > Signed-off-by: René Rebe <rene@exactco.de>
> > > > > > > >
> > > > > > > > --- a/drivers/block/ps3disk.c 2023-08-07 17:40:55.200957746 +0200
> > > > > > > > +++ b/drivers/block/ps3disk.c 2023-08-07 17:46:39.702964129 +0200
> > > > > > > > @@ -85,10 +93,14 @@
> > > > > > > > struct bio_vec bvec;
> > > > > > > >
> > > > > > > > rq_for_each_segment(bvec, req, iter) {
> > > > > > > > + dev_dbg(&dev->sbd.core, "%s:%u: %u sectors from %llu\n",
> > > > > > > > + __func__, __LINE__, bio_sectors(iter.bio),
> > > > > > >
> > > > > > > In the future, please note that all dev_dbg() calls have __func__ built
> > > > > > > into them, and I think __LINE__ as well, so there's no need to ever add
> > > > > > > that to the output. Take a look at the documentation for more details
> > > > > > > on how to see that in the log.
> > > > > >
> > > > > > That does not seem to be true in v6.18-rc6: the function name and line
> > > > > > number are not included.
> > > > >
> > > > > What changed to remove them? Look at
> > > > > Documentation/admin-guide/dynamic-debug-howto.rst for how to enable both
> > > > > of these for any dynamic debug message in the kernel.
> > > > >
> > > > > Does that not work properly here?
> > > >
> > > > Oh, dynamic debug, dunno ;-)
> > > > It does not work with #define DEBUG at the top of a source file.
> > >
> > > Nor should it, that's a really old scheme :)
> >
> > That method is still documented in
> > Documentation/admin-guide/dynamic-debug-howto.rst, and I use it a lot.
> >
> > It does mean "all dev_dbg() calls have __func__ built in to them" is false...
>
> But it's there, just set it from the debugfs api, right?
Only with dynamic debug, not with -DDEBUG, so "all" is not appropriate.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
Hi,
> On 17. Nov 2025, at 10:47, Geert Uytterhoeven <geert@linux-m68k.org> wrote:
>
> Hi Greg,
>
> On Sun, 16 Nov 2025 at 13:14, Greg Kroah-Hartman
> <gregkh@linuxfoundation.org> wrote:
>> On Fri, Nov 14, 2025 at 03:30:33PM +0100, Rene Rebe wrote:
>>> With 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec) converting
>>> ps3disk to new bvec helpers, incrementing the offset was accidently
>>> lost, corrupting consecutive buffers. Restore index for non-corrupted
>>> data transfers.
>>>
>>> Fixes: 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec)
>>>
>>> Signed-off-by: René Rebe <rene@exactco.de>
>>>
>>> --- a/drivers/block/ps3disk.c 2023-08-07 17:40:55.200957746 +0200
>>> +++ b/drivers/block/ps3disk.c 2023-08-07 17:46:39.702964129 +0200
>>> @@ -85,10 +93,14 @@
>>> struct bio_vec bvec;
>>>
>>> rq_for_each_segment(bvec, req, iter) {
>>> + dev_dbg(&dev->sbd.core, "%s:%u: %u sectors from %llu\n",
>>> + __func__, __LINE__, bio_sectors(iter.bio),
>>
>> In the future, please note that all dev_dbg() calls have __func__ built
>> into them, and I think __LINE__ as well, so there's no need to ever add
>> that to the output. Take a look at the documentation for more details
>> on how to see that in the log.
>
> That does not seem to be true in v6.18-rc6: the function name and line
> number are not included.
Thanks, I will keep that in mind for future patches, in this case I only
Exactly restored what was deleted by the broken commit.
René
--
https://exactco.de - https://t2linux.com - https://rene.rebe.de
On Fri, 14 Nov 2025 15:30:33 +0100, Rene Rebe wrote:
> With 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec) converting
> ps3disk to new bvec helpers, incrementing the offset was accidently
> lost, corrupting consecutive buffers. Restore index for non-corrupted
> data transfers.
>
> Fixes: 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec)
>
> [...]
Applied, thanks!
[1/1] Fix ps3disk: use memcpy_{from,to}_bvec index
(no commit info)
Best regards,
--
Jens Axboe
On Fri, Nov 14, 2025 at 03:30:33PM +0100, Rene Rebe wrote:
> With 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec) converting
> ps3disk to new bvec helpers, incrementing the offset was accidently
> lost, corrupting consecutive buffers. Restore index for non-corrupted
> data transfers.
>
> Fixes: 6e0a48552b8c (ps3disk: use memcpy_{from,to}_bvec)
>
> Signed-off-by: René Rebe <rene@exactco.de>
Looks good:
Reviewed-by: Christoph Hellwig <hch@lst.de>
Note that there should be no empty line after the fixes tag,
but maybe Jens can fix this when applying the patch.
© 2016 - 2026 Red Hat, Inc.