2:20 a.m.
Hi, Alexander
Thank you for the patch set.
On 11/29/23 22:12, devel-request(a)lists.crash-utility.osci.io wrote:
Date: Wed, 29 Nov 2023 13:47:33 +0100
From: Alexander Gordeev <agordeev(a)linux.ibm.com>
Subject: [Crash-utility] [PATCH 0/3] s390x: uncouple physical and
virtual memory spaces
To: devel(a)lists.crash-utility.osci.io
Message-ID: <cover.1701256901.git.agordeev(a)linux.ibm.com>
Hi all,
Currently physical and virtual addresses are the same on S390X,
but in the future it is going to be uncoupled just like on any
other architecture. This series supports that feature, but it
does not impact the current and older kernel versions.
I did the test, and the patches can work on old kernel versions.
However, since VA and PA are the same on S390x, it's hard to determine
if all changes have been made.
In addition, I can not test the patch 3, seems that the current kernel
does not support this feature(I tried it on the kernel 6.7, did not see
the different VA and PA).
Anyway, If you still want to have these patches in crash-utility for
now, I can help ack this one.
Patch 2 is basically NOP and only fix semantics.
Patch 3 uses the feature if it is available in kernel.
For patch 3, I saw a warning:
s390x.c: In function ‘is_read_proc_kcore’:
s390x.c:671:13: warning: unused variable ‘kcore_fd’ [-Wunused-variable]
671 | int kcore_fd, fd;
| ^~~~~~~~
The variable kcore_fd is not used in the patch 3, it would be good to
remove it.
Thanks
Lianbo
Thanks!
Alexander Gordeev (3):
Fix identity_map_base value dump on S390
s390x: fix virtual vs physical address confusion
s390x: uncouple physical and virtual memory spaces
defs.h | 20 ++++-
s390.c | 2 +-
s390x.c | 245 ++++++++++++++++++++++++++++++++++++++++++++++++++++++--
3 files changed, 255 insertions(+), 12 deletions(-)
9:02 a.m.
New subject: [PATCH 0/3] s390x: uncouple physical and virtual memory spaces
On Mon, Dec 04, 2023 at 04:20:03PM +0800, Lianbo Jiang wrote:
Hi, Alexander
Hi Lianbo!
Thank you for the patch set.
...
I did the test, and the patches can work on old kernel versions.
However,
since VA and PA are the same on S390x, it's hard to determine if all changes
have been made.
In addition, I can not test the patch 3, seems that the current kernel does
not support this feature(I tried it on the kernel 6.7, did not see the
different VA and PA).
That is correct. There is no version that supports this feature yet.
Anyway, If you still want to have these patches in crash-utility for
now, I
can help ack this one.
Yes, I would prefer to pick it now.
...
For patch 3, I saw a warning:
s390x.c: In function ‘is_read_proc_kcore’:
s390x.c:671:13: warning: unused variable ‘kcore_fd’ [-Wunused-variable]
671 | int kcore_fd, fd;
| ^~~~~~~~
The variable kcore_fd is not used in the patch 3, it would be good to remove
it.
I will send v2 in reply to this mail.
Thank you, Lianbo!
> Thanks
>
> Lianbo
9:12 p.m.
New subject: [PATCH 0/3] s390x: uncouple physical and virtual memory spaces
On 12/4/23 23:02, Alexander Gordeev wrote:
On Mon, Dec 04, 2023 at 04:20:03PM +0800, Lianbo Jiang wrote:
> Hi, Alexander
Hi Lianbo!
> Thank you for the patch set.
...
> I did the test, and the patches can work on old kernel versions. However,
> since VA and PA are the same on S390x, it's hard to determine if all changes
> have been made.
>
> In addition, I can not test the patch 3, seems that the current kernel does
> not support this feature(I tried it on the kernel 6.7, did not see the
> different VA and PA).
That is correct. There is no version that supports this feature yet.
> Anyway, If you still want to have these patches in crash-utility for now, I
> can help ack this one.
Yes, I would prefer to pick it now.
Ok.
...
> For patch 3, I saw a warning:
>
> s390x.c: In function ‘is_read_proc_kcore’:
> s390x.c:671:13: warning: unused variable ‘kcore_fd’ [-Wunused-variable]
> 671 | int kcore_fd, fd;
> | ^~~~~~~~
>
> The variable kcore_fd is not used in the patch 3, it would be good to remove
> it.
I will send v2 in reply to this mail.
The Makefile was modified in the v2, it seems unnecessary.
So I would prefer the v1 series [PATCH 0/3] with the above warning
fixes. Given that, for the v1: Ack
Thanks
Lianbo
Thank you, Lianbo!
> Thanks
>
> Lianbo
209
days inactive
210
days old
devel@lists.crash-utility.osci.io
2 comments
2 participants
participants (2)
-
Alexander Gordeev -
Lianbo Jiang