From: Sourabh Jain <sourabhjain@linux.ibm.com>
To: Aneesh Kumar K V <aneesh.kumar@kernel.org>,
Michael Ellerman <mpe@ellerman.id.au>,
Aneesh Kumar K V <aneesh.kumar@linux.ibm.com>,
linuxppc-dev@ozlabs.org
Cc: Aditya Gupta <adityag@linux.ibm.com>,
Mahesh Salgaonkar <mahesh@linux.ibm.com>,
Hari Bathini <hbathini@linux.ibm.com>
Subject: Re: [PATCH v5 1/3] powerpc: make fadump resilient with memory add/remove events
Date: Fri, 24 Nov 2023 10:50:14 +0530 [thread overview]
Message-ID: <a5cf8f63-2dda-41d5-9231-aca25b7e6960@linux.ibm.com> (raw)
In-Reply-To: <47a119c1-05e9-47fb-9cc7-85aa7e1e4ae7@kernel.org>
Hello Aneesh,
On 22/11/23 17:50, Aneesh Kumar K V wrote:
> On 11/22/23 4:05 PM, Sourabh Jain wrote:
>> Hello Michael,
>>
>>
>> On 22/11/23 10:47, Michael Ellerman wrote:
>>> Aneesh Kumar K V <aneesh.kumar@linux.ibm.com> writes:
>>> ...
>>>> I am not sure whether we need to add all the complexity to enable supporting different fadump kernel
>>>> version. Is that even a possible use case with fadump? Can't we always assume that with fadump the
>>>> crash kernel and fadump kernel will be same version?
>>> How sure are we of that?
>>>
>>> Don't we go through grub when we boot into the 2nd kernel. And so
>>> couldn't it choose to boot a different kernel, for whatever reason.
>>>
>>> I don't think we need to support different pt_reg / cpumask sizes, but
>>> requiring the exact same kernel version is too strict I think.
>> Agree.
>>> But maybe I'm wrong. Would be good to hear what distro folks think.
>> How about checking fadump crash info header compatibility in the following way?
>>
>> static bool is_fadump_header_compatible(struct fadump_crash_info_header *fdh)
>> {
>> if (fdh->magic_number == FADUMP_CRASH_INFO_MAGIC_OLD) {
>> pr_err("Old magic number, can't process the dump.");
>> return false;
>> }
>>
>> if (fdh->magic_number != FADUMP_CRASH_INFO_MAGIC) {
>> pr_err("Fadump header is corrupted.");
>> return false;
>> }
>>
>> /*
>> * If the kernel version of the first/crashed kernel and the second/fadump
>> * kernel is not same, then only collect the dump if the size of all
>> * non-primitive type members of the fadump header is the same across kernels.
>> */
>> if (strcmp(fdh->kernel_version, init_uts_ns.name.release)) {
>> if (fdh->pt_regs_sz != sizeof(struct pt_regs) || fdh->cpu_mask_sz != sizeof(struct cpumask)) {
>> pr_err("Fadump header size mismatch.\n")
>> return false;
>> } else
>> pr_warn("Kernel version mismatch; dump data is unreliable.\n");
>> }
>>
> You also want a fdh->version check?
Even though we don't have any action against an fdh->version right now,
I think I should
check the fadump header version. Currently, if the version doesn't
match, it means the
header is corrupted.
> I am still not sure you need the kernel_version check. IMHO that is too strict
> and can hit that check where you have kexec kernel which is not installed in /boot crashing?
If the kernel versions mismatch, we still collect the dump if the
`pt_regs` and `cpu_mask`
sizes are the same across the kernels. The kernel version check is just
to warn users that
the collected dump may be unreliable.
Should I remove the kernel_version filed from fadump crash info header
and remove the
the kernel version check while processing the kernel dump?
Thanks,
Sourabh Jain
next prev parent reply other threads:[~2023-11-24 5:21 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-10-29 12:45 [PATCH v5 0/3] powerpc: make fadump resilient with memory add/remove events Sourabh Jain
2023-10-29 12:45 ` [PATCH v5 1/3] " Sourabh Jain
2023-11-09 12:14 ` Michael Ellerman
2023-11-13 6:42 ` Sourabh Jain
2023-11-15 4:44 ` Aneesh Kumar K.V
2023-11-17 4:33 ` Sourabh Jain
2023-11-17 5:31 ` Aneesh Kumar K V
2023-11-17 6:14 ` Hari Bathini
2023-11-22 5:17 ` Michael Ellerman
2023-11-22 10:35 ` Sourabh Jain
2023-11-22 12:20 ` Aneesh Kumar K V
2023-11-24 5:20 ` Sourabh Jain [this message]
2023-11-22 12:52 ` Michael Ellerman
2023-11-24 7:21 ` Sourabh Jain
2023-10-29 12:45 ` [PATCH v5 2/3] powerpc/fadump: add hotplug_ready sysfs interface Sourabh Jain
2023-10-29 12:45 ` [PATCH v5 3/3] Documentation/powerpc: update fadump implementation details Sourabh Jain
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=a5cf8f63-2dda-41d5-9231-aca25b7e6960@linux.ibm.com \
--to=sourabhjain@linux.ibm.com \
--cc=adityag@linux.ibm.com \
--cc=aneesh.kumar@kernel.org \
--cc=aneesh.kumar@linux.ibm.com \
--cc=hbathini@linux.ibm.com \
--cc=linuxppc-dev@ozlabs.org \
--cc=mahesh@linux.ibm.com \
--cc=mpe@ellerman.id.au \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).