linux-kernel - Re: [PATCH v8 7/7] x86/crash: Add x86 crash hotplug support for kexec

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <fed45ff2-0f1d-6c55-244f-357e103fd775@oracle.com>
Date:   Tue, 31 May 2022 17:18:34 -0500
From:   Eric DeVolder <eric.devolder@...cle.com>
To:     Sourabh Jain <sourabhjain@...ux.ibm.com>,
        linux-kernel@...r.kernel.org, x86@...nel.org,
        kexec@...ts.infradead.org, ebiederm@...ssion.com,
        dyoung@...hat.com, bhe@...hat.com, vgoyal@...hat.com
Cc:     tglx@...utronix.de, mingo@...hat.com, bp@...en8.de,
        dave.hansen@...ux.intel.com, hpa@...or.com,
        nramas@...ux.microsoft.com, thomas.lendacky@....com,
        robh@...nel.org, efault@....de, rppt@...nel.org, david@...hat.com,
        konrad.wilk@...cle.com, boris.ostrovsky@...cle.com
Subject: Re: [PATCH v8 7/7] x86/crash: Add x86 crash hotplug support for
 kexec_load



On 5/25/22 09:26, Sourabh Jain wrote:
> Hello Eric,
> 
> On 06/05/22 00:16, Eric DeVolder wrote:
>> For kexec_file_load support, the loading of the crash kernel occurs
>> entirely within the kernel, and as such the elfcorehdr is readily
>> identified (so that it can be modified upon hotplug events).
>>
>> This change enables support for kexec_load by identifying the
>> elfcorehdr segment in the arch_crash_handle_hotplug_event(),
>> if it has not already been identified.
>>
>> Signed-off-by: Eric DeVolder <eric.devolder@...cle.com>
>> Acked-by: Baoquan He <bhe@...hat.com>
>> ---
>>   arch/x86/kernel/crash.c | 24 ++++++++++++++++++++++++
>>   1 file changed, 24 insertions(+)
>>
>> diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c
>> index 951ef365f0a7..845d7c77854d 100644
>> --- a/arch/x86/kernel/crash.c
>> +++ b/arch/x86/kernel/crash.c
>> @@ -485,6 +485,30 @@ void arch_crash_handle_hotplug_event(struct kimage *image,
>>       void *elfbuf = NULL;
>>       unsigned long mem, memsz;
>> +    /*
>> +     * When the struct kimage is alloced, it is wiped to zero, so
>> +     * the elfcorehdr_index_valid defaults to false. It is set on the
>> +     * kexec_file_load path, or here for kexec_load, if not already
>> +     * identified.
>> +     */
>> +    if (!image->elfcorehdr_index_valid) {
>> +        unsigned int n;
>> +
>> +        for (n = 0; n < image->nr_segments; n++) {
>> +            mem = image->segment[n].mem;
>> +            memsz = image->segment[n].memsz;
>> +            ptr = map_crash_pages(mem, memsz);
>> +            if (ptr) {
>> +                /* The segment containing elfcorehdr */
>> +                if (memcmp(ptr, ELFMAG, SELFMAG) == 0) {
>> +                    image->elfcorehdr_index = (int)n;
>> +                    image->elfcorehdr_index_valid = true;
> 
> How about finding elfcorehdr index on kexec_load path post kimage_load_segment call in
> do_kexec_load (kernel/kexec.c) or other suitable place? This way we can avoid checking for
> elfcorehdr index for every hotplug. Also we might not need image->elfcorehdr_index_valid.

That would be a viable place to put it. However, a couple of notes.

This code actually works for both kexec_load and kexec_file_load paths (how I originally used this 
code); so if we were to move this to a more common location, I'd be in favor of a location that 
serves both paths (ie handle_hotplug_event() makes sense to me). That would eliminate the need for 
setting the index in the kexec_file_load paths.

Also, I don't see a scenario where elfcorehdr_index_valid can be eliminated; the index isn't always 
valid. What are you thinking?

And the check is rather low cost being the check of a boolean.

eric
> 
> Thanks,
> Sourabh Jain
> 
>