linux-kernel - Re: [RFC PATCH 10/10] powerpc/fsl_booke/kaslr: dump out kernel offset information on panic

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: <599f9ff9-9799-b57e-4c88-cf9c60b94a46@huawei.com>
Date:   Mon, 29 Jul 2019 22:08:30 +0800
From:   Jason Yan <yanaijie@...wei.com>
To:     Christophe Leroy <christophe.leroy@....fr>, <mpe@...erman.id.au>,
        <linuxppc-dev@...ts.ozlabs.org>, <diana.craciun@....com>,
        <benh@...nel.crashing.org>, <paulus@...ba.org>,
        <npiggin@...il.com>, <keescook@...omium.org>,
        <kernel-hardening@...ts.openwall.com>
CC:     <linux-kernel@...r.kernel.org>, <wangkefeng.wang@...wei.com>,
        <yebin10@...wei.com>, <thunder.leizhen@...wei.com>,
        <jingxiangfeng@...wei.com>, <fanchengyang@...wei.com>
Subject: Re: [RFC PATCH 10/10] powerpc/fsl_booke/kaslr: dump out kernel offset
 information on panic


On 2019/7/29 19:43, Christophe Leroy wrote:
> 
> 
> Le 17/07/2019 à 10:06, Jason Yan a écrit :
>> When kaslr is enabled, the kernel offset is different for every boot.
>> This brings some difficult to debug the kernel. Dump out the kernel
>> offset when panic so that we can easily debug the kernel.
>>
>> Signed-off-by: Jason Yan <yanaijie@...wei.com>
>> Cc: Diana Craciun <diana.craciun@....com>
>> Cc: Michael Ellerman <mpe@...erman.id.au>
>> Cc: Christophe Leroy <christophe.leroy@....fr>
>> Cc: Benjamin Herrenschmidt <benh@...nel.crashing.org>
>> Cc: Paul Mackerras <paulus@...ba.org>
>> Cc: Nicholas Piggin <npiggin@...il.com>
>> Cc: Kees Cook <keescook@...omium.org>
>> ---
>>   arch/powerpc/include/asm/page.h     |  5 +++++
>>   arch/powerpc/kernel/machine_kexec.c |  1 +
>>   arch/powerpc/kernel/setup-common.c  | 23 +++++++++++++++++++++++
>>   3 files changed, 29 insertions(+)
>>
>> diff --git a/arch/powerpc/include/asm/page.h 
>> b/arch/powerpc/include/asm/page.h
>> index 60a68d3a54b1..cd3ac530e58d 100644
>> --- a/arch/powerpc/include/asm/page.h
>> +++ b/arch/powerpc/include/asm/page.h
>> @@ -317,6 +317,11 @@ struct vm_area_struct;
>>   extern unsigned long kimage_vaddr;
>> +static inline unsigned long kaslr_offset(void)
>> +{
>> +    return kimage_vaddr - KERNELBASE;
>> +}
>> +
>>   #include <asm-generic/memory_model.h>
>>   #endif /* __ASSEMBLY__ */
>>   #include <asm/slice.h>
>> diff --git a/arch/powerpc/kernel/machine_kexec.c 
>> b/arch/powerpc/kernel/machine_kexec.c
>> index c4ed328a7b96..078fe3d76feb 100644
>> --- a/arch/powerpc/kernel/machine_kexec.c
>> +++ b/arch/powerpc/kernel/machine_kexec.c
>> @@ -86,6 +86,7 @@ void arch_crash_save_vmcoreinfo(void)
>>       VMCOREINFO_STRUCT_SIZE(mmu_psize_def);
>>       VMCOREINFO_OFFSET(mmu_psize_def, shift);
>>   #endif
>> +    vmcoreinfo_append_str("KERNELOFFSET=%lx\n", kaslr_offset());
>>   }
>>   /*
>> diff --git a/arch/powerpc/kernel/setup-common.c 
>> b/arch/powerpc/kernel/setup-common.c
>> index 1f8db666468d..49e540c0adeb 100644
>> --- a/arch/powerpc/kernel/setup-common.c
>> +++ b/arch/powerpc/kernel/setup-common.c
>> @@ -715,12 +715,35 @@ static struct notifier_block ppc_panic_block = {
>>       .priority = INT_MIN /* may not return; must be done last */
>>   };
>> +/*
>> + * Dump out kernel offset information on panic.
>> + */
>> +static int dump_kernel_offset(struct notifier_block *self, unsigned 
>> long v,
>> +                  void *p)
>> +{
>> +    const unsigned long offset = kaslr_offset();
>> +
>> +    if (IS_ENABLED(CONFIG_RANDOMIZE_BASE) && offset > 0)
>> +        pr_emerg("Kernel Offset: 0x%lx from 0x%lx\n",
>> +             offset, KERNELBASE);
>> +    else
>> +        pr_emerg("Kernel Offset: disabled\n");
> 
> Do we really need that else branch ?
> 
> Why not just make the below atomic_notifier_chain_register() 
> conditionnal to IS_ENABLED(CONFIG_RANDOMIZE_BASE) && offset > 0
> and not print anything otherwise ?
> 

I'm trying to keep the same fashion as x86/arm64 do. But I agree
with you that it's simpler to not print anything else if not randomized.

> Christophe
> 
>> +
>> +    return 0;
>> +}
>> +
>> +static struct notifier_block kernel_offset_notifier = {
>> +    .notifier_call = dump_kernel_offset
>> +};
>> +
>>   void __init setup_panic(void)
>>   {
>>       /* PPC64 always does a hard irq disable in its panic handler */
>>       if (!IS_ENABLED(CONFIG_PPC64) && !ppc_md.panic)
>>           return;
>>       atomic_notifier_chain_register(&panic_notifier_list, 
>> &ppc_panic_block);
>> +    atomic_notifier_chain_register(&panic_notifier_list,
>> +                       &kernel_offset_notifier);
>>   }
>>   #ifdef CONFIG_CHECK_CACHE_COHERENCY
>>
> 
> .
>