From: chenzhou <chenzhou10@huawei.com>
To: Catalin Marinas <catalin.marinas@arm.com>, <bhe@redhat.com>
Cc: <mingo@redhat.com>, <tglx@linutronix.de>, <rppt@kernel.org>,
<dyoung@redhat.com>, <will@kernel.org>, <nsaenzjulienne@suse.de>,
<corbet@lwn.net>, <John.P.donnelly@oracle.com>,
<bhsharma@redhat.com>, <prabhakar.pkin@gmail.com>,
<horms@verge.net.au>, <robh+dt@kernel.org>, <arnd@arndb.de>,
<james.morse@arm.com>, <xiexiuqi@huawei.com>,
<guohanjun@huawei.com>, <huawei.libin@huawei.com>,
<wangkefeng.wang@huawei.com>, <linux-doc@vger.kernel.org>,
<linux-arm-kernel@lists.infradead.org>,
<linux-kernel@vger.kernel.org>, <kexec@lists.infradead.org>
Subject: Re: [PATCH v14 08/11] arm64: kdump: reimplement crashkernel=X
Date: Fri, 26 Feb 2021 18:31:06 +0800 [thread overview]
Message-ID: <94cc9191-4eff-355f-ff02-1c5da416960e@huawei.com> (raw)
In-Reply-To: <20210224160408.GC28965@arm.com>
On 2021/2/25 0:04, Catalin Marinas wrote:
> On Sat, Jan 30, 2021 at 03:10:22PM +0800, Chen Zhou wrote:
>> There are following issues in arm64 kdump:
>> 1. We use crashkernel=X to reserve crashkernel below 4G, which
>> will fail when there is no enough low memory.
>> 2. If reserving crashkernel above 4G, in this case, crash dump
>> kernel will boot failure because there is no low memory available
>> for allocation.
>>
>> To solve these issues, change the behavior of crashkernel=X and
>> introduce crashkernel=X,[high,low]. crashkernel=X tries low allocation
>> in DMA zone, and fall back to high allocation if it fails.
>> We can also use "crashkernel=X,high" to select a region above DMA zone,
>> which also tries to allocate at least 256M in DMA zone automatically.
>> "crashkernel=Y,low" can be used to allocate specified size low memory.
>>
>> Another minor change, there may be two regions reserved for crash
>> dump kernel, in order to distinct from the high region and make no
>> effect to the use of existing kexec-tools, rename the low region as
>> "Crash kernel (low)".
> I think we discussed this but I don't remember the conclusion. Is this
> only renamed conditionally so that we don't break current kexec-tools?
Yes.
>
> IOW, assuming that the full crashkernel region is reserved below 4GB,
> does the "(low)" suffix still appear or it's only if a high region is
> additionally reserved?
Suffix "low" only appear if a high region is additionally reserved.
>
>> diff --git a/arch/arm64/include/asm/kexec.h b/arch/arm64/include/asm/kexec.h
>> index 3f6ecae0bc68..f0caed0cb5e1 100644
>> --- a/arch/arm64/include/asm/kexec.h
>> +++ b/arch/arm64/include/asm/kexec.h
>> @@ -96,6 +96,10 @@ static inline void crash_prepare_suspend(void) {}
>> static inline void crash_post_resume(void) {}
>> #endif
>>
>> +#ifdef CONFIG_KEXEC_CORE
>> +extern void __init reserve_crashkernel(void);
>> +#endif
> Why not have this in some generic header?
>
>> diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c
>> index c18aacde8bb0..69c592c546de 100644
>> --- a/arch/arm64/kernel/setup.c
>> +++ b/arch/arm64/kernel/setup.c
>> @@ -238,7 +238,18 @@ static void __init request_standard_resources(void)
>> kernel_data.end <= res->end)
>> request_resource(res, &kernel_data);
>> #ifdef CONFIG_KEXEC_CORE
>> - /* Userspace will find "Crash kernel" region in /proc/iomem. */
>> + /*
>> + * Userspace will find "Crash kernel" or "Crash kernel (low)"
>> + * region in /proc/iomem.
>> + * In order to distinct from the high region and make no effect
>> + * to the use of existing kexec-tools, rename the low region as
>> + * "Crash kernel (low)".
>> + */
>> + if (crashk_low_res.end && crashk_low_res.start >= res->start &&
>> + crashk_low_res.end <= res->end) {
>> + crashk_low_res.name = "Crash kernel (low)";
>> + request_resource(res, &crashk_low_res);
>> + }
>> if (crashk_res.end && crashk_res.start >= res->start &&
>> crashk_res.end <= res->end)
>> request_resource(res, &crashk_res);
> My reading of the new generic reserve_crashkernel() is that
> crashk_low_res will only be populated if crask_res is above 4GB. If
> that's correct, I'm fine with the renaming here since current systems
> would not get a renamed low reservation (as long as they don't change
> the kernel cmdline).
>
>> diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
>> index 912f64f505f7..d20f5c444ebf 100644
>> --- a/arch/arm64/mm/init.c
>> +++ b/arch/arm64/mm/init.c
>> @@ -35,6 +35,7 @@
>> #include <asm/fixmap.h>
>> #include <asm/kasan.h>
>> #include <asm/kernel-pgtable.h>
>> +#include <asm/kexec.h>
>> #include <asm/memory.h>
>> #include <asm/numa.h>
>> #include <asm/sections.h>
>> @@ -61,66 +62,11 @@ EXPORT_SYMBOL(memstart_addr);
>> */
>> phys_addr_t arm64_dma_phys_limit __ro_after_init;
>>
>> -#ifdef CONFIG_KEXEC_CORE
>> -/*
>> - * reserve_crashkernel() - reserves memory for crash kernel
>> - *
>> - * This function reserves memory area given in "crashkernel=" kernel command
>> - * line parameter. The memory reserved is used by dump capture kernel when
>> - * primary kernel is crashing.
>> - */
>> +#ifndef CONFIG_KEXEC_CORE
>> static void __init reserve_crashkernel(void)
>> {
> [...]
>> }
>> +#endif
> Can we not have the dummy reserve_crashkernel() in the generic code as
> well and avoid the #ifndef here?
You mean put the dummy reserve_crashkernel() and the relate function declaration in some generic header?
Baoquan also mentioned about this.
Now all the arch that support kdump have the dummy reserve_crashkernel() and
function declaration, such as arm/arm64/ppc/s390..
But currently different arch may have different CONFIG and different function declaration about this,
for example,
for s390,
static void __init reserve_crashkernel(void)
{
#ifdef CONFIG_CRASH_DUMP
...
#endif
}
for ppc,
#ifdef CONFIG_KEXEC_CORE
extern void reserve_crashkernel(void);
#else
static inline void reserve_crashkernel(void) { ; }
#endif
If we move these to generic header we need think about:
1. the related config in different arch
2. function declaration(static/non static)
As Baoquan said in patch 9, how about leave with it for now and i try to solve this later?
>
>> #ifdef CONFIG_CRASH_DUMP
>> static int __init early_init_dt_scan_elfcorehdr(unsigned long node,
>> @@ -446,6 +392,14 @@ void __init bootmem_init(void)
>> * reserved, so do it here.
>> */
>> reserve_crashkernel();
>> +#ifdef CONFIG_KEXEC_CORE
>> + /*
>> + * The low region is intended to be used for crash dump kernel devices,
>> + * just mark the low region as "nomap" simply.
>> + */
>> + if (crashk_low_res.end)
>> + memblock_mark_nomap(crashk_low_res.start, resource_size(&crashk_low_res));
>> +#endif
> Do we do something similar for crashk_res?
Not. In the primary kernel(production kernel), we need to use crashk_res memory for crash kernel
elf core header, initrd...
Different with this, the crashk_low_res is only for crash dump kernel devices.
>
> Also, I can see we call crash_exclude_mem_range() only for crashk_res.
> Do we need to do this for crashk_low_res as well?
You are right, i missed about this. Will do in next version.
Thanks,
Chen Zhou
>
next prev parent reply other threads:[~2021-02-26 10:35 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-01-30 7:10 [PATCH v14 00/11] support reserving crashkernel above 4G on arm64 kdump Chen Zhou
2021-01-30 7:10 ` [PATCH v14 01/11] x86: kdump: replace the hard-coded alignment with macro CRASH_ALIGN Chen Zhou
2021-02-18 3:29 ` Baoquan He
2021-02-24 14:19 ` Catalin Marinas
2021-02-25 7:25 ` Baoquan He
2021-02-26 6:45 ` chenzhou
2021-02-26 15:38 ` Eric W. Biederman
2021-03-02 7:43 ` Baoquan He
2021-03-29 2:34 ` chenzhou
2021-01-30 7:10 ` [PATCH v14 02/11] x86: kdump: make the lower bound of crash kernel reservation consistent Chen Zhou
2021-02-18 3:33 ` Baoquan He
2021-02-24 14:35 ` Catalin Marinas
2021-02-25 7:08 ` Baoquan He
2021-02-25 14:42 ` Catalin Marinas
2021-02-25 15:44 ` Baoquan He
2021-02-26 7:32 ` chenzhou
2021-01-30 7:10 ` [PATCH v14 03/11] x86: kdump: use macro CRASH_ADDR_LOW_MAX in functions reserve_crashkernel() Chen Zhou
2021-02-18 8:23 ` Baoquan He
2021-01-30 7:10 ` [PATCH v14 04/11] x86: kdump: move xen_pv_domain() check and insert_resource() to setup_arch() Chen Zhou
2021-02-18 4:14 ` Baoquan He
2021-01-30 7:10 ` [PATCH v14 05/11] x86: kdump: move reserve_crashkernel[_low]() into crash_core.c Chen Zhou
2021-01-30 7:10 ` [PATCH v14 06/11] x86/elf: Move vmcore_elf_check_arch_cross to arch/x86/include/asm/elf.h Chen Zhou
2021-02-18 6:31 ` Baoquan He
2021-02-18 7:05 ` chenzhou
2021-01-30 7:10 ` [PATCH v14 07/11] arm64: kdump: introduce some macroes for crash kernel reservation Chen Zhou
2021-02-04 16:20 ` Nicolas Saenz Julienne
2021-02-04 16:27 ` Nicolas Saenz Julienne
2021-01-30 7:10 ` [PATCH v14 08/11] arm64: kdump: reimplement crashkernel=X Chen Zhou
2021-02-24 16:04 ` Catalin Marinas
2021-02-26 10:31 ` chenzhou [this message]
2021-02-26 10:43 ` chenzhou
2021-01-30 7:10 ` [PATCH v14 09/11] x86, arm64: Add ARCH_WANT_RESERVE_CRASH_KERNEL config Chen Zhou
2021-02-18 7:31 ` Baoquan He
2021-02-18 7:40 ` Baoquan He
2021-02-18 8:35 ` Baoquan He
2021-02-20 3:22 ` chenzhou
2021-01-30 7:10 ` [PATCH v14 10/11] arm64: kdump: add memory for devices by DT property linux,usable-memory-range Chen Zhou
2021-01-30 7:10 ` [PATCH v14 11/11] kdump: update Documentation about crashkernel Chen Zhou
2021-01-30 17:53 ` Randy Dunlap
2021-02-04 1:53 ` chenzhou
2021-02-18 8:40 ` Baoquan He
2021-02-20 3:25 ` chenzhou
2021-02-08 6:46 ` [PATCH v14 00/11] support reserving crashkernel above 4G on arm64 kdump chenzhou
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=94cc9191-4eff-355f-ff02-1c5da416960e@huawei.com \
--to=chenzhou10@huawei.com \
--cc=John.P.donnelly@oracle.com \
--cc=arnd@arndb.de \
--cc=bhe@redhat.com \
--cc=bhsharma@redhat.com \
--cc=catalin.marinas@arm.com \
--cc=corbet@lwn.net \
--cc=dyoung@redhat.com \
--cc=guohanjun@huawei.com \
--cc=horms@verge.net.au \
--cc=huawei.libin@huawei.com \
--cc=james.morse@arm.com \
--cc=kexec@lists.infradead.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=nsaenzjulienne@suse.de \
--cc=prabhakar.pkin@gmail.com \
--cc=robh+dt@kernel.org \
--cc=rppt@kernel.org \
--cc=tglx@linutronix.de \
--cc=wangkefeng.wang@huawei.com \
--cc=will@kernel.org \
--cc=xiexiuqi@huawei.com \
/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).