* [PATCH v21 1/5] kdump: return -ENOENT if required cmdline option does not exist
2022-02-27 3:07 [PATCH v21 0/5] support reserving crashkernel above 4G on arm64 kdump Zhen Lei
@ 2022-02-27 3:07 ` Zhen Lei
2022-03-15 11:57 ` Baoquan He
2022-03-16 5:39 ` Baoquan He
2022-02-27 3:07 ` [PATCH v21 2/5] arm64: Use insert_resource() to simplify code Zhen Lei
` (4 subsequent siblings)
5 siblings, 2 replies; 27+ messages in thread
From: Zhen Lei @ 2022-02-27 3:07 UTC (permalink / raw)
To: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Baoquan He,
Vivek Goyal, Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc
Cc: Zhen Lei, Randy Dunlap, Feng Zhou, Kefeng Wang, Chen Zhou,
John Donnelly, Dave Kleikamp
The crashkernel=Y,low is an optional command-line option. When it doesn't
exist, kernel will try to allocate minimum required memory below 4G
automatically. Give it a unique error code to distinguish it from other
error scenarios.
Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
---
kernel/crash_core.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/kernel/crash_core.c b/kernel/crash_core.c
index 256cf6db573cd09..4d57c03714f4e13 100644
--- a/kernel/crash_core.c
+++ b/kernel/crash_core.c
@@ -243,9 +243,8 @@ static int __init __parse_crashkernel(char *cmdline,
*crash_base = 0;
ck_cmdline = get_last_crashkernel(cmdline, name, suffix);
-
if (!ck_cmdline)
- return -EINVAL;
+ return -ENOENT;
ck_cmdline += strlen(name);
--
2.25.1
^ permalink raw reply related [flat|nested] 27+ messages in thread
* Re: [PATCH v21 1/5] kdump: return -ENOENT if required cmdline option does not exist
2022-02-27 3:07 ` [PATCH v21 1/5] kdump: return -ENOENT if required cmdline option does not exist Zhen Lei
@ 2022-03-15 11:57 ` Baoquan He
2022-03-15 12:21 ` Baoquan He
2022-03-16 5:39 ` Baoquan He
1 sibling, 1 reply; 27+ messages in thread
From: Baoquan He @ 2022-03-15 11:57 UTC (permalink / raw)
To: Zhen Lei
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 02/27/22 at 11:07am, Zhen Lei wrote:
> The crashkernel=Y,low is an optional command-line option. When it doesn't
> exist, kernel will try to allocate minimum required memory below 4G
> automatically. Give it a unique error code to distinguish it from other
> error scenarios.
>
> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
> ---
> kernel/crash_core.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/kernel/crash_core.c b/kernel/crash_core.c
> index 256cf6db573cd09..4d57c03714f4e13 100644
> --- a/kernel/crash_core.c
> +++ b/kernel/crash_core.c
> @@ -243,9 +243,8 @@ static int __init __parse_crashkernel(char *cmdline,
> *crash_base = 0;
>
> ck_cmdline = get_last_crashkernel(cmdline, name, suffix);
> -
> if (!ck_cmdline)
> - return -EINVAL;
> + return -ENOENT;
Firstly, I am not sure if '-ENOENT' is a right value to return. From the
code comment of ENOENT, it's used for file or dir?
#define ENOENT 2 /* No such file or directory */
Secondly, we ever discussed the case including
- no crashkernel=,low is provided;
- messy code is provied, e.g crashkernel=aaaaaabbbb,low
The 2nd one is not handled in this patchset. How about taking the
handling into another round of patches. This patchset just adds
crashkernel=,high purely.
>
> ck_cmdline += strlen(name);
>
> --
> 2.25.1
>
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 1/5] kdump: return -ENOENT if required cmdline option does not exist
2022-03-15 11:57 ` Baoquan He
@ 2022-03-15 12:21 ` Baoquan He
2022-03-15 13:32 ` Leizhen (ThunderTown)
0 siblings, 1 reply; 27+ messages in thread
From: Baoquan He @ 2022-03-15 12:21 UTC (permalink / raw)
To: Zhen Lei
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 03/15/22 at 07:57pm, Baoquan He wrote:
> On 02/27/22 at 11:07am, Zhen Lei wrote:
> > The crashkernel=Y,low is an optional command-line option. When it doesn't
> > exist, kernel will try to allocate minimum required memory below 4G
> > automatically. Give it a unique error code to distinguish it from other
> > error scenarios.
> >
> > Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
> > ---
> > kernel/crash_core.c | 3 +--
> > 1 file changed, 1 insertion(+), 2 deletions(-)
> >
> > diff --git a/kernel/crash_core.c b/kernel/crash_core.c
> > index 256cf6db573cd09..4d57c03714f4e13 100644
> > --- a/kernel/crash_core.c
> > +++ b/kernel/crash_core.c
> > @@ -243,9 +243,8 @@ static int __init __parse_crashkernel(char *cmdline,
> > *crash_base = 0;
> >
> > ck_cmdline = get_last_crashkernel(cmdline, name, suffix);
> > -
> > if (!ck_cmdline)
> > - return -EINVAL;
> > + return -ENOENT;
>
> Firstly, I am not sure if '-ENOENT' is a right value to return. From the
> code comment of ENOENT, it's used for file or dir?
> #define ENOENT 2 /* No such file or directory */
>
> Secondly, we ever discussed the case including
> - no crashkernel=,low is provided;
> - messy code is provied, e.g crashkernel=aaaaaabbbb,low
Checking the 3rd pach, this is handled. Take back my below words,
continue reviewing.
>
> The 2nd one is not handled in this patchset. How about taking the
> handling into another round of patches. This patchset just adds
> crashkernel=,high purely.
>
> >
> > ck_cmdline += strlen(name);
> >
> > --
> > 2.25.1
> >
>
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 1/5] kdump: return -ENOENT if required cmdline option does not exist
2022-03-15 12:21 ` Baoquan He
@ 2022-03-15 13:32 ` Leizhen (ThunderTown)
2022-03-16 5:17 ` Baoquan He
0 siblings, 1 reply; 27+ messages in thread
From: Leizhen (ThunderTown) @ 2022-03-15 13:32 UTC (permalink / raw)
To: Baoquan He
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 2022/3/15 20:21, Baoquan He wrote:
> On 03/15/22 at 07:57pm, Baoquan He wrote:
>> On 02/27/22 at 11:07am, Zhen Lei wrote:
>>> The crashkernel=Y,low is an optional command-line option. When it doesn't
>>> exist, kernel will try to allocate minimum required memory below 4G
>>> automatically. Give it a unique error code to distinguish it from other
>>> error scenarios.
>>>
>>> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
>>> ---
>>> kernel/crash_core.c | 3 +--
>>> 1 file changed, 1 insertion(+), 2 deletions(-)
>>>
>>> diff --git a/kernel/crash_core.c b/kernel/crash_core.c
>>> index 256cf6db573cd09..4d57c03714f4e13 100644
>>> --- a/kernel/crash_core.c
>>> +++ b/kernel/crash_core.c
>>> @@ -243,9 +243,8 @@ static int __init __parse_crashkernel(char *cmdline,
>>> *crash_base = 0;
>>>
>>> ck_cmdline = get_last_crashkernel(cmdline, name, suffix);
>>> -
>>> if (!ck_cmdline)
>>> - return -EINVAL;
>>> + return -ENOENT;
>>
>> Firstly, I am not sure if '-ENOENT' is a right value to return. From the
>> code comment of ENOENT, it's used for file or dir?
>> #define ENOENT 2 /* No such file or directory */
This error code does not return to user mode, so there is no problem.
There are a lot of places in the kernel that are used this way. For example:
int stop_one_cpu(unsigned int cpu, cpu_stop_fn_t fn, void *arg)
{
if (!cpu_stop_queue_work(cpu, &work))
return -ENOENT;
>>
>> Secondly, we ever discussed the case including
>> - no crashkernel=,low is provided;
>> - messy code is provied, e.g crashkernel=aaaaaabbbb,low
>
> Checking the 3rd pach, this is handled. Take back my below words,
> continue reviewing.
Yes.
>
>>
>> The 2nd one is not handled in this patchset. How about taking the
>> handling into another round of patches. This patchset just adds
>> crashkernel=,high purely.
>>
>>>
>>> ck_cmdline += strlen(name);
>>>
>>> --
>>> 2.25.1
>>>
>>
>
> .
>
--
Regards,
Zhen Lei
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 1/5] kdump: return -ENOENT if required cmdline option does not exist
2022-03-15 13:32 ` Leizhen (ThunderTown)
@ 2022-03-16 5:17 ` Baoquan He
0 siblings, 0 replies; 27+ messages in thread
From: Baoquan He @ 2022-03-16 5:17 UTC (permalink / raw)
To: Leizhen (ThunderTown)
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 03/15/22 at 09:32pm, Leizhen (ThunderTown) wrote:
>
>
> On 2022/3/15 20:21, Baoquan He wrote:
> > On 03/15/22 at 07:57pm, Baoquan He wrote:
> >> On 02/27/22 at 11:07am, Zhen Lei wrote:
> >>> The crashkernel=Y,low is an optional command-line option. When it doesn't
> >>> exist, kernel will try to allocate minimum required memory below 4G
> >>> automatically. Give it a unique error code to distinguish it from other
> >>> error scenarios.
> >>>
> >>> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
> >>> ---
> >>> kernel/crash_core.c | 3 +--
> >>> 1 file changed, 1 insertion(+), 2 deletions(-)
> >>>
> >>> diff --git a/kernel/crash_core.c b/kernel/crash_core.c
> >>> index 256cf6db573cd09..4d57c03714f4e13 100644
> >>> --- a/kernel/crash_core.c
> >>> +++ b/kernel/crash_core.c
> >>> @@ -243,9 +243,8 @@ static int __init __parse_crashkernel(char *cmdline,
> >>> *crash_base = 0;
> >>>
> >>> ck_cmdline = get_last_crashkernel(cmdline, name, suffix);
> >>> -
> >>> if (!ck_cmdline)
> >>> - return -EINVAL;
> >>> + return -ENOENT;
> >>
> >> Firstly, I am not sure if '-ENOENT' is a right value to return. From the
> >> code comment of ENOENT, it's used for file or dir?
> >> #define ENOENT 2 /* No such file or directory */
>
> This error code does not return to user mode, so there is no problem.
> There are a lot of places in the kernel that are used this way. For example:
>
> int stop_one_cpu(unsigned int cpu, cpu_stop_fn_t fn, void *arg)
> {
> if (!cpu_stop_queue_work(cpu, &work))
> return -ENOENT;
OK, it's fine to me. Thanks for the investigation.
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 1/5] kdump: return -ENOENT if required cmdline option does not exist
2022-02-27 3:07 ` [PATCH v21 1/5] kdump: return -ENOENT if required cmdline option does not exist Zhen Lei
2022-03-15 11:57 ` Baoquan He
@ 2022-03-16 5:39 ` Baoquan He
2022-03-16 6:15 ` Leizhen (ThunderTown)
1 sibling, 1 reply; 27+ messages in thread
From: Baoquan He @ 2022-03-16 5:39 UTC (permalink / raw)
To: Zhen Lei
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 02/27/22 at 11:07am, Zhen Lei wrote:
> The crashkernel=Y,low is an optional command-line option. When it doesn't
> exist, kernel will try to allocate minimum required memory below 4G
> automatically. Give it a unique error code to distinguish it from other
> error scenarios.
This log is a little confusing. __parse_crashkernel() has three callers.
- parse_crashkernel()
- parse_crashkernel_high()
- parse_crashkernel_low()
How about tuning the git log as below:
==================
According to the current crashkernel=Y,low support in other ARCHes, it's
an optional command-line option. When it doesn't exist, kernel will try
to allocate minimum required memory below 4G automatically.
However, __parse_crashkernel() returns '-EINVAL' for all error cases. It
can't distinguish the nonexistent option from invalid option.
Change __parse_crashkernel() to return '-ENOENT' for the nonexistent option
case. With this change, crashkernel,low memory will take the default
value if crashkernel=,low is not specified; while crashkernel reservation
will fail and bail out if an invalid option is specified.
==================
>
> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
> ---
> kernel/crash_core.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
>
> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
> ---
> kernel/crash_core.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/kernel/crash_core.c b/kernel/crash_core.c
> index 256cf6db573cd09..4d57c03714f4e13 100644
> --- a/kernel/crash_core.c
> +++ b/kernel/crash_core.c
> @@ -243,9 +243,8 @@ static int __init __parse_crashkernel(char *cmdline,
> *crash_base = 0;
>
> ck_cmdline = get_last_crashkernel(cmdline, name, suffix);
> -
> if (!ck_cmdline)
> - return -EINVAL;
> + return -ENOENT;
>
> ck_cmdline += strlen(name);
>
> --
> 2.25.1
>
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 1/5] kdump: return -ENOENT if required cmdline option does not exist
2022-03-16 5:39 ` Baoquan He
@ 2022-03-16 6:15 ` Leizhen (ThunderTown)
0 siblings, 0 replies; 27+ messages in thread
From: Leizhen (ThunderTown) @ 2022-03-16 6:15 UTC (permalink / raw)
To: Baoquan He
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 2022/3/16 13:39, Baoquan He wrote:
> On 02/27/22 at 11:07am, Zhen Lei wrote:
>> The crashkernel=Y,low is an optional command-line option. When it doesn't
>> exist, kernel will try to allocate minimum required memory below 4G
>> automatically. Give it a unique error code to distinguish it from other
>> error scenarios.
>
> This log is a little confusing. __parse_crashkernel() has three callers.
> - parse_crashkernel()
> - parse_crashkernel_high()
> - parse_crashkernel_low()
>
> How about tuning the git log as below:
Sure. Your description is much clearer than mine.
>
> ==================
> According to the current crashkernel=Y,low support in other ARCHes, it's
> an optional command-line option. When it doesn't exist, kernel will try
> to allocate minimum required memory below 4G automatically.
>
> However, __parse_crashkernel() returns '-EINVAL' for all error cases. It
> can't distinguish the nonexistent option from invalid option.
>
> Change __parse_crashkernel() to return '-ENOENT' for the nonexistent option
> case. With this change, crashkernel,low memory will take the default
> value if crashkernel=,low is not specified; while crashkernel reservation
> will fail and bail out if an invalid option is specified.
> ==================
>
>>
>> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
>> ---
>> kernel/crash_core.c | 3 +--
>> 1 file changed, 1 insertion(+), 2 deletions(-)
>>
>>
>> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
>> ---
>> kernel/crash_core.c | 3 +--
>> 1 file changed, 1 insertion(+), 2 deletions(-)
>>
>> diff --git a/kernel/crash_core.c b/kernel/crash_core.c
>> index 256cf6db573cd09..4d57c03714f4e13 100644
>> --- a/kernel/crash_core.c
>> +++ b/kernel/crash_core.c
>> @@ -243,9 +243,8 @@ static int __init __parse_crashkernel(char *cmdline,
>> *crash_base = 0;
>>
>> ck_cmdline = get_last_crashkernel(cmdline, name, suffix);
>> -
>> if (!ck_cmdline)
>> - return -EINVAL;
>> + return -ENOENT;
>>
>> ck_cmdline += strlen(name);
>>
>> --
>> 2.25.1
>>
>
> .
>
--
Regards,
Zhen Lei
^ permalink raw reply [flat|nested] 27+ messages in thread
* [PATCH v21 2/5] arm64: Use insert_resource() to simplify code
2022-02-27 3:07 [PATCH v21 0/5] support reserving crashkernel above 4G on arm64 kdump Zhen Lei
2022-02-27 3:07 ` [PATCH v21 1/5] kdump: return -ENOENT if required cmdline option does not exist Zhen Lei
@ 2022-02-27 3:07 ` Zhen Lei
2022-02-27 3:07 ` [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X Zhen Lei
` (3 subsequent siblings)
5 siblings, 0 replies; 27+ messages in thread
From: Zhen Lei @ 2022-02-27 3:07 UTC (permalink / raw)
To: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Baoquan He,
Vivek Goyal, Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc
Cc: Zhen Lei, Randy Dunlap, Feng Zhou, Kefeng Wang, Chen Zhou,
John Donnelly, Dave Kleikamp
insert_resource() traverses the subtree layer by layer from the root node
until a proper location is found. Compared with request_resource(), the
parent node does not need to be determined in advance.
In addition, move the insertion of node 'crashk_res' into function
reserve_crashkernel() to make the associated code close together.
Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Acked-by: John Donnelly <john.p.donnelly@oracle.com>
Acked-by: Baoquan He <bhe@redhat.com>
---
arch/arm64/kernel/setup.c | 17 +++--------------
arch/arm64/mm/init.c | 1 +
2 files changed, 4 insertions(+), 14 deletions(-)
diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c
index f70573928f1bff0..a81efcc359e4e78 100644
--- a/arch/arm64/kernel/setup.c
+++ b/arch/arm64/kernel/setup.c
@@ -225,6 +225,8 @@ static void __init request_standard_resources(void)
kernel_code.end = __pa_symbol(__init_begin - 1);
kernel_data.start = __pa_symbol(_sdata);
kernel_data.end = __pa_symbol(_end - 1);
+ insert_resource(&iomem_resource, &kernel_code);
+ insert_resource(&iomem_resource, &kernel_data);
num_standard_resources = memblock.memory.cnt;
res_size = num_standard_resources * sizeof(*standard_resources);
@@ -246,20 +248,7 @@ static void __init request_standard_resources(void)
res->end = __pfn_to_phys(memblock_region_memory_end_pfn(region)) - 1;
}
- request_resource(&iomem_resource, res);
-
- if (kernel_code.start >= res->start &&
- kernel_code.end <= res->end)
- request_resource(res, &kernel_code);
- if (kernel_data.start >= res->start &&
- kernel_data.end <= res->end)
- request_resource(res, &kernel_data);
-#ifdef CONFIG_KEXEC_CORE
- /* Userspace will find "Crash kernel" region in /proc/iomem. */
- if (crashk_res.end && crashk_res.start >= res->start &&
- crashk_res.end <= res->end)
- request_resource(res, &crashk_res);
-#endif
+ insert_resource(&iomem_resource, res);
}
}
diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
index db63cc885771a52..90f276d46b93bc6 100644
--- a/arch/arm64/mm/init.c
+++ b/arch/arm64/mm/init.c
@@ -109,6 +109,7 @@ static void __init reserve_crashkernel(void)
kmemleak_ignore_phys(crash_base);
crashk_res.start = crash_base;
crashk_res.end = crash_base + crash_size - 1;
+ insert_resource(&iomem_resource, &crashk_res);
}
#else
static void __init reserve_crashkernel(void)
--
2.25.1
^ permalink raw reply related [flat|nested] 27+ messages in thread
* [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X
2022-02-27 3:07 [PATCH v21 0/5] support reserving crashkernel above 4G on arm64 kdump Zhen Lei
2022-02-27 3:07 ` [PATCH v21 1/5] kdump: return -ENOENT if required cmdline option does not exist Zhen Lei
2022-02-27 3:07 ` [PATCH v21 2/5] arm64: Use insert_resource() to simplify code Zhen Lei
@ 2022-02-27 3:07 ` Zhen Lei
2022-03-16 12:11 ` Baoquan He
` (2 more replies)
2022-02-27 3:07 ` [PATCH v21 4/5] of: fdt: Add memory for devices by DT property "linux,usable-memory-range" Zhen Lei
` (2 subsequent siblings)
5 siblings, 3 replies; 27+ messages in thread
From: Zhen Lei @ 2022-02-27 3:07 UTC (permalink / raw)
To: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Baoquan He,
Vivek Goyal, Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc
Cc: Zhen Lei, Randy Dunlap, Feng Zhou, Kefeng Wang, Chen Zhou,
John Donnelly, Dave Kleikamp
From: Chen Zhou <chenzhou10@huawei.com>
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.
Signed-off-by: Chen Zhou <chenzhou10@huawei.com>
Co-developed-by: Zhen Lei <thunder.leizhen@huawei.com>
Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
---
arch/arm64/kernel/machine_kexec.c | 9 ++-
arch/arm64/kernel/machine_kexec_file.c | 12 ++-
arch/arm64/mm/init.c | 106 +++++++++++++++++++++++--
3 files changed, 115 insertions(+), 12 deletions(-)
diff --git a/arch/arm64/kernel/machine_kexec.c b/arch/arm64/kernel/machine_kexec.c
index e16b248699d5c3c..19c2d487cb08feb 100644
--- a/arch/arm64/kernel/machine_kexec.c
+++ b/arch/arm64/kernel/machine_kexec.c
@@ -329,8 +329,13 @@ bool crash_is_nosave(unsigned long pfn)
/* in reserved memory? */
addr = __pfn_to_phys(pfn);
- if ((addr < crashk_res.start) || (crashk_res.end < addr))
- return false;
+ if ((addr < crashk_res.start) || (crashk_res.end < addr)) {
+ if (!crashk_low_res.end)
+ return false;
+
+ if ((addr < crashk_low_res.start) || (crashk_low_res.end < addr))
+ return false;
+ }
if (!kexec_crash_image)
return true;
diff --git a/arch/arm64/kernel/machine_kexec_file.c b/arch/arm64/kernel/machine_kexec_file.c
index 59c648d51848886..889951291cc0f9c 100644
--- a/arch/arm64/kernel/machine_kexec_file.c
+++ b/arch/arm64/kernel/machine_kexec_file.c
@@ -65,10 +65,18 @@ static int prepare_elf_headers(void **addr, unsigned long *sz)
/* Exclude crashkernel region */
ret = crash_exclude_mem_range(cmem, crashk_res.start, crashk_res.end);
+ if (ret)
+ goto out;
+
+ if (crashk_low_res.end) {
+ ret = crash_exclude_mem_range(cmem, crashk_low_res.start, crashk_low_res.end);
+ if (ret)
+ goto out;
+ }
- if (!ret)
- ret = crash_prepare_elf64_headers(cmem, true, addr, sz);
+ ret = crash_prepare_elf64_headers(cmem, true, addr, sz);
+out:
kfree(cmem);
return ret;
}
diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
index 90f276d46b93bc6..30ae6638ff54c47 100644
--- a/arch/arm64/mm/init.c
+++ b/arch/arm64/mm/init.c
@@ -65,6 +65,44 @@ EXPORT_SYMBOL(memstart_addr);
phys_addr_t arm64_dma_phys_limit __ro_after_init;
#ifdef CONFIG_KEXEC_CORE
+/* Current arm64 boot protocol requires 2MB alignment */
+#define CRASH_ALIGN SZ_2M
+
+#define CRASH_ADDR_LOW_MAX arm64_dma_phys_limit
+#define CRASH_ADDR_HIGH_MAX memblock.current_limit
+
+/*
+ * This is an empirical value in x86_64 and taken here directly. Please
+ * refer to the code comment in reserve_crashkernel_low() of x86_64 for more
+ * details.
+ */
+#define DEFAULT_CRASH_KERNEL_LOW_SIZE \
+ max(swiotlb_size_or_default() + (8UL << 20), 256UL << 20)
+
+static int __init reserve_crashkernel_low(unsigned long long low_size)
+{
+ unsigned long long low_base;
+
+ /* passed with crashkernel=0,low ? */
+ if (!low_size)
+ return 0;
+
+ low_base = memblock_phys_alloc_range(low_size, CRASH_ALIGN, 0, CRASH_ADDR_LOW_MAX);
+ if (!low_base) {
+ pr_err("cannot allocate crashkernel low memory (size:0x%llx).\n", low_size);
+ return -ENOMEM;
+ }
+
+ pr_info("crashkernel low memory reserved: 0x%08llx - 0x%08llx (%lld MB)\n",
+ low_base, low_base + low_size, low_size >> 20);
+
+ crashk_low_res.start = low_base;
+ crashk_low_res.end = low_base + low_size - 1;
+ insert_resource(&iomem_resource, &crashk_low_res);
+
+ return 0;
+}
+
/*
* reserve_crashkernel() - reserves memory for crash kernel
*
@@ -75,30 +113,79 @@ phys_addr_t arm64_dma_phys_limit __ro_after_init;
static void __init reserve_crashkernel(void)
{
unsigned long long crash_base, crash_size;
- unsigned long long crash_max = arm64_dma_phys_limit;
+ unsigned long long crash_low_size;
+ unsigned long long crash_max = CRASH_ADDR_LOW_MAX;
int ret;
+ bool fixed_base, high = false;
+ char *cmdline = boot_command_line;
- ret = parse_crashkernel(boot_command_line, memblock_phys_mem_size(),
+ /* crashkernel=X[@offset] */
+ ret = parse_crashkernel(cmdline, memblock_phys_mem_size(),
&crash_size, &crash_base);
- /* no crashkernel= or invalid value specified */
- if (ret || !crash_size)
- return;
+ if (ret || !crash_size) {
+ /* crashkernel=X,high */
+ ret = parse_crashkernel_high(cmdline, 0, &crash_size, &crash_base);
+ if (ret || !crash_size)
+ return;
+
+ /* crashkernel=Y,low */
+ ret = parse_crashkernel_low(cmdline, 0, &crash_low_size, &crash_base);
+ if (ret == -ENOENT)
+ /*
+ * crashkernel=Y,low is not specified explicitly, use
+ * default size automatically.
+ */
+ crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
+ else if (ret)
+ /* crashkernel=Y,low is specified but Y is invalid */
+ return;
+
+ /* Mark crashkernel=X,high is specified */
+ high = true;
+ crash_max = CRASH_ADDR_HIGH_MAX;
+ }
+ fixed_base = !!crash_base;
crash_size = PAGE_ALIGN(crash_size);
/* User specifies base address explicitly. */
- if (crash_base)
+ if (fixed_base)
crash_max = crash_base + crash_size;
- /* Current arm64 boot protocol requires 2MB alignment */
- crash_base = memblock_phys_alloc_range(crash_size, SZ_2M,
+retry:
+ crash_base = memblock_phys_alloc_range(crash_size, CRASH_ALIGN,
crash_base, crash_max);
if (!crash_base) {
+ /*
+ * Attempt to fully allocate low memory failed, fall back
+ * to high memory, the minimum required low memory will be
+ * reserved later.
+ */
+ if (!fixed_base && (crash_max == CRASH_ADDR_LOW_MAX)) {
+ crash_max = CRASH_ADDR_HIGH_MAX;
+ goto retry;
+ }
+
pr_warn("cannot allocate crashkernel (size:0x%llx)\n",
crash_size);
return;
}
+ if (crash_base >= SZ_4G) {
+ /*
+ * For case crashkernel=X, low memory is not enough and fall
+ * back to reserve specified size of memory above 4G, try to
+ * allocate minimum required memory below 4G again.
+ */
+ if (!high)
+ crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
+
+ if (reserve_crashkernel_low(crash_low_size)) {
+ memblock_phys_free(crash_base, crash_size);
+ return;
+ }
+ }
+
pr_info("crashkernel reserved: 0x%016llx - 0x%016llx (%lld MB)\n",
crash_base, crash_base + crash_size, crash_size >> 20);
@@ -107,6 +194,9 @@ static void __init reserve_crashkernel(void)
* map. Inform kmemleak so that it won't try to access it.
*/
kmemleak_ignore_phys(crash_base);
+ if (crashk_low_res.end)
+ kmemleak_ignore_phys(crashk_low_res.start);
+
crashk_res.start = crash_base;
crashk_res.end = crash_base + crash_size - 1;
insert_resource(&iomem_resource, &crashk_res);
--
2.25.1
^ permalink raw reply related [flat|nested] 27+ messages in thread
* Re: [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X
2022-02-27 3:07 ` [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X Zhen Lei
@ 2022-03-16 12:11 ` Baoquan He
2022-03-16 13:11 ` Leizhen (ThunderTown)
2022-03-17 2:38 ` Baoquan He
2022-03-21 13:29 ` John Donnelly
2 siblings, 1 reply; 27+ messages in thread
From: Baoquan He @ 2022-03-16 12:11 UTC (permalink / raw)
To: Zhen Lei
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 02/27/22 at 11:07am, Zhen Lei wrote:
> From: Chen Zhou <chenzhou10@huawei.com>
>
> 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.
>
> Signed-off-by: Chen Zhou <chenzhou10@huawei.com>
> Co-developed-by: Zhen Lei <thunder.leizhen@huawei.com>
> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
> ---
> arch/arm64/kernel/machine_kexec.c | 9 ++-
> arch/arm64/kernel/machine_kexec_file.c | 12 ++-
> arch/arm64/mm/init.c | 106 +++++++++++++++++++++++--
> 3 files changed, 115 insertions(+), 12 deletions(-)
>
> diff --git a/arch/arm64/kernel/machine_kexec.c b/arch/arm64/kernel/machine_kexec.c
> index e16b248699d5c3c..19c2d487cb08feb 100644
> --- a/arch/arm64/kernel/machine_kexec.c
> +++ b/arch/arm64/kernel/machine_kexec.c
> @@ -329,8 +329,13 @@ bool crash_is_nosave(unsigned long pfn)
>
> /* in reserved memory? */
> addr = __pfn_to_phys(pfn);
> - if ((addr < crashk_res.start) || (crashk_res.end < addr))
> - return false;
> + if ((addr < crashk_res.start) || (crashk_res.end < addr)) {
> + if (!crashk_low_res.end)
> + return false;
> +
> + if ((addr < crashk_low_res.start) || (crashk_low_res.end < addr))
> + return false;
> + }
>
> if (!kexec_crash_image)
> return true;
> diff --git a/arch/arm64/kernel/machine_kexec_file.c b/arch/arm64/kernel/machine_kexec_file.c
> index 59c648d51848886..889951291cc0f9c 100644
> --- a/arch/arm64/kernel/machine_kexec_file.c
> +++ b/arch/arm64/kernel/machine_kexec_file.c
> @@ -65,10 +65,18 @@ static int prepare_elf_headers(void **addr, unsigned long *sz)
>
> /* Exclude crashkernel region */
> ret = crash_exclude_mem_range(cmem, crashk_res.start, crashk_res.end);
> + if (ret)
> + goto out;
> +
> + if (crashk_low_res.end) {
> + ret = crash_exclude_mem_range(cmem, crashk_low_res.start, crashk_low_res.end);
> + if (ret)
> + goto out;
> + }
>
> - if (!ret)
> - ret = crash_prepare_elf64_headers(cmem, true, addr, sz);
> + ret = crash_prepare_elf64_headers(cmem, true, addr, sz);
>
> +out:
> kfree(cmem);
> return ret;
> }
> diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
> index 90f276d46b93bc6..30ae6638ff54c47 100644
> --- a/arch/arm64/mm/init.c
> +++ b/arch/arm64/mm/init.c
> @@ -65,6 +65,44 @@ EXPORT_SYMBOL(memstart_addr);
> phys_addr_t arm64_dma_phys_limit __ro_after_init;
>
> #ifdef CONFIG_KEXEC_CORE
> +/* Current arm64 boot protocol requires 2MB alignment */
> +#define CRASH_ALIGN SZ_2M
> +
> +#define CRASH_ADDR_LOW_MAX arm64_dma_phys_limit
> +#define CRASH_ADDR_HIGH_MAX memblock.current_limit
> +
> +/*
> + * This is an empirical value in x86_64 and taken here directly. Please
> + * refer to the code comment in reserve_crashkernel_low() of x86_64 for more
> + * details.
> + */
> +#define DEFAULT_CRASH_KERNEL_LOW_SIZE \
> + max(swiotlb_size_or_default() + (8UL << 20), 256UL << 20)
> +
> +static int __init reserve_crashkernel_low(unsigned long long low_size)
> +{
> + unsigned long long low_base;
> +
> + /* passed with crashkernel=0,low ? */
> + if (!low_size)
> + return 0;
> +
> + low_base = memblock_phys_alloc_range(low_size, CRASH_ALIGN, 0, CRASH_ADDR_LOW_MAX);
> + if (!low_base) {
> + pr_err("cannot allocate crashkernel low memory (size:0x%llx).\n", low_size);
> + return -ENOMEM;
> + }
> +
> + pr_info("crashkernel low memory reserved: 0x%08llx - 0x%08llx (%lld MB)\n",
> + low_base, low_base + low_size, low_size >> 20);
> +
> + crashk_low_res.start = low_base;
> + crashk_low_res.end = low_base + low_size - 1;
> + insert_resource(&iomem_resource, &crashk_low_res);
> +
> + return 0;
> +}
> +
> /*
> * reserve_crashkernel() - reserves memory for crash kernel
> *
> @@ -75,30 +113,79 @@ phys_addr_t arm64_dma_phys_limit __ro_after_init;
> static void __init reserve_crashkernel(void)
> {
> unsigned long long crash_base, crash_size;
> - unsigned long long crash_max = arm64_dma_phys_limit;
> + unsigned long long crash_low_size;
> + unsigned long long crash_max = CRASH_ADDR_LOW_MAX;
> int ret;
Even though reverse xmas tree style is not enforced, this 'int ret;' is
really annoying to look at. Maybe move it down two lines.
> + bool fixed_base, high = false;
> + char *cmdline = boot_command_line;
>
> - ret = parse_crashkernel(boot_command_line, memblock_phys_mem_size(),
> + /* crashkernel=X[@offset] */
> + ret = parse_crashkernel(cmdline, memblock_phys_mem_size(),
> &crash_size, &crash_base);
> - /* no crashkernel= or invalid value specified */
> - if (ret || !crash_size)
> - return;
> + if (ret || !crash_size) {
> + /* crashkernel=X,high */
> + ret = parse_crashkernel_high(cmdline, 0, &crash_size, &crash_base);
> + if (ret || !crash_size)
> + return;
> +
> + /* crashkernel=Y,low */
> + ret = parse_crashkernel_low(cmdline, 0, &crash_low_size, &crash_base);
> + if (ret == -ENOENT)
> + /*
> + * crashkernel=Y,low is not specified explicitly, use
> + * default size automatically.
> + */
> + crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
> + else if (ret)
> + /* crashkernel=Y,low is specified but Y is invalid */
> + return;
> +
> + /* Mark crashkernel=X,high is specified */
> + high = true;
> + crash_max = CRASH_ADDR_HIGH_MAX;
> + }
>
> + fixed_base = !!crash_base;
> crash_size = PAGE_ALIGN(crash_size);
>
> /* User specifies base address explicitly. */
This is over commenting, can't see why it's needed.
> - if (crash_base)
> + if (fixed_base)
> crash_max = crash_base + crash_size;
Hi leizhen,
I made change on reserve_crashkenrel(), inline comment may be slow.
Please check and consider if they can be taken.
diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
index 30ae6638ff54..f96351da1e3e 100644
--- a/arch/arm64/mm/init.c
+++ b/arch/arm64/mm/init.c
@@ -109,38 +109,43 @@ static int __init reserve_crashkernel_low(unsigned long long low_size)
* 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.
+ *
+ * NOTE: Reservation of crashkernel,low is special since its existence
+ * is not independent, need rely on the existence of crashkernel,high.
+ * Hence there are different cases for crashkernel,low reservation:
+ * 1) crashkernel=Y,low is specified explicitly, crashkernel,low takes Y;
+ * 2) crashkernel=,low is not given, while crashkernel=,high is specified,
+ * take the default crashkernel,low value;
+ * 3) crashkernel=X is specified, while fallback to get a memory region
+ * in high memory, take the default crashkernel,low value;
+ * 4) crashkernel='invalid value',low is specified, failed the whole
+ * crashkernel reservation and bail out.
*/
static void __init reserve_crashkernel(void)
{
unsigned long long crash_base, crash_size;
unsigned long long crash_low_size;
unsigned long long crash_max = CRASH_ADDR_LOW_MAX;
- int ret;
bool fixed_base, high = false;
char *cmdline = boot_command_line;
+ int ret;
/* crashkernel=X[@offset] */
ret = parse_crashkernel(cmdline, memblock_phys_mem_size(),
&crash_size, &crash_base);
if (ret || !crash_size) {
- /* crashkernel=X,high */
ret = parse_crashkernel_high(cmdline, 0, &crash_size, &crash_base);
if (ret || !crash_size)
return;
- /* crashkernel=Y,low */
ret = parse_crashkernel_low(cmdline, 0, &crash_low_size, &crash_base);
if (ret == -ENOENT)
- /*
- * crashkernel=Y,low is not specified explicitly, use
- * default size automatically.
- */
+ /* case #2 of crashkernel,low reservation */
crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
else if (ret)
- /* crashkernel=Y,low is specified but Y is invalid */
+ /* case #4 of crashkernel,low reservation */
return;
- /* Mark crashkernel=X,high is specified */
high = true;
crash_max = CRASH_ADDR_HIGH_MAX;
}
@@ -148,7 +153,6 @@ static void __init reserve_crashkernel(void)
fixed_base = !!crash_base;
crash_size = PAGE_ALIGN(crash_size);
- /* User specifies base address explicitly. */
if (fixed_base)
crash_max = crash_base + crash_size;
@@ -172,11 +176,7 @@ static void __init reserve_crashkernel(void)
}
if (crash_base >= SZ_4G) {
- /*
- * For case crashkernel=X, low memory is not enough and fall
- * back to reserve specified size of memory above 4G, try to
- * allocate minimum required memory below 4G again.
- */
+ /* case #3 of crashkernel,low reservation */
if (!high)
crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
>
> - /* Current arm64 boot protocol requires 2MB alignment */
> - crash_base = memblock_phys_alloc_range(crash_size, SZ_2M,
> +retry:
> + crash_base = memblock_phys_alloc_range(crash_size, CRASH_ALIGN,
> crash_base, crash_max);
> if (!crash_base) {
> + /*
> + * Attempt to fully allocate low memory failed, fall back
> + * to high memory, the minimum required low memory will be
> + * reserved later.
> + */
> + if (!fixed_base && (crash_max == CRASH_ADDR_LOW_MAX)) {
> + crash_max = CRASH_ADDR_HIGH_MAX;
> + goto retry;
> + }
> +
> pr_warn("cannot allocate crashkernel (size:0x%llx)\n",
> crash_size);
> return;
> }
>
> + if (crash_base >= SZ_4G) {
> + /*
> + * For case crashkernel=X, low memory is not enough and fall
> + * back to reserve specified size of memory above 4G, try to
> + * allocate minimum required memory below 4G again.
> + */
> + if (!high)
> + crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
> +
> + if (reserve_crashkernel_low(crash_low_size)) {
> + memblock_phys_free(crash_base, crash_size);
> + return;
> + }
> + }
> +
> pr_info("crashkernel reserved: 0x%016llx - 0x%016llx (%lld MB)\n",
> crash_base, crash_base + crash_size, crash_size >> 20);
>
> @@ -107,6 +194,9 @@ static void __init reserve_crashkernel(void)
> * map. Inform kmemleak so that it won't try to access it.
> */
> kmemleak_ignore_phys(crash_base);
> + if (crashk_low_res.end)
> + kmemleak_ignore_phys(crashk_low_res.start);
> +
> crashk_res.start = crash_base;
> crashk_res.end = crash_base + crash_size - 1;
> insert_resource(&iomem_resource, &crashk_res);
> --
> 2.25.1
>
^ permalink raw reply related [flat|nested] 27+ messages in thread
* Re: [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X
2022-03-16 12:11 ` Baoquan He
@ 2022-03-16 13:11 ` Leizhen (ThunderTown)
2022-03-17 2:36 ` Baoquan He
0 siblings, 1 reply; 27+ messages in thread
From: Leizhen (ThunderTown) @ 2022-03-16 13:11 UTC (permalink / raw)
To: Baoquan He
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 2022/3/16 20:11, Baoquan He wrote:
> On 02/27/22 at 11:07am, Zhen Lei wrote:
>> From: Chen Zhou <chenzhou10@huawei.com>
>>
>> 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.
>>
>> Signed-off-by: Chen Zhou <chenzhou10@huawei.com>
>> Co-developed-by: Zhen Lei <thunder.leizhen@huawei.com>
>> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
>> ---
>> arch/arm64/kernel/machine_kexec.c | 9 ++-
>> arch/arm64/kernel/machine_kexec_file.c | 12 ++-
>> arch/arm64/mm/init.c | 106 +++++++++++++++++++++++--
>> 3 files changed, 115 insertions(+), 12 deletions(-)
>>
>> diff --git a/arch/arm64/kernel/machine_kexec.c b/arch/arm64/kernel/machine_kexec.c
>> index e16b248699d5c3c..19c2d487cb08feb 100644
>> --- a/arch/arm64/kernel/machine_kexec.c
>> +++ b/arch/arm64/kernel/machine_kexec.c
>> @@ -329,8 +329,13 @@ bool crash_is_nosave(unsigned long pfn)
>>
>> /* in reserved memory? */
>> addr = __pfn_to_phys(pfn);
>> - if ((addr < crashk_res.start) || (crashk_res.end < addr))
>> - return false;
>> + if ((addr < crashk_res.start) || (crashk_res.end < addr)) {
>> + if (!crashk_low_res.end)
>> + return false;
>> +
>> + if ((addr < crashk_low_res.start) || (crashk_low_res.end < addr))
>> + return false;
>> + }
>>
>> if (!kexec_crash_image)
>> return true;
>> diff --git a/arch/arm64/kernel/machine_kexec_file.c b/arch/arm64/kernel/machine_kexec_file.c
>> index 59c648d51848886..889951291cc0f9c 100644
>> --- a/arch/arm64/kernel/machine_kexec_file.c
>> +++ b/arch/arm64/kernel/machine_kexec_file.c
>> @@ -65,10 +65,18 @@ static int prepare_elf_headers(void **addr, unsigned long *sz)
>>
>> /* Exclude crashkernel region */
>> ret = crash_exclude_mem_range(cmem, crashk_res.start, crashk_res.end);
>> + if (ret)
>> + goto out;
>> +
>> + if (crashk_low_res.end) {
>> + ret = crash_exclude_mem_range(cmem, crashk_low_res.start, crashk_low_res.end);
>> + if (ret)
>> + goto out;
>> + }
>>
>> - if (!ret)
>> - ret = crash_prepare_elf64_headers(cmem, true, addr, sz);
>> + ret = crash_prepare_elf64_headers(cmem, true, addr, sz);
>>
>> +out:
>> kfree(cmem);
>> return ret;
>> }
>> diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
>> index 90f276d46b93bc6..30ae6638ff54c47 100644
>> --- a/arch/arm64/mm/init.c
>> +++ b/arch/arm64/mm/init.c
>> @@ -65,6 +65,44 @@ EXPORT_SYMBOL(memstart_addr);
>> phys_addr_t arm64_dma_phys_limit __ro_after_init;
>>
>> #ifdef CONFIG_KEXEC_CORE
>> +/* Current arm64 boot protocol requires 2MB alignment */
>> +#define CRASH_ALIGN SZ_2M
>> +
>> +#define CRASH_ADDR_LOW_MAX arm64_dma_phys_limit
>> +#define CRASH_ADDR_HIGH_MAX memblock.current_limit
>> +
>> +/*
>> + * This is an empirical value in x86_64 and taken here directly. Please
>> + * refer to the code comment in reserve_crashkernel_low() of x86_64 for more
>> + * details.
>> + */
>> +#define DEFAULT_CRASH_KERNEL_LOW_SIZE \
>> + max(swiotlb_size_or_default() + (8UL << 20), 256UL << 20)
>> +
>> +static int __init reserve_crashkernel_low(unsigned long long low_size)
>> +{
>> + unsigned long long low_base;
>> +
>> + /* passed with crashkernel=0,low ? */
>> + if (!low_size)
>> + return 0;
>> +
>> + low_base = memblock_phys_alloc_range(low_size, CRASH_ALIGN, 0, CRASH_ADDR_LOW_MAX);
>> + if (!low_base) {
>> + pr_err("cannot allocate crashkernel low memory (size:0x%llx).\n", low_size);
>> + return -ENOMEM;
>> + }
>> +
>> + pr_info("crashkernel low memory reserved: 0x%08llx - 0x%08llx (%lld MB)\n",
>> + low_base, low_base + low_size, low_size >> 20);
>> +
>> + crashk_low_res.start = low_base;
>> + crashk_low_res.end = low_base + low_size - 1;
>> + insert_resource(&iomem_resource, &crashk_low_res);
>> +
>> + return 0;
>> +}
>> +
>> /*
>> * reserve_crashkernel() - reserves memory for crash kernel
>> *
>> @@ -75,30 +113,79 @@ phys_addr_t arm64_dma_phys_limit __ro_after_init;
>> static void __init reserve_crashkernel(void)
>> {
>> unsigned long long crash_base, crash_size;
>> - unsigned long long crash_max = arm64_dma_phys_limit;
>> + unsigned long long crash_low_size;
>> + unsigned long long crash_max = CRASH_ADDR_LOW_MAX;
>> int ret;
>
> Even though reverse xmas tree style is not enforced, this 'int ret;' is
> really annoying to look at. Maybe move it down two lines.
>
>> + bool fixed_base, high = false;
>> + char *cmdline = boot_command_line;
>>
>> - ret = parse_crashkernel(boot_command_line, memblock_phys_mem_size(),
>> + /* crashkernel=X[@offset] */
>> + ret = parse_crashkernel(cmdline, memblock_phys_mem_size(),
>> &crash_size, &crash_base);
>> - /* no crashkernel= or invalid value specified */
>> - if (ret || !crash_size)
>> - return;
>> + if (ret || !crash_size) {
>> + /* crashkernel=X,high */
>> + ret = parse_crashkernel_high(cmdline, 0, &crash_size, &crash_base);
>> + if (ret || !crash_size)
>> + return;
>> +
>> + /* crashkernel=Y,low */
>> + ret = parse_crashkernel_low(cmdline, 0, &crash_low_size, &crash_base);
>> + if (ret == -ENOENT)
>> + /*
>> + * crashkernel=Y,low is not specified explicitly, use
>> + * default size automatically.
>> + */
>> + crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
>> + else if (ret)
>> + /* crashkernel=Y,low is specified but Y is invalid */
>> + return;
>> +
>> + /* Mark crashkernel=X,high is specified */
>> + high = true;
>> + crash_max = CRASH_ADDR_HIGH_MAX;
>> + }
>>
>> + fixed_base = !!crash_base;
>> crash_size = PAGE_ALIGN(crash_size);
>>
>> /* User specifies base address explicitly. */
> This is over commenting, can't see why it's needed.
>> - if (crash_base)
>> + if (fixed_base)
>> crash_max = crash_base + crash_size;
>
> Hi leizhen,
>
> I made change on reserve_crashkenrel(), inline comment may be slow.
> Please check and consider if they can be taken.
That's great. Thank you very much.
>
> diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
> index 30ae6638ff54..f96351da1e3e 100644
> --- a/arch/arm64/mm/init.c
> +++ b/arch/arm64/mm/init.c
> @@ -109,38 +109,43 @@ static int __init reserve_crashkernel_low(unsigned long long low_size)
> * 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.
> + *
> + * NOTE: Reservation of crashkernel,low is special since its existence
> + * is not independent, need rely on the existence of crashkernel,high.
> + * Hence there are different cases for crashkernel,low reservation:
> + * 1) crashkernel=Y,low is specified explicitly, crashkernel,low takes Y;
> + * 2) crashkernel=,low is not given, while crashkernel=,high is specified,
> + * take the default crashkernel,low value;
> + * 3) crashkernel=X is specified, while fallback to get a memory region
> + * in high memory, take the default crashkernel,low value;
> + * 4) crashkernel='invalid value',low is specified, failed the whole
> + * crashkernel reservation and bail out.
> */
> static void __init reserve_crashkernel(void)
> {
> unsigned long long crash_base, crash_size;
> unsigned long long crash_low_size;
> unsigned long long crash_max = CRASH_ADDR_LOW_MAX;
> - int ret;
> bool fixed_base, high = false;
> char *cmdline = boot_command_line;
> + int ret;
>
> /* crashkernel=X[@offset] */
> ret = parse_crashkernel(cmdline, memblock_phys_mem_size(),
> &crash_size, &crash_base);
> if (ret || !crash_size) {
> - /* crashkernel=X,high */
> ret = parse_crashkernel_high(cmdline, 0, &crash_size, &crash_base);
> if (ret || !crash_size)
> return;
>
> - /* crashkernel=Y,low */
> ret = parse_crashkernel_low(cmdline, 0, &crash_low_size, &crash_base);
> if (ret == -ENOENT)
> - /*
> - * crashkernel=Y,low is not specified explicitly, use
> - * default size automatically.
> - */
> + /* case #2 of crashkernel,low reservation */
> crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
> else if (ret)
> - /* crashkernel=Y,low is specified but Y is invalid */
> + /* case #4 of crashkernel,low reservation */
> return;
>
> - /* Mark crashkernel=X,high is specified */
> high = true;
> crash_max = CRASH_ADDR_HIGH_MAX;
> }
> @@ -148,7 +153,6 @@ static void __init reserve_crashkernel(void)
> fixed_base = !!crash_base;
> crash_size = PAGE_ALIGN(crash_size);
>
> - /* User specifies base address explicitly. */
> if (fixed_base)
> crash_max = crash_base + crash_size;
>
> @@ -172,11 +176,7 @@ static void __init reserve_crashkernel(void)
> }
>
> if (crash_base >= SZ_4G) {
> - /*
> - * For case crashkernel=X, low memory is not enough and fall
> - * back to reserve specified size of memory above 4G, try to
> - * allocate minimum required memory below 4G again.
> - */
> + /* case #3 of crashkernel,low reservation */
> if (!high)
> crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
>
>
>>
>> - /* Current arm64 boot protocol requires 2MB alignment */
>> - crash_base = memblock_phys_alloc_range(crash_size, SZ_2M,
>> +retry:
>> + crash_base = memblock_phys_alloc_range(crash_size, CRASH_ALIGN,
>> crash_base, crash_max);
>> if (!crash_base) {
>> + /*
>> + * Attempt to fully allocate low memory failed, fall back
>> + * to high memory, the minimum required low memory will be
>> + * reserved later.
>> + */
>> + if (!fixed_base && (crash_max == CRASH_ADDR_LOW_MAX)) {
>> + crash_max = CRASH_ADDR_HIGH_MAX;
>> + goto retry;
>> + }
>> +
>> pr_warn("cannot allocate crashkernel (size:0x%llx)\n",
>> crash_size);
>> return;
>> }
>>
>> + if (crash_base >= SZ_4G) {
>> + /*
>> + * For case crashkernel=X, low memory is not enough and fall
>> + * back to reserve specified size of memory above 4G, try to
>> + * allocate minimum required memory below 4G again.
>> + */
>> + if (!high)
>> + crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
>> +
>> + if (reserve_crashkernel_low(crash_low_size)) {
>> + memblock_phys_free(crash_base, crash_size);
>> + return;
>> + }
>> + }
>> +
>> pr_info("crashkernel reserved: 0x%016llx - 0x%016llx (%lld MB)\n",
>> crash_base, crash_base + crash_size, crash_size >> 20);
>>
>> @@ -107,6 +194,9 @@ static void __init reserve_crashkernel(void)
>> * map. Inform kmemleak so that it won't try to access it.
>> */
>> kmemleak_ignore_phys(crash_base);
>> + if (crashk_low_res.end)
>> + kmemleak_ignore_phys(crashk_low_res.start);
>> +
>> crashk_res.start = crash_base;
>> crashk_res.end = crash_base + crash_size - 1;
>> insert_resource(&iomem_resource, &crashk_res);
>> --
>> 2.25.1
>>
>
> .
>
--
Regards,
Zhen Lei
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X
2022-03-16 13:11 ` Leizhen (ThunderTown)
@ 2022-03-17 2:36 ` Baoquan He
2022-03-17 3:19 ` Leizhen (ThunderTown)
0 siblings, 1 reply; 27+ messages in thread
From: Baoquan He @ 2022-03-17 2:36 UTC (permalink / raw)
To: Leizhen (ThunderTown)
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 03/16/22 at 09:11pm, Leizhen (ThunderTown) wrote:
>
>
> On 2022/3/16 20:11, Baoquan He wrote:
> > On 02/27/22 at 11:07am, Zhen Lei wrote:
......
> > Hi leizhen,
> >
> > I made change on reserve_crashkenrel(), inline comment may be slow.
> > Please check and consider if they can be taken.
>
> That's great. Thank you very much.
>
> >
> > diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
> > index 30ae6638ff54..f96351da1e3e 100644
> > --- a/arch/arm64/mm/init.c
> > +++ b/arch/arm64/mm/init.c
> > @@ -109,38 +109,43 @@ static int __init reserve_crashkernel_low(unsigned long long low_size)
> > * 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.
> > + *
> > + * NOTE: Reservation of crashkernel,low is special since its existence
> > + * is not independent, need rely on the existence of crashkernel,high.
> > + * Hence there are different cases for crashkernel,low reservation:
Considering to update the 3rd line as below:
* NOTE: Reservation of crashkernel,low is special since its existence
* is not independent, need rely on the existence of crashkernel,high.
* Here, four cases of crashkernel,low reservation are summarized:
> > + * 1) crashkernel=Y,low is specified explicitly, crashkernel,low takes Y;
> > + * 2) crashkernel=,low is not given, while crashkernel=,high is specified,
> > + * take the default crashkernel,low value;
> > + * 3) crashkernel=X is specified, while fallback to get a memory region
> > + * in high memory, take the default crashkernel,low value;
> > + * 4) crashkernel='invalid value',low is specified, failed the whole
> > + * crashkernel reservation and bail out.
> > */
> > static void __init reserve_crashkernel(void)
> > {
> > unsigned long long crash_base, crash_size;
> > unsigned long long crash_low_size;
> > unsigned long long crash_max = CRASH_ADDR_LOW_MAX;
> > - int ret;
> > bool fixed_base, high = false;
> > char *cmdline = boot_command_line;
> > + int ret;
> >
> > /* crashkernel=X[@offset] */
> > ret = parse_crashkernel(cmdline, memblock_phys_mem_size(),
> > &crash_size, &crash_base);
> > if (ret || !crash_size) {
> > - /* crashkernel=X,high */
> > ret = parse_crashkernel_high(cmdline, 0, &crash_size, &crash_base);
> > if (ret || !crash_size)
> > return;
> >
> > - /* crashkernel=Y,low */
> > ret = parse_crashkernel_low(cmdline, 0, &crash_low_size, &crash_base);
> > if (ret == -ENOENT)
> > - /*
> > - * crashkernel=Y,low is not specified explicitly, use
> > - * default size automatically.
> > - */
> > + /* case #2 of crashkernel,low reservation */
> > crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
> > else if (ret)
> > - /* crashkernel=Y,low is specified but Y is invalid */
> > + /* case #4 of crashkernel,low reservation */
> > return;
> >
> > - /* Mark crashkernel=X,high is specified */
> > high = true;
> > crash_max = CRASH_ADDR_HIGH_MAX;
> > }
> > @@ -148,7 +153,6 @@ static void __init reserve_crashkernel(void)
> > fixed_base = !!crash_base;
> > crash_size = PAGE_ALIGN(crash_size);
> >
> > - /* User specifies base address explicitly. */
> > if (fixed_base)
> > crash_max = crash_base + crash_size;
> >
> > @@ -172,11 +176,7 @@ static void __init reserve_crashkernel(void)
> > }
> >
> > if (crash_base >= SZ_4G) {
> > - /*
> > - * For case crashkernel=X, low memory is not enough and fall
> > - * back to reserve specified size of memory above 4G, try to
> > - * allocate minimum required memory below 4G again.
> > - */
> > + /* case #3 of crashkernel,low reservation */
> > if (!high)
> > crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
> >
> >
> >>
> >> - /* Current arm64 boot protocol requires 2MB alignment */
> >> - crash_base = memblock_phys_alloc_range(crash_size, SZ_2M,
> >> +retry:
> >> + crash_base = memblock_phys_alloc_range(crash_size, CRASH_ALIGN,
> >> crash_base, crash_max);
> >> if (!crash_base) {
> >> + /*
> >> + * Attempt to fully allocate low memory failed, fall back
> >> + * to high memory, the minimum required low memory will be
> >> + * reserved later.
> >> + */
> >> + if (!fixed_base && (crash_max == CRASH_ADDR_LOW_MAX)) {
> >> + crash_max = CRASH_ADDR_HIGH_MAX;
> >> + goto retry;
> >> + }
> >> +
> >> pr_warn("cannot allocate crashkernel (size:0x%llx)\n",
> >> crash_size);
> >> return;
> >> }
> >>
> >> + if (crash_base >= SZ_4G) {
> >> + /*
> >> + * For case crashkernel=X, low memory is not enough and fall
> >> + * back to reserve specified size of memory above 4G, try to
> >> + * allocate minimum required memory below 4G again.
> >> + */
> >> + if (!high)
> >> + crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
> >> +
> >> + if (reserve_crashkernel_low(crash_low_size)) {
> >> + memblock_phys_free(crash_base, crash_size);
> >> + return;
> >> + }
> >> + }
> >> +
> >> pr_info("crashkernel reserved: 0x%016llx - 0x%016llx (%lld MB)\n",
> >> crash_base, crash_base + crash_size, crash_size >> 20);
> >>
> >> @@ -107,6 +194,9 @@ static void __init reserve_crashkernel(void)
> >> * map. Inform kmemleak so that it won't try to access it.
> >> */
> >> kmemleak_ignore_phys(crash_base);
> >> + if (crashk_low_res.end)
> >> + kmemleak_ignore_phys(crashk_low_res.start);
> >> +
> >> crashk_res.start = crash_base;
> >> crashk_res.end = crash_base + crash_size - 1;
> >> insert_resource(&iomem_resource, &crashk_res);
> >> --
> >> 2.25.1
> >>
> >
> > .
> >
>
> --
> Regards,
> Zhen Lei
>
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X
2022-03-17 2:36 ` Baoquan He
@ 2022-03-17 3:19 ` Leizhen (ThunderTown)
2022-03-17 3:47 ` Baoquan He
0 siblings, 1 reply; 27+ messages in thread
From: Leizhen (ThunderTown) @ 2022-03-17 3:19 UTC (permalink / raw)
To: Baoquan He
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 2022/3/17 10:36, Baoquan He wrote:
> On 03/16/22 at 09:11pm, Leizhen (ThunderTown) wrote:
>>
>>
>> On 2022/3/16 20:11, Baoquan He wrote:
>>> On 02/27/22 at 11:07am, Zhen Lei wrote:
> ......
>
>>> Hi leizhen,
>>>
>>> I made change on reserve_crashkenrel(), inline comment may be slow.
>>> Please check and consider if they can be taken.
>>
>> That's great. Thank you very much.
>>
>>>
>>> diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
>>> index 30ae6638ff54..f96351da1e3e 100644
>>> --- a/arch/arm64/mm/init.c
>>> +++ b/arch/arm64/mm/init.c
>>> @@ -109,38 +109,43 @@ static int __init reserve_crashkernel_low(unsigned long long low_size)
>>> * 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.
>>> + *
>>> + * NOTE: Reservation of crashkernel,low is special since its existence
>>> + * is not independent, need rely on the existence of crashkernel,high.
>>> + * Hence there are different cases for crashkernel,low reservation:
>
> Considering to update the 3rd line as below:
>
> * NOTE: Reservation of crashkernel,low is special since its existence
> * is not independent, need rely on the existence of crashkernel,high.
> * Here, four cases of crashkernel,low reservation are summarized:
OK. How about change "crashkernel,low" to "crashkernel low memory"?
"crashkernel=Y,low", "crashkernel=,low" and "crashkernel,low" are very similar,
may dazzle the reader.
>
>>> + * 1) crashkernel=Y,low is specified explicitly, crashkernel,low takes Y;
>>> + * 2) crashkernel=,low is not given, while crashkernel=,high is specified,
>>> + * take the default crashkernel,low value;
>>> + * 3) crashkernel=X is specified, while fallback to get a memory region
>>> + * in high memory, take the default crashkernel,low value;
>>> + * 4) crashkernel='invalid value',low is specified, failed the whole
>>> + * crashkernel reservation and bail out.
>>> */
>>> static void __init reserve_crashkernel(void)
>>> {
>>> unsigned long long crash_base, crash_size;
>>> unsigned long long crash_low_size;
>>> unsigned long long crash_max = CRASH_ADDR_LOW_MAX;
>>> - int ret;
>>> bool fixed_base, high = false;
>>> char *cmdline = boot_command_line;
>>> + int ret;
>>>
>>> /* crashkernel=X[@offset] */
>>> ret = parse_crashkernel(cmdline, memblock_phys_mem_size(),
>>> &crash_size, &crash_base);
>>> if (ret || !crash_size) {
>>> - /* crashkernel=X,high */
>>> ret = parse_crashkernel_high(cmdline, 0, &crash_size, &crash_base);
>>> if (ret || !crash_size)
>>> return;
>>>
>>> - /* crashkernel=Y,low */
>>> ret = parse_crashkernel_low(cmdline, 0, &crash_low_size, &crash_base);
>>> if (ret == -ENOENT)
>>> - /*
>>> - * crashkernel=Y,low is not specified explicitly, use
>>> - * default size automatically.
>>> - */
>>> + /* case #2 of crashkernel,low reservation */
>>> crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
>>> else if (ret)
>>> - /* crashkernel=Y,low is specified but Y is invalid */
>>> + /* case #4 of crashkernel,low reservation */
>>> return;
>>>
>>> - /* Mark crashkernel=X,high is specified */
>>> high = true;
>>> crash_max = CRASH_ADDR_HIGH_MAX;
>>> }
>>> @@ -148,7 +153,6 @@ static void __init reserve_crashkernel(void)
>>> fixed_base = !!crash_base;
>>> crash_size = PAGE_ALIGN(crash_size);
>>>
>>> - /* User specifies base address explicitly. */
>>> if (fixed_base)
>>> crash_max = crash_base + crash_size;
>>>
>>> @@ -172,11 +176,7 @@ static void __init reserve_crashkernel(void)
>>> }
>>>
>>> if (crash_base >= SZ_4G) {
>>> - /*
>>> - * For case crashkernel=X, low memory is not enough and fall
>>> - * back to reserve specified size of memory above 4G, try to
>>> - * allocate minimum required memory below 4G again.
>>> - */
>>> + /* case #3 of crashkernel,low reservation */
>>> if (!high)
>>> crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
>>>
>>>
>>>>
>>>> - /* Current arm64 boot protocol requires 2MB alignment */
>>>> - crash_base = memblock_phys_alloc_range(crash_size, SZ_2M,
>>>> +retry:
>>>> + crash_base = memblock_phys_alloc_range(crash_size, CRASH_ALIGN,
>>>> crash_base, crash_max);
>>>> if (!crash_base) {
>>>> + /*
>>>> + * Attempt to fully allocate low memory failed, fall back
>>>> + * to high memory, the minimum required low memory will be
>>>> + * reserved later.
>>>> + */
>>>> + if (!fixed_base && (crash_max == CRASH_ADDR_LOW_MAX)) {
>>>> + crash_max = CRASH_ADDR_HIGH_MAX;
>>>> + goto retry;
>>>> + }
>>>> +
>>>> pr_warn("cannot allocate crashkernel (size:0x%llx)\n",
>>>> crash_size);
>>>> return;
>>>> }
>>>>
>>>> + if (crash_base >= SZ_4G) {
>>>> + /*
>>>> + * For case crashkernel=X, low memory is not enough and fall
>>>> + * back to reserve specified size of memory above 4G, try to
>>>> + * allocate minimum required memory below 4G again.
>>>> + */
>>>> + if (!high)
>>>> + crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
>>>> +
>>>> + if (reserve_crashkernel_low(crash_low_size)) {
>>>> + memblock_phys_free(crash_base, crash_size);
>>>> + return;
>>>> + }
>>>> + }
>>>> +
>>>> pr_info("crashkernel reserved: 0x%016llx - 0x%016llx (%lld MB)\n",
>>>> crash_base, crash_base + crash_size, crash_size >> 20);
>>>>
>>>> @@ -107,6 +194,9 @@ static void __init reserve_crashkernel(void)
>>>> * map. Inform kmemleak so that it won't try to access it.
>>>> */
>>>> kmemleak_ignore_phys(crash_base);
>>>> + if (crashk_low_res.end)
>>>> + kmemleak_ignore_phys(crashk_low_res.start);
>>>> +
>>>> crashk_res.start = crash_base;
>>>> crashk_res.end = crash_base + crash_size - 1;
>>>> insert_resource(&iomem_resource, &crashk_res);
>>>> --
>>>> 2.25.1
>>>>
>>>
>>> .
>>>
>>
>> --
>> Regards,
>> Zhen Lei
>>
>
> .
>
--
Regards,
Zhen Lei
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X
2022-03-17 3:19 ` Leizhen (ThunderTown)
@ 2022-03-17 3:47 ` Baoquan He
2022-03-17 7:30 ` Leizhen (ThunderTown)
0 siblings, 1 reply; 27+ messages in thread
From: Baoquan He @ 2022-03-17 3:47 UTC (permalink / raw)
To: Leizhen (ThunderTown)
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 03/17/22 at 11:19am, Leizhen (ThunderTown) wrote:
>
>
> On 2022/3/17 10:36, Baoquan He wrote:
> > On 03/16/22 at 09:11pm, Leizhen (ThunderTown) wrote:
> >>
> >>
> >> On 2022/3/16 20:11, Baoquan He wrote:
> >>> On 02/27/22 at 11:07am, Zhen Lei wrote:
> > ......
> >
> >>> Hi leizhen,
> >>>
> >>> I made change on reserve_crashkenrel(), inline comment may be slow.
> >>> Please check and consider if they can be taken.
> >>
> >> That's great. Thank you very much.
> >>
> >>>
> >>> diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
> >>> index 30ae6638ff54..f96351da1e3e 100644
> >>> --- a/arch/arm64/mm/init.c
> >>> +++ b/arch/arm64/mm/init.c
> >>> @@ -109,38 +109,43 @@ static int __init reserve_crashkernel_low(unsigned long long low_size)
> >>> * 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.
> >>> + *
> >>> + * NOTE: Reservation of crashkernel,low is special since its existence
> >>> + * is not independent, need rely on the existence of crashkernel,high.
> >>> + * Hence there are different cases for crashkernel,low reservation:
> >
> > Considering to update the 3rd line as below:
> >
> > * NOTE: Reservation of crashkernel,low is special since its existence
> > * is not independent, need rely on the existence of crashkernel,high.
> > * Here, four cases of crashkernel,low reservation are summarized:
>
> OK. How about change "crashkernel,low" to "crashkernel low memory"?
> "crashkernel=Y,low", "crashkernel=,low" and "crashkernel,low" are very similar,
> may dazzle the reader.
Fine by me. 'crashkernel low memory' is formal, just make sentence a
little longer. Please take what you think fitter.
>
> >
> >>> + * 1) crashkernel=Y,low is specified explicitly, crashkernel,low takes Y;
> >>> + * 2) crashkernel=,low is not given, while crashkernel=,high is specified,
> >>> + * take the default crashkernel,low value;
> >>> + * 3) crashkernel=X is specified, while fallback to get a memory region
> >>> + * in high memory, take the default crashkernel,low value;
> >>> + * 4) crashkernel='invalid value',low is specified, failed the whole
> >>> + * crashkernel reservation and bail out.
> >>> */
> >>> static void __init reserve_crashkernel(void)
> >>> {
> >>> unsigned long long crash_base, crash_size;
> >>> unsigned long long crash_low_size;
> >>> unsigned long long crash_max = CRASH_ADDR_LOW_MAX;
> >>> - int ret;
> >>> bool fixed_base, high = false;
> >>> char *cmdline = boot_command_line;
> >>> + int ret;
> >>>
> >>> /* crashkernel=X[@offset] */
> >>> ret = parse_crashkernel(cmdline, memblock_phys_mem_size(),
> >>> &crash_size, &crash_base);
> >>> if (ret || !crash_size) {
> >>> - /* crashkernel=X,high */
> >>> ret = parse_crashkernel_high(cmdline, 0, &crash_size, &crash_base);
> >>> if (ret || !crash_size)
> >>> return;
> >>>
> >>> - /* crashkernel=Y,low */
> >>> ret = parse_crashkernel_low(cmdline, 0, &crash_low_size, &crash_base);
> >>> if (ret == -ENOENT)
> >>> - /*
> >>> - * crashkernel=Y,low is not specified explicitly, use
> >>> - * default size automatically.
> >>> - */
> >>> + /* case #2 of crashkernel,low reservation */
> >>> crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
> >>> else if (ret)
> >>> - /* crashkernel=Y,low is specified but Y is invalid */
> >>> + /* case #4 of crashkernel,low reservation */
> >>> return;
> >>>
> >>> - /* Mark crashkernel=X,high is specified */
> >>> high = true;
> >>> crash_max = CRASH_ADDR_HIGH_MAX;
> >>> }
> >>> @@ -148,7 +153,6 @@ static void __init reserve_crashkernel(void)
> >>> fixed_base = !!crash_base;
> >>> crash_size = PAGE_ALIGN(crash_size);
> >>>
> >>> - /* User specifies base address explicitly. */
> >>> if (fixed_base)
> >>> crash_max = crash_base + crash_size;
> >>>
> >>> @@ -172,11 +176,7 @@ static void __init reserve_crashkernel(void)
> >>> }
> >>>
> >>> if (crash_base >= SZ_4G) {
> >>> - /*
> >>> - * For case crashkernel=X, low memory is not enough and fall
> >>> - * back to reserve specified size of memory above 4G, try to
> >>> - * allocate minimum required memory below 4G again.
> >>> - */
> >>> + /* case #3 of crashkernel,low reservation */
> >>> if (!high)
> >>> crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
> >>>
> >>>
> >>>>
> >>>> - /* Current arm64 boot protocol requires 2MB alignment */
> >>>> - crash_base = memblock_phys_alloc_range(crash_size, SZ_2M,
> >>>> +retry:
> >>>> + crash_base = memblock_phys_alloc_range(crash_size, CRASH_ALIGN,
> >>>> crash_base, crash_max);
> >>>> if (!crash_base) {
> >>>> + /*
> >>>> + * Attempt to fully allocate low memory failed, fall back
> >>>> + * to high memory, the minimum required low memory will be
> >>>> + * reserved later.
> >>>> + */
> >>>> + if (!fixed_base && (crash_max == CRASH_ADDR_LOW_MAX)) {
> >>>> + crash_max = CRASH_ADDR_HIGH_MAX;
> >>>> + goto retry;
> >>>> + }
> >>>> +
> >>>> pr_warn("cannot allocate crashkernel (size:0x%llx)\n",
> >>>> crash_size);
> >>>> return;
> >>>> }
> >>>>
> >>>> + if (crash_base >= SZ_4G) {
> >>>> + /*
> >>>> + * For case crashkernel=X, low memory is not enough and fall
> >>>> + * back to reserve specified size of memory above 4G, try to
> >>>> + * allocate minimum required memory below 4G again.
> >>>> + */
> >>>> + if (!high)
> >>>> + crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
> >>>> +
> >>>> + if (reserve_crashkernel_low(crash_low_size)) {
> >>>> + memblock_phys_free(crash_base, crash_size);
> >>>> + return;
> >>>> + }
> >>>> + }
> >>>> +
> >>>> pr_info("crashkernel reserved: 0x%016llx - 0x%016llx (%lld MB)\n",
> >>>> crash_base, crash_base + crash_size, crash_size >> 20);
> >>>>
> >>>> @@ -107,6 +194,9 @@ static void __init reserve_crashkernel(void)
> >>>> * map. Inform kmemleak so that it won't try to access it.
> >>>> */
> >>>> kmemleak_ignore_phys(crash_base);
> >>>> + if (crashk_low_res.end)
> >>>> + kmemleak_ignore_phys(crashk_low_res.start);
> >>>> +
> >>>> crashk_res.start = crash_base;
> >>>> crashk_res.end = crash_base + crash_size - 1;
> >>>> insert_resource(&iomem_resource, &crashk_res);
> >>>> --
> >>>> 2.25.1
> >>>>
> >>>
> >>> .
> >>>
> >>
> >> --
> >> Regards,
> >> Zhen Lei
> >>
> >
> > .
> >
>
> --
> Regards,
> Zhen Lei
>
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X
2022-03-17 3:47 ` Baoquan He
@ 2022-03-17 7:30 ` Leizhen (ThunderTown)
0 siblings, 0 replies; 27+ messages in thread
From: Leizhen (ThunderTown) @ 2022-03-17 7:30 UTC (permalink / raw)
To: Baoquan He
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 2022/3/17 11:47, Baoquan He wrote:
> On 03/17/22 at 11:19am, Leizhen (ThunderTown) wrote:
>>
>>
>> On 2022/3/17 10:36, Baoquan He wrote:
>>> On 03/16/22 at 09:11pm, Leizhen (ThunderTown) wrote:
>>>>
>>>>
>>>> On 2022/3/16 20:11, Baoquan He wrote:
>>>>> On 02/27/22 at 11:07am, Zhen Lei wrote:
>>> ......
>>>
>>>>> Hi leizhen,
>>>>>
>>>>> I made change on reserve_crashkenrel(), inline comment may be slow.
>>>>> Please check and consider if they can be taken.
>>>>
>>>> That's great. Thank you very much.
>>>>
>>>>>
>>>>> diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
>>>>> index 30ae6638ff54..f96351da1e3e 100644
>>>>> --- a/arch/arm64/mm/init.c
>>>>> +++ b/arch/arm64/mm/init.c
>>>>> @@ -109,38 +109,43 @@ static int __init reserve_crashkernel_low(unsigned long long low_size)
>>>>> * 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.
>>>>> + *
>>>>> + * NOTE: Reservation of crashkernel,low is special since its existence
>>>>> + * is not independent, need rely on the existence of crashkernel,high.
>>>>> + * Hence there are different cases for crashkernel,low reservation:
>>>
>>> Considering to update the 3rd line as below:
>>>
>>> * NOTE: Reservation of crashkernel,low is special since its existence
>>> * is not independent, need rely on the existence of crashkernel,high.
>>> * Here, four cases of crashkernel,low reservation are summarized:
>>
>> OK. How about change "crashkernel,low" to "crashkernel low memory"?
>> "crashkernel=Y,low", "crashkernel=,low" and "crashkernel,low" are very similar,
>> may dazzle the reader.
>
> Fine by me. 'crashkernel low memory' is formal, just make sentence a
> little longer. Please take what you think fitter.
OK, I will send v22 after v5.18-rc1.
>
>>
>>>
>>>>> + * 1) crashkernel=Y,low is specified explicitly, crashkernel,low takes Y;
>>>>> + * 2) crashkernel=,low is not given, while crashkernel=,high is specified,
>>>>> + * take the default crashkernel,low value;
>>>>> + * 3) crashkernel=X is specified, while fallback to get a memory region
>>>>> + * in high memory, take the default crashkernel,low value;
>>>>> + * 4) crashkernel='invalid value',low is specified, failed the whole
>>>>> + * crashkernel reservation and bail out.
>>>>> */
>>>>> static void __init reserve_crashkernel(void)
>>>>> {
>>>>> unsigned long long crash_base, crash_size;
>>>>> unsigned long long crash_low_size;
>>>>> unsigned long long crash_max = CRASH_ADDR_LOW_MAX;
>>>>> - int ret;
>>>>> bool fixed_base, high = false;
>>>>> char *cmdline = boot_command_line;
>>>>> + int ret;
>>>>>
>>>>> /* crashkernel=X[@offset] */
>>>>> ret = parse_crashkernel(cmdline, memblock_phys_mem_size(),
>>>>> &crash_size, &crash_base);
>>>>> if (ret || !crash_size) {
>>>>> - /* crashkernel=X,high */
>>>>> ret = parse_crashkernel_high(cmdline, 0, &crash_size, &crash_base);
>>>>> if (ret || !crash_size)
>>>>> return;
>>>>>
>>>>> - /* crashkernel=Y,low */
>>>>> ret = parse_crashkernel_low(cmdline, 0, &crash_low_size, &crash_base);
>>>>> if (ret == -ENOENT)
>>>>> - /*
>>>>> - * crashkernel=Y,low is not specified explicitly, use
>>>>> - * default size automatically.
>>>>> - */
>>>>> + /* case #2 of crashkernel,low reservation */
>>>>> crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
>>>>> else if (ret)
>>>>> - /* crashkernel=Y,low is specified but Y is invalid */
>>>>> + /* case #4 of crashkernel,low reservation */
>>>>> return;
>>>>>
>>>>> - /* Mark crashkernel=X,high is specified */
>>>>> high = true;
>>>>> crash_max = CRASH_ADDR_HIGH_MAX;
>>>>> }
>>>>> @@ -148,7 +153,6 @@ static void __init reserve_crashkernel(void)
>>>>> fixed_base = !!crash_base;
>>>>> crash_size = PAGE_ALIGN(crash_size);
>>>>>
>>>>> - /* User specifies base address explicitly. */
>>>>> if (fixed_base)
>>>>> crash_max = crash_base + crash_size;
>>>>>
>>>>> @@ -172,11 +176,7 @@ static void __init reserve_crashkernel(void)
>>>>> }
>>>>>
>>>>> if (crash_base >= SZ_4G) {
>>>>> - /*
>>>>> - * For case crashkernel=X, low memory is not enough and fall
>>>>> - * back to reserve specified size of memory above 4G, try to
>>>>> - * allocate minimum required memory below 4G again.
>>>>> - */
>>>>> + /* case #3 of crashkernel,low reservation */
>>>>> if (!high)
>>>>> crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
>>>>>
>>>>>
>>>>>>
>>>>>> - /* Current arm64 boot protocol requires 2MB alignment */
>>>>>> - crash_base = memblock_phys_alloc_range(crash_size, SZ_2M,
>>>>>> +retry:
>>>>>> + crash_base = memblock_phys_alloc_range(crash_size, CRASH_ALIGN,
>>>>>> crash_base, crash_max);
>>>>>> if (!crash_base) {
>>>>>> + /*
>>>>>> + * Attempt to fully allocate low memory failed, fall back
>>>>>> + * to high memory, the minimum required low memory will be
>>>>>> + * reserved later.
>>>>>> + */
>>>>>> + if (!fixed_base && (crash_max == CRASH_ADDR_LOW_MAX)) {
>>>>>> + crash_max = CRASH_ADDR_HIGH_MAX;
>>>>>> + goto retry;
>>>>>> + }
>>>>>> +
>>>>>> pr_warn("cannot allocate crashkernel (size:0x%llx)\n",
>>>>>> crash_size);
>>>>>> return;
>>>>>> }
>>>>>>
>>>>>> + if (crash_base >= SZ_4G) {
>>>>>> + /*
>>>>>> + * For case crashkernel=X, low memory is not enough and fall
>>>>>> + * back to reserve specified size of memory above 4G, try to
>>>>>> + * allocate minimum required memory below 4G again.
>>>>>> + */
>>>>>> + if (!high)
>>>>>> + crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
>>>>>> +
>>>>>> + if (reserve_crashkernel_low(crash_low_size)) {
>>>>>> + memblock_phys_free(crash_base, crash_size);
>>>>>> + return;
>>>>>> + }
>>>>>> + }
>>>>>> +
>>>>>> pr_info("crashkernel reserved: 0x%016llx - 0x%016llx (%lld MB)\n",
>>>>>> crash_base, crash_base + crash_size, crash_size >> 20);
>>>>>>
>>>>>> @@ -107,6 +194,9 @@ static void __init reserve_crashkernel(void)
>>>>>> * map. Inform kmemleak so that it won't try to access it.
>>>>>> */
>>>>>> kmemleak_ignore_phys(crash_base);
>>>>>> + if (crashk_low_res.end)
>>>>>> + kmemleak_ignore_phys(crashk_low_res.start);
>>>>>> +
>>>>>> crashk_res.start = crash_base;
>>>>>> crashk_res.end = crash_base + crash_size - 1;
>>>>>> insert_resource(&iomem_resource, &crashk_res);
>>>>>> --
>>>>>> 2.25.1
>>>>>>
>>>>>
>>>>> .
>>>>>
>>>>
>>>> --
>>>> Regards,
>>>> Zhen Lei
>>>>
>>>
>>> .
>>>
>>
>> --
>> Regards,
>> Zhen Lei
>>
>
> .
>
--
Regards,
Zhen Lei
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X
2022-02-27 3:07 ` [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X Zhen Lei
2022-03-16 12:11 ` Baoquan He
@ 2022-03-17 2:38 ` Baoquan He
2022-03-17 3:23 ` Leizhen (ThunderTown)
2022-03-21 13:29 ` John Donnelly
2 siblings, 1 reply; 27+ messages in thread
From: Baoquan He @ 2022-03-17 2:38 UTC (permalink / raw)
To: Zhen Lei
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 02/27/22 at 11:07am, Zhen Lei wrote:
> From: Chen Zhou <chenzhou10@huawei.com>
>
> 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
~~ change it to "get boot failure" or "fail to boot"
> 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.
>
> Signed-off-by: Chen Zhou <chenzhou10@huawei.com>
> Co-developed-by: Zhen Lei <thunder.leizhen@huawei.com>
> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X
2022-03-17 2:38 ` Baoquan He
@ 2022-03-17 3:23 ` Leizhen (ThunderTown)
0 siblings, 0 replies; 27+ messages in thread
From: Leizhen (ThunderTown) @ 2022-03-17 3:23 UTC (permalink / raw)
To: Baoquan He
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 2022/3/17 10:38, Baoquan He wrote:
> On 02/27/22 at 11:07am, Zhen Lei wrote:
>> From: Chen Zhou <chenzhou10@huawei.com>
>>
>> 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
> ~~ change it to "get boot failure" or "fail to boot"
OK. I'm going to use "fail to boot".
>> 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.
>>
>> Signed-off-by: Chen Zhou <chenzhou10@huawei.com>
>> Co-developed-by: Zhen Lei <thunder.leizhen@huawei.com>
>> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
>
> .
>
--
Regards,
Zhen Lei
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X
2022-02-27 3:07 ` [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X Zhen Lei
2022-03-16 12:11 ` Baoquan He
2022-03-17 2:38 ` Baoquan He
@ 2022-03-21 13:29 ` John Donnelly
2022-03-21 14:09 ` Dave Kleikamp
2022-03-22 1:58 ` Leizhen (ThunderTown)
2 siblings, 2 replies; 27+ messages in thread
From: John Donnelly @ 2022-03-21 13:29 UTC (permalink / raw)
To: Zhen Lei, Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Baoquan He,
Vivek Goyal, Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, John Donnelly
Cc: Randy Dunlap, Feng Zhou, Kefeng Wang, Chen Zhou, Dave Kleikamp
On 2/26/22 9:07 PM, Zhen Lei wrote:
> From: Chen Zhou <chenzhou10@huawei.com>
>
> 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.
" Not enough "
> 2. If reserving crashkernel above 4G, in this case, crash dump
> kernel will boot failure because there is no low memory available
> for allocation.
We can't have a "boot failure". If the requested reservation
can not be met, the kdump configuration is not setup.
>
> 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.
Is there going to be documentation on what values certain Arm platforms
are going to use this on ?
>
> Signed-off-by: Chen Zhou <chenzhou10@huawei.com>
> Co-developed-by: Zhen Lei <thunder.leizhen@huawei.com>
> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
> ---
> arch/arm64/kernel/machine_kexec.c | 9 ++-
> arch/arm64/kernel/machine_kexec_file.c | 12 ++-
> arch/arm64/mm/init.c | 106 +++++++++++++++++++++++--
> 3 files changed, 115 insertions(+), 12 deletions(-)
>
> diff --git a/arch/arm64/kernel/machine_kexec.c b/arch/arm64/kernel/machine_kexec.c
> index e16b248699d5c3c..19c2d487cb08feb 100644
> --- a/arch/arm64/kernel/machine_kexec.c
> +++ b/arch/arm64/kernel/machine_kexec.c
> @@ -329,8 +329,13 @@ bool crash_is_nosave(unsigned long pfn)
>
> /* in reserved memory? */
> addr = __pfn_to_phys(pfn);
> - if ((addr < crashk_res.start) || (crashk_res.end < addr))
> - return false;
> + if ((addr < crashk_res.start) || (crashk_res.end < addr)) {
> + if (!crashk_low_res.end)
> + return false;
> +
> + if ((addr < crashk_low_res.start) || (crashk_low_res.end < addr))
> + return false;
> + }
>
> if (!kexec_crash_image)
> return true;
> diff --git a/arch/arm64/kernel/machine_kexec_file.c b/arch/arm64/kernel/machine_kexec_file.c
> index 59c648d51848886..889951291cc0f9c 100644
> --- a/arch/arm64/kernel/machine_kexec_file.c
> +++ b/arch/arm64/kernel/machine_kexec_file.c
> @@ -65,10 +65,18 @@ static int prepare_elf_headers(void **addr, unsigned long *sz)
>
> /* Exclude crashkernel region */
> ret = crash_exclude_mem_range(cmem, crashk_res.start, crashk_res.end);
> + if (ret)
> + goto out;
> +
> + if (crashk_low_res.end) {
> + ret = crash_exclude_mem_range(cmem, crashk_low_res.start, crashk_low_res.end);
> + if (ret)
> + goto out;
> + }
>
> - if (!ret)
> - ret = crash_prepare_elf64_headers(cmem, true, addr, sz);
> + ret = crash_prepare_elf64_headers(cmem, true, addr, sz);
>
> +out:
> kfree(cmem);
> return ret;
> }
> diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
> index 90f276d46b93bc6..30ae6638ff54c47 100644
> --- a/arch/arm64/mm/init.c
> +++ b/arch/arm64/mm/init.c
> @@ -65,6 +65,44 @@ EXPORT_SYMBOL(memstart_addr);
> phys_addr_t arm64_dma_phys_limit __ro_after_init;
>
> #ifdef CONFIG_KEXEC_CORE
> +/* Current arm64 boot protocol requires 2MB alignment */
> +#define CRASH_ALIGN SZ_2M
> +
> +#define CRASH_ADDR_LOW_MAX arm64_dma_phys_limit
> +#define CRASH_ADDR_HIGH_MAX memblock.current_limit
> +
> +/*
> + * This is an empirical value in x86_64 and taken here directly. Please
> + * refer to the code comment in reserve_crashkernel_low() of x86_64 for more
> + * details.
> + */
> +#define DEFAULT_CRASH_KERNEL_LOW_SIZE \
> + max(swiotlb_size_or_default() + (8UL << 20), 256UL << 20)
> +
> +static int __init reserve_crashkernel_low(unsigned long long low_size)
> +{
> + unsigned long long low_base;
> +
> + /* passed with crashkernel=0,low ? */
> + if (!low_size)
> + return 0;
> +
> + low_base = memblock_phys_alloc_range(low_size, CRASH_ALIGN, 0, CRASH_ADDR_LOW_MAX);
> + if (!low_base) {
> + pr_err("cannot allocate crashkernel low memory (size:0x%llx).\n", low_size);
> + return -ENOMEM;
> + }
> +
> + pr_info("crashkernel low memory reserved: 0x%08llx - 0x%08llx (%lld MB)\n",
> + low_base, low_base + low_size, low_size >> 20);
> +
> + crashk_low_res.start = low_base;
> + crashk_low_res.end = low_base + low_size - 1;
> + insert_resource(&iomem_resource, &crashk_low_res);
> +
> + return 0;
> +}
> +
> /*
> * reserve_crashkernel() - reserves memory for crash kernel
> *
> @@ -75,30 +113,79 @@ phys_addr_t arm64_dma_phys_limit __ro_after_init;
> static void __init reserve_crashkernel(void)
> {
> unsigned long long crash_base, crash_size;
> - unsigned long long crash_max = arm64_dma_phys_limit;
> + unsigned long long crash_low_size;
> + unsigned long long crash_max = CRASH_ADDR_LOW_MAX;
> int ret;
> + bool fixed_base, high = false;
> + char *cmdline = boot_command_line;
>
> - ret = parse_crashkernel(boot_command_line, memblock_phys_mem_size(),
> + /* crashkernel=X[@offset] */
> + ret = parse_crashkernel(cmdline, memblock_phys_mem_size(),
> &crash_size, &crash_base);
> - /* no crashkernel= or invalid value specified */
> - if (ret || !crash_size)
> - return;
> + if (ret || !crash_size) {
> + /* crashkernel=X,high */
> + ret = parse_crashkernel_high(cmdline, 0, &crash_size, &crash_base);
> + if (ret || !crash_size)
> + return;
> +
> + /* crashkernel=Y,low */
> + ret = parse_crashkernel_low(cmdline, 0, &crash_low_size, &crash_base);
> + if (ret == -ENOENT)
> + /*
> + * crashkernel=Y,low is not specified explicitly, use
> + * default size automatically.
> + */
> + crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
> + else if (ret)
> + /* crashkernel=Y,low is specified but Y is invalid */
> + return;
> +
> + /* Mark crashkernel=X,high is specified */
> + high = true;
> + crash_max = CRASH_ADDR_HIGH_MAX;
> + }
>
> + fixed_base = !!crash_base;
> crash_size = PAGE_ALIGN(crash_size);
>
> /* User specifies base address explicitly. */
> - if (crash_base)
> + if (fixed_base)
> crash_max = crash_base + crash_size;
>
> - /* Current arm64 boot protocol requires 2MB alignment */
> - crash_base = memblock_phys_alloc_range(crash_size, SZ_2M,
> +retry:
> + crash_base = memblock_phys_alloc_range(crash_size, CRASH_ALIGN,
> crash_base, crash_max);
> if (!crash_base) {
> + /*
> + * Attempt to fully allocate low memory failed, fall back
> + * to high memory, the minimum required low memory will be
> + * reserved later.
> + */
> + if (!fixed_base && (crash_max == CRASH_ADDR_LOW_MAX)) {
> + crash_max = CRASH_ADDR_HIGH_MAX;
> + goto retry;
> + }
> +
> pr_warn("cannot allocate crashkernel (size:0x%llx)\n",
> crash_size);
> return;
> }
>
> + if (crash_base >= SZ_4G) {
> + /*
> + * For case crashkernel=X, low memory is not enough and fall
> + * back to reserve specified size of memory above 4G, try to
> + * allocate minimum required memory below 4G again.
> + */
> + if (!high)
> + crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
> +
> + if (reserve_crashkernel_low(crash_low_size)) {
> + memblock_phys_free(crash_base, crash_size);
> + return;
> + }
> + }
> +
> pr_info("crashkernel reserved: 0x%016llx - 0x%016llx (%lld MB)\n",
> crash_base, crash_base + crash_size, crash_size >> 20);
>
> @@ -107,6 +194,9 @@ static void __init reserve_crashkernel(void)
> * map. Inform kmemleak so that it won't try to access it.
> */
> kmemleak_ignore_phys(crash_base);
> + if (crashk_low_res.end)
> + kmemleak_ignore_phys(crashk_low_res.start);
> +
> crashk_res.start = crash_base;
> crashk_res.end = crash_base + crash_size - 1;
> insert_resource(&iomem_resource, &crashk_res);
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X
2022-03-21 13:29 ` John Donnelly
@ 2022-03-21 14:09 ` Dave Kleikamp
2022-03-22 1:58 ` Leizhen (ThunderTown)
1 sibling, 0 replies; 27+ messages in thread
From: Dave Kleikamp @ 2022-03-21 14:09 UTC (permalink / raw)
To: John Donnelly, Zhen Lei, Thomas Gleixner, Ingo Molnar,
Borislav Petkov, x86, H . Peter Anvin, linux-kernel, Dave Young,
Baoquan He, Vivek Goyal, Eric Biederman, kexec, Catalin Marinas,
Will Deacon, linux-arm-kernel, Rob Herring, Frank Rowand,
devicetree, Jonathan Corbet, linux-doc
Cc: Randy Dunlap, Feng Zhou, Kefeng Wang, Chen Zhou
On 3/21/22 8:29AM, John Donnelly wrote:
> On 2/26/22 9:07 PM, Zhen Lei wrote:
>> From: Chen Zhou <chenzhou10@huawei.com>
>>
>> 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.
>
> " Not enough "
>> 2. If reserving crashkernel above 4G, in this case, crash dump
>> kernel will boot failure because there is no low memory available
>> for allocation.
>
> We can't have a "boot failure". If the requested reservation
> can not be met, the kdump configuration is not setup.
I think you misread this. Without these patches, if only high memory is
reserved for the crash kernel, then the crash kernel will fail to boot.
>>
>> 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.
>
> Is there going to be documentation on what values certain Arm platforms
> are going to use this on ?
>
>>
>> Signed-off-by: Chen Zhou <chenzhou10@huawei.com>
>> Co-developed-by: Zhen Lei <thunder.leizhen@huawei.com>
>> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X
2022-03-21 13:29 ` John Donnelly
2022-03-21 14:09 ` Dave Kleikamp
@ 2022-03-22 1:58 ` Leizhen (ThunderTown)
1 sibling, 0 replies; 27+ messages in thread
From: Leizhen (ThunderTown) @ 2022-03-22 1:58 UTC (permalink / raw)
To: John Donnelly, Thomas Gleixner, Ingo Molnar, Borislav Petkov,
x86, H . Peter Anvin, linux-kernel, Dave Young, Baoquan He,
Vivek Goyal, Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc
Cc: Randy Dunlap, Feng Zhou, Kefeng Wang, Chen Zhou, Dave Kleikamp
On 2022/3/21 21:29, John Donnelly wrote:
> On 2/26/22 9:07 PM, Zhen Lei wrote:
>> From: Chen Zhou <chenzhou10@huawei.com>
>>
>> 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.
>
> " Not enough "
OK, thanks
>> 2. If reserving crashkernel above 4G, in this case, crash dump
>> kernel will boot failure because there is no low memory available
>> for allocation.
>
> We can't have a "boot failure". If the requested reservation
> can not be met, the kdump configuration is not setup.
>>
>> 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.
>
> Is there going to be documentation on what values certain Arm platforms are going to use this on ?
There is no exact formula.
>
>>
>> Signed-off-by: Chen Zhou <chenzhou10@huawei.com>
>> Co-developed-by: Zhen Lei <thunder.leizhen@huawei.com>
>> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
>> ---
>> arch/arm64/kernel/machine_kexec.c | 9 ++-
>> arch/arm64/kernel/machine_kexec_file.c | 12 ++-
>> arch/arm64/mm/init.c | 106 +++++++++++++++++++++++--
>> 3 files changed, 115 insertions(+), 12 deletions(-)
>>
>> diff --git a/arch/arm64/kernel/machine_kexec.c b/arch/arm64/kernel/machine_kexec.c
>> index e16b248699d5c3c..19c2d487cb08feb 100644
>> --- a/arch/arm64/kernel/machine_kexec.c
>> +++ b/arch/arm64/kernel/machine_kexec.c
>> @@ -329,8 +329,13 @@ bool crash_is_nosave(unsigned long pfn)
>> /* in reserved memory? */
>> addr = __pfn_to_phys(pfn);
>> - if ((addr < crashk_res.start) || (crashk_res.end < addr))
>> - return false;
>> + if ((addr < crashk_res.start) || (crashk_res.end < addr)) {
>> + if (!crashk_low_res.end)
>> + return false;
>> +
>> + if ((addr < crashk_low_res.start) || (crashk_low_res.end < addr))
>> + return false;
>> + }
>> if (!kexec_crash_image)
>> return true;
>> diff --git a/arch/arm64/kernel/machine_kexec_file.c b/arch/arm64/kernel/machine_kexec_file.c
>> index 59c648d51848886..889951291cc0f9c 100644
>> --- a/arch/arm64/kernel/machine_kexec_file.c
>> +++ b/arch/arm64/kernel/machine_kexec_file.c
>> @@ -65,10 +65,18 @@ static int prepare_elf_headers(void **addr, unsigned long *sz)
>> /* Exclude crashkernel region */
>> ret = crash_exclude_mem_range(cmem, crashk_res.start, crashk_res.end);
>> + if (ret)
>> + goto out;
>> +
>> + if (crashk_low_res.end) {
>> + ret = crash_exclude_mem_range(cmem, crashk_low_res.start, crashk_low_res.end);
>> + if (ret)
>> + goto out;
>> + }
>> - if (!ret)
>> - ret = crash_prepare_elf64_headers(cmem, true, addr, sz);
>> + ret = crash_prepare_elf64_headers(cmem, true, addr, sz);
>> +out:
>> kfree(cmem);
>> return ret;
>> }
>> diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
>> index 90f276d46b93bc6..30ae6638ff54c47 100644
>> --- a/arch/arm64/mm/init.c
>> +++ b/arch/arm64/mm/init.c
>> @@ -65,6 +65,44 @@ EXPORT_SYMBOL(memstart_addr);
>> phys_addr_t arm64_dma_phys_limit __ro_after_init;
>> #ifdef CONFIG_KEXEC_CORE
>> +/* Current arm64 boot protocol requires 2MB alignment */
>> +#define CRASH_ALIGN SZ_2M
>> +
>> +#define CRASH_ADDR_LOW_MAX arm64_dma_phys_limit
>> +#define CRASH_ADDR_HIGH_MAX memblock.current_limit
>> +
>> +/*
>> + * This is an empirical value in x86_64 and taken here directly. Please
>> + * refer to the code comment in reserve_crashkernel_low() of x86_64 for more
>> + * details.
>> + */
>> +#define DEFAULT_CRASH_KERNEL_LOW_SIZE \
>> + max(swiotlb_size_or_default() + (8UL << 20), 256UL << 20)
>> +
>> +static int __init reserve_crashkernel_low(unsigned long long low_size)
>> +{
>> + unsigned long long low_base;
>> +
>> + /* passed with crashkernel=0,low ? */
>> + if (!low_size)
>> + return 0;
>> +
>> + low_base = memblock_phys_alloc_range(low_size, CRASH_ALIGN, 0, CRASH_ADDR_LOW_MAX);
>> + if (!low_base) {
>> + pr_err("cannot allocate crashkernel low memory (size:0x%llx).\n", low_size);
>> + return -ENOMEM;
>> + }
>> +
>> + pr_info("crashkernel low memory reserved: 0x%08llx - 0x%08llx (%lld MB)\n",
>> + low_base, low_base + low_size, low_size >> 20);
>> +
>> + crashk_low_res.start = low_base;
>> + crashk_low_res.end = low_base + low_size - 1;
>> + insert_resource(&iomem_resource, &crashk_low_res);
>> +
>> + return 0;
>> +}
>> +
>> /*
>> * reserve_crashkernel() - reserves memory for crash kernel
>> *
>> @@ -75,30 +113,79 @@ phys_addr_t arm64_dma_phys_limit __ro_after_init;
>> static void __init reserve_crashkernel(void)
>> {
>> unsigned long long crash_base, crash_size;
>> - unsigned long long crash_max = arm64_dma_phys_limit;
>> + unsigned long long crash_low_size;
>> + unsigned long long crash_max = CRASH_ADDR_LOW_MAX;
>> int ret;
>> + bool fixed_base, high = false;
>> + char *cmdline = boot_command_line;
>> - ret = parse_crashkernel(boot_command_line, memblock_phys_mem_size(),
>> + /* crashkernel=X[@offset] */
>> + ret = parse_crashkernel(cmdline, memblock_phys_mem_size(),
>> &crash_size, &crash_base);
>> - /* no crashkernel= or invalid value specified */
>> - if (ret || !crash_size)
>> - return;
>> + if (ret || !crash_size) {
>> + /* crashkernel=X,high */
>> + ret = parse_crashkernel_high(cmdline, 0, &crash_size, &crash_base);
>> + if (ret || !crash_size)
>> + return;
>> +
>> + /* crashkernel=Y,low */
>> + ret = parse_crashkernel_low(cmdline, 0, &crash_low_size, &crash_base);
>> + if (ret == -ENOENT)
>> + /*
>> + * crashkernel=Y,low is not specified explicitly, use
>> + * default size automatically.
>> + */
>> + crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
>> + else if (ret)
>> + /* crashkernel=Y,low is specified but Y is invalid */
>> + return;
>> +
>> + /* Mark crashkernel=X,high is specified */
>> + high = true;
>> + crash_max = CRASH_ADDR_HIGH_MAX;
>> + }
>> + fixed_base = !!crash_base;
>> crash_size = PAGE_ALIGN(crash_size);
>> /* User specifies base address explicitly. */
>> - if (crash_base)
>> + if (fixed_base)
>> crash_max = crash_base + crash_size;
>> - /* Current arm64 boot protocol requires 2MB alignment */
>> - crash_base = memblock_phys_alloc_range(crash_size, SZ_2M,
>> +retry:
>> + crash_base = memblock_phys_alloc_range(crash_size, CRASH_ALIGN,
>> crash_base, crash_max);
>> if (!crash_base) {
>> + /*
>> + * Attempt to fully allocate low memory failed, fall back
>> + * to high memory, the minimum required low memory will be
>> + * reserved later.
>> + */
>> + if (!fixed_base && (crash_max == CRASH_ADDR_LOW_MAX)) {
>> + crash_max = CRASH_ADDR_HIGH_MAX;
>> + goto retry;
>> + }
>> +
>> pr_warn("cannot allocate crashkernel (size:0x%llx)\n",
>> crash_size);
>> return;
>> }
>> + if (crash_base >= SZ_4G) {
>> + /*
>> + * For case crashkernel=X, low memory is not enough and fall
>> + * back to reserve specified size of memory above 4G, try to
>> + * allocate minimum required memory below 4G again.
>> + */
>> + if (!high)
>> + crash_low_size = DEFAULT_CRASH_KERNEL_LOW_SIZE;
>> +
>> + if (reserve_crashkernel_low(crash_low_size)) {
>> + memblock_phys_free(crash_base, crash_size);
>> + return;
>> + }
>> + }
>> +
>> pr_info("crashkernel reserved: 0x%016llx - 0x%016llx (%lld MB)\n",
>> crash_base, crash_base + crash_size, crash_size >> 20);
>> @@ -107,6 +194,9 @@ static void __init reserve_crashkernel(void)
>> * map. Inform kmemleak so that it won't try to access it.
>> */
>> kmemleak_ignore_phys(crash_base);
>> + if (crashk_low_res.end)
>> + kmemleak_ignore_phys(crashk_low_res.start);
>> +
>> crashk_res.start = crash_base;
>> crashk_res.end = crash_base + crash_size - 1;
>> insert_resource(&iomem_resource, &crashk_res);
>
> .
>
--
Regards,
Zhen Lei
^ permalink raw reply [flat|nested] 27+ messages in thread
* [PATCH v21 4/5] of: fdt: Add memory for devices by DT property "linux,usable-memory-range"
2022-02-27 3:07 [PATCH v21 0/5] support reserving crashkernel above 4G on arm64 kdump Zhen Lei
` (2 preceding siblings ...)
2022-02-27 3:07 ` [PATCH v21 3/5] arm64: kdump: reimplement crashkernel=X Zhen Lei
@ 2022-02-27 3:07 ` Zhen Lei
2022-02-27 3:07 ` [PATCH v21 5/5] docs: kdump: Update the crashkernel description for arm64 Zhen Lei
2022-04-08 9:32 ` [PATCH v21 0/5] support reserving crashkernel above 4G on arm64 kdump Baoquan He
5 siblings, 0 replies; 27+ messages in thread
From: Zhen Lei @ 2022-02-27 3:07 UTC (permalink / raw)
To: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Baoquan He,
Vivek Goyal, Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc
Cc: Zhen Lei, Randy Dunlap, Feng Zhou, Kefeng Wang, Chen Zhou,
John Donnelly, Dave Kleikamp
From: Chen Zhou <chenzhou10@huawei.com>
When reserving crashkernel in high memory, some low memory is reserved
for crash dump kernel devices and never mapped by the first kernel.
This memory range is advertised to crash dump kernel via DT property
under /chosen,
linux,usable-memory-range = <BASE1 SIZE1 [BASE2 SIZE2]>
We reused the DT property linux,usable-memory-range and made the low
memory region as the second range "BASE2 SIZE2", which keeps compatibility
with existing user-space and older kdump kernels.
Crash dump kernel reads this property at boot time and call memblock_add()
to add the low memory region after memblock_cap_memory_range() has been
called.
Signed-off-by: Chen Zhou <chenzhou10@huawei.com>
Co-developed-by: Zhen Lei <thunder.leizhen@huawei.com>
Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Tested-by: Dave Kleikamp <dave.kleikamp@oracle.com>
---
drivers/of/fdt.c | 33 +++++++++++++++++++++++----------
1 file changed, 23 insertions(+), 10 deletions(-)
diff --git a/drivers/of/fdt.c b/drivers/of/fdt.c
index ec315b060cd50d2..2f248d0acc04830 100644
--- a/drivers/of/fdt.c
+++ b/drivers/of/fdt.c
@@ -973,16 +973,24 @@ static void __init early_init_dt_check_for_elfcorehdr(unsigned long node)
static unsigned long chosen_node_offset = -FDT_ERR_NOTFOUND;
+/*
+ * The main usage of linux,usable-memory-range is for crash dump kernel.
+ * Originally, the number of usable-memory regions is one. Now there may
+ * be two regions, low region and high region.
+ * To make compatibility with existing user-space and older kdump, the low
+ * region is always the last range of linux,usable-memory-range if exist.
+ */
+#define MAX_USABLE_RANGES 2
+
/**
* early_init_dt_check_for_usable_mem_range - Decode usable memory range
* location from flat tree
*/
void __init early_init_dt_check_for_usable_mem_range(void)
{
- const __be32 *prop;
- int len;
- phys_addr_t cap_mem_addr;
- phys_addr_t cap_mem_size;
+ struct memblock_region rgn[MAX_USABLE_RANGES] = {0};
+ const __be32 *prop, *endp;
+ int len, i;
unsigned long node = chosen_node_offset;
if ((long)node < 0)
@@ -991,16 +999,21 @@ void __init early_init_dt_check_for_usable_mem_range(void)
pr_debug("Looking for usable-memory-range property... ");
prop = of_get_flat_dt_prop(node, "linux,usable-memory-range", &len);
- if (!prop || (len < (dt_root_addr_cells + dt_root_size_cells)))
+ if (!prop || (len % (dt_root_addr_cells + dt_root_size_cells)))
return;
- cap_mem_addr = dt_mem_next_cell(dt_root_addr_cells, &prop);
- cap_mem_size = dt_mem_next_cell(dt_root_size_cells, &prop);
+ endp = prop + (len / sizeof(__be32));
+ for (i = 0; i < MAX_USABLE_RANGES && prop < endp; i++) {
+ rgn[i].base = dt_mem_next_cell(dt_root_addr_cells, &prop);
+ rgn[i].size = dt_mem_next_cell(dt_root_size_cells, &prop);
- pr_debug("cap_mem_start=%pa cap_mem_size=%pa\n", &cap_mem_addr,
- &cap_mem_size);
+ pr_debug("cap_mem_regions[%d]: base=%pa, size=%pa\n",
+ i, &rgn[i].base, &rgn[i].size);
+ }
- memblock_cap_memory_range(cap_mem_addr, cap_mem_size);
+ memblock_cap_memory_range(rgn[0].base, rgn[0].size);
+ for (i = 1; i < MAX_USABLE_RANGES && rgn[i].size; i++)
+ memblock_add(rgn[i].base, rgn[i].size);
}
#ifdef CONFIG_SERIAL_EARLYCON
--
2.25.1
^ permalink raw reply related [flat|nested] 27+ messages in thread
* [PATCH v21 5/5] docs: kdump: Update the crashkernel description for arm64
2022-02-27 3:07 [PATCH v21 0/5] support reserving crashkernel above 4G on arm64 kdump Zhen Lei
` (3 preceding siblings ...)
2022-02-27 3:07 ` [PATCH v21 4/5] of: fdt: Add memory for devices by DT property "linux,usable-memory-range" Zhen Lei
@ 2022-02-27 3:07 ` Zhen Lei
2022-03-15 11:59 ` Baoquan He
2022-04-08 9:32 ` [PATCH v21 0/5] support reserving crashkernel above 4G on arm64 kdump Baoquan He
5 siblings, 1 reply; 27+ messages in thread
From: Zhen Lei @ 2022-02-27 3:07 UTC (permalink / raw)
To: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Baoquan He,
Vivek Goyal, Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc
Cc: Zhen Lei, Randy Dunlap, Feng Zhou, Kefeng Wang, Chen Zhou,
John Donnelly, Dave Kleikamp
Now arm64 has added support for "crashkernel=X,high" and
"crashkernel=Y,low", and implements "crashkernel=X[@offset]" in the
same way as x86. So update the Documentation.
Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
---
Documentation/admin-guide/kernel-parameters.txt | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
index f5a27f067db9ed9..63098786c93828c 100644
--- a/Documentation/admin-guide/kernel-parameters.txt
+++ b/Documentation/admin-guide/kernel-parameters.txt
@@ -789,7 +789,7 @@
memory region [offset, offset + size] for that kernel
image. If '@offset' is omitted, then a suitable offset
is selected automatically.
- [KNL, X86-64] Select a region under 4G first, and
+ [KNL, X86-64, ARM64] Select a region under 4G first, and
fall back to reserve region above 4G when '@offset'
hasn't been specified.
See Documentation/admin-guide/kdump/kdump.rst for further details.
@@ -802,20 +802,20 @@
Documentation/admin-guide/kdump/kdump.rst for an example.
crashkernel=size[KMG],high
- [KNL, X86-64] range could be above 4G. Allow kernel
+ [KNL, X86-64, ARM64] range could be above 4G. Allow kernel
to allocate physical memory region from top, so could
be above 4G if system have more than 4G ram installed.
Otherwise memory region will be allocated below 4G, if
available.
It will be ignored if crashkernel=X is specified.
crashkernel=size[KMG],low
- [KNL, X86-64] range under 4G. When crashkernel=X,high
+ [KNL, X86-64, ARM64] range under 4G. When crashkernel=X,high
is passed, kernel could allocate physical memory region
above 4G, that cause second kernel crash on system
that require some amount of low memory, e.g. swiotlb
requires at least 64M+32K low memory, also enough extra
low memory is needed to make sure DMA buffers for 32-bit
- devices won't run out. Kernel would try to allocate at
+ devices won't run out. Kernel would try to allocate
at least 256M below 4G automatically.
This one let user to specify own low range under 4G
for second kernel instead.
--
2.25.1
^ permalink raw reply related [flat|nested] 27+ messages in thread
* Re: [PATCH v21 5/5] docs: kdump: Update the crashkernel description for arm64
2022-02-27 3:07 ` [PATCH v21 5/5] docs: kdump: Update the crashkernel description for arm64 Zhen Lei
@ 2022-03-15 11:59 ` Baoquan He
0 siblings, 0 replies; 27+ messages in thread
From: Baoquan He @ 2022-03-15 11:59 UTC (permalink / raw)
To: Zhen Lei
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 02/27/22 at 11:07am, Zhen Lei wrote:
> Now arm64 has added support for "crashkernel=X,high" and
> "crashkernel=Y,low", and implements "crashkernel=X[@offset]" in the
> same way as x86. So update the Documentation.
>
> Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Looks good to me, thx.
Acked-by: Baoquan He <bhe@redhat.com>
> ---
> Documentation/admin-guide/kernel-parameters.txt | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
> index f5a27f067db9ed9..63098786c93828c 100644
> --- a/Documentation/admin-guide/kernel-parameters.txt
> +++ b/Documentation/admin-guide/kernel-parameters.txt
> @@ -789,7 +789,7 @@
> memory region [offset, offset + size] for that kernel
> image. If '@offset' is omitted, then a suitable offset
> is selected automatically.
> - [KNL, X86-64] Select a region under 4G first, and
> + [KNL, X86-64, ARM64] Select a region under 4G first, and
> fall back to reserve region above 4G when '@offset'
> hasn't been specified.
> See Documentation/admin-guide/kdump/kdump.rst for further details.
> @@ -802,20 +802,20 @@
> Documentation/admin-guide/kdump/kdump.rst for an example.
>
> crashkernel=size[KMG],high
> - [KNL, X86-64] range could be above 4G. Allow kernel
> + [KNL, X86-64, ARM64] range could be above 4G. Allow kernel
> to allocate physical memory region from top, so could
> be above 4G if system have more than 4G ram installed.
> Otherwise memory region will be allocated below 4G, if
> available.
> It will be ignored if crashkernel=X is specified.
> crashkernel=size[KMG],low
> - [KNL, X86-64] range under 4G. When crashkernel=X,high
> + [KNL, X86-64, ARM64] range under 4G. When crashkernel=X,high
> is passed, kernel could allocate physical memory region
> above 4G, that cause second kernel crash on system
> that require some amount of low memory, e.g. swiotlb
> requires at least 64M+32K low memory, also enough extra
> low memory is needed to make sure DMA buffers for 32-bit
> - devices won't run out. Kernel would try to allocate at
> + devices won't run out. Kernel would try to allocate
> at least 256M below 4G automatically.
> This one let user to specify own low range under 4G
> for second kernel instead.
> --
> 2.25.1
>
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 0/5] support reserving crashkernel above 4G on arm64 kdump
2022-02-27 3:07 [PATCH v21 0/5] support reserving crashkernel above 4G on arm64 kdump Zhen Lei
` (4 preceding siblings ...)
2022-02-27 3:07 ` [PATCH v21 5/5] docs: kdump: Update the crashkernel description for arm64 Zhen Lei
@ 2022-04-08 9:32 ` Baoquan He
2022-04-08 9:47 ` Leizhen (ThunderTown)
5 siblings, 1 reply; 27+ messages in thread
From: Baoquan He @ 2022-04-08 9:32 UTC (permalink / raw)
To: Zhen Lei
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
Hi, Lei
On 02/27/22 at 11:07am, Zhen Lei wrote:
> Changes since [v20]:
> 1. Check whether crashkernel=Y,low is incorrectly configured or not configured. Do different processing.
> 2. Share the existing description of x86. The configuration of arm64 is the same as that of x86.
> 3. Define the value of macro CRASH_ADDR_HIGH_MAX as memblock.current_limit, instead of MEMBLOCK_ALLOC_ACCESSIBLE.
> 4. To improve readability, some lightweight code adjustments have been made to reserve_craskernel(), including comments.
> 5. The defined value of DEFAULT_CRASH_KERNEL_LOW_SIZE reconsiders swiotlb, just like x86, to share documents.
5.18 rc1 is already done, do you have plan to post a new version for
reviewing?
Thanks
Baoquan
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 0/5] support reserving crashkernel above 4G on arm64 kdump
2022-04-08 9:32 ` [PATCH v21 0/5] support reserving crashkernel above 4G on arm64 kdump Baoquan He
@ 2022-04-08 9:47 ` Leizhen (ThunderTown)
2022-04-11 2:56 ` Baoquan He
0 siblings, 1 reply; 27+ messages in thread
From: Leizhen (ThunderTown) @ 2022-04-08 9:47 UTC (permalink / raw)
To: Baoquan He
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 2022/4/8 17:32, Baoquan He wrote:
> Hi, Lei
>
> On 02/27/22 at 11:07am, Zhen Lei wrote:
>> Changes since [v20]:
>> 1. Check whether crashkernel=Y,low is incorrectly configured or not configured. Do different processing.
>> 2. Share the existing description of x86. The configuration of arm64 is the same as that of x86.
>> 3. Define the value of macro CRASH_ADDR_HIGH_MAX as memblock.current_limit, instead of MEMBLOCK_ALLOC_ACCESSIBLE.
>> 4. To improve readability, some lightweight code adjustments have been made to reserve_craskernel(), including comments.
>> 5. The defined value of DEFAULT_CRASH_KERNEL_LOW_SIZE reconsiders swiotlb, just like x86, to share documents.
>
> 5.18 rc1 is already done, do you have plan to post a new version for
> reviewing?
Yes, v5.18-rc1 has added a new patch
commit 031495635b46 ("arm64: Do not defer reserve_crashkernel() for platforms with no DMA memory zones")
to allow block mapping again, so my patches need to be modified. It should be post next week.
>
> Thanks
> Baoquan
>
> .
>
--
Regards,
Zhen Lei
^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [PATCH v21 0/5] support reserving crashkernel above 4G on arm64 kdump
2022-04-08 9:47 ` Leizhen (ThunderTown)
@ 2022-04-11 2:56 ` Baoquan He
0 siblings, 0 replies; 27+ messages in thread
From: Baoquan He @ 2022-04-11 2:56 UTC (permalink / raw)
To: Leizhen (ThunderTown)
Cc: Thomas Gleixner, Ingo Molnar, Borislav Petkov, x86,
H . Peter Anvin, linux-kernel, Dave Young, Vivek Goyal,
Eric Biederman, kexec, Catalin Marinas, Will Deacon,
linux-arm-kernel, Rob Herring, Frank Rowand, devicetree,
Jonathan Corbet, linux-doc, Randy Dunlap, Feng Zhou, Kefeng Wang,
Chen Zhou, John Donnelly, Dave Kleikamp
On 04/08/22 at 05:47pm, Leizhen (ThunderTown) wrote:
>
>
> On 2022/4/8 17:32, Baoquan He wrote:
> > Hi, Lei
> >
> > On 02/27/22 at 11:07am, Zhen Lei wrote:
> >> Changes since [v20]:
> >> 1. Check whether crashkernel=Y,low is incorrectly configured or not configured. Do different processing.
> >> 2. Share the existing description of x86. The configuration of arm64 is the same as that of x86.
> >> 3. Define the value of macro CRASH_ADDR_HIGH_MAX as memblock.current_limit, instead of MEMBLOCK_ALLOC_ACCESSIBLE.
> >> 4. To improve readability, some lightweight code adjustments have been made to reserve_craskernel(), including comments.
> >> 5. The defined value of DEFAULT_CRASH_KERNEL_LOW_SIZE reconsiders swiotlb, just like x86, to share documents.
> >
> > 5.18 rc1 is already done, do you have plan to post a new version for
> > reviewing?
>
> Yes, v5.18-rc1 has added a new patch
> commit 031495635b46 ("arm64: Do not defer reserve_crashkernel() for platforms with no DMA memory zones")
> to allow block mapping again, so my patches need to be modified. It should be post next week.
Sounds great, thanks. Just a reminder, please take your time.
^ permalink raw reply [flat|nested] 27+ messages in thread