[PATCH v4] powerpc/kdump: Add support for crashkernel CMA reservation

[PATCH v4] powerpc/kdump: Add support for crashkernel CMA reservation

[Sourabh Jain]

Commit 35c18f2933c5 ("Add a new optional ",cma" suffix to the
crashkernel= command line option") and commit ab475510e042 ("kdump:
implement reserve_crashkernel_cma") added CMA support for kdump
crashkernel reservation.

Extend crashkernel CMA reservation support to powerpc.

The following changes are made to enable CMA reservation on powerpc:

- Parse and obtain the CMA reservation size along with other crashkernel
  parameters
- Call reserve_crashkernel_cma() to allocate the CMA region for kdump
- Include the CMA-reserved ranges in the usable memory ranges for the
  kdump kernel to use.
- Exclude the CMA-reserved ranges from the crash kernel memory to
  prevent them from being exported through /proc/vmcore.

With the introduction of the CMA crashkernel regions,
crash_exclude_mem_range() needs to be called multiple times to exclude
both crashk_res and crashk_cma_ranges from the crash memory ranges. To
avoid repetitive logic for validating mem_ranges size and handling
reallocation when required, this functionality is moved to a new wrapper
function crash_exclude_mem_range_guarded().

To ensure proper CMA reservation, reserve_crashkernel_cma() is called
after pageblock_order is initialized.

Cc: Baoquan he <bhe@redhat.com>
Cc: Jiri Bohac <jbohac@suse.cz>
Cc: Hari Bathini <hbathini@linux.ibm.com>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Mahesh Salgaonkar <mahesh@linux.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Cc: Shivang Upadhyay <shivangu@linux.ibm.com>
Cc: linuxppc-dev@lists.ozlabs.org
Signed-off-by: Sourabh Jain <sourabhjain@linux.ibm.com>
---
Changlog:

v3 -> v4
 - Removed repeated initialization to tmem in
   crash_exclude_mem_range_guarded()
 - Call crash_exclude_mem_range() with right crashk ranges

---
 arch/powerpc/include/asm/kexec.h   |  2 ++
 arch/powerpc/kernel/setup-common.c |  4 ++-
 arch/powerpc/kexec/core.c          | 10 ++++++-
 arch/powerpc/kexec/ranges.c        | 43 ++++++++++++++++++++++--------
 4 files changed, 46 insertions(+), 13 deletions(-)

diff --git a/arch/powerpc/include/asm/kexec.h b/arch/powerpc/include/asm/kexec.h
index 4bbf9f699aaa..bd4a6c42a5f3 100644
--- a/arch/powerpc/include/asm/kexec.h
+++ b/arch/powerpc/include/asm/kexec.h
@@ -115,9 +115,11 @@ int setup_new_fdt_ppc64(const struct kimage *image, void *fdt, struct crash_mem
 #ifdef CONFIG_CRASH_RESERVE
 int __init overlaps_crashkernel(unsigned long start, unsigned long size);
 extern void arch_reserve_crashkernel(void);
+extern void kdump_cma_reserve(void);
 #else
 static inline void arch_reserve_crashkernel(void) {}
 static inline int overlaps_crashkernel(unsigned long start, unsigned long size) { return 0; }
+static inline void kdump_cma_reserve(void) { }
 #endif
 
 #if defined(CONFIG_CRASH_DUMP)
diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c
index 68d47c53876c..c8c42b419742 100644
--- a/arch/powerpc/kernel/setup-common.c
+++ b/arch/powerpc/kernel/setup-common.c
@@ -35,6 +35,7 @@
 #include <linux/of_irq.h>
 #include <linux/hugetlb.h>
 #include <linux/pgtable.h>
+#include <asm/kexec.h>
 #include <asm/io.h>
 #include <asm/paca.h>
 #include <asm/processor.h>
@@ -995,11 +996,12 @@ void __init setup_arch(char **cmdline_p)
 	initmem_init();
 
 	/*
-	 * Reserve large chunks of memory for use by CMA for fadump, KVM and
+	 * Reserve large chunks of memory for use by CMA for kdump, fadump, KVM and
 	 * hugetlb. These must be called after initmem_init(), so that
 	 * pageblock_order is initialised.
 	 */
 	fadump_cma_init();
+	kdump_cma_reserve();
 	kvm_cma_reserve();
 	gigantic_hugetlb_cma_reserve();
 
diff --git a/arch/powerpc/kexec/core.c b/arch/powerpc/kexec/core.c
index d1a2d755381c..25744737eff5 100644
--- a/arch/powerpc/kexec/core.c
+++ b/arch/powerpc/kexec/core.c
@@ -33,6 +33,8 @@ void machine_kexec_cleanup(struct kimage *image)
 {
 }
 
+unsigned long long cma_size;
+
 /*
  * Do not allocate memory (or fail in any way) in machine_kexec().
  * We are past the point of no return, committed to rebooting now.
@@ -110,7 +112,7 @@ void __init arch_reserve_crashkernel(void)
 
 	/* use common parsing */
 	ret = parse_crashkernel(boot_command_line, total_mem_sz, &crash_size,
-				&crash_base, NULL, NULL, NULL);
+				&crash_base, NULL, &cma_size, NULL);
 
 	if (ret)
 		return;
@@ -130,6 +132,12 @@ void __init arch_reserve_crashkernel(void)
 	reserve_crashkernel_generic(crash_size, crash_base, 0, false);
 }
 
+void __init kdump_cma_reserve(void)
+{
+	if (cma_size)
+		reserve_crashkernel_cma(cma_size);
+}
+
 int __init overlaps_crashkernel(unsigned long start, unsigned long size)
 {
 	return (start + size) > crashk_res.start && start <= crashk_res.end;
diff --git a/arch/powerpc/kexec/ranges.c b/arch/powerpc/kexec/ranges.c
index 3702b0bdab14..3bd27c38726b 100644
--- a/arch/powerpc/kexec/ranges.c
+++ b/arch/powerpc/kexec/ranges.c
@@ -515,7 +515,7 @@ int get_exclude_memory_ranges(struct crash_mem **mem_ranges)
  */
 int get_usable_memory_ranges(struct crash_mem **mem_ranges)
 {
-	int ret;
+	int ret, i;
 
 	/*
 	 * Early boot failure observed on guests when low memory (first memory
@@ -528,6 +528,13 @@ int get_usable_memory_ranges(struct crash_mem **mem_ranges)
 	if (ret)
 		goto out;
 
+	for (i = 0; i < crashk_cma_cnt; i++) {
+		ret = add_mem_range(mem_ranges, crashk_cma_ranges[i].start,
+				    crashk_cma_ranges[i].end - crashk_cma_ranges[i].start + 1);
+		if (ret)
+			goto out;
+	}
+
 	ret = add_rtas_mem_range(mem_ranges);
 	if (ret)
 		goto out;
@@ -546,6 +553,22 @@ int get_usable_memory_ranges(struct crash_mem **mem_ranges)
 #endif /* CONFIG_KEXEC_FILE */
 
 #ifdef CONFIG_CRASH_DUMP
+static int crash_exclude_mem_range_guarded(struct crash_mem **mem_ranges,
+					   unsigned long long mstart,
+					   unsigned long long mend)
+{
+	struct crash_mem *tmem = *mem_ranges;
+
+	/* Reallocate memory ranges if there is no space to split ranges */
+	if (tmem && (tmem->nr_ranges == tmem->max_nr_ranges)) {
+		tmem = realloc_mem_ranges(mem_ranges);
+		if (!tmem)
+			return -ENOMEM;
+	}
+
+	return crash_exclude_mem_range(tmem, mstart, mend);
+}
+
 /**
  * get_crash_memory_ranges - Get crash memory ranges. This list includes
  *                           first/crashing kernel's memory regions that
@@ -557,7 +580,6 @@ int get_usable_memory_ranges(struct crash_mem **mem_ranges)
 int get_crash_memory_ranges(struct crash_mem **mem_ranges)
 {
 	phys_addr_t base, end;
-	struct crash_mem *tmem;
 	u64 i;
 	int ret;
 
@@ -582,19 +604,18 @@ int get_crash_memory_ranges(struct crash_mem **mem_ranges)
 			sort_memory_ranges(*mem_ranges, true);
 	}
 
-	/* Reallocate memory ranges if there is no space to split ranges */
-	tmem = *mem_ranges;
-	if (tmem && (tmem->nr_ranges == tmem->max_nr_ranges)) {
-		tmem = realloc_mem_ranges(mem_ranges);
-		if (!tmem)
-			goto out;
-	}
-
 	/* Exclude crashkernel region */
-	ret = crash_exclude_mem_range(tmem, crashk_res.start, crashk_res.end);
+	ret = crash_exclude_mem_range_guarded(mem_ranges, crashk_res.start, crashk_res.end);
 	if (ret)
 		goto out;
 
+	for (i = 0; i < crashk_cma_cnt; ++i) {
+		ret = crash_exclude_mem_range_guarded(mem_ranges, crashk_cma_ranges[i].start,
+					      crashk_cma_ranges[i].end);
+		if (ret)
+			goto out;
+	}
+
 	/*
 	 * FIXME: For now, stay in parity with kexec-tools but if RTAS/OPAL
 	 *        regions are exported to save their context at the time of
-- 
2.51.0

Re: [PATCH v4] powerpc/kdump: Add support for crashkernel CMA reservation

[Ritesh Harjani]

Sourabh Jain <sourabhjain@linux.ibm.com> writes:

> Commit 35c18f2933c5 ("Add a new optional ",cma" suffix to the
> crashkernel= command line option") and commit ab475510e042 ("kdump:
> implement reserve_crashkernel_cma") added CMA support for kdump
> crashkernel reservation.
>
> Extend crashkernel CMA reservation support to powerpc.
>
> The following changes are made to enable CMA reservation on powerpc:
>
> - Parse and obtain the CMA reservation size along with other crashkernel
>   parameters
> - Call reserve_crashkernel_cma() to allocate the CMA region for kdump
> - Include the CMA-reserved ranges in the usable memory ranges for the
>   kdump kernel to use.
> - Exclude the CMA-reserved ranges from the crash kernel memory to
>   prevent them from being exported through /proc/vmcore.
>
> With the introduction of the CMA crashkernel regions,
> crash_exclude_mem_range() needs to be called multiple times to exclude
> both crashk_res and crashk_cma_ranges from the crash memory ranges. To
> avoid repetitive logic for validating mem_ranges size and handling
> reallocation when required, this functionality is moved to a new wrapper
> function crash_exclude_mem_range_guarded().
>
> To ensure proper CMA reservation, reserve_crashkernel_cma() is called
> after pageblock_order is initialized.
>
> Cc: Baoquan he <bhe@redhat.com>
> Cc: Jiri Bohac <jbohac@suse.cz>
> Cc: Hari Bathini <hbathini@linux.ibm.com>
> Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
> Cc: Mahesh Salgaonkar <mahesh@linux.ibm.com>
> Cc: Michael Ellerman <mpe@ellerman.id.au>
> Cc: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
> Cc: Shivang Upadhyay <shivangu@linux.ibm.com>
> Cc: linuxppc-dev@lists.ozlabs.org
> Signed-off-by: Sourabh Jain <sourabhjain@linux.ibm.com>
> ---
> Changlog:
>
> v3 -> v4
>  - Removed repeated initialization to tmem in
>    crash_exclude_mem_range_guarded()
>  - Call crash_exclude_mem_range() with right crashk ranges
>
> ---
>  arch/powerpc/include/asm/kexec.h   |  2 ++
>  arch/powerpc/kernel/setup-common.c |  4 ++-
>  arch/powerpc/kexec/core.c          | 10 ++++++-
>  arch/powerpc/kexec/ranges.c        | 43 ++++++++++++++++++++++--------
>  4 files changed, 46 insertions(+), 13 deletions(-)


I am not much familiar with the crash kernel workings but was curious
about the following query related to this patch:

As I understand this patch allows for the remaining crash kernel
memory to come from CMA region. But do we limit the CMA region to be lower
than 4G? Is this patch dependent over your other patch series [1] which
supports high crashkernel reservation?

[1]: https://lore.kernel.org/linuxppc-dev/20251027151338.819957-1-sourabhjain@linux.ibm.com/


A minor comment: I guess we could reflect the support for ,cma for PPC
too in Documentation/admin-guide/kernel-parameters.txt like how it is
done for x86.


-ritesh

Re: [PATCH v4] powerpc/kdump: Add support for crashkernel CMA reservation

[Sourabh Jain]

Hi Ritesh,

On 31/10/25 09:27, Ritesh Harjani (IBM) wrote:
> Sourabh Jain <sourabhjain@linux.ibm.com> writes:
>
>> Commit 35c18f2933c5 ("Add a new optional ",cma" suffix to the
>> crashkernel= command line option") and commit ab475510e042 ("kdump:
>> implement reserve_crashkernel_cma") added CMA support for kdump
>> crashkernel reservation.
>>
>> Extend crashkernel CMA reservation support to powerpc.
>>
>> The following changes are made to enable CMA reservation on powerpc:
>>
>> - Parse and obtain the CMA reservation size along with other crashkernel
>>    parameters
>> - Call reserve_crashkernel_cma() to allocate the CMA region for kdump
>> - Include the CMA-reserved ranges in the usable memory ranges for the
>>    kdump kernel to use.
>> - Exclude the CMA-reserved ranges from the crash kernel memory to
>>    prevent them from being exported through /proc/vmcore.
>>
>> With the introduction of the CMA crashkernel regions,
>> crash_exclude_mem_range() needs to be called multiple times to exclude
>> both crashk_res and crashk_cma_ranges from the crash memory ranges. To
>> avoid repetitive logic for validating mem_ranges size and handling
>> reallocation when required, this functionality is moved to a new wrapper
>> function crash_exclude_mem_range_guarded().
>>
>> To ensure proper CMA reservation, reserve_crashkernel_cma() is called
>> after pageblock_order is initialized.
>>
>> Cc: Baoquan he <bhe@redhat.com>
>> Cc: Jiri Bohac <jbohac@suse.cz>
>> Cc: Hari Bathini <hbathini@linux.ibm.com>
>> Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
>> Cc: Mahesh Salgaonkar <mahesh@linux.ibm.com>
>> Cc: Michael Ellerman <mpe@ellerman.id.au>
>> Cc: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
>> Cc: Shivang Upadhyay <shivangu@linux.ibm.com>
>> Cc: linuxppc-dev@lists.ozlabs.org
>> Signed-off-by: Sourabh Jain <sourabhjain@linux.ibm.com>
>> ---
>> Changlog:
>>
>> v3 -> v4
>>   - Removed repeated initialization to tmem in
>>     crash_exclude_mem_range_guarded()
>>   - Call crash_exclude_mem_range() with right crashk ranges
>>
>> ---
>>   arch/powerpc/include/asm/kexec.h   |  2 ++
>>   arch/powerpc/kernel/setup-common.c |  4 ++-
>>   arch/powerpc/kexec/core.c          | 10 ++++++-
>>   arch/powerpc/kexec/ranges.c        | 43 ++++++++++++++++++++++--------
>>   4 files changed, 46 insertions(+), 13 deletions(-)
>
> I am not much familiar with the crash kernel workings but was curious
> about the following query related to this patch:
>
> As I understand this patch allows for the remaining crash kernel
> memory to come from CMA region. But do we limit the CMA region to be lower
> than 4G?

No we are not and we don't need to.

> Is this patch dependent over your other patch series [1] which
> supports high crashkernel reservation?
>
> [1]: https://lore.kernel.org/linuxppc-dev/20251027151338.819957-1-sourabhjain@linux.ibm.com/

No, this is an independent patch.

>
>
> A minor comment: I guess we could reflect the support for ,cma for PPC
> too in Documentation/admin-guide/kernel-parameters.txt like how it is
> done for x86.

Yes we should. I will make the change in the next version.

Thanks for the review.

- Sourabh Jain

Re: [PATCH v4] powerpc/kdump: Add support for crashkernel CMA reservation

[Ritesh Harjani]

Sourabh Jain <sourabhjain@linux.ibm.com> writes:

> Hi Ritesh,
>
> On 31/10/25 09:27, Ritesh Harjani (IBM) wrote:
>> Sourabh Jain <sourabhjain@linux.ibm.com> writes:
>>
>>> Commit 35c18f2933c5 ("Add a new optional ",cma" suffix to the
>>> crashkernel= command line option") and commit ab475510e042 ("kdump:
>>> implement reserve_crashkernel_cma") added CMA support for kdump
>>> crashkernel reservation.
>>>
>>> Extend crashkernel CMA reservation support to powerpc.
>>>
>>> The following changes are made to enable CMA reservation on powerpc:
>>>
>>> - Parse and obtain the CMA reservation size along with other crashkernel
>>>    parameters
>>> - Call reserve_crashkernel_cma() to allocate the CMA region for kdump
>>> - Include the CMA-reserved ranges in the usable memory ranges for the
>>>    kdump kernel to use.
>>> - Exclude the CMA-reserved ranges from the crash kernel memory to
>>>    prevent them from being exported through /proc/vmcore.
>>>
>>> With the introduction of the CMA crashkernel regions,
>>> crash_exclude_mem_range() needs to be called multiple times to exclude
>>> both crashk_res and crashk_cma_ranges from the crash memory ranges. To
>>> avoid repetitive logic for validating mem_ranges size and handling
>>> reallocation when required, this functionality is moved to a new wrapper
>>> function crash_exclude_mem_range_guarded().
>>>
>>> To ensure proper CMA reservation, reserve_crashkernel_cma() is called
>>> after pageblock_order is initialized.
>>>
>>> Cc: Baoquan he <bhe@redhat.com>
>>> Cc: Jiri Bohac <jbohac@suse.cz>
>>> Cc: Hari Bathini <hbathini@linux.ibm.com>
>>> Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
>>> Cc: Mahesh Salgaonkar <mahesh@linux.ibm.com>
>>> Cc: Michael Ellerman <mpe@ellerman.id.au>
>>> Cc: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
>>> Cc: Shivang Upadhyay <shivangu@linux.ibm.com>
>>> Cc: linuxppc-dev@lists.ozlabs.org
>>> Signed-off-by: Sourabh Jain <sourabhjain@linux.ibm.com>
>>> ---
>>> Changlog:
>>>
>>> v3 -> v4
>>>   - Removed repeated initialization to tmem in
>>>     crash_exclude_mem_range_guarded()
>>>   - Call crash_exclude_mem_range() with right crashk ranges
>>>
>>> ---
>>>   arch/powerpc/include/asm/kexec.h   |  2 ++
>>>   arch/powerpc/kernel/setup-common.c |  4 ++-
>>>   arch/powerpc/kexec/core.c          | 10 ++++++-
>>>   arch/powerpc/kexec/ranges.c        | 43 ++++++++++++++++++++++--------
>>>   4 files changed, 46 insertions(+), 13 deletions(-)
>>
>> I am not much familiar with the crash kernel workings but was curious
>> about the following query related to this patch:
>>
>> As I understand this patch allows for the remaining crash kernel
>> memory to come from CMA region. But do we limit the CMA region to be lower
>> than 4G?
>
> No we are not and we don't need to.
>
>> Is this patch dependent over your other patch series [1] which
>> supports high crashkernel reservation?
>>
>> [1]: https://lore.kernel.org/linuxppc-dev/20251027151338.819957-1-sourabhjain@linux.ibm.com/
>
> No, this is an independent patch.
>

Say, if we are in Hash mode and if the CMA reservations have come from
higher addresses. Will that work with kdump kernel when it boots with Hash
mmu? Because memory region beyond RMA is not accessible in Hash correct?

-ritesh

>>
>>
>> A minor comment: I guess we could reflect the support for ,cma for PPC
>> too in Documentation/admin-guide/kernel-parameters.txt like how it is
>> done for x86.
>
> Yes we should. I will make the change in the next version.
>
> Thanks for the review.
>
> - Sourabh Jain

Re: [PATCH v4] powerpc/kdump: Add support for crashkernel CMA reservation

[Ritesh Harjani]

Ritesh Harjani (IBM) <ritesh.list@gmail.com> writes:

>>>
>>> I am not much familiar with the crash kernel workings but was curious
>>> about the following query related to this patch:
>>>
>>> As I understand this patch allows for the remaining crash kernel
>>> memory to come from CMA region. But do we limit the CMA region to be lower
>>> than 4G?
>>
>> No we are not and we don't need to.
>>
>>> Is this patch dependent over your other patch series [1] which
>>> supports high crashkernel reservation?
>>>
>>> [1]: https://lore.kernel.org/linuxppc-dev/20251027151338.819957-1-sourabhjain@linux.ibm.com/
>>
>> No, this is an independent patch.
>>
>
> Say, if we are in Hash mode and if the CMA reservations have come from
> higher addresses. Will that work with kdump kernel when it boots with Hash
> mmu? Because memory region beyond RMA is not accessible in Hash correct?
>

Oh sorry my bad! I think I got the answer to above question now. 
So this feature allows us to reserve the "extra memory" using CMA which
is mainly used to serve the kdump kernel's memory allocation requests.
So we will have two memory reservations i.e.

crashkernel=64M,crashkernel=1G,cma. 

So the second 1G cma reservation is mainly to serve the kdump kernel's
memory allocation requests to avoid the ooms. And this will only be
required once the MMU is initialized, so we don't have those RMA
restrictions which are only during early init time (before Hash is
initialized).

-ritesh

Re: [PATCH v4] powerpc/kdump: Add support for crashkernel CMA reservation

[Sourabh Jain]

On 31/10/25 16:48, Ritesh Harjani (IBM) wrote:
> Ritesh Harjani (IBM) <ritesh.list@gmail.com> writes:
>
>>>> I am not much familiar with the crash kernel workings but was curious
>>>> about the following query related to this patch:
>>>>
>>>> As I understand this patch allows for the remaining crash kernel
>>>> memory to come from CMA region. But do we limit the CMA region to be lower
>>>> than 4G?
>>> No we are not and we don't need to.
>>>
>>>> Is this patch dependent over your other patch series [1] which
>>>> supports high crashkernel reservation?
>>>>
>>>> [1]: https://lore.kernel.org/linuxppc-dev/20251027151338.819957-1-sourabhjain@linux.ibm.com/
>>> No, this is an independent patch.
>>>
>> Say, if we are in Hash mode and if the CMA reservations have come from
>> higher addresses. Will that work with kdump kernel when it boots with Hash
>> mmu? Because memory region beyond RMA is not accessible in Hash correct?
>>
> Oh sorry my bad! I think I got the answer to above question now.
> So this feature allows us to reserve the "extra memory" using CMA which
> is mainly used to serve the kdump kernel's memory allocation requests.
> So we will have two memory reservations i.e.
>
> crashkernel=64M,crashkernel=1G,cma.
>
> So the second 1G cma reservation is mainly to serve the kdump kernel's
> memory allocation requests to avoid the ooms. And this will only be
> required once the MMU is initialized, so we don't have those RMA
> restrictions which are only during early init time (before Hash is
> initialized).
Yes, exactly. The regular reservation will be done as usual, and the kdump
image (vmlinux, initrd, dtb, etc.) will be loaded in the regular crashkernel
reserved region. So, no impact is expected due to the MMU type in use.

- Sourabh Jain