[PATCH v6 0/8] crash: Kernel handling of CPU and memory hot un/plug

[PATCH v6 0/8] crash: Kernel handling of CPU and memory hot un/plug

[Eric DeVolder]

When the kdump service is loaded, if a CPU or memory is hot
un/plugged, the crash elfcorehdr (for x86), which describes the CPUs
and memory in the system, must also be updated, else the resulting
vmcore is inaccurate (eg. missing either CPU context or memory
regions).

The current solution utilizes udev to initiate an unload-then-reload
of the kdump image (e. kernel, initrd, boot_params, puratory and
elfcorehdr) by the userspace kexec utility. In previous posts I have
outlined the significant performance problems related to offloading
this activity to userspace.

This patchset introduces a generic crash hot un/plug handler that
registers with the CPU and memory notifiers. Upon CPU or memory
changes, this generic handler is invoked and performs important
housekeeping, for example obtaining the appropriate lock, and then
invokes an architecture specific handler to do the appropriate
updates.

In the case of x86_64, the arch specific handler generates a new
elfcorehdr, and overwrites the old one in memory. No involvement
with userspace needed.

To realize the benefits/test this patchset, one must make a couple
of minor changes to userspace:

 - Disable the udev rule for updating kdump on hot un/plug changes.
   Add the following as the first two lines to the udev rule file
   /usr/lib/udev/rules.d/98-kexec.rules:

# For x86_64, the kernel handles updates to crash elfcorehdr
CONST{arch}=="x86-64", GOTO="kdump_reload_end"

   These two lines will cause cpu and memory hot un/plug events
   to be skipped within this rule file.

 - Change to the kexec_file_load for loading the kdump kernel:
   Eg. on RHEL: in /usr/bin/kdumpctl, change to:
    standard_kexec_args="-p -d -s"
   which adds the -s to select kexec_file_load syscall.

This patchset supports kexec_load with a modified kexec userspace
utility, and a working changeset to the kexec userspace utility
is provided here (and to use, the above change to standard_kexec_args
would be, for example, to append --hotplug-size=131072 instead of -s).

 diff --git a/kexec/arch/i386/crashdump-x86.c b/kexec/arch/i386/crashdump-x86.c
 index 9826f6d..06adb7e 100644
 --- a/kexec/arch/i386/crashdump-x86.c
 +++ b/kexec/arch/i386/crashdump-x86.c
 @@ -48,6 +48,7 @@
  #include <x86/x86-linux.h>
  
  extern struct arch_options_t arch_options;
 +extern unsigned long long hotplug_size;
  
  static int get_kernel_page_offset(struct kexec_info *UNUSED(info),
  				  struct crash_elf_info *elf_info)
 @@ -975,6 +976,13 @@ int load_crashdump_segments(struct kexec_info *info, char* mod_cmdline,
  	} else {
  		memsz = bufsz;
  	}
 +
 +    /* If hotplug support enabled, use that size */
 +    if (hotplug_size) {
 +        memsz = hotplug_size;
 +    }
 +
 +    info->elfcorehdr =
  	elfcorehdr = add_buffer(info, tmp, bufsz, memsz, align, min_base,
  							max_addr, -1);
  	dbgprintf("Created elf header segment at 0x%lx\n", elfcorehdr);
 diff --git a/kexec/kexec.c b/kexec/kexec.c
 index f63b36b..9569d9a 100644
 --- a/kexec/kexec.c
 +++ b/kexec/kexec.c
 @@ -58,6 +58,7 @@
  
  unsigned long long mem_min = 0;
  unsigned long long mem_max = ULONG_MAX;
 +unsigned long long hotplug_size = 0;
  static unsigned long kexec_flags = 0;
  /* Flags for kexec file (fd) based syscall */
  static unsigned long kexec_file_flags = 0;
 @@ -672,6 +673,12 @@ static void update_purgatory(struct kexec_info *info)
  		if (info->segment[i].mem == (void *)info->rhdr.rel_addr) {
  			continue;
  		}
 +        /* Don't include elfcorehdr in the checksum, if hotplug
 +         * support enabled.
 +         */
 +        if (hotplug_size && (info->segment[i].mem == (void *)info->elfcorehdr)) {
 +			continue;
 +		}
  		sha256_update(&ctx, info->segment[i].buf,
  			      info->segment[i].bufsz);
  		nullsz = info->segment[i].memsz - info->segment[i].bufsz;
 @@ -1504,6 +1511,17 @@ int main(int argc, char *argv[])
  		case OPT_PRINT_CKR_SIZE:
  			print_crashkernel_region_size();
  			return 0;
 +		case OPT_HOTPLUG_SIZE:
 +            /* Reserved the specified size for hotplug growth */
 +			hotplug_size = strtoul(optarg, &endptr, 0);
 +			if (*endptr) {
 +				fprintf(stderr,
 +					"Bad option value in --hotplug-size=%s\n",
 +					optarg);
 +				usage();
 +				return 1;
 +			}
 +			break;
  		default:
  			break;
  		}
 diff --git a/kexec/kexec.h b/kexec/kexec.h
 index 595dd68..b30dda4 100644
 --- a/kexec/kexec.h
 +++ b/kexec/kexec.h
 @@ -169,6 +169,7 @@ struct kexec_info {
  	int command_line_len;
  
  	int skip_checks;
 +    unsigned long elfcorehdr;
   };
  
  struct arch_map_entry {
 @@ -231,7 +232,8 @@ extern int file_types;
  #define OPT_PRINT_CKR_SIZE	262
  #define OPT_LOAD_LIVE_UPDATE	263
  #define OPT_EXEC_LIVE_UPDATE	264
 -#define OPT_MAX			265
 +#define OPT_HOTPLUG_SIZE	265
 +#define OPT_MAX			266
  #define KEXEC_OPTIONS \
  	{ "help",		0, 0, OPT_HELP }, \
  	{ "version",		0, 0, OPT_VERSION }, \
 @@ -258,6 +260,7 @@ extern int file_types;
  	{ "debug",		0, 0, OPT_DEBUG }, \
  	{ "status",		0, 0, OPT_STATUS }, \
  	{ "print-ckr-size",     0, 0, OPT_PRINT_CKR_SIZE }, \
 +	{ "hotplug-size",     2, 0, OPT_HOTPLUG_SIZE }, \
  
  #define KEXEC_OPT_STR "h?vdfixyluet:pscaS"
 

Regards,
eric
---
v6: 1apr202
 - Reword commit messages and some comment cleanup per Baoquan.
 - Changed elf_index to elfcorehdr_index for clarity.
 - Minor code changes per Baoquan.

v5: 3mar2022
 https://lkml.org/lkml/2022/3/3/674
 - Reworded description of CRASH_HOTPLUG_ELFCOREHDR_SZ, per
   David Hildenbrand.
 - Refactored slightly a few patches per Baoquan recommendation.

v4: 9feb2022
 https://lkml.org/lkml/2022/2/9/1406
 - Refactored patches per Baoquan suggestsions.
 - A few corrections, per Baoquan.

v3: 10jan2022
 https://lkml.org/lkml/2022/1/10/1212
 - Rebasing per Baoquan He request.
 - Changed memory notifier per David Hildenbrand.
 - Providing example kexec userspace change in cover letter.

RFC v2: 7dec2021
 https://lkml.org/lkml/2021/12/7/1088
 - Acting upon Baoquan He suggestion of removing elfcorehdr from
   the purgatory list of segments, removed purgatory code from
   patchset, and it is signficiantly simpler now.

RFC v1: 18nov2021
 https://lkml.org/lkml/2021/11/18/845
 - working patchset demonstrating kernel handling of hotplug
   updates to x86 elfcorehdr for kexec_file_load

RFC: 14dec2020
 https://lkml.org/lkml/2020/12/14/532
 - proposed concept of allowing kernel to handle hotplug update
   of elfcorehdr
---


Eric DeVolder (8):
  x86/crash: fix minor typo/bug in debug message
  x86/crash: Introduce new options to support cpu and memory hotplug
  crash: prototype change for crash_prepare_elf64_headers
  crash: add generic infrastructure for crash hotplug support
  kexec: exclude elfcorehdr from the segment digest
  kexec: exclude hot remove cpu from elfcorehdr notes
  x86/crash: Add x86 crash hotplug support for kexec_file_load
  x86/crash: Add x86 crash hotplug support for kexec_load

 arch/arm64/kernel/machine_kexec_file.c |   6 +-
 arch/powerpc/kexec/file_load_64.c      |   2 +-
 arch/x86/Kconfig                       |  26 +++++
 arch/x86/kernel/crash.c                | 147 ++++++++++++++++++++++++-
 include/linux/kexec.h                  |  21 +++-
 kernel/crash_core.c                    | 118 ++++++++++++++++++++
 kernel/kexec_file.c                    |  15 ++-
 7 files changed, 325 insertions(+), 10 deletions(-)

-- 
2.27.0

[PATCH v6 1/8] x86/crash: fix minor typo/bug in debug message

[Eric DeVolder]

The pr_debug() intends to display the memsz member, but the
parameter is actually the bufsz member (which is already
displayed). Correct this to display memsz value.

Signed-off-by: Eric DeVolder <eric.devolder@oracle.com>
Acked-by: Baoquan He <bhe@redhat.com>
---
 arch/x86/kernel/crash.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c
index e8326a8d1c5d..9730c88530fc 100644
--- a/arch/x86/kernel/crash.c
+++ b/arch/x86/kernel/crash.c
@@ -407,7 +407,7 @@ int crash_load_segments(struct kimage *image)
 	}
 	image->elf_load_addr = kbuf.mem;
 	pr_debug("Loaded ELF headers at 0x%lx bufsz=0x%lx memsz=0x%lx\n",
-		 image->elf_load_addr, kbuf.bufsz, kbuf.bufsz);
+		 image->elf_load_addr, kbuf.bufsz, kbuf.memsz);
 
 	return ret;
 }
-- 
2.27.0

[PATCH v6 2/8] x86/crash: Introduce new options to support cpu and memory hotplug

[Eric DeVolder]

CRASH_HOTPLUG is to enable cpu and memory hotplug support of crash.

CRASH_HOTPLUG_ELFCOREHDR_SZ is used to specify the maximum size of
the elfcorehdr buffer/segment.

This is a preparation for later usage.

Signed-off-by: Eric DeVolder <eric.devolder@oracle.com>
---
 arch/x86/Kconfig | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)

diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 7340d9f01b62..8b51d3196b82 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -2072,6 +2072,32 @@ config CRASH_DUMP
 	  (CONFIG_RELOCATABLE=y).
 	  For more details see Documentation/admin-guide/kdump/kdump.rst
 
+config CRASH_HOTPLUG
+	bool "kernel updates of crash elfcorehdr"
+	depends on CRASH_DUMP && (HOTPLUG_CPU || MEMORY_HOTPLUG) && KEXEC_FILE
+	help
+	  Enable the kernel to update the crash elfcorehdr (which contains
+	  the list of CPUs and memory regions) directly when hot plug/unplug
+	  of CPUs or memory. Otherwise userspace must monitor these hot
+	  plug/unplug change notifications via udev in order to
+	  unload-then-reload the crash kernel so that the list of CPUs and
+	  memory regions is kept up-to-date. Note that the udev CPU and
+	  memory change notifications still occur (however, userspace is not
+	  required to monitor for crash dump purposes).
+
+config CRASH_HOTPLUG_ELFCOREHDR_SZ
+	depends on CRASH_HOTPLUG
+	int
+	default 131072
+	help
+	  Specify the maximum size of the elfcorehdr buffer/segment.
+	  The 128KiB default is sized so that it can accommodate 2048
+	  Elf64_Phdr, where each Phdr represents either a CPU or a
+	  region of memory.
+	  For example, this size can accommodate a machine with up to 1024
+	  CPUs and up to 1024 memory regions, eg. as represented by the
+	  'System RAM' entries in /proc/iomem.
+
 config KEXEC_JUMP
 	bool "kexec jump"
 	depends on KEXEC && HIBERNATION
-- 
2.27.0

[PATCH v6 3/8] crash: prototype change for crash_prepare_elf64_headers

[Eric DeVolder]

From within crash_prepare_elf64_headers() there is a need to
reference the struct kimage hotplug members. As such, this
change passes the struct kimage as a parameter to the
crash_prepare_elf64_headers().

This is preparation for later patch, no functionality change.

Signed-off-by: Eric DeVolder <eric.devolder@oracle.com>
Acked-by: Baoquan He <bhe@redhat.com>
---
 arch/arm64/kernel/machine_kexec_file.c | 6 +++---
 arch/powerpc/kexec/file_load_64.c      | 2 +-
 arch/x86/kernel/crash.c                | 3 ++-
 include/linux/kexec.h                  | 5 +++--
 kernel/kexec_file.c                    | 4 ++--
 5 files changed, 11 insertions(+), 9 deletions(-)

diff --git a/arch/arm64/kernel/machine_kexec_file.c b/arch/arm64/kernel/machine_kexec_file.c
index 59c648d51848..7dbafb42ecf2 100644
--- a/arch/arm64/kernel/machine_kexec_file.c
+++ b/arch/arm64/kernel/machine_kexec_file.c
@@ -39,7 +39,7 @@ int arch_kimage_file_post_load_cleanup(struct kimage *image)
 	return kexec_image_post_load_cleanup_default(image);
 }
 
-static int prepare_elf_headers(void **addr, unsigned long *sz)
+static int prepare_elf_headers(struct kimage *image, void **addr, unsigned long *sz)
 {
 	struct crash_mem *cmem;
 	unsigned int nr_ranges;
@@ -67,7 +67,7 @@ static int prepare_elf_headers(void **addr, unsigned long *sz)
 	ret = crash_exclude_mem_range(cmem, crashk_res.start, crashk_res.end);
 
 	if (!ret)
-		ret =  crash_prepare_elf64_headers(cmem, true, addr, sz);
+		ret =  crash_prepare_elf64_headers(image, cmem, true, addr, sz);
 
 	kfree(cmem);
 	return ret;
@@ -96,7 +96,7 @@ int load_other_segments(struct kimage *image,
 
 	/* load elf core header */
 	if (image->type == KEXEC_TYPE_CRASH) {
-		ret = prepare_elf_headers(&headers, &headers_sz);
+		ret = prepare_elf_headers(image, &headers, &headers_sz);
 		if (ret) {
 			pr_err("Preparing elf core header failed\n");
 			goto out_err;
diff --git a/arch/powerpc/kexec/file_load_64.c b/arch/powerpc/kexec/file_load_64.c
index b4981b651d9a..07da6bf1cf24 100644
--- a/arch/powerpc/kexec/file_load_64.c
+++ b/arch/powerpc/kexec/file_load_64.c
@@ -797,7 +797,7 @@ static int load_elfcorehdr_segment(struct kimage *image, struct kexec_buf *kbuf)
 		goto out;
 
 	/* Setup elfcorehdr segment */
-	ret = crash_prepare_elf64_headers(cmem, false, &headers, &headers_sz);
+	ret = crash_prepare_elf64_headers(image, cmem, false, &headers, &headers_sz);
 	if (ret) {
 		pr_err("Failed to prepare elf headers for the core\n");
 		goto out;
diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c
index 9730c88530fc..9db41cce8d97 100644
--- a/arch/x86/kernel/crash.c
+++ b/arch/x86/kernel/crash.c
@@ -265,7 +265,8 @@ static int prepare_elf_headers(struct kimage *image, void **addr,
 		goto out;
 
 	/* By default prepare 64bit headers */
-	ret =  crash_prepare_elf64_headers(cmem, IS_ENABLED(CONFIG_X86_64), addr, sz);
+	ret =  crash_prepare_elf64_headers(image, cmem,
+				IS_ENABLED(CONFIG_X86_64), addr, sz);
 
 out:
 	vfree(cmem);
diff --git a/include/linux/kexec.h b/include/linux/kexec.h
index 58d1b58a971e..f93f2591fc1e 100644
--- a/include/linux/kexec.h
+++ b/include/linux/kexec.h
@@ -227,8 +227,9 @@ struct crash_mem {
 extern int crash_exclude_mem_range(struct crash_mem *mem,
 				   unsigned long long mstart,
 				   unsigned long long mend);
-extern int crash_prepare_elf64_headers(struct crash_mem *mem, int kernel_map,
-				       void **addr, unsigned long *sz);
+extern int crash_prepare_elf64_headers(struct kimage *image,
+	struct crash_mem *mem, int kernel_map,
+	void **addr, unsigned long *sz);
 #endif /* CONFIG_KEXEC_FILE */
 
 #ifdef CONFIG_KEXEC_ELF
diff --git a/kernel/kexec_file.c b/kernel/kexec_file.c
index 8347fc158d2b..801d0d0a5012 100644
--- a/kernel/kexec_file.c
+++ b/kernel/kexec_file.c
@@ -1260,8 +1260,8 @@ int crash_exclude_mem_range(struct crash_mem *mem,
 	return 0;
 }
 
-int crash_prepare_elf64_headers(struct crash_mem *mem, int kernel_map,
-			  void **addr, unsigned long *sz)
+int crash_prepare_elf64_headers(struct kimage *image, struct crash_mem *mem,
+	int kernel_map, void **addr, unsigned long *sz)
 {
 	Elf64_Ehdr *ehdr;
 	Elf64_Phdr *phdr;
-- 
2.27.0

[PATCH v6 4/8] crash: add generic infrastructure for crash hotplug support

[Eric DeVolder]

Upon CPU and memory changes, a generic crash_hotplug_handler()
dispatches the hot plug/unplug event to the architecture specific
arch_crash_hotplug_handler(). During the process, the kexec_mutex
is held.

To support cpu hotplug, a callback is registered to capture the
CPUHP_AP_ONLINE_DYN online and ofline events via
cpuhp_setup_state_nocalls().

To support memory hotplug, a notifier is registered to capture the
MEM_ONLINE and MEM_OFFLINE events via register_memory_notifier().

The cpu callback and memory notifier then call crash_hotplug_handler()
to handle the hot plug/unplug event.

Signed-off-by: Eric DeVolder <eric.devolder@oracle.com>
---
 include/linux/kexec.h |  16 +++++++
 kernel/crash_core.c   | 108 ++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 124 insertions(+)

diff --git a/include/linux/kexec.h b/include/linux/kexec.h
index f93f2591fc1e..40e426cfd795 100644
--- a/include/linux/kexec.h
+++ b/include/linux/kexec.h
@@ -306,6 +306,13 @@ struct kimage {
 
 	/* Information for loading purgatory */
 	struct purgatory_info purgatory_info;
+
+#ifdef CONFIG_CRASH_HOTPLUG
+	bool hotplug_event;
+	int offlinecpu;
+	bool elfcorehdr_index_valid;
+	int elfcorehdr_index;
+#endif
 #endif
 
 #ifdef CONFIG_IMA_KEXEC
@@ -322,6 +329,15 @@ struct kimage {
 	unsigned long elf_load_addr;
 };
 
+#ifdef CONFIG_CRASH_HOTPLUG
+void arch_crash_hotplug_handler(struct kimage *image,
+	unsigned int hp_action, unsigned long a, unsigned long b);
+#define KEXEC_CRASH_HP_REMOVE_CPU   0
+#define KEXEC_CRASH_HP_ADD_CPU      1
+#define KEXEC_CRASH_HP_REMOVE_MEMORY 2
+#define KEXEC_CRASH_HP_ADD_MEMORY   3
+#endif /* CONFIG_CRASH_HOTPLUG */
+
 /* kexec interface functions */
 extern void machine_kexec(struct kimage *image);
 extern int machine_kexec_prepare(struct kimage *image);
diff --git a/kernel/crash_core.c b/kernel/crash_core.c
index 256cf6db573c..76959d440f71 100644
--- a/kernel/crash_core.c
+++ b/kernel/crash_core.c
@@ -9,12 +9,17 @@
 #include <linux/init.h>
 #include <linux/utsname.h>
 #include <linux/vmalloc.h>
+#include <linux/highmem.h>
+#include <linux/memory.h>
+#include <linux/cpuhotplug.h>
 
 #include <asm/page.h>
 #include <asm/sections.h>
 
 #include <crypto/sha1.h>
 
+#include "kexec_internal.h"
+
 /* vmcoreinfo stuff */
 unsigned char *vmcoreinfo_data;
 size_t vmcoreinfo_size;
@@ -491,3 +496,106 @@ static int __init crash_save_vmcoreinfo_init(void)
 }
 
 subsys_initcall(crash_save_vmcoreinfo_init);
+
+#ifdef CONFIG_CRASH_HOTPLUG
+void __weak arch_crash_hotplug_handler(struct kimage *image,
+	unsigned int hp_action, unsigned long a, unsigned long b)
+{
+	pr_warn("crash hp: %s not implemented", __func__);
+}
+
+static void crash_hotplug_handler(unsigned int hp_action,
+	unsigned long a, unsigned long b)
+{
+	/* Obtain lock while changing crash information */
+	if (!mutex_trylock(&kexec_mutex))
+		return;
+
+	/* Check kdump is loaded */
+	if (kexec_crash_image) {
+		pr_debug("crash hp: hp_action %u, a %lu, b %lu", hp_action,
+			a, b);
+
+		/* Needed in order for the segments to be updated */
+		arch_kexec_unprotect_crashkres();
+
+		/* Flag to differentiate between normal load and hotplug */
+		kexec_crash_image->hotplug_event = true;
+
+		/* Now invoke arch-specific update handler */
+		arch_crash_hotplug_handler(kexec_crash_image, hp_action, a, b);
+
+		/* No longer handling a hotplug event */
+		kexec_crash_image->hotplug_event = false;
+
+		/* Change back to read-only */
+		arch_kexec_protect_crashkres();
+	}
+
+	/* Release lock now that update complete */
+	mutex_unlock(&kexec_mutex);
+}
+
+#if defined(CONFIG_MEMORY_HOTPLUG)
+static int crash_memhp_notifier(struct notifier_block *nb,
+	unsigned long val, void *v)
+{
+	struct memory_notify *mhp = v;
+	unsigned long start, end;
+
+	start = mhp->start_pfn << PAGE_SHIFT;
+	end = ((mhp->start_pfn + mhp->nr_pages) << PAGE_SHIFT) - 1;
+
+	switch (val) {
+	case MEM_ONLINE:
+		crash_hotplug_handler(KEXEC_CRASH_HP_ADD_MEMORY,
+			start, end-start);
+		break;
+
+	case MEM_OFFLINE:
+		crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_MEMORY,
+			start, end-start);
+		break;
+	}
+	return NOTIFY_OK;
+}
+
+static struct notifier_block crash_memhp_nb = {
+	.notifier_call = crash_memhp_notifier,
+	.priority = 0
+};
+#endif
+
+#if defined(CONFIG_HOTPLUG_CPU)
+static int crash_cpuhp_online(unsigned int cpu)
+{
+	crash_hotplug_handler(KEXEC_CRASH_HP_ADD_CPU, cpu, 0);
+	return 0;
+}
+
+static int crash_cpuhp_offline(unsigned int cpu)
+{
+	crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_CPU, cpu, 0);
+	return 0;
+}
+#endif
+
+static int __init crash_hotplug_init(void)
+{
+	int result = 0;
+
+#if defined(CONFIG_MEMORY_HOTPLUG)
+	register_memory_notifier(&crash_memhp_nb);
+#endif
+
+#if defined(CONFIG_HOTPLUG_CPU)
+	result = cpuhp_setup_state_nocalls(CPUHP_AP_ONLINE_DYN,
+				"crash/cpuhp",
+				crash_cpuhp_online, crash_cpuhp_offline);
+#endif
+
+	return result;
+}
+
+subsys_initcall(crash_hotplug_init);
+#endif /* CONFIG_CRASH_HOTPLUG */
-- 
2.27.0

[PATCH v6 5/8] kexec: exclude elfcorehdr from the segment digest

[Eric DeVolder]

When a crash kernel is loaded via the kexec_file_load syscall, the
kernel places the various segments (ie crash kernel, crash initrd,
boot_params, elfcorehdr, purgatory, etc) in memory. For those
architectures that utilize purgatory, a hash digest of the segments
is calculated for integrity checking. This digest is embedded into
the purgatory image prior to placing purgatory in memory.

Since hotplug events cause changes to the elfcorehdr, purgatory
integrity checking fails (at crash time, and no kdump created).
As a result, this change explicitly excludes the elfcorehdr segment
from the list of segments used to create the digest. By doing so,
this permits changes to the elfcorehdr in response to hotplug events,
without having to also reload purgatory due to the change to the
digest.

Signed-off-by: Eric DeVolder <eric.devolder@oracle.com>
Acked-by: Baoquan He <bhe@redhat.com>
---
 kernel/kexec_file.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/kernel/kexec_file.c b/kernel/kexec_file.c
index 801d0d0a5012..94a459209111 100644
--- a/kernel/kexec_file.c
+++ b/kernel/kexec_file.c
@@ -765,6 +765,12 @@ static int kexec_calculate_store_digests(struct kimage *image)
 	for (j = i = 0; i < image->nr_segments; i++) {
 		struct kexec_segment *ksegment;
 
+#ifdef CONFIG_CRASH_HOTPLUG
+		/* This segment excluded to allow future changes via hotplug */
+		if (image->elfcorehdr_index_valid && (j == image->elfcorehdr_index))
+			continue;
+#endif
+
 		ksegment = &image->segment[i];
 		/*
 		 * Skip purgatory as it will be modified once we put digest
-- 
2.27.0

[PATCH v6 6/8] kexec: exclude hot remove cpu from elfcorehdr notes

[Eric DeVolder]

Due to use of CPUHP_AP_ONLINE_DYN, upon CPU unplug, the CPU is
still in the for_each_present_cpu() list when within the
crash_hotplug_handler(). Thus the CPU must be explicitly excluded
when building the new list of CPUs.

This change identifies in crash_hotplug_handler() the CPU to be
excluded, and the check for excluding the CPU in
crash_prepare_elf64_headers().

Signed-off-by: Eric DeVolder <eric.devolder@oracle.com>
Acked-by: Baoquan He <bhe@redhat.com>
---
 kernel/crash_core.c | 10 ++++++++++
 kernel/kexec_file.c |  5 +++++
 2 files changed, 15 insertions(+)

diff --git a/kernel/crash_core.c b/kernel/crash_core.c
index 76959d440f71..d3fd51f2195c 100644
--- a/kernel/crash_core.c
+++ b/kernel/crash_core.c
@@ -522,6 +522,16 @@ static void crash_hotplug_handler(unsigned int hp_action,
 		/* Flag to differentiate between normal load and hotplug */
 		kexec_crash_image->hotplug_event = true;
 
+		/*
+		 * Due to use of CPUHP_AP_ONLINE_DYN, upon unplug and during
+		 * this callback, the CPU is still in the for_each_present_cpu()
+		 * list. Must explicitly look to exclude this CPU when building
+		 * new list.
+		 */
+		kexec_crash_image->offlinecpu =
+			(hp_action == KEXEC_CRASH_HP_REMOVE_CPU) ?
+				(unsigned int)a : ~0U;
+
 		/* Now invoke arch-specific update handler */
 		arch_crash_hotplug_handler(kexec_crash_image, hp_action, a, b);
 
diff --git a/kernel/kexec_file.c b/kernel/kexec_file.c
index 94a459209111..9d5c4eea0179 100644
--- a/kernel/kexec_file.c
+++ b/kernel/kexec_file.c
@@ -1314,6 +1314,11 @@ int crash_prepare_elf64_headers(struct kimage *image, struct crash_mem *mem,
 
 	/* Prepare one phdr of type PT_NOTE for each present CPU */
 	for_each_present_cpu(cpu) {
+#ifdef CONFIG_CRASH_HOTPLUG
+		/* Skip the soon-to-be offlined cpu */
+		if (image->hotplug_event && (cpu == image->offlinecpu))
+			continue;
+#endif
 		phdr->p_type = PT_NOTE;
 		notes_addr = per_cpu_ptr_to_phys(per_cpu_ptr(crash_notes, cpu));
 		phdr->p_offset = phdr->p_paddr = notes_addr;
-- 
2.27.0

[PATCH v6 7/8] x86/crash: Add x86 crash hotplug support for kexec_file_load

[Eric DeVolder]

For x86_64, when CPU or memory is hot un/plugged, the crash
elfcorehdr, which describes the CPUs and memory in the system,
must also be updated.

To update the elfcorehdr for x86_64, a new elfcorehdr must be
generated from the available CPUs and memory. The new elfcorehdr
is prepared into a buffer, and then installed over the top of
the existing elfcorehdr.

In the patch 'kexec: exclude elfcorehdr from the segment digest'
the need to update purgatory due to the change in elfcorehdr was
eliminated.  As a result, no changes to purgatory or boot_params
(as the elfcorehdr= kernel command line parameter pointer
remains unchanged and correct) are needed, just elfcorehdr.

To accommodate a growing number of resources via hotplug, the
elfcorehdr segment must be sufficiently large enough to accommodate
changes, see the CRASH_HOTPLUG_ELFCOREHDR_SZ configure item.

With this change, crash hotplug for kexec_file_load syscall
is supported. When loading the crash kernel via kexec_file_load,
the elfcorehdr is identified at load time in crash_load_segments().

Signed-off-by: Eric DeVolder <eric.devolder@oracle.com>
---
 arch/x86/kernel/crash.c | 118 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 118 insertions(+)

diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c
index 9db41cce8d97..101c030ac038 100644
--- a/arch/x86/kernel/crash.c
+++ b/arch/x86/kernel/crash.c
@@ -25,6 +25,7 @@
 #include <linux/slab.h>
 #include <linux/vmalloc.h>
 #include <linux/memblock.h>
+#include <linux/highmem.h>
 
 #include <asm/processor.h>
 #include <asm/hardirq.h>
@@ -398,7 +399,17 @@ int crash_load_segments(struct kimage *image)
 	image->elf_headers = kbuf.buffer;
 	image->elf_headers_sz = kbuf.bufsz;
 
+#ifdef CONFIG_CRASH_HOTPLUG
+	/* Ensure elfcorehdr segment large enough for hotplug changes */
+	kbuf.memsz = CONFIG_CRASH_HOTPLUG_ELFCOREHDR_SZ;
+	/* For marking as usable to crash kernel */
+	image->elf_headers_sz = kbuf.memsz;
+	/* Record the index of the elfcorehdr segment */
+	image->elfcorehdr_index = image->nr_segments;
+	image->elfcorehdr_index_valid = true;
+#else
 	kbuf.memsz = kbuf.bufsz;
+#endif
 	kbuf.buf_align = ELF_CORE_HEADER_ALIGN;
 	kbuf.mem = KEXEC_BUF_MEM_UNKNOWN;
 	ret = kexec_add_buffer(&kbuf);
@@ -413,3 +424,110 @@ int crash_load_segments(struct kimage *image)
 	return ret;
 }
 #endif /* CONFIG_KEXEC_FILE */
+
+#ifdef CONFIG_CRASH_HOTPLUG
+static void *map_crash_pages(unsigned long paddr, unsigned long size)
+{
+	/*
+	 * NOTE: The addresses and sizes passed to this routine have
+	 * already been fully aligned on page boundaries. There is no
+	 * need for massaging the address or size.
+	 */
+	void *ptr = NULL;
+
+	/* NOTE: requires arch_kexec_[un]protect_crashkres() for write access */
+	if (size > 0) {
+		struct page *page = pfn_to_page(paddr >> PAGE_SHIFT);
+
+		ptr = kmap(page);
+	}
+
+	return ptr;
+}
+
+static void unmap_crash_pages(void **ptr)
+{
+	if (ptr) {
+		if (*ptr)
+			kunmap(*ptr);
+		*ptr = NULL;
+	}
+}
+
+/**
+ * arch_crash_hotplug_handler() - Handle hotplug elfcorehdr changes
+ * @image: the active struct kimage
+ * @hp_action: the hot un/plug action being handled
+ * @a: first parameter dependent upon hp_action
+ * @b: first parameter dependent upon hp_action
+ *
+ * To accurately reflect hot un/plug changes, the elfcorehdr (which
+ * is passed to the crash kernel via the elfcorehdr= parameter)
+ * must be updated with the new list of CPUs and memories. The new
+ * elfcorehdr is prepared in a kernel buffer, and then it is
+ * written on top of the existing/old elfcorehdr.
+ *
+ * For hotplug changes to elfcorehdr to work, two conditions are
+ * needed:
+ * First, the segment containing the elfcorehdr must be large enough
+ * to permit a growing number of resources. See
+ * CONFIG_CRASH_HOTPLUG_ELFCOREHDR_SZ.
+ * Second, purgatory must explicitly exclude the elfcorehdr from the
+ * list of segments it checks (since the elfcorehdr changes and thus
+ * would require an update to purgatory itself to update the digest).
+ *
+ */
+void arch_crash_hotplug_handler(struct kimage *image,
+	unsigned int hp_action, unsigned long a, unsigned long b)
+{
+	struct kexec_segment *ksegment;
+	unsigned char *ptr = NULL;
+	unsigned long elfsz = 0;
+	void *elfbuf = NULL;
+	unsigned long mem, memsz;
+
+	if (!image->elfcorehdr_index_valid) {
+		pr_err("crash hp: unable to locate elfcorehdr segment");
+		goto out;
+	}
+
+	ksegment = &image->segment[image->elfcorehdr_index];
+	mem = ksegment->mem;
+	memsz = ksegment->memsz;
+
+	/*
+	 * Create the new elfcorehdr reflecting the changes to CPU and/or
+	 * memory resources.
+	 */
+	if (prepare_elf_headers(image, &elfbuf, &elfsz)) {
+		pr_err("crash hp: unable to prepare elfcore headers");
+		goto out;
+	}
+	if (elfsz > memsz) {
+		pr_err("crash hp: update elfcorehdr elfsz %lu > memsz %lu",
+			elfsz, memsz);
+		goto out;
+	}
+
+	/*
+	 * At this point, we are all but assured of success.
+	 * Copy new elfcorehdr into destination.
+	 */
+	ptr = map_crash_pages(mem, memsz);
+	if (ptr) {
+		/*
+		 * Temporarily invalidate the crash image while the
+		 * elfcorehdr is updated.
+		 */
+		xchg(&kexec_crash_image, NULL);
+		memcpy_flushcache((void *)ptr, elfbuf, elfsz);
+		xchg(&kexec_crash_image, image);
+	}
+	unmap_crash_pages((void **)&ptr);
+	pr_debug("crash hp: re-loaded elfcorehdr at 0x%lx\n", mem);
+
+out:
+	if (elfbuf)
+		vfree(elfbuf);
+}
+#endif /* CONFIG_CRASH_HOTPLUG */
-- 
2.27.0

[PATCH v6 8/8] x86/crash: Add x86 crash hotplug support for kexec_load

[Eric DeVolder]

For kexec_file_load support, the loading of the crash kernel occurs
entirely within the kernel, and as such the elfcorehdr is readily
identified (so that it can be modified upon hotplug events).

This change enables support for kexec_load by identifying the
elfcorehdr segment in the arch_crash_hotplug_handler(), if it has
not already been identified.

Signed-off-by: Eric DeVolder <eric.devolder@oracle.com>
---
 arch/x86/kernel/crash.c | 24 ++++++++++++++++++++++++
 1 file changed, 24 insertions(+)

diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c
index 101c030ac038..682f130b83e3 100644
--- a/arch/x86/kernel/crash.c
+++ b/arch/x86/kernel/crash.c
@@ -486,6 +486,30 @@ void arch_crash_hotplug_handler(struct kimage *image,
 	void *elfbuf = NULL;
 	unsigned long mem, memsz;
 
+	/*
+	 * When the struct kimage is alloced, it is wiped to zero, so
+	 * the elfcorehdr_index_valid defaults to false. It is set on the
+	 * kexec_file_load path, or here for kexec_load, if not already
+	 * identified.
+	 */
+	if (!image->elfcorehdr_index_valid) {
+		unsigned int n;
+
+		for (n = 0; n < image->nr_segments; n++) {
+			mem = image->segment[n].mem;
+			memsz = image->segment[n].memsz;
+			ptr = map_crash_pages(mem, memsz);
+			if (ptr) {
+				/* The segment containing elfcorehdr */
+				if (memcmp(ptr, ELFMAG, SELFMAG) == 0) {
+					image->elfcorehdr_index = (int)n;
+					image->elfcorehdr_index_valid = true;
+				}
+			}
+			unmap_crash_pages((void **)&ptr);
+		}
+	}
+
 	if (!image->elfcorehdr_index_valid) {
 		pr_err("crash hp: unable to locate elfcorehdr segment");
 		goto out;
-- 
2.27.0

Re: [PATCH v6 2/8] x86/crash: Introduce new options to support cpu and memory hotplug

[Baoquan He]

On 04/01/22 at 02:30pm, Eric DeVolder wrote:
> CRASH_HOTPLUG is to enable cpu and memory hotplug support of crash.
> 
> CRASH_HOTPLUG_ELFCOREHDR_SZ is used to specify the maximum size of
> the elfcorehdr buffer/segment.
> 
> This is a preparation for later usage.
> 
> Signed-off-by: Eric DeVolder <eric.devolder@oracle.com>

LGTM,

Acked-by: Baoquan He <bhe@redhat.com>

> ---
>  arch/x86/Kconfig | 26 ++++++++++++++++++++++++++
>  1 file changed, 26 insertions(+)
> 
> diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
> index 7340d9f01b62..8b51d3196b82 100644
> --- a/arch/x86/Kconfig
> +++ b/arch/x86/Kconfig
> @@ -2072,6 +2072,32 @@ config CRASH_DUMP
>  	  (CONFIG_RELOCATABLE=y).
>  	  For more details see Documentation/admin-guide/kdump/kdump.rst
>  
> +config CRASH_HOTPLUG
> +	bool "kernel updates of crash elfcorehdr"
> +	depends on CRASH_DUMP && (HOTPLUG_CPU || MEMORY_HOTPLUG) && KEXEC_FILE
> +	help
> +	  Enable the kernel to update the crash elfcorehdr (which contains
> +	  the list of CPUs and memory regions) directly when hot plug/unplug
> +	  of CPUs or memory. Otherwise userspace must monitor these hot
> +	  plug/unplug change notifications via udev in order to
> +	  unload-then-reload the crash kernel so that the list of CPUs and
> +	  memory regions is kept up-to-date. Note that the udev CPU and
> +	  memory change notifications still occur (however, userspace is not
> +	  required to monitor for crash dump purposes).
> +
> +config CRASH_HOTPLUG_ELFCOREHDR_SZ
> +	depends on CRASH_HOTPLUG
> +	int
> +	default 131072
> +	help
> +	  Specify the maximum size of the elfcorehdr buffer/segment.
> +	  The 128KiB default is sized so that it can accommodate 2048
> +	  Elf64_Phdr, where each Phdr represents either a CPU or a
> +	  region of memory.
> +	  For example, this size can accommodate a machine with up to 1024
> +	  CPUs and up to 1024 memory regions, eg. as represented by the
> +	  'System RAM' entries in /proc/iomem.
> +
>  config KEXEC_JUMP
>  	bool "kexec jump"
>  	depends on KEXEC && HIBERNATION
> -- 
> 2.27.0
> 

Re: [PATCH v6 4/8] crash: add generic infrastructure for crash hotplug support

[Baoquan He]

Hi Eric,

On 04/01/22 at 02:30pm, Eric DeVolder wrote:
... ...

> +static void crash_hotplug_handler(unsigned int hp_action,
> +	unsigned long a, unsigned long b)

I am still struggling to consider if these unused parameters should be
kept or removed. Do you foresee or feel on which ARCH they could be used?

Considering our elfcorehdr updating method, once memory or cpu changed,
we will update elfcorehdr and cpu notes to reflect all existing memory
regions and cpu in the current system. We could end up with having them
but never being used. Then we may finally need to clean them up.

If you have investigated and foresee or feel they could be used on a
certain architecture, we can keep them for the time being.

> +{
> +	/* Obtain lock while changing crash information */
> +	if (!mutex_trylock(&kexec_mutex))
> +		return;
> +
> +	/* Check kdump is loaded */
> +	if (kexec_crash_image) {
> +		pr_debug("crash hp: hp_action %u, a %lu, b %lu", hp_action,
> +			a, b);
> +
> +		/* Needed in order for the segments to be updated */
> +		arch_kexec_unprotect_crashkres();
> +
> +		/* Flag to differentiate between normal load and hotplug */
> +		kexec_crash_image->hotplug_event = true;
> +
> +		/* Now invoke arch-specific update handler */
> +		arch_crash_hotplug_handler(kexec_crash_image, hp_action, a, b);
> +
> +		/* No longer handling a hotplug event */
> +		kexec_crash_image->hotplug_event = false;
> +
> +		/* Change back to read-only */
> +		arch_kexec_protect_crashkres();
> +	}
> +
> +	/* Release lock now that update complete */
> +	mutex_unlock(&kexec_mutex);
> +}
> +
> +#if defined(CONFIG_MEMORY_HOTPLUG)
> +static int crash_memhp_notifier(struct notifier_block *nb,
> +	unsigned long val, void *v)
> +{
> +	struct memory_notify *mhp = v;
> +	unsigned long start, end;
> +
> +	start = mhp->start_pfn << PAGE_SHIFT;
> +	end = ((mhp->start_pfn + mhp->nr_pages) << PAGE_SHIFT) - 1;
> +
> +	switch (val) {
> +	case MEM_ONLINE:
> +		crash_hotplug_handler(KEXEC_CRASH_HP_ADD_MEMORY,
> +			start, end-start);
> +		break;
> +
> +	case MEM_OFFLINE:
> +		crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_MEMORY,
> +			start, end-start);
> +		break;
> +	}
> +	return NOTIFY_OK;
> +}
> +
> +static struct notifier_block crash_memhp_nb = {
> +	.notifier_call = crash_memhp_notifier,
> +	.priority = 0
> +};
> +#endif
> +
> +#if defined(CONFIG_HOTPLUG_CPU)
> +static int crash_cpuhp_online(unsigned int cpu)
> +{
> +	crash_hotplug_handler(KEXEC_CRASH_HP_ADD_CPU, cpu, 0);
> +	return 0;
> +}
> +
> +static int crash_cpuhp_offline(unsigned int cpu)
> +{
> +	crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_CPU, cpu, 0);
> +	return 0;
> +}
> +#endif
> +
> +static int __init crash_hotplug_init(void)
> +{
> +	int result = 0;
> +
> +#if defined(CONFIG_MEMORY_HOTPLUG)
> +	register_memory_notifier(&crash_memhp_nb);
> +#endif
> +
> +#if defined(CONFIG_HOTPLUG_CPU)
> +	result = cpuhp_setup_state_nocalls(CPUHP_AP_ONLINE_DYN,
> +				"crash/cpuhp",
> +				crash_cpuhp_online, crash_cpuhp_offline);
> +#endif
> +
> +	return result;
> +}
> +
> +subsys_initcall(crash_hotplug_init);
> +#endif /* CONFIG_CRASH_HOTPLUG */
> -- 
> 2.27.0
> 

Re: [PATCH v6 4/8] crash: add generic infrastructure for crash hotplug support

[Eric DeVolder]

On 4/11/22 04:20, Baoquan He wrote:
> Hi Eric,
> 
> On 04/01/22 at 02:30pm, Eric DeVolder wrote:
> ... ...
> 
>> +static void crash_hotplug_handler(unsigned int hp_action,
>> +	unsigned long a, unsigned long b)
> 
> I am still struggling to consider if these unused parameters should be
> kept or removed. Do you foresee or feel on which ARCH they could be used?
> 
> Considering our elfcorehdr updating method, once memory or cpu changed,
> we will update elfcorehdr and cpu notes to reflect all existing memory
> regions and cpu in the current system. We could end up with having them
> but never being used. Then we may finally need to clean them up.
> 
> If you have investigated and foresee or feel they could be used on a
> certain architecture, we can keep them for the time being.

So 'hp_action' and 'a' are used within the existing patch series.
In crash_core.c, there is this bit of code:

+       kexec_crash_image->offlinecpu =
+           (hp_action == KEXEC_CRASH_HP_REMOVE_CPU) ?
+               (unsigned int)a : ~0U;

which is referencing both 'hp_action' and using 'a' from the cpu notifier handler.
I looked into removing 'a' and setting offlinecpu directly, but I thought
it better that offlinecpu be set within the safety of the kexec_mutex.
Also, Sourabh Jain's work with PowerPC utilizing this framework directly
references hp_action in the arch-specific handler.

The cpu and memory notifier handlers set hp_action accordingly. For cpu handler,
the 'a' is set with the impacted cpu. For memory handler, 'a' and 'b' form the
impacted memory range. I agree it looks like the memory range is currently
not useful.

Let me know what you think.
eric

> 
>> +{
>> +	/* Obtain lock while changing crash information */
>> +	if (!mutex_trylock(&kexec_mutex))
>> +		return;
>> +
>> +	/* Check kdump is loaded */
>> +	if (kexec_crash_image) {
>> +		pr_debug("crash hp: hp_action %u, a %lu, b %lu", hp_action,
>> +			a, b);
>> +
>> +		/* Needed in order for the segments to be updated */
>> +		arch_kexec_unprotect_crashkres();
>> +
>> +		/* Flag to differentiate between normal load and hotplug */
>> +		kexec_crash_image->hotplug_event = true;
>> +
>> +		/* Now invoke arch-specific update handler */
>> +		arch_crash_hotplug_handler(kexec_crash_image, hp_action, a, b);
>> +
>> +		/* No longer handling a hotplug event */
>> +		kexec_crash_image->hotplug_event = false;
>> +
>> +		/* Change back to read-only */
>> +		arch_kexec_protect_crashkres();
>> +	}
>> +
>> +	/* Release lock now that update complete */
>> +	mutex_unlock(&kexec_mutex);
>> +}
>> +
>> +#if defined(CONFIG_MEMORY_HOTPLUG)
>> +static int crash_memhp_notifier(struct notifier_block *nb,
>> +	unsigned long val, void *v)
>> +{
>> +	struct memory_notify *mhp = v;
>> +	unsigned long start, end;
>> +
>> +	start = mhp->start_pfn << PAGE_SHIFT;
>> +	end = ((mhp->start_pfn + mhp->nr_pages) << PAGE_SHIFT) - 1;
>> +
>> +	switch (val) {
>> +	case MEM_ONLINE:
>> +		crash_hotplug_handler(KEXEC_CRASH_HP_ADD_MEMORY,
>> +			start, end-start);
>> +		break;
>> +
>> +	case MEM_OFFLINE:
>> +		crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_MEMORY,
>> +			start, end-start);
>> +		break;
>> +	}
>> +	return NOTIFY_OK;
>> +}
>> +
>> +static struct notifier_block crash_memhp_nb = {
>> +	.notifier_call = crash_memhp_notifier,
>> +	.priority = 0
>> +};
>> +#endif
>> +
>> +#if defined(CONFIG_HOTPLUG_CPU)
>> +static int crash_cpuhp_online(unsigned int cpu)
>> +{
>> +	crash_hotplug_handler(KEXEC_CRASH_HP_ADD_CPU, cpu, 0);
>> +	return 0;
>> +}
>> +
>> +static int crash_cpuhp_offline(unsigned int cpu)
>> +{
>> +	crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_CPU, cpu, 0);
>> +	return 0;
>> +}
>> +#endif
>> +
>> +static int __init crash_hotplug_init(void)
>> +{
>> +	int result = 0;
>> +
>> +#if defined(CONFIG_MEMORY_HOTPLUG)
>> +	register_memory_notifier(&crash_memhp_nb);
>> +#endif
>> +
>> +#if defined(CONFIG_HOTPLUG_CPU)
>> +	result = cpuhp_setup_state_nocalls(CPUHP_AP_ONLINE_DYN,
>> +				"crash/cpuhp",
>> +				crash_cpuhp_online, crash_cpuhp_offline);
>> +#endif
>> +
>> +	return result;
>> +}
>> +
>> +subsys_initcall(crash_hotplug_init);
>> +#endif /* CONFIG_CRASH_HOTPLUG */
>> -- 
>> 2.27.0
>>
> 

Re: [PATCH v6 2/8] x86/crash: Introduce new options to support cpu and memory hotplug

[Eric DeVolder]

On 4/8/22 03:07, Baoquan He wrote:
> On 04/01/22 at 02:30pm, Eric DeVolder wrote:
>> CRASH_HOTPLUG is to enable cpu and memory hotplug support of crash.
>>
>> CRASH_HOTPLUG_ELFCOREHDR_SZ is used to specify the maximum size of
>> the elfcorehdr buffer/segment.
>>
>> This is a preparation for later usage.
>>
>> Signed-off-by: Eric DeVolder <eric.devolder@oracle.com>
> 
> LGTM,
> 
> Acked-by: Baoquan He <bhe@redhat.com>

Excellent, thanks!
eric

> 
>> ---
>>   arch/x86/Kconfig | 26 ++++++++++++++++++++++++++
>>   1 file changed, 26 insertions(+)
>>
>> diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
>> index 7340d9f01b62..8b51d3196b82 100644
>> --- a/arch/x86/Kconfig
>> +++ b/arch/x86/Kconfig
>> @@ -2072,6 +2072,32 @@ config CRASH_DUMP
>>   	  (CONFIG_RELOCATABLE=y).
>>   	  For more details see Documentation/admin-guide/kdump/kdump.rst
>>   
>> +config CRASH_HOTPLUG
>> +	bool "kernel updates of crash elfcorehdr"
>> +	depends on CRASH_DUMP && (HOTPLUG_CPU || MEMORY_HOTPLUG) && KEXEC_FILE
>> +	help
>> +	  Enable the kernel to update the crash elfcorehdr (which contains
>> +	  the list of CPUs and memory regions) directly when hot plug/unplug
>> +	  of CPUs or memory. Otherwise userspace must monitor these hot
>> +	  plug/unplug change notifications via udev in order to
>> +	  unload-then-reload the crash kernel so that the list of CPUs and
>> +	  memory regions is kept up-to-date. Note that the udev CPU and
>> +	  memory change notifications still occur (however, userspace is not
>> +	  required to monitor for crash dump purposes).
>> +
>> +config CRASH_HOTPLUG_ELFCOREHDR_SZ
>> +	depends on CRASH_HOTPLUG
>> +	int
>> +	default 131072
>> +	help
>> +	  Specify the maximum size of the elfcorehdr buffer/segment.
>> +	  The 128KiB default is sized so that it can accommodate 2048
>> +	  Elf64_Phdr, where each Phdr represents either a CPU or a
>> +	  region of memory.
>> +	  For example, this size can accommodate a machine with up to 1024
>> +	  CPUs and up to 1024 memory regions, eg. as represented by the
>> +	  'System RAM' entries in /proc/iomem.
>> +
>>   config KEXEC_JUMP
>>   	bool "kexec jump"
>>   	depends on KEXEC && HIBERNATION
>> -- 
>> 2.27.0
>>
> 

Re: [PATCH v6 4/8] crash: add generic infrastructure for crash hotplug support

[Baoquan He]

On 04/11/22 at 08:54am, Eric DeVolder wrote:
> 
> 
> On 4/11/22 04:20, Baoquan He wrote:
> > Hi Eric,
> > 
> > On 04/01/22 at 02:30pm, Eric DeVolder wrote:
> > ... ...
> > 
> > > +static void crash_hotplug_handler(unsigned int hp_action,
> > > +	unsigned long a, unsigned long b)
> > 
> > I am still struggling to consider if these unused parameters should be
> > kept or removed. Do you foresee or feel on which ARCH they could be used?
> > 
> > Considering our elfcorehdr updating method, once memory or cpu changed,
> > we will update elfcorehdr and cpu notes to reflect all existing memory
> > regions and cpu in the current system. We could end up with having them
> > but never being used. Then we may finally need to clean them up.
> > 
> > If you have investigated and foresee or feel they could be used on a
> > certain architecture, we can keep them for the time being.
> 
> So 'hp_action' and 'a' are used within the existing patch series.
> In crash_core.c, there is this bit of code:
> 
> +       kexec_crash_image->offlinecpu =
> +           (hp_action == KEXEC_CRASH_HP_REMOVE_CPU) ?
> +               (unsigned int)a : ~0U;
> 
> which is referencing both 'hp_action' and using 'a' from the cpu notifier handler.
> I looked into removing 'a' and setting offlinecpu directly, but I thought
> it better that offlinecpu be set within the safety of the kexec_mutex.
> Also, Sourabh Jain's work with PowerPC utilizing this framework directly
> references hp_action in the arch-specific handler.
> 
> The cpu and memory notifier handlers set hp_action accordingly. For cpu handler,
> the 'a' is set with the impacted cpu. For memory handler, 'a' and 'b' form the
> impacted memory range. I agree it looks like the memory range is currently
> not useful.

OK, memory handler doesn't need the action, memory regions. While cpu
handler needs it to exclude the hot plugged cpu.

We could have two ways to acheive this as below. How do you think about
them?

static void crash_hotplug_handler(unsigned int hp_action,
        unsigned long cpu)

static int crash_memhp_notifier(struct notifier_block *nb,
        unsigned long val, void *v)
{
......
        switch (val) {
        case MEM_ONLINE:
                crash_hotplug_handler(KEXEC_CRASH_HP_ADD_MEMORY,
                        -1UL);
                break;

        case MEM_OFFLINE:
                crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_MEMORY,
                        -1UL);
                break;
        }
        return NOTIFY_OK;
}

static int crash_cpuhp_online(unsigned int cpu)
{
        crash_hotplug_handler(KEXEC_CRASH_HP_ADD_CPU, cpu);
        return 0;
}

static int crash_cpuhp_offline(unsigned int cpu)
{
        crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_CPU, cpu);
        return 0;
}

OR,

static void crash_hotplug_handler(unsigned int hp_action,
        int* cpu)

static int crash_cpuhp_online(unsigned int cpu)
{
        crash_hotplug_handler(KEXEC_CRASH_HP_ADD_CPU, NULL);
        return 0;
}

static int crash_cpuhp_offline(unsigned int cpu)
{
	int dead_cpu = cpu;
        crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_CPU, &cpu);
        return 0;
}

Re: [PATCH v6 4/8] crash: add generic infrastructure for crash hotplug support

[Eric DeVolder]

On 4/12/22 21:41, Baoquan He wrote:
> On 04/11/22 at 08:54am, Eric DeVolder wrote:
>>
>>
>> On 4/11/22 04:20, Baoquan He wrote:
>>> Hi Eric,
>>>
>>> On 04/01/22 at 02:30pm, Eric DeVolder wrote:
>>> ... ...
>>>
>>>> +static void crash_hotplug_handler(unsigned int hp_action,
>>>> +	unsigned long a, unsigned long b)
>>>
>>> I am still struggling to consider if these unused parameters should be
>>> kept or removed. Do you foresee or feel on which ARCH they could be used?
>>>
>>> Considering our elfcorehdr updating method, once memory or cpu changed,
>>> we will update elfcorehdr and cpu notes to reflect all existing memory
>>> regions and cpu in the current system. We could end up with having them
>>> but never being used. Then we may finally need to clean them up.
>>>
>>> If you have investigated and foresee or feel they could be used on a
>>> certain architecture, we can keep them for the time being.
>>
>> So 'hp_action' and 'a' are used within the existing patch series.
>> In crash_core.c, there is this bit of code:
>>
>> +       kexec_crash_image->offlinecpu =
>> +           (hp_action == KEXEC_CRASH_HP_REMOVE_CPU) ?
>> +               (unsigned int)a : ~0U;
>>
>> which is referencing both 'hp_action' and using 'a' from the cpu notifier handler.
>> I looked into removing 'a' and setting offlinecpu directly, but I thought
>> it better that offlinecpu be set within the safety of the kexec_mutex.
>> Also, Sourabh Jain's work with PowerPC utilizing this framework directly
>> references hp_action in the arch-specific handler.
>>
>> The cpu and memory notifier handlers set hp_action accordingly. For cpu handler,
>> the 'a' is set with the impacted cpu. For memory handler, 'a' and 'b' form the
>> impacted memory range. I agree it looks like the memory range is currently
>> not useful.
> 
> OK, memory handler doesn't need the action, memory regions. While cpu
> handler needs it to exclude the hot plugged cpu.
> 
> We could have two ways to acheive this as below. How do you think about
> them?
> 
> static void crash_hotplug_handler(unsigned int hp_action,
>          unsigned long cpu)
> 
> static int crash_memhp_notifier(struct notifier_block *nb,
>          unsigned long val, void *v)
> {
> ......
>          switch (val) {
>          case MEM_ONLINE:
>                  crash_hotplug_handler(KEXEC_CRASH_HP_ADD_MEMORY,
>                          -1UL);
>                  break;
> 
>          case MEM_OFFLINE:
>                  crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_MEMORY,
>                          -1UL);
>                  break;
>          }
>          return NOTIFY_OK;
> }
> 
> static int crash_cpuhp_online(unsigned int cpu)
> {
>          crash_hotplug_handler(KEXEC_CRASH_HP_ADD_CPU, cpu);
>          return 0;
> }
> 
> static int crash_cpuhp_offline(unsigned int cpu)
> {
>          crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_CPU, cpu);
>          return 0;
> }

I'm OK with the above. Shall I post v7 or are you still looking at patches 7 and 8?
Thanks!
Eric

> 
> OR,
> 
> static void crash_hotplug_handler(unsigned int hp_action,
>          int* cpu)
> 
> static int crash_cpuhp_online(unsigned int cpu)
> {
>          crash_hotplug_handler(KEXEC_CRASH_HP_ADD_CPU, NULL);
>          return 0;
> }
> 
> static int crash_cpuhp_offline(unsigned int cpu)
> {
> 	int dead_cpu = cpu;
>          crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_CPU, &cpu);
>          return 0;
> }
> 

Re: [PATCH v6 8/8] x86/crash: Add x86 crash hotplug support for kexec_load

[Baoquan He]

On 04/01/22 at 02:30pm, Eric DeVolder wrote:
> For kexec_file_load support, the loading of the crash kernel occurs
> entirely within the kernel, and as such the elfcorehdr is readily
> identified (so that it can be modified upon hotplug events).
> 
> This change enables support for kexec_load by identifying the
> elfcorehdr segment in the arch_crash_hotplug_handler(), if it has
> not already been identified.
> 
> Signed-off-by: Eric DeVolder <eric.devolder@oracle.com>

LGTM,

Acked-by: Baoquan He <bhe@redhat.com>

> ---
>  arch/x86/kernel/crash.c | 24 ++++++++++++++++++++++++
>  1 file changed, 24 insertions(+)
> 
> diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c
> index 101c030ac038..682f130b83e3 100644
> --- a/arch/x86/kernel/crash.c
> +++ b/arch/x86/kernel/crash.c
> @@ -486,6 +486,30 @@ void arch_crash_hotplug_handler(struct kimage *image,
>  	void *elfbuf = NULL;
>  	unsigned long mem, memsz;
>  
> +	/*
> +	 * When the struct kimage is alloced, it is wiped to zero, so
> +	 * the elfcorehdr_index_valid defaults to false. It is set on the
> +	 * kexec_file_load path, or here for kexec_load, if not already
> +	 * identified.
> +	 */
> +	if (!image->elfcorehdr_index_valid) {
> +		unsigned int n;
> +
> +		for (n = 0; n < image->nr_segments; n++) {
> +			mem = image->segment[n].mem;
> +			memsz = image->segment[n].memsz;
> +			ptr = map_crash_pages(mem, memsz);
> +			if (ptr) {
> +				/* The segment containing elfcorehdr */
> +				if (memcmp(ptr, ELFMAG, SELFMAG) == 0) {
> +					image->elfcorehdr_index = (int)n;
> +					image->elfcorehdr_index_valid = true;
> +				}
> +			}
> +			unmap_crash_pages((void **)&ptr);
> +		}
> +	}
> +
>  	if (!image->elfcorehdr_index_valid) {
>  		pr_err("crash hp: unable to locate elfcorehdr segment");
>  		goto out;
> -- 
> 2.27.0
> 

Re: [PATCH v6 4/8] crash: add generic infrastructure for crash hotplug support

[Baoquan He]

On 04/13/22 at 07:37am, Eric DeVolder wrote:
> 
> 
> On 4/12/22 21:41, Baoquan He wrote:
> > On 04/11/22 at 08:54am, Eric DeVolder wrote:
> > > 
> > > 
> > > On 4/11/22 04:20, Baoquan He wrote:
> > > > Hi Eric,
> > > > 
> > > > On 04/01/22 at 02:30pm, Eric DeVolder wrote:
> > > > ... ...
> > > > 
> > > > > +static void crash_hotplug_handler(unsigned int hp_action,
> > > > > +	unsigned long a, unsigned long b)
> > > > 
> > > > I am still struggling to consider if these unused parameters should be
> > > > kept or removed. Do you foresee or feel on which ARCH they could be used?
> > > > 
> > > > Considering our elfcorehdr updating method, once memory or cpu changed,
> > > > we will update elfcorehdr and cpu notes to reflect all existing memory
> > > > regions and cpu in the current system. We could end up with having them
> > > > but never being used. Then we may finally need to clean them up.
> > > > 
> > > > If you have investigated and foresee or feel they could be used on a
> > > > certain architecture, we can keep them for the time being.
> > > 
> > > So 'hp_action' and 'a' are used within the existing patch series.
> > > In crash_core.c, there is this bit of code:
> > > 
> > > +       kexec_crash_image->offlinecpu =
> > > +           (hp_action == KEXEC_CRASH_HP_REMOVE_CPU) ?
> > > +               (unsigned int)a : ~0U;
> > > 
> > > which is referencing both 'hp_action' and using 'a' from the cpu notifier handler.
> > > I looked into removing 'a' and setting offlinecpu directly, but I thought
> > > it better that offlinecpu be set within the safety of the kexec_mutex.
> > > Also, Sourabh Jain's work with PowerPC utilizing this framework directly
> > > references hp_action in the arch-specific handler.
> > > 
> > > The cpu and memory notifier handlers set hp_action accordingly. For cpu handler,
> > > the 'a' is set with the impacted cpu. For memory handler, 'a' and 'b' form the
> > > impacted memory range. I agree it looks like the memory range is currently
> > > not useful.
> > 
> > OK, memory handler doesn't need the action, memory regions. While cpu
> > handler needs it to exclude the hot plugged cpu.
> > 
> > We could have two ways to acheive this as below. How do you think about
> > them?
> > 
> > static void crash_hotplug_handler(unsigned int hp_action,
> >          unsigned long cpu)
> > 
> > static int crash_memhp_notifier(struct notifier_block *nb,
> >          unsigned long val, void *v)
> > {
> > ......
> >          switch (val) {
> >          case MEM_ONLINE:
> >                  crash_hotplug_handler(KEXEC_CRASH_HP_ADD_MEMORY,
> >                          -1UL);
> >                  break;
> > 
> >          case MEM_OFFLINE:
> >                  crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_MEMORY,
> >                          -1UL);
> >                  break;
> >          }
> >          return NOTIFY_OK;
> > }
> > 
> > static int crash_cpuhp_online(unsigned int cpu)
> > {
> >          crash_hotplug_handler(KEXEC_CRASH_HP_ADD_CPU, cpu);
> >          return 0;
> > }
> > 
> > static int crash_cpuhp_offline(unsigned int cpu)
> > {
> >          crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_CPU, cpu);
> >          return 0;
> > }
> 
> I'm OK with the above. Shall I post v7 or are you still looking at patches 7 and 8?
> Thanks!

Just acked patch 8. Patch 7 need be updated too, so will check in v7.

> > 
> > OR,
> > 
> > static void crash_hotplug_handler(unsigned int hp_action,
> >          int* cpu)
> > 
> > static int crash_cpuhp_online(unsigned int cpu)
> > {
> >          crash_hotplug_handler(KEXEC_CRASH_HP_ADD_CPU, NULL);
> >          return 0;
> > }
> > 
> > static int crash_cpuhp_offline(unsigned int cpu)
> > {
> > 	int dead_cpu = cpu;
> >          crash_hotplug_handler(KEXEC_CRASH_HP_REMOVE_CPU, &cpu);
> >          return 0;
> > }
> > 
>