kexec.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: Anthony Yznaga <anthony.yznaga@oracle.com>
To: Coiby Xu <coxu@redhat.com>
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org,
	tglx@linutronix.de, mingo@redhat.com, bp@alien8.de,
	x86@kernel.org, hpa@zytor.com, dave.hansen@linux.intel.com,
	luto@kernel.org, peterz@infradead.org, rppt@kernel.org,
	akpm@linux-foundation.org, ebiederm@xmission.com,
	keescook@chromium.org, graf@amazon.com, jason.zeng@intel.com,
	lei.l.li@intel.com, steven.sistare@oracle.com,
	fam.zheng@bytedance.com, mgalaxy@akamai.com,
	kexec@lists.infradead.org
Subject: Re: [RFC v3 07/21] mm: PKRAM: introduce super block
Date: Mon, 5 Jun 2023 19:01:56 -0700	[thread overview]
Message-ID: <f287334c-4054-525e-52ea-a0bbf077b653@oracle.com> (raw)
In-Reply-To: <upgf3qmqxn7xzbnwwt6nxau4ugqkq7szdaheewzbok27xwch4e@6tdnhkqspok4>

Hi Coiby,

On 6/4/23 7:40 PM, Coiby Xu wrote:
> Hi Anthony,
>
> On Wed, Apr 26, 2023 at 05:08:43PM -0700, Anthony Yznaga wrote:
>> The PKRAM super block is the starting point for restoring preserved
>> memory. By providing the super block to the new kernel at boot time,
>> preserved memory can be reserved and made available to be restored.
>> To point the kernel to the location of the super block, one passes
>> its pfn via the 'pkram' boot param. 
>
> I'm curious to ask how will the 'pkram' boot param be passed. It seems I
> can't find the answer in this patch set.

The pfn of the super block read from /sys/kernel/pkram is passed to

the next kernel by adding the boot parameter, pkram=<super block pfn>.

The next kernel picks it up through the early_param("pkram", 
parse_pkram_sb_pfn)

in this patch below.


Anthony


>
>
>> For that purpose, the pkram super
>> block pfn is exported via /sys/kernel/pkram. If none is passed, any
>> preserved memory will not be kept, and a new super block will be
>> allocated.
>>
>> Originally-by: Vladimir Davydov <vdavydov.dev@gmail.com>
>> Signed-off-by: Anthony Yznaga <anthony.yznaga@oracle.com>
>> ---
>> mm/pkram.c | 102 
>> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++--
>> 1 file changed, 100 insertions(+), 2 deletions(-)
>>
>> diff --git a/mm/pkram.c b/mm/pkram.c
>> index da166cb6afb7..c66b2ae4d520 100644
>> --- a/mm/pkram.c
>> +++ b/mm/pkram.c
>> @@ -5,15 +5,18 @@
>> #include <linux/init.h>
>> #include <linux/io.h>
>> #include <linux/kernel.h>
>> +#include <linux/kobject.h>
>> #include <linux/list.h>
>> #include <linux/mm.h>
>> #include <linux/module.h>
>> #include <linux/mutex.h>
>> #include <linux/notifier.h>
>> +#include <linux/pfn.h>
>> #include <linux/pkram.h>
>> #include <linux/reboot.h>
>> #include <linux/sched.h>
>> #include <linux/string.h>
>> +#include <linux/sysfs.h>
>> #include <linux/types.h>
>>
>> #include "internal.h"
>> @@ -82,12 +85,38 @@ struct pkram_node {
>> #define PKRAM_ACCMODE_MASK    3
>>
>> /*
>> + * The PKRAM super block contains data needed to restore the 
>> preserved memory
>> + * structure on boot. The pointer to it (pfn) should be passed via 
>> the 'pkram'
>> + * boot param if one wants to restore preserved data saved by the 
>> previously
>> + * executing kernel. For that purpose the kernel exports the pfn via
>> + * /sys/kernel/pkram. If none is passed, preserved memory if any 
>> will not be
>> + * preserved and a new clean page will be allocated for the super 
>> block.
>> + *
>> + * The structure occupies a memory page.
>> + */
>> +struct pkram_super_block {
>> +    __u64    node_pfn;        /* first element of the node list */
>> +};
>> +
>> +static unsigned long pkram_sb_pfn __initdata;
>> +static struct pkram_super_block *pkram_sb;
>> +
>> +/*
>>  * For convenience sake PKRAM nodes are kept in an auxiliary 
>> doubly-linked list
>>  * connected through the lru field of the page struct.
>>  */
>> static LIST_HEAD(pkram_nodes);            /* linked through page::lru */
>> static DEFINE_MUTEX(pkram_mutex);        /* serializes open/close */
>>
>> +/*
>> + * The PKRAM super block pfn, see above.
>> + */
>> +static int __init parse_pkram_sb_pfn(char *arg)
>> +{
>> +    return kstrtoul(arg, 16, &pkram_sb_pfn);
>> +}
>> +early_param("pkram", parse_pkram_sb_pfn);
>> +
>> static inline struct page *pkram_alloc_page(gfp_t gfp_mask)
>> {
>>     return alloc_page(gfp_mask);
>> @@ -270,6 +299,7 @@ static void pkram_stream_init(struct pkram_stream 
>> *ps,
>>  * @gfp_mask specifies the memory allocation mask to be used when 
>> saving data.
>>  *
>>  * Error values:
>> + *    %ENODEV: PKRAM not available
>>  *    %ENAMETOOLONG: name len >= PKRAM_NAME_MAX
>>  *    %ENOMEM: insufficient memory available
>>  *    %EEXIST: node with specified name already exists
>> @@ -285,6 +315,9 @@ int pkram_prepare_save(struct pkram_stream *ps, 
>> const char *name, gfp_t gfp_mask
>>     struct pkram_node *node;
>>     int err = 0;
>>
>> +    if (!pkram_sb)
>> +        return -ENODEV;
>> +
>>     if (strlen(name) >= PKRAM_NAME_MAX)
>>         return -ENAMETOOLONG;
>>
>> @@ -404,6 +437,7 @@ void pkram_discard_save(struct pkram_stream *ps)
>>  * Returns 0 on success, -errno on failure.
>>  *
>>  * Error values:
>> + *    %ENODEV: PKRAM not available
>>  *    %ENOENT: node with specified name does not exist
>>  *    %EBUSY: save to required node has not finished yet
>>  *
>> @@ -414,6 +448,9 @@ int pkram_prepare_load(struct pkram_stream *ps, 
>> const char *name)
>>     struct pkram_node *node;
>>     int err = 0;
>>
>> +    if (!pkram_sb)
>> +        return -ENODEV;
>> +
>>     mutex_lock(&pkram_mutex);
>>     node = pkram_find_node(name);
>>     if (!node) {
>> @@ -825,6 +862,13 @@ static void __pkram_reboot(void)
>>         node->node_pfn = node_pfn;
>>         node_pfn = page_to_pfn(page);
>>     }
>> +
>> +    /*
>> +     * Zero out pkram_sb completely since it may have been passed from
>> +     * the previous boot.
>> +     */
>> +    memset(pkram_sb, 0, PAGE_SIZE);
>> +    pkram_sb->node_pfn = node_pfn;
>> }
>>
>> static int pkram_reboot(struct notifier_block *notifier,
>> @@ -832,7 +876,8 @@ static int pkram_reboot(struct notifier_block 
>> *notifier,
>> {
>>     if (val != SYS_RESTART)
>>         return NOTIFY_DONE;
>> -    __pkram_reboot();
>> +    if (pkram_sb)
>> +        __pkram_reboot();
>>     return NOTIFY_OK;
>> }
>>
>> @@ -840,9 +885,62 @@ static int pkram_reboot(struct notifier_block 
>> *notifier,
>>     .notifier_call = pkram_reboot,
>> };
>>
>> +static ssize_t show_pkram_sb_pfn(struct kobject *kobj,
>> +        struct kobj_attribute *attr, char *buf)
>> +{
>> +    unsigned long pfn = pkram_sb ? PFN_DOWN(__pa(pkram_sb)) : 0;
>> +
>> +    return sprintf(buf, "%lx\n", pfn);
>> +}
>> +
>> +static struct kobj_attribute pkram_sb_pfn_attr =
>> +    __ATTR(pkram, 0444, show_pkram_sb_pfn, NULL);
>> +
>> +static struct attribute *pkram_attrs[] = {
>> +    &pkram_sb_pfn_attr.attr,
>> +    NULL,
>> +};
>> +
>> +static struct attribute_group pkram_attr_group = {
>> +    .attrs = pkram_attrs,
>> +};
>> +
>> +/* returns non-zero on success */
>> +static int __init pkram_init_sb(void)
>> +{
>> +    unsigned long pfn;
>> +    struct pkram_node *node;
>> +
>> +    if (!pkram_sb) {
>> +        struct page *page;
>> +
>> +        page = pkram_alloc_page(GFP_KERNEL | __GFP_ZERO);
>> +        if (!page) {
>> +            pr_err("PKRAM: Failed to allocate super block\n");
>> +            return 0;
>> +        }
>> +        pkram_sb = page_address(page);
>> +    }
>> +
>> +    /*
>> +     * Build auxiliary doubly-linked list of nodes connected through
>> +     * page::lru for convenience sake.
>> +     */
>> +    pfn = pkram_sb->node_pfn;
>> +    while (pfn) {
>> +        node = pfn_to_kaddr(pfn);
>> +        pkram_insert_node(node);
>> +        pfn = node->node_pfn;
>> +    }
>> +    return 1;
>> +}
>> +
>> static int __init pkram_init(void)
>> {
>> -    register_reboot_notifier(&pkram_reboot_notifier);
>> +    if (pkram_init_sb()) {
>> +        register_reboot_notifier(&pkram_reboot_notifier);
>> +        sysfs_update_group(kernel_kobj, &pkram_attr_group);
>> +    }
>>     return 0;
>> }
>> module_init(pkram_init);
>> -- 
>> 1.9.4
>>
>>
>> _______________________________________________
>> kexec mailing list
>> kexec@lists.infradead.org
>> http://lists.infradead.org/mailman/listinfo/kexec
>>
>

_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

  reply	other threads:[~2023-06-06  2:03 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-04-27  0:08 [RFC v3 00/21] Preserved-over-Kexec RAM Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 01/21] mm: add PKRAM API stubs and Kconfig Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 02/21] mm: PKRAM: implement node load and save functions Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 03/21] mm: PKRAM: implement object " Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 04/21] mm: PKRAM: implement folio stream operations Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 05/21] mm: PKRAM: implement byte " Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 06/21] mm: PKRAM: link nodes by pfn before reboot Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 07/21] mm: PKRAM: introduce super block Anthony Yznaga
2023-06-05  2:40   ` Coiby Xu
2023-06-06  2:01     ` Anthony Yznaga [this message]
2023-06-06  2:55       ` Coiby Xu
2023-06-06  3:12         ` Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 08/21] PKRAM: track preserved pages in a physical mapping pagetable Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 09/21] PKRAM: pass a list of preserved ranges to the next kernel Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 10/21] PKRAM: prepare for adding preserved ranges to memblock reserved Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 11/21] mm: PKRAM: reserve preserved memory at boot Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 12/21] PKRAM: free the preserved ranges list Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 13/21] PKRAM: prevent inadvertent use of a stale superblock Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 14/21] PKRAM: provide a way to ban pages from use by PKRAM Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 15/21] kexec: PKRAM: prevent kexec clobbering preserved pages in some cases Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 16/21] PKRAM: provide a way to check if a memory range has preserved pages Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 17/21] kexec: PKRAM: avoid clobbering already " Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 18/21] mm: PKRAM: allow preserved memory to be freed from userspace Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 19/21] PKRAM: disable feature when running the kdump kernel Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 20/21] x86/KASLR: PKRAM: support physical kaslr Anthony Yznaga
2023-04-27  0:08 ` [RFC v3 21/21] x86/boot/compressed/64: use 1GB pages for mappings Anthony Yznaga
2023-04-27 18:40   ` H. Peter Anvin
2023-04-27 22:38     ` Anthony Yznaga
2023-05-26 13:57 ` [RFC v3 00/21] Preserved-over-Kexec RAM Gowans, James
2023-05-31 23:14   ` Anthony Yznaga
2023-06-01  2:15 ` Baoquan He
2023-06-01 23:58   ` Anthony Yznaga

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=f287334c-4054-525e-52ea-a0bbf077b653@oracle.com \
    --to=anthony.yznaga@oracle.com \
    --cc=akpm@linux-foundation.org \
    --cc=bp@alien8.de \
    --cc=coxu@redhat.com \
    --cc=dave.hansen@linux.intel.com \
    --cc=ebiederm@xmission.com \
    --cc=fam.zheng@bytedance.com \
    --cc=graf@amazon.com \
    --cc=hpa@zytor.com \
    --cc=jason.zeng@intel.com \
    --cc=keescook@chromium.org \
    --cc=kexec@lists.infradead.org \
    --cc=lei.l.li@intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=luto@kernel.org \
    --cc=mgalaxy@akamai.com \
    --cc=mingo@redhat.com \
    --cc=peterz@infradead.org \
    --cc=rppt@kernel.org \
    --cc=steven.sistare@oracle.com \
    --cc=tglx@linutronix.de \
    --cc=x86@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).