From: Roberto Ricci <io@r-ricci.it>
To: Baoquan He <bhe@redhat.com>
Cc: Dave Young <dyoung@redhat.com>,
ebiederm@xmission.com, rafael@kernel.org, pavel@ucw.cz,
ytcoode@gmail.com, kexec@lists.infradead.org,
linux-pm@vger.kernel.org, akpm@linux-foundation.org,
regressions@lists.linux.dev, linux-kernel@vger.kernel.org,
msizanoen@qtmlabs.xyz, rafael.j.wysocki@intel.com,
yu.c.chen@intel.com
Subject: Re: [REGRESSION] Kernel booted via kexec fails to resume from hibernation
Date: Sat, 5 Apr 2025 01:31:23 +0200 [thread overview]
Message-ID: <Z_Bry0JcB1jqPztR@desktop0a> (raw)
In-Reply-To: <Z-hYWc9LtBU1Yhtg@desktop0a>
On 2025-03-29 21:30 +0100, Roberto Ricci wrote:
> On 2025-03-29 09:44 +0800, Baoquan He wrote:
> > [snip]
> > 3) If answer to 1) and 2) is yes, does kexec_load works for you? Asking
> > this because kexec_load interface defaults to put kexec kernel on top of
> > system RAM which is equivalent to applying commit b3ba234171cd.
>
> No, it doesn't. While hibernation alone works, kexec + hibernation
> results in the system just rebooting without resuming the hibernation
> image, but no crash or other weird behaviour occurs.
> Initially I decided to focus on kexec_file_load in order to narrow
> things down, but that was before noticing that the bug could manifest
> itself in different forms.
> It is possible, indeed, that both syscalls are affected by the same
> problem, which is not caused by commit b3ba234171cd.
> I tried to test kexec_load with some older kernels, but I got build
> errors, so I tested longterm releases where such errors have been fixed.
> With v4.9.337, kexec (via kexec_load) + hibernation works.
> With v5.4.291 it doesn't.
> I'm not sure how bisection could be done in this case.
> [snip]
I've bisected this other bug with kexec_load. I found commit
62a03defeabd PM / hibernate: Verify the consistent of e820 memory map by md5 digest
Reverting it on v6.14 fixes kexec_load, but not kexec_file_load.
Also applying the patch suggested by msizanoen fixes kexec_file_load, too:
https://lore.kernel.org/all/Z_BDbwmFV6wxDPV1@desktop0a/
FYI, this is how I reverted that commit (I had to manually resolve
conflicts):
diff --git a/arch/x86/power/hibernate.c b/arch/x86/power/hibernate.c
index 5b81d19cd114..f2021a515bad 100644
--- a/arch/x86/power/hibernate.c
+++ b/arch/x86/power/hibernate.c
@@ -40,20 +40,6 @@ unsigned long restore_cr3 __visible;
unsigned long temp_pgt __visible;
unsigned long relocated_restore_code __visible;
-/**
- * pfn_is_nosave - check if given pfn is in the 'nosave' section
- */
-int pfn_is_nosave(unsigned long pfn)
-{
- unsigned long nosave_begin_pfn;
- unsigned long nosave_end_pfn;
-
- nosave_begin_pfn = __pa_symbol(&__nosave_begin) >> PAGE_SHIFT;
- nosave_end_pfn = PAGE_ALIGN(__pa_symbol(&__nosave_end)) >> PAGE_SHIFT;
-
- return pfn >= nosave_begin_pfn && pfn < nosave_end_pfn;
-}
-
struct restore_data_record {
unsigned long jump_address;
unsigned long jump_address_phys;
@@ -83,69 +69,6 @@ static inline u32 compute_e820_crc32(struct e820_table *table)
#define RESTORE_MAGIC 0x12345679UL
#endif
-/**
- * arch_hibernation_header_save - populate the architecture specific part
- * of a hibernation image header
- * @addr: address to save the data at
- */
-int arch_hibernation_header_save(void *addr, unsigned int max_size)
-{
- struct restore_data_record *rdr = addr;
-
- if (max_size < sizeof(struct restore_data_record))
- return -EOVERFLOW;
- rdr->magic = RESTORE_MAGIC;
- rdr->jump_address = (unsigned long)restore_registers;
- rdr->jump_address_phys = __pa_symbol(restore_registers);
-
- /*
- * The restore code fixes up CR3 and CR4 in the following sequence:
- *
- * [in hibernation asm]
- * 1. CR3 <= temporary page tables
- * 2. CR4 <= mmu_cr4_features (from the kernel that restores us)
- * 3. CR3 <= rdr->cr3
- * 4. CR4 <= mmu_cr4_features (from us, i.e. the image kernel)
- * [in restore_processor_state()]
- * 5. CR4 <= saved CR4
- * 6. CR3 <= saved CR3
- *
- * Our mmu_cr4_features has CR4.PCIDE=0, and toggling
- * CR4.PCIDE while CR3's PCID bits are nonzero is illegal, so
- * rdr->cr3 needs to point to valid page tables but must not
- * have any of the PCID bits set.
- */
- rdr->cr3 = restore_cr3 & ~CR3_PCID_MASK;
-
- rdr->e820_checksum = compute_e820_crc32(e820_table_firmware);
- return 0;
-}
-
-/**
- * arch_hibernation_header_restore - read the architecture specific data
- * from the hibernation image header
- * @addr: address to read the data from
- */
-int arch_hibernation_header_restore(void *addr)
-{
- struct restore_data_record *rdr = addr;
-
- if (rdr->magic != RESTORE_MAGIC) {
- pr_crit("Unrecognized hibernate image header format!\n");
- return -EINVAL;
- }
-
- restore_jump_address = rdr->jump_address;
- jump_address_phys = rdr->jump_address_phys;
- restore_cr3 = rdr->cr3;
-
- if (rdr->e820_checksum != compute_e820_crc32(e820_table_firmware)) {
- pr_crit("Hibernate inconsistent memory map detected!\n");
- return -ENODEV;
- }
-
- return 0;
-}
int relocate_restore_code(void)
{
diff --git a/arch/x86/power/hibernate_64.c b/arch/x86/power/hibernate_64.c
index a595953f1d6d..924420fdaab4 100644
--- a/arch/x86/power/hibernate_64.c
+++ b/arch/x86/power/hibernate_64.c
@@ -140,3 +140,56 @@ asmlinkage int swsusp_arch_resume(void)
restore_image();
return 0;
}
+
+/*
+ * pfn_is_nosave - check if given pfn is in the 'nosave' section
+ */
+
+int pfn_is_nosave(unsigned long pfn)
+{
+ unsigned long nosave_begin_pfn = __pa_symbol(&__nosave_begin) >> PAGE_SHIFT;
+ unsigned long nosave_end_pfn = PAGE_ALIGN(__pa_symbol(&__nosave_end)) >> PAGE_SHIFT;
+ return (pfn >= nosave_begin_pfn) && (pfn < nosave_end_pfn);
+}
+
+struct restore_data_record {
+ unsigned long jump_address;
+ unsigned long jump_address_phys;
+ unsigned long cr3;
+ unsigned long magic;
+};
+
+#define RESTORE_MAGIC 0x123456789ABCDEF0UL
+
+/**
+ * arch_hibernation_header_save - populate the architecture specific part
+ * of a hibernation image header
+ * @addr: address to save the data at
+ */
+int arch_hibernation_header_save(void *addr, unsigned int max_size)
+{
+ struct restore_data_record *rdr = addr;
+
+ if (max_size < sizeof(struct restore_data_record))
+ return -EOVERFLOW;
+ rdr->jump_address = (unsigned long)&restore_registers;
+ rdr->jump_address_phys = __pa_symbol(&restore_registers);
+ rdr->cr3 = restore_cr3;
+ rdr->magic = RESTORE_MAGIC;
+ return 0;
+}
+
+/**
+ * arch_hibernation_header_restore - read the architecture specific data
+ * from the hibernation image header
+ * @addr: address to read the data from
+ */
+int arch_hibernation_header_restore(void *addr)
+{
+ struct restore_data_record *rdr = addr;
+
+ restore_jump_address = rdr->jump_address;
+ jump_address_phys = rdr->jump_address_phys;
+ restore_cr3 = rdr->cr3;
+ return (rdr->magic == RESTORE_MAGIC) ? 0 : -EINVAL;
+}
next prev parent reply other threads:[~2025-04-04 23:31 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-01-13 21:28 [REGRESSION] Kernel booted via kexec fails to resume from hibernation Roberto Ricci
2025-01-13 21:31 ` Roberto Ricci
2025-01-14 3:42 ` Baoquan He
2025-04-01 12:59 ` msizanoen
2025-04-03 22:00 ` Roberto Ricci
2025-04-04 2:54 ` msizanoen
2025-04-04 4:56 ` msizanoen
2025-04-04 5:50 ` msizanoen
2025-04-04 20:39 ` Roberto Ricci
2025-04-05 5:15 ` msizanoen
2025-04-04 20:00 ` Roberto Ricci
2025-01-13 21:32 ` Roberto Ricci
2025-01-13 23:17 ` Andrew Morton
2025-01-14 13:19 ` Roberto Ricci
2025-01-14 13:16 ` Roberto Ricci
2025-01-15 4:04 ` Baoquan He
2025-01-15 12:00 ` Roberto Ricci
2025-01-16 11:52 ` Roberto Ricci
2025-01-17 1:55 ` Baoquan He
2025-01-17 3:41 ` Baoquan He
2025-01-17 7:52 ` Roberto Ricci
2025-01-16 9:54 ` Yuntao Wang
2025-01-22 9:45 ` RuiRui Yang
2025-01-22 13:01 ` Roberto Ricci
2025-01-27 2:39 ` Dave Young
2025-01-27 2:42 ` Dave Young
2025-03-09 17:09 ` Donald
2025-03-29 0:14 ` Roberto Ricci
2025-03-29 0:14 ` Roberto Ricci
2025-03-29 0:15 ` Roberto Ricci
2025-03-29 1:44 ` Baoquan He
2025-03-29 20:30 ` Roberto Ricci
2025-03-29 20:33 ` Roberto Ricci
2025-03-31 3:22 ` Dave Young
2025-04-03 21:59 ` Roberto Ricci
2025-04-04 23:31 ` Roberto Ricci [this message]
2025-04-04 23:37 ` Roberto Ricci
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=Z_Bry0JcB1jqPztR@desktop0a \
--to=io@r-ricci.it \
--cc=akpm@linux-foundation.org \
--cc=bhe@redhat.com \
--cc=dyoung@redhat.com \
--cc=ebiederm@xmission.com \
--cc=kexec@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=msizanoen@qtmlabs.xyz \
--cc=pavel@ucw.cz \
--cc=rafael.j.wysocki@intel.com \
--cc=rafael@kernel.org \
--cc=regressions@lists.linux.dev \
--cc=ytcoode@gmail.com \
--cc=yu.c.chen@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).