From: Nathan Chancellor <nathan@kernel.org>
To: David Woodhouse <dwmw2@infradead.org>
Cc: "Ning, Hongyu" <hongyu.ning@linux.intel.com>,
kexec@lists.infradead.org, Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
Dave Hansen <dave.hansen@linux.intel.com>,
x86@kernel.org, "H. Peter Anvin" <hpa@zytor.com>,
"Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>,
Kai Huang <kai.huang@intel.com>,
Nikolay Borisov <nik.borisov@suse.com>,
linux-kernel@vger.kernel.org, Simon Horman <horms@kernel.org>,
Dave Young <dyoung@redhat.com>,
Peter Zijlstra <peterz@infradead.org>,
jpoimboe@kernel.org, bsz@amazon.de
Subject: Re: [PATCH] x86/kexec: Only write through identity mapping of control page
Date: Thu, 12 Dec 2024 13:31:32 -0700 [thread overview]
Message-ID: <20241212203132.GA1627368@ax162> (raw)
In-Reply-To: <9c68688625f409104b16164da30aa6d3eb494e5d.camel@infradead.org>
On Thu, Dec 12, 2024 at 08:11:19PM +0000, David Woodhouse wrote:
> From: David Woodhouse <dwmw@amazon.co.uk>
>
> The virtual mapping of the control page may have been _PAGE_GLOBAL and
> thus its PTE might not have been flushed on the %cr3 switch and it might
> effectively still be read-only. Move the writes to it down into the
> identity_mapped() function where the same %rip-relative addressing will
> get the new mapping.
>
> The stack is fine, as that's using the identity mapped address anyway.
>
> Fixes: 5a82223e0743 ("x86/kexec: Mark relocate_kernel page as ROX instead of RWX")
> Reported-by: Nathan Chancellor <nathan@kernel.org>
> Reported-by: "Ning, Hongyu" <hongyu.ning@linux.intel.com>
> Closes: https://bugzilla.kernel.org/show_bug.cgi?id=219592
> Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Tested-by: Nathan Chancellor <nathan@kernel.org>
> ---
> arch/x86/kernel/relocate_kernel_64.S | 32 +++++++++++++++++-----------
> 1 file changed, 20 insertions(+), 12 deletions(-)
>
> diff --git a/arch/x86/kernel/relocate_kernel_64.S b/arch/x86/kernel/relocate_kernel_64.S
> index 553d67845b84..b9c80b3091c8 100644
> --- a/arch/x86/kernel/relocate_kernel_64.S
> +++ b/arch/x86/kernel/relocate_kernel_64.S
> @@ -90,22 +90,17 @@ SYM_CODE_START_NOALIGN(relocate_kernel)
> movq kexec_pa_table_page(%rip), %r9
> movq %r9, %cr3
>
> - /* Save %rsp and CRs. */
> - movq %rsp, saved_rsp(%rip)
> - movq %rax, saved_cr3(%rip)
> - movq %cr0, %rax
> - movq %rax, saved_cr0(%rip)
> - /* Leave CR4 in %r13 to enable the right paging mode later. */
> - movq %cr4, %r13
> - movq %r13, saved_cr4(%rip)
> -
> - /* save indirection list for jumping back */
> - movq %rdi, pa_backup_pages_map(%rip)
> + /*
> + * The control page still might not be writable because the original
> + * kernel PTE may have had the _PAGE_GLOBAL bit set. Don't write to
> + * it except through the *identmap* address.
> + */
>
> /* Save the preserve_context to %r11 as swap_pages clobbers %rcx. */
> movq %rcx, %r11
>
> /* setup a new stack at the end of the physical control page */
> + movq %rsp, %rbp
> lea PAGE_SIZE(%rsi), %rsp
>
> /* jump to identity mapped page */
> @@ -118,6 +113,19 @@ SYM_CODE_END(relocate_kernel)
>
> SYM_CODE_START_LOCAL_NOALIGN(identity_mapped)
> UNWIND_HINT_END_OF_STACK
> +
> + /* Save original %rsp and CRs. */
> + movq %rbp, saved_rsp(%rip)
> + movq %rax, saved_cr3(%rip)
> + movq %cr0, %rax
> + movq %rax, saved_cr0(%rip)
> + /* Leave CR4 in %r13 to enable the right paging mode later. */
> + movq %cr4, %r13
> + movq %r13, saved_cr4(%rip)
> +
> + /* save indirection list for jumping back */
> + movq %rdi, pa_backup_pages_map(%rip)
> +
> /*
> * %rdi indirection page
> * %rdx start address
> @@ -185,7 +193,7 @@ SYM_CODE_START_LOCAL_NOALIGN(identity_mapped)
> * - Machine check exception on TDX guest, if it was enabled before.
> * Clearing MCE might not be allowed in TDX guests, depending on setup.
> *
> - * Use R13 that contains the original CR4 value, read in relocate_kernel().
> + * Use R13 that contains the original CR4 value
> * PAE is always set in the original CR4.
> */
> andl $(X86_CR4_PAE | X86_CR4_LA57), %r13d
> --
> 2.47.0
>
>
next prev parent reply other threads:[~2024-12-12 20:31 UTC|newest]
Thread overview: 68+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-12-05 15:05 [PATCH v5 00/20] x86/kexec: Add exception handling for relocate_kernel and further yak-shaving David Woodhouse
2024-12-05 15:05 ` [PATCH v5 01/20] x86/kexec: Restore GDT on return from preserve_context kexec David Woodhouse
2024-12-06 10:16 ` [tip: x86/urgent] x86/kexec: Restore GDT on return from ::preserve_context kexec tip-bot2 for David Woodhouse
2024-12-05 15:05 ` [PATCH v5 02/20] x86/kexec: Clean up and document register use in relocate_kernel_64.S David Woodhouse
2024-12-06 10:16 ` [tip: x86/boot] " tip-bot2 for David Woodhouse
2024-12-05 15:05 ` [PATCH v5 03/20] x86/kexec: Use named labels in swap_pages " David Woodhouse
2024-12-06 10:16 ` [tip: x86/boot] " tip-bot2 for David Woodhouse
2024-12-05 15:05 ` [PATCH v5 04/20] x86/kexec: Only swap pages for preserve_context mode David Woodhouse
2024-12-06 10:16 ` [tip: x86/boot] x86/kexec: Only swap pages for ::preserve_context mode tip-bot2 for David Woodhouse
2024-12-05 15:05 ` [PATCH v5 05/20] x86/kexec: Allocate PGD for x86_64 transition page tables separately David Woodhouse
2024-12-06 10:16 ` [tip: x86/boot] " tip-bot2 for David Woodhouse
2024-12-05 15:05 ` [PATCH v5 06/20] x86/kexec: Copy control page into place in machine_kexec_prepare() David Woodhouse
2024-12-06 10:16 ` [tip: x86/boot] " tip-bot2 for David Woodhouse
2024-12-05 15:05 ` [PATCH v5 07/20] x86/kexec: Invoke copy of relocate_kernel() instead of the original David Woodhouse
2024-12-06 10:16 ` [tip: x86/boot] " tip-bot2 for David Woodhouse
2024-12-14 23:08 ` [PATCH v5 07/20] " Nathan Chancellor
2024-12-15 7:19 ` David Woodhouse
2024-12-15 10:09 ` David Woodhouse
2024-12-16 5:49 ` Nathan Chancellor
2024-12-16 8:13 ` David Woodhouse
2024-12-16 12:09 ` David Woodhouse
2024-12-17 12:03 ` David Woodhouse
2024-12-18 9:03 ` Josh Poimboeuf
2024-12-18 9:44 ` David Woodhouse
2024-12-18 21:23 ` Josh Poimboeuf
2024-12-18 22:27 ` David Woodhouse
2024-12-19 0:20 ` Josh Poimboeuf
2024-12-19 10:02 ` David Woodhouse
2024-12-19 22:28 ` Josh Poimboeuf
2024-12-05 15:05 ` [PATCH v5 08/20] x86/kexec: Move relocate_kernel to kernel .data section David Woodhouse
2024-12-06 10:16 ` [tip: x86/boot] " tip-bot2 for David Woodhouse
2024-12-05 15:05 ` [PATCH v5 09/20] x86/kexec: Add data section to relocate_kernel David Woodhouse
2024-12-06 10:16 ` [tip: x86/boot] " tip-bot2 for David Woodhouse
2024-12-05 15:05 ` [PATCH v5 10/20] x86/kexec: Drop page_list argument from relocate_kernel() David Woodhouse
2024-12-06 10:16 ` [tip: x86/boot] " tip-bot2 for David Woodhouse
2024-12-05 15:05 ` [PATCH v5 11/20] x86/kexec: Eliminate writes through kernel mapping of relocate_kernel page David Woodhouse
2024-12-06 10:16 ` [tip: x86/boot] " tip-bot2 for David Woodhouse
2024-12-05 15:05 ` [PATCH v5 12/20] x86/kexec: Clean up register usage in relocate_kernel() David Woodhouse
2024-12-06 10:16 ` [tip: x86/boot] " tip-bot2 for David Woodhouse
2024-12-05 15:05 ` [PATCH v5 13/20] x86/kexec: Mark relocate_kernel page as ROX instead of RWX David Woodhouse
2024-12-06 10:16 ` [tip: x86/boot] " tip-bot2 for David Woodhouse
2024-12-12 1:44 ` [PATCH v5 13/20] " Nathan Chancellor
2024-12-12 10:30 ` David Woodhouse
2024-12-12 15:04 ` Nathan Chancellor
2024-12-12 17:00 ` David Woodhouse
2024-12-12 17:42 ` Nathan Chancellor
2024-12-12 19:31 ` David Woodhouse
2024-12-12 20:11 ` [PATCH] x86/kexec: Only write through identity mapping of control page David Woodhouse
2024-12-12 20:31 ` Nathan Chancellor [this message]
2024-12-12 21:18 ` Dave Hansen
2024-12-12 21:32 ` David Woodhouse
2024-12-12 21:43 ` Dave Hansen
2024-12-12 21:59 ` David Woodhouse
2024-12-12 23:08 ` [PATCH] x86/kexec: Disable global pages before writing to " David Woodhouse
2024-12-13 7:51 ` Ning, Hongyu
2024-12-13 6:47 ` [PATCH] x86/kexec: Only write through identity mapping of " Ning, Hongyu
2024-12-12 3:03 ` [PATCH v5 13/20] x86/kexec: Mark relocate_kernel page as ROX instead of RWX Ning, Hongyu
2024-12-12 10:13 ` David Woodhouse
2024-12-13 6:45 ` Ning, Hongyu
2024-12-13 7:01 ` David Woodhouse
2024-12-13 7:41 ` Ning, Hongyu
2024-12-05 15:05 ` [PATCH v5 14/20] x86/kexec: Add CONFIG_KEXEC_DEBUG option David Woodhouse
2024-12-05 15:05 ` [PATCH v5 15/20] x86/kexec: Debugging support: load a GDT David Woodhouse
2024-12-05 15:05 ` [PATCH v5 16/20] x86/kexec: Debugging support: Load an IDT and basic exception entry points David Woodhouse
2024-12-05 15:05 ` [PATCH v5 17/20] x86/kexec: Debugging support: Dump registers on exception David Woodhouse
2024-12-05 15:05 ` [PATCH v5 18/20] x86/kexec: Add 8250 serial port output David Woodhouse
2024-12-05 15:05 ` [PATCH v5 19/20] x86/kexec: Add 8250 MMIO " David Woodhouse
2024-12-05 15:05 ` [PATCH v5 20/20] [DO NOT MERGE] x86/kexec: Add int3 in kexec path for testing David Woodhouse
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=20241212203132.GA1627368@ax162 \
--to=nathan@kernel.org \
--cc=bp@alien8.de \
--cc=bsz@amazon.de \
--cc=dave.hansen@linux.intel.com \
--cc=dwmw2@infradead.org \
--cc=dyoung@redhat.com \
--cc=hongyu.ning@linux.intel.com \
--cc=horms@kernel.org \
--cc=hpa@zytor.com \
--cc=jpoimboe@kernel.org \
--cc=kai.huang@intel.com \
--cc=kexec@lists.infradead.org \
--cc=kirill.shutemov@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=nik.borisov@suse.com \
--cc=peterz@infradead.org \
--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