LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: "H. Peter Anvin" <hpa@zytor.com>
To: Lai Jiangshan <jiangshanlai@gmail.com>, linux-kernel@vger.kernel.org
Cc: Lai Jiangshan <laijs@linux.alibaba.com>,
	Andy Lutomirski <luto@kernel.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
	x86@kernel.org, Dave Hansen <dave.hansen@linux.intel.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Dave Jiang <dave.jiang@intel.com>,
	Ben Widawsky <ben.widawsky@intel.com>,
	Dan Williams <dan.j.williams@intel.com>,
	Arvind Sankar <nivedita@alum.mit.edu>
Subject: Re: [PATCH 25/24] x86/traps: Rewrite native_load_gs_index in C code
Date: Tue, 7 Sep 2021 21:42:43 -0700	[thread overview]
Message-ID: <f84c2b3c-a880-502f-4f51-4624b3800a05@zytor.com> (raw)
In-Reply-To: <9fdb04b1-dbb8-069d-f5ef-d4e8c0f2a83e@zytor.com>



On 9/7/21 6:38 PM, H. Peter Anvin wrote:
> On 9/2/21 3:50 AM, Lai Jiangshan wrote:
>> From: Lai Jiangshan <laijs@linux.alibaba.com>
>>
>> There is no constrain/limition to force native_load_gs_index() to be in
>> ASM code.
> 
> Hi,
> 
> First of all, let me say I really like your patchset, and I will try to
> review it in detail ASAP (most of the initial read pass looks very sane
> to me.
> 
> However, I would like to object in part this specific patch. It adds a
> fair bit of extra code to the exception path, and adds jumps between
> files which makes the code much harder to read.
> 
> You end up doing one swapgs in assembly and one in C, which would seem
> to be a very good indication that really isn't an improvement.
> 
> Note that this entire sequence is scheduled to be obsoleted by a single
> atomic hardware instruction, LKGS, which will replace ALL of
> native_load_gs_index(); it will no longer be necessary even to disable
> interrupts as there is no non-atomic state. In that sense, doing this as
> an out-of-line C function (with some inline assembly) is great, because
> it makes it far easier to use LKGS as an alternative; the only (small)
> disadvantage is that it ends up clobbering a lot of registers
> unnecessarily (in assembly it can be implemented clobbering only two
> registers; one if one uses pushf/popf to save the interrupt flag.)
> 

OK, here is a version which actually compiles:

(It makes me wonder if it would be useful to have an _ASM_EXTABLE_GOTO() 
macro to highlight the use of an asm goto C label.)

noinstr void native_load_gs_index(unsigned int selector)
{
	unsigned long flags;

	local_irq_save(flags);
	native_swapgs();

do_mov_gs:
	asm_volatile_goto("1: mov %[seg],%%gs\n"
			  "2:\n"
			  _ASM_EXTABLE(1b,%l[bad_seg])
			  : : [seg] "r" (selector) : : bad_seg);

	alternative("", "mfence", X86_BUG_SWAPGS_FENCE);
	native_swapgs();
	local_irq_restore(flags);
	return;

bad_seg:
	selector = 0;

	/* The exception dispatch will have restored kernel GS */
	native_swapgs();

	if (static_cpu_has_bug(X86_BUG_NULL_SEG))
		asm volatile("mov %[seg],%%gs"
			     : : [seg] "r" (__USER_DS));
	goto do_mov_gs;
}


  reply	other threads:[~2021-09-08  4:43 UTC|newest]

Thread overview: 72+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-08-31 17:50 [PATCH 00/24] x86/entry/64: Convert a bunch of ASM entry code into " Lai Jiangshan
2021-08-31 17:50 ` [PATCH 01/24] x86/traps: Remove stack-protector from traps.c Lai Jiangshan
2021-08-31 17:50 ` [PATCH 02/24] x86/traps: Move arch/x86/kernel/traps.c to arch/x86/entry/ Lai Jiangshan
2021-09-02  8:09   ` Joerg Roedel
2021-09-02  9:21     ` Lai Jiangshan
2021-09-02 10:50       ` Peter Zijlstra
2021-09-02 11:54         ` Lai Jiangshan
2021-09-02 12:05           ` Peter Zijlstra
2021-09-02 13:34             ` Peter Zijlstra
2021-09-02 17:05               ` Nick Desaulniers
2021-09-02 17:19                 ` Miguel Ojeda
2021-09-02 17:23                   ` Nick Desaulniers
2021-09-03  7:36                 ` Martin Liška
2021-09-07 21:12                   ` Nick Desaulniers
2021-09-08  7:33                     ` Martin Liška
2021-08-31 17:50 ` [PATCH 03/24] x86/traps: Move declaration of native_irq_return_iret up Lai Jiangshan
2021-08-31 17:50 ` [PATCH 04/24] x86/entry: Expose the address of .Lgs_change to traps.c Lai Jiangshan
2021-09-02  9:14   ` Peter Zijlstra
2021-09-02  9:20     ` Lai Jiangshan
2021-08-31 17:50 ` [PATCH 05/24] x86/entry: Introduce __entry_text for entry code written in C Lai Jiangshan
2021-08-31 19:34   ` Peter Zijlstra
2021-09-01  0:23     ` Lai Jiangshan
2021-08-31 17:50 ` [PATCH 06/24] x86/entry: Move PTI_USER_* to arch/x86/include/asm/processor-flags.h Lai Jiangshan
2021-08-31 17:50 ` [PATCH 07/24] x86: Mark __native_read_cr3() & native_write_cr3() as __always_inline Lai Jiangshan
2021-08-31 17:50 ` [PATCH 08/24] x86/traps: Add C verion of SWITCH_TO_KERNEL_CR3 as switch_to_kernel_cr3() Lai Jiangshan
2021-08-31 17:50 ` [PATCH 09/24] x86/traps: Add fence_swapgs_{user,kernel}_entry() Lai Jiangshan
2021-09-02  9:25   ` Peter Zijlstra
2021-08-31 17:50 ` [PATCH 10/24] x86/traps: Move pt_regs only in fixup_bad_iret() Lai Jiangshan
2021-08-31 17:50 ` [PATCH 11/24] x86/entry: Replace the most of asm code of error_entry to C code Lai Jiangshan
2021-09-02 10:16   ` Peter Zijlstra
2021-09-02 12:08     ` Lai Jiangshan
2021-08-31 17:50 ` [PATCH 12/24] x86/traps: Reconstruct pt_regs on task stack directly in fixup_bad_iret() Lai Jiangshan
2021-08-31 17:50 ` [PATCH 13/24] x86/traps: Mark sync_regs() and fixup_bad_iret() as static __always_inline Lai Jiangshan
2021-08-31 17:50 ` [PATCH 14/24] x86/entry: Make paranoid_exit() callable Lai Jiangshan
2021-08-31 17:50 ` [PATCH 15/24] x86/entry: Call paranoid_exit() in asm_exc_nmi() Lai Jiangshan
2021-08-31 17:50 ` [PATCH 16/24] x86/entry: Use skip_rdi instead of save_ret for PUSH_AND_CLEAR_REGS Lai Jiangshan
2021-08-31 17:50 ` [PATCH 17/24] x86/entry: Introduce struct ist_regs Lai Jiangshan
2021-09-10  0:18   ` Lai Jiangshan
2021-09-10  0:36     ` Lai Jiangshan
2021-09-10  4:31     ` H. Peter Anvin
2021-09-10  7:13       ` Lai Jiangshan
2021-09-10  7:14         ` H. Peter Anvin
2021-09-10  4:50     ` H. Peter Anvin
2021-09-10  4:51       ` H. Peter Anvin
2021-08-31 17:50 ` [PATCH 18/24] x86/entry: Add the C version ist_switch_to_kernel_cr3() Lai Jiangshan
2021-08-31 17:50 ` [PATCH 19/24] x86/entry: Add the C version ist_restore_cr3() Lai Jiangshan
2021-08-31 17:50 ` [PATCH 20/24] x86/entry: Add the C version get_percpu_base() Lai Jiangshan
2021-08-31 17:50 ` [PATCH 21/24] x86/entry: Add the C version ist_switch_to_kernel_gsbase() Lai Jiangshan
2021-08-31 17:50 ` [PATCH 22/24] x86/entry: Implement and use do_paranoid_entry() and paranoid_exit() Lai Jiangshan
2021-09-02 10:33   ` Peter Zijlstra
2021-09-02 10:42     ` Lai Jiangshan
2021-09-02 12:02       ` Peter Zijlstra
2021-09-02 11:58     ` Lai Jiangshan
2021-09-02 12:29       ` Joerg Roedel
2021-08-31 17:50 ` [PATCH 23/24] x86/entry: Remove the unused ASM macros Lai Jiangshan
2021-08-31 17:50 ` [PATCH 24/24] x86/syscall/64: Move the checking for sysret to C code Lai Jiangshan
2021-09-10  7:20   ` Nikolay Borisov
2021-09-10  7:30     ` Lai Jiangshan
2021-08-31 20:44 ` [PATCH 00/24] x86/entry/64: Convert a bunch of ASM entry code into " Peter Zijlstra
2021-09-02  6:28   ` Lai Jiangshan
2021-09-02  7:44     ` Peter Zijlstra
2021-09-02 10:50 ` [PATCH 25/24] x86/traps: Rewrite native_load_gs_index in " Lai Jiangshan
2021-09-08  1:38   ` H. Peter Anvin
2021-09-08  4:42     ` H. Peter Anvin [this message]
2021-09-08  5:00       ` H. Peter Anvin
2021-09-08  7:12         ` Lai Jiangshan
2021-09-09 23:16           ` H. Peter Anvin
2021-09-13 20:01   ` Andy Lutomirski
2021-09-14  2:04     ` Lai Jiangshan
2021-09-14  8:14       ` Peter Zijlstra
2021-09-14  8:17         ` Borislav Petkov
2021-09-14  8:40         ` Lai Jiangshan

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=f84c2b3c-a880-502f-4f51-4624b3800a05@zytor.com \
    --to=hpa@zytor.com \
    --cc=ben.widawsky@intel.com \
    --cc=bp@alien8.de \
    --cc=dan.j.williams@intel.com \
    --cc=dave.hansen@linux.intel.com \
    --cc=dave.jiang@intel.com \
    --cc=jiangshanlai@gmail.com \
    --cc=laijs@linux.alibaba.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=luto@kernel.org \
    --cc=mingo@redhat.com \
    --cc=nivedita@alum.mit.edu \
    --cc=peterz@infradead.org \
    --cc=tglx@linutronix.de \
    --cc=x86@kernel.org \
    --subject='Re: [PATCH 25/24] x86/traps: Rewrite native_load_gs_index in C code' \
    /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

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).