From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id E5762C43334 for ; Wed, 22 Jun 2022 15:06:01 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender: Content-Transfer-Encoding:Content-Type:List-Subscribe:List-Help:List-Post: List-Archive:List-Unsubscribe:List-Id:In-Reply-To:MIME-Version:References: Message-ID:Subject:Cc:To:From:Date:Reply-To:Content-ID:Content-Description: Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID: List-Owner; bh=+UmYF27Kaaf53IKAOJ6LXmblfzWD3yXWOmbCXySyLAc=; b=PDiHPJmTUrXfqw HgLZlG8tijwAw7ExfSlJV21sWwSQxa3iDMGlPU1cDa2zJ+y2swbh7h+y2BXF/lhCiB7xaggXrT523 k4Sc8Bc/a+WS3Uqefj66o0Gg74+gtygrCv6/JxDASBMJIQvtpfyB5ZQips2BfxbwV5MC1+ipbHGDj a9emq4B+8Oxg0oKUx/P8E9WWiwULBhYFQRQgcr+hX5Tk7Qz2zsucIxljRQRMgrPF6geOzkRO1DazP BIpMM14/4/z2ZyByCqTIXNBbKhIpA6iH7f+W55MLc1+eu3Bsr87jxnTcUtSKcmITrXireuDKH0EXd Ad8gNWU7e781IbQiO2Cw==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1o41ue-00B6SA-Di; Wed, 22 Jun 2022 15:05:00 +0000 Received: from foss.arm.com ([217.140.110.172]) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1o41ua-00B6Qg-FH for linux-arm-kernel@lists.infradead.org; Wed, 22 Jun 2022 15:04:58 +0000 Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 00F5412FC; Wed, 22 Jun 2022 08:04:55 -0700 (PDT) Received: from FVFF77S0Q05N.cambridge.arm.com (FVFF77S0Q05N.cambridge.arm.com [10.1.37.125]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPSA id 0785D3F792; Wed, 22 Jun 2022 08:04:53 -0700 (PDT) Date: Wed, 22 Jun 2022 16:04:51 +0100 From: Mark Rutland To: Ard Biesheuvel Cc: Linux ARM , Will Deacon , Catalin Marinas , James Morse Subject: Re: [PATCH] arm64: entry: simplify trampoline data page Message-ID: References: <20220427102255.1336386-1-ardb@kernel.org> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20220622_080456_656196_2C124A6E X-CRM114-Status: GOOD ( 32.33 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org On Wed, Jun 22, 2022 at 04:41:41PM +0200, Ard Biesheuvel wrote: > On Wed, 27 Apr 2022 at 12:23, Ard Biesheuvel wrote: > > > > Get rid of some clunky open coded arithmetic on section addresses, by > > emitting the trampoline data variables into a separate, dedicated r/o > > data section, and putting it at the next page boundary. This way, we can > > access the literals via single LDR instruction. > > > > While at it, get rid of other, implicit literals, and use ADRP/ADD or > > MOVZ/MOVK sequences, as appropriate. Note that the latter are only > > supported for CONFIG_RELOCATABLE=n (which is usually the case if > > CONFIG_RANDOMIZE_BASE=n), so update the CPP conditionals to reflect > > this. > > > > Cc: James Morse > > Signed-off-by: Ard Biesheuvel > > Ping, in case this one slipped behind the desk. Sorry for the delay. THis has been on my queue of things to look at with a bunch of other stuff, and I've had some difficulty prioritizing all that. > > > --- > > arch/arm64/include/asm/fixmap.h | 4 +- > > arch/arm64/kernel/entry.S | 45 ++++++-------------- > > arch/arm64/kernel/vmlinux.lds.S | 3 +- > > arch/arm64/mm/mmu.c | 10 ++--- > > 4 files changed, 22 insertions(+), 40 deletions(-) > > > > diff --git a/arch/arm64/include/asm/fixmap.h b/arch/arm64/include/asm/fixmap.h > > index daff882883f9..71ed5fdf718b 100644 > > --- a/arch/arm64/include/asm/fixmap.h > > +++ b/arch/arm64/include/asm/fixmap.h > > @@ -62,10 +62,12 @@ enum fixed_addresses { > > #endif /* CONFIG_ACPI_APEI_GHES */ > > > > #ifdef CONFIG_UNMAP_KERNEL_AT_EL0 > > +#ifdef CONFIG_RELOCATABLE > > + FIX_ENTRY_TRAMP_TEXT4, /* one extra slot for the data page */ > > +#endif > > FIX_ENTRY_TRAMP_TEXT3, > > FIX_ENTRY_TRAMP_TEXT2, > > FIX_ENTRY_TRAMP_TEXT1, > > - FIX_ENTRY_TRAMP_DATA, > > #define TRAMP_VALIAS (__fix_to_virt(FIX_ENTRY_TRAMP_TEXT1)) > > #endif /* CONFIG_UNMAP_KERNEL_AT_EL0 */ > > __end_of_permanent_fixed_addresses, > > diff --git a/arch/arm64/kernel/entry.S b/arch/arm64/kernel/entry.S > > index ede028dee81b..aed2b41e05aa 100644 > > --- a/arch/arm64/kernel/entry.S > > +++ b/arch/arm64/kernel/entry.S > > @@ -636,18 +636,20 @@ alternative_else_nop_endif > > */ > > .endm > > > > - .macro tramp_data_page dst > > - adr_l \dst, .entry.tramp.text > > - sub \dst, \dst, PAGE_SIZE > > - .endm > > - > > - .macro tramp_data_read_var dst, var > > -#ifdef CONFIG_RANDOMIZE_BASE > > - tramp_data_page \dst > > - add \dst, \dst, #:lo12:__entry_tramp_data_\var > > - ldr \dst, [\dst] > > + .macro tramp_data_read_var dst, var > > +#ifdef CONFIG_RELOCATABLE > > + ldr \dst, .L__tramp_data_\var > > + .ifndef .L__tramp_data_\var > > + .pushsection ".entry.tramp.rodata", "a", %progbits > > + .align 3 > > +.L__tramp_data_\var: > > + .quad \var > > + .popsection > > + .endif > > #else > > - ldr \dst, =\var > > + movz \dst, :abs_g2_s:\var > > + movk \dst, :abs_g1_nc:\var > > + movk \dst, :abs_g0_nc:\var > > #endif > > .endm Given the lack of a g3 reloc, I assumme `var` is always an address, and we're assuming it's always in the upper 48-bits? I think it'd be worth a comment as to why this is safe, or just use a g3 reloc since then it's always good per inspection. I'm a bit confused that we've put the var into the literal; I thought the idea here was that it was secret and needed to be placed in a page not mapped during userspace. Is the assumption there that it's pointless for !RELOCATABLE kernels since it can be known anyway, have I misunderstood, or something else? Otherwise this all looks good superficially; I just haven't had the time to page it all in. Mark. > > > > @@ -695,7 +697,7 @@ alternative_else_nop_endif > > msr vbar_el1, x30 > > isb > > .else > > - ldr x30, =vectors > > + adr_l x30, vectors > > .endif // \kpti == 1 > > > > .if \bhb == BHB_MITIGATION_FW > > @@ -764,24 +766,7 @@ SYM_CODE_END(tramp_exit_native) > > SYM_CODE_START(tramp_exit_compat) > > tramp_exit 32 > > SYM_CODE_END(tramp_exit_compat) > > - > > - .ltorg > > .popsection // .entry.tramp.text > > -#ifdef CONFIG_RANDOMIZE_BASE > > - .pushsection ".rodata", "a" > > - .align PAGE_SHIFT > > -SYM_DATA_START(__entry_tramp_data_start) > > -__entry_tramp_data_vectors: > > - .quad vectors > > -#ifdef CONFIG_ARM_SDE_INTERFACE > > -__entry_tramp_data___sdei_asm_handler: > > - .quad __sdei_asm_handler > > -#endif /* CONFIG_ARM_SDE_INTERFACE */ > > -__entry_tramp_data_this_cpu_vector: > > - .quad this_cpu_vector > > -SYM_DATA_END(__entry_tramp_data_start) > > - .popsection // .rodata > > -#endif /* CONFIG_RANDOMIZE_BASE */ > > #endif /* CONFIG_UNMAP_KERNEL_AT_EL0 */ > > > > /* > > @@ -932,7 +917,6 @@ NOKPROBE(call_on_irq_stack) > > * This clobbers x4, __sdei_handler() will restore this from firmware's > > * copy. > > */ > > -.ltorg > > .pushsection ".entry.tramp.text", "ax" > > SYM_CODE_START(__sdei_asm_entry_trampoline) > > mrs x4, ttbr1_el1 > > @@ -967,7 +951,6 @@ SYM_CODE_START(__sdei_asm_exit_trampoline) > > 1: sdei_handler_exit exit_mode=x2 > > SYM_CODE_END(__sdei_asm_exit_trampoline) > > NOKPROBE(__sdei_asm_exit_trampoline) > > - .ltorg > > .popsection // .entry.tramp.text > > #endif /* CONFIG_UNMAP_KERNEL_AT_EL0 */ > > > > diff --git a/arch/arm64/kernel/vmlinux.lds.S b/arch/arm64/kernel/vmlinux.lds.S > > index edaf0faf766f..17e554be9198 100644 > > --- a/arch/arm64/kernel/vmlinux.lds.S > > +++ b/arch/arm64/kernel/vmlinux.lds.S > > @@ -117,7 +117,8 @@ jiffies = jiffies_64; > > __entry_tramp_text_start = .; \ > > *(.entry.tramp.text) \ > > . = ALIGN(PAGE_SIZE); \ > > - __entry_tramp_text_end = .; > > + __entry_tramp_text_end = .; \ > > + *(.entry.tramp.rodata) > > #else > > #define TRAMP_TEXT > > #endif > > diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c > > index 626ec32873c6..be4d6c3f5692 100644 > > --- a/arch/arm64/mm/mmu.c > > +++ b/arch/arm64/mm/mmu.c > > @@ -665,13 +665,9 @@ static int __init map_entry_trampoline(void) > > __set_fixmap(FIX_ENTRY_TRAMP_TEXT1 - i, > > pa_start + i * PAGE_SIZE, prot); > > > > - if (IS_ENABLED(CONFIG_RANDOMIZE_BASE)) { > > - extern char __entry_tramp_data_start[]; > > - > > - __set_fixmap(FIX_ENTRY_TRAMP_DATA, > > - __pa_symbol(__entry_tramp_data_start), > > - PAGE_KERNEL_RO); > > - } > > + if (IS_ENABLED(CONFIG_RELOCATABLE)) > > + __set_fixmap(FIX_ENTRY_TRAMP_TEXT1 - i, > > + pa_start + i * PAGE_SIZE, PAGE_KERNEL_RO); > > > > return 0; > > } > > -- > > 2.30.2 > > _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel