Received: by 2002:a25:868d:0:0:0:0:0 with SMTP id z13csp2780390ybk; Tue, 12 May 2020 08:01:27 -0700 (PDT) X-Google-Smtp-Source: APiQypIcdQvArIjR8LUv1+rWBvzwjgfH+zMNeXls7YbWFdXhQIhKUC0xQpNrhAuX0Eij6tPDBa4E X-Received: by 2002:a17:906:c7cc:: with SMTP id dc12mr12277588ejb.263.1589295687515; Tue, 12 May 2020 08:01:27 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1589295687; cv=none; d=google.com; s=arc-20160816; b=ALpF+cKAAcwuiN77tDtD88gWHgS4HI0NkUvP3ftBa3fhFnZWJSvLqgkdloFU/JiHX7 2wqfdhTzxT9+WS7XJy8zCfdCRlNXWBr65mvy9zI2JkvV0Ohtvg8k9U8vPab0aaMlORFy rytjIN+qUcvfY/y0MthmO+lxNc3ZwF6NZGSTiSCS1iVaYeNhZzLfbKHnkStbDXb0KrAn 4L8aquST0vFwxruA3KTGqVhrRt6hVdq9EPOk1MUo+I+N2wVPgCSpjr4cFKUYG/MfwJa4 QXEvxQHGi662HbdLW0rDB8sDxYaIS+e9GU6eH51CS0MEfF6N9nWmRNtiDT6L0B9hD+07 O1wg== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:sender:content-transfer-encoding:mime-version :references:in-reply-to:message-id:date:subject:cc:to:from :ironport-sdr:ironport-sdr; bh=qxqcZqDI+03Y9pXULLRxXfDCwvFB+YaIN/vzRKAUDpA=; b=02DZPOGCGIODkwyteEMcjczhhMkVJT/ujZuVuX3nPP3vo598WsbOnNFVM9rWutxqDi 0yvPJjmGaABeOzSFWSsf007o0ryk7p+3HgYOvXCioFiLz2CtqOXqiUGtMunxEJZVIs2q FIE8WboCHpBdTaJmJgDefUR1PqM2cymm5FPAINUY7zyEphNAkOF4cPFkmwidb/AraBdv t7G9yvbpWHBbdJVxfwyi/m/nFY9bQUiv0h/8prvZGVG/5uLapikJSSv3R3wkqzVY1V5/ dH/v4y9NWICznqPS+h/UhF8eNOybSRGkRoWPgVSryhGVyVSTqFP8nASRfgcp9YrNauPM yo9A== ARC-Authentication-Results: i=1; mx.google.com; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=intel.com Return-Path: Received: from vger.kernel.org (vger.kernel.org. [23.128.96.18]) by mx.google.com with ESMTP id cd9si7851225ejb.159.2020.05.12.08.01.03; Tue, 12 May 2020 08:01:27 -0700 (PDT) Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18; Authentication-Results: mx.google.com; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=intel.com Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1730406AbgELO5L (ORCPT + 99 others); Tue, 12 May 2020 10:57:11 -0400 Received: from mga05.intel.com ([192.55.52.43]:29538 "EHLO mga05.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1725929AbgELO5K (ORCPT ); Tue, 12 May 2020 10:57:10 -0400 IronPort-SDR: WmsWh8AFPmJDKA4Ad5ldEfPMINEG+d78Yj8+Zywp0E4ECASjGpcizeZ2DW0X8Xum4KeyBjDygr uZsRj31cwjpA== X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from fmsmga001.fm.intel.com ([10.253.24.23]) by fmsmga105.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 12 May 2020 07:55:08 -0700 IronPort-SDR: LyNbA3Z7F7reJm6OO5SEVYP9TBiBk5IELcQlq0KtRKX7DHUfvhlFZ6jSPu9tV44VKaP+9xHXlM mRLbMV9KLl4A== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.73,384,1583222400"; d="scan'208";a="371587808" Received: from yyu32-desk.sc.intel.com ([143.183.136.146]) by fmsmga001.fm.intel.com with ESMTP; 12 May 2020 07:55:08 -0700 From: Yu-cheng Yu To: linux-kernel@vger.kernel.org, x86@kernel.org, "H. Peter Anvin" , Thomas Gleixner , Ingo Molnar , Dave Hansen , Tony Luck , Andy Lutomirski , Borislav Petkov , Rik van Riel , "Ravi V. Shankar" , Sebastian Andrzej Siewior , Fenghua Yu , Peter Zijlstra Cc: Yu-cheng Yu Subject: [PATCH v4 03/10] x86/fpu/xstate: Separate user and supervisor xfeatures mask Date: Tue, 12 May 2020 07:54:37 -0700 Message-Id: <20200512145444.15483-4-yu-cheng.yu@intel.com> X-Mailer: git-send-email 2.21.0 In-Reply-To: <20200512145444.15483-1-yu-cheng.yu@intel.com> References: <20200512145444.15483-1-yu-cheng.yu@intel.com> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Before the introduction of XSAVES supervisor states, 'xfeatures_mask' is used at various places to determine XSAVE buffer components and XCR0 bits. It contains only user xstates. To support supervisor xstates, it is necessary to separate user and supervisor xstates: - First, change 'xfeatures_mask' to 'xfeatures_mask_all', which represents the full set of bits that should ever be set in a kernel XSAVE buffer. - Introduce xfeatures_mask_supervisor() and xfeatures_mask_user() to extract relevant xfeatures from xfeatures_mask_all. Co-developed-by: Fenghua Yu Signed-off-by: Fenghua Yu Signed-off-by: Yu-cheng Yu Reviewed-by: Dave Hansen Reviewed-by: Tony Luck --- v3: - Change xfeature_enabled() type from static int to static bool while at it. v2: - Fix typo in commit log. - Move xfeatures_mask_supervisor() from xstate.c to xstate.h. - Remove printing of user xstates from fpu__init_system_xstate(). arch/x86/include/asm/fpu/internal.h | 2 +- arch/x86/include/asm/fpu/xstate.h | 13 ++++- arch/x86/kernel/fpu/signal.c | 16 +++++-- arch/x86/kernel/fpu/xstate.c | 73 +++++++++++++++++------------ 4 files changed, 67 insertions(+), 37 deletions(-) diff --git a/arch/x86/include/asm/fpu/internal.h b/arch/x86/include/asm/fpu/internal.h index 44c48e34d799..ccb1bb32ad7d 100644 --- a/arch/x86/include/asm/fpu/internal.h +++ b/arch/x86/include/asm/fpu/internal.h @@ -92,7 +92,7 @@ static inline void fpstate_init_xstate(struct xregs_state *xsave) * XRSTORS requires these bits set in xcomp_bv, or it will * trigger #GP: */ - xsave->header.xcomp_bv = XCOMP_BV_COMPACTED_FORMAT | xfeatures_mask; + xsave->header.xcomp_bv = XCOMP_BV_COMPACTED_FORMAT | xfeatures_mask_all; } static inline void fpstate_init_fxstate(struct fxregs_state *fx) diff --git a/arch/x86/include/asm/fpu/xstate.h b/arch/x86/include/asm/fpu/xstate.h index b08fa823425f..92104b298d77 100644 --- a/arch/x86/include/asm/fpu/xstate.h +++ b/arch/x86/include/asm/fpu/xstate.h @@ -51,7 +51,18 @@ #define REX_PREFIX #endif -extern u64 xfeatures_mask; +extern u64 xfeatures_mask_all; + +static inline u64 xfeatures_mask_supervisor(void) +{ + return xfeatures_mask_all & XFEATURE_MASK_SUPERVISOR_SUPPORTED; +} + +static inline u64 xfeatures_mask_user(void) +{ + return xfeatures_mask_all & XFEATURE_MASK_USER_SUPPORTED; +} + extern u64 xstate_fx_sw_bytes[USER_XSTATE_FX_SW_WORDS]; extern void __init update_regset_xstate_info(unsigned int size, diff --git a/arch/x86/kernel/fpu/signal.c b/arch/x86/kernel/fpu/signal.c index 585e3651b98f..3df0cfae535f 100644 --- a/arch/x86/kernel/fpu/signal.c +++ b/arch/x86/kernel/fpu/signal.c @@ -252,13 +252,17 @@ sanitize_restored_xstate(union fpregs_state *state, */ static int copy_user_to_fpregs_zeroing(void __user *buf, u64 xbv, int fx_only) { + u64 init_bv; + if (use_xsave()) { if (fx_only) { - u64 init_bv = xfeatures_mask & ~XFEATURE_MASK_FPSSE; + init_bv = xfeatures_mask_user() & ~XFEATURE_MASK_FPSSE; + copy_kernel_to_xregs(&init_fpstate.xsave, init_bv); return copy_user_to_fxregs(buf); } else { - u64 init_bv = xfeatures_mask & ~xbv; + init_bv = xfeatures_mask_user() & ~xbv; + if (unlikely(init_bv)) copy_kernel_to_xregs(&init_fpstate.xsave, init_bv); return copy_user_to_xregs(buf, xbv); @@ -358,7 +362,7 @@ static int __fpu__restore_sig(void __user *buf, void __user *buf_fx, int size) if (use_xsave() && !fx_only) { - u64 init_bv = xfeatures_mask & ~xfeatures; + u64 init_bv = xfeatures_mask_user() & ~xfeatures; if (using_compacted_format()) { ret = copy_user_to_xstate(&fpu->state.xsave, buf_fx); @@ -389,7 +393,9 @@ static int __fpu__restore_sig(void __user *buf, void __user *buf_fx, int size) fpregs_lock(); if (use_xsave()) { - u64 init_bv = xfeatures_mask & ~XFEATURE_MASK_FPSSE; + u64 init_bv; + + init_bv = xfeatures_mask_user() & ~XFEATURE_MASK_FPSSE; copy_kernel_to_xregs(&init_fpstate.xsave, init_bv); } @@ -465,7 +471,7 @@ void fpu__init_prepare_fx_sw_frame(void) fx_sw_reserved.magic1 = FP_XSTATE_MAGIC1; fx_sw_reserved.extended_size = size; - fx_sw_reserved.xfeatures = xfeatures_mask; + fx_sw_reserved.xfeatures = xfeatures_mask_user(); fx_sw_reserved.xstate_size = fpu_user_xstate_size; if (IS_ENABLED(CONFIG_IA32_EMULATION) || diff --git a/arch/x86/kernel/fpu/xstate.c b/arch/x86/kernel/fpu/xstate.c index 9997df717339..fa71af643025 100644 --- a/arch/x86/kernel/fpu/xstate.c +++ b/arch/x86/kernel/fpu/xstate.c @@ -54,9 +54,10 @@ static short xsave_cpuid_features[] __initdata = { }; /* - * Mask of xstate features supported by the CPU and the kernel: + * This represents the full set of bits that should ever be set in a kernel + * XSAVE buffer, both supervisor and user xstates. */ -u64 xfeatures_mask __read_mostly; +u64 xfeatures_mask_all __read_mostly; static unsigned int xstate_offsets[XFEATURE_MAX] = { [ 0 ... XFEATURE_MAX - 1] = -1}; static unsigned int xstate_sizes[XFEATURE_MAX] = { [ 0 ... XFEATURE_MAX - 1] = -1}; @@ -76,7 +77,7 @@ unsigned int fpu_user_xstate_size; */ int cpu_has_xfeatures(u64 xfeatures_needed, const char **feature_name) { - u64 xfeatures_missing = xfeatures_needed & ~xfeatures_mask; + u64 xfeatures_missing = xfeatures_needed & ~xfeatures_mask_all; if (unlikely(feature_name)) { long xfeature_idx, max_idx; @@ -150,7 +151,7 @@ void fpstate_sanitize_xstate(struct fpu *fpu) * None of the feature bits are in init state. So nothing else * to do for us, as the memory layout is up to date. */ - if ((xfeatures & xfeatures_mask) == xfeatures_mask) + if ((xfeatures & xfeatures_mask_all) == xfeatures_mask_all) return; /* @@ -177,7 +178,7 @@ void fpstate_sanitize_xstate(struct fpu *fpu) * in a special way already: */ feature_bit = 0x2; - xfeatures = (xfeatures_mask & ~xfeatures) >> 2; + xfeatures = (xfeatures_mask_user() & ~xfeatures) >> 2; /* * Update all the remaining memory layouts according to their @@ -205,19 +206,28 @@ void fpstate_sanitize_xstate(struct fpu *fpu) */ void fpu__init_cpu_xstate(void) { - if (!boot_cpu_has(X86_FEATURE_XSAVE) || !xfeatures_mask) + u64 unsup_bits; + + if (!boot_cpu_has(X86_FEATURE_XSAVE) || !xfeatures_mask_all) return; /* * Unsupported supervisor xstates should not be found in * the xfeatures mask. */ - WARN_ONCE((xfeatures_mask & XFEATURE_MASK_SUPERVISOR_UNSUPPORTED), - "x86/fpu: Found unsupported supervisor xstates.\n"); + unsup_bits = xfeatures_mask_all & XFEATURE_MASK_SUPERVISOR_UNSUPPORTED; + WARN_ONCE(unsup_bits, "x86/fpu: Found unsupported supervisor xstates: 0x%llx\n", + unsup_bits); - xfeatures_mask &= ~XFEATURE_MASK_SUPERVISOR_UNSUPPORTED; + xfeatures_mask_all &= ~XFEATURE_MASK_SUPERVISOR_UNSUPPORTED; cr4_set_bits(X86_CR4_OSXSAVE); - xsetbv(XCR_XFEATURE_ENABLED_MASK, xfeatures_mask); + + /* + * XCR_XFEATURE_ENABLED_MASK (aka. XCR0) sets user features + * managed by XSAVE{C, OPT, S} and XRSTOR{S}. Only XSAVE user + * states can be set here. + */ + xsetbv(XCR_XFEATURE_ENABLED_MASK, xfeatures_mask_user()); } /* @@ -225,9 +235,9 @@ void fpu__init_cpu_xstate(void) * functions here: one for user xstates and the other for * system xstates. For now, they are the same. */ -static int xfeature_enabled(enum xfeature xfeature) +static bool xfeature_enabled(enum xfeature xfeature) { - return !!(xfeatures_mask & (1UL << xfeature)); + return xfeatures_mask_all & BIT_ULL(xfeature); } /* @@ -414,7 +424,7 @@ static void __init setup_init_fpu_buf(void) if (boot_cpu_has(X86_FEATURE_XSAVES)) init_fpstate.xsave.header.xcomp_bv = XCOMP_BV_COMPACTED_FORMAT | - xfeatures_mask; + xfeatures_mask_all; /* * Init all the features state with header.xfeatures being 0x0 @@ -474,7 +484,7 @@ int using_compacted_format(void) int validate_user_xstate_header(const struct xstate_header *hdr) { /* No unknown or supervisor features may be set */ - if (hdr->xfeatures & ~(xfeatures_mask & XFEATURE_MASK_USER_SUPPORTED)) + if (hdr->xfeatures & ~xfeatures_mask_user()) return -EINVAL; /* Userspace must use the uncompacted format */ @@ -609,7 +619,7 @@ static void do_extra_xstate_size_checks(void) /* - * Get total size of enabled xstates in XCR0/xfeatures_mask. + * Get total size of enabled xstates in XCR0 | IA32_XSS. * * Note the SDM's wording here. "sub-function 0" only enumerates * the size of the *user* states. If we use it to size a buffer @@ -699,7 +709,7 @@ static int __init init_xstate_size(void) */ static void fpu__init_disable_system_xstate(void) { - xfeatures_mask = 0; + xfeatures_mask_all = 0; cr4_clear_bits(X86_CR4_OSXSAVE); setup_clear_cpu_cap(X86_FEATURE_XSAVE); } @@ -734,16 +744,21 @@ void __init fpu__init_system_xstate(void) return; } + /* + * Find user xstates supported by the processor. + */ cpuid_count(XSTATE_CPUID, 0, &eax, &ebx, &ecx, &edx); - xfeatures_mask = eax + ((u64)edx << 32); + xfeatures_mask_all = eax + ((u64)edx << 32); - if ((xfeatures_mask & XFEATURE_MASK_FPSSE) != XFEATURE_MASK_FPSSE) { + /* Place supervisor features in xfeatures_mask_all here */ + if ((xfeatures_mask_user() & XFEATURE_MASK_FPSSE) != XFEATURE_MASK_FPSSE) { /* * This indicates that something really unexpected happened * with the enumeration. Disable XSAVE and try to continue * booting without it. This is too early to BUG(). */ - pr_err("x86/fpu: FP/SSE not present amongst the CPU's xstate features: 0x%llx.\n", xfeatures_mask); + pr_err("x86/fpu: FP/SSE not present amongst the CPU's xstate features: 0x%llx.\n", + xfeatures_mask_all); goto out_disable; } @@ -752,10 +767,10 @@ void __init fpu__init_system_xstate(void) */ for (i = 0; i < ARRAY_SIZE(xsave_cpuid_features); i++) { if (!boot_cpu_has(xsave_cpuid_features[i])) - xfeatures_mask &= ~BIT(i); + xfeatures_mask_all &= ~BIT_ULL(i); } - xfeatures_mask &= fpu__get_supported_xfeatures_mask(); + xfeatures_mask_all &= fpu__get_supported_xfeatures_mask(); /* Enable xstate instructions to be able to continue with initialization: */ fpu__init_cpu_xstate(); @@ -767,8 +782,7 @@ void __init fpu__init_system_xstate(void) * Update info used for ptrace frames; use standard-format size and no * supervisor xstates: */ - update_regset_xstate_info(fpu_user_xstate_size, - xfeatures_mask & XFEATURE_MASK_USER_SUPPORTED); + update_regset_xstate_info(fpu_user_xstate_size, xfeatures_mask_user()); fpu__init_prepare_fx_sw_frame(); setup_init_fpu_buf(); @@ -776,7 +790,7 @@ void __init fpu__init_system_xstate(void) print_xstate_offset_size(); pr_info("x86/fpu: Enabled xstate features 0x%llx, context size is %d bytes, using '%s' format.\n", - xfeatures_mask, + xfeatures_mask_all, fpu_kernel_xstate_size, boot_cpu_has(X86_FEATURE_XSAVES) ? "compacted" : "standard"); return; @@ -795,7 +809,7 @@ void fpu__resume_cpu(void) * Restore XCR0 on xsave capable CPUs: */ if (boot_cpu_has(X86_FEATURE_XSAVE)) - xsetbv(XCR_XFEATURE_ENABLED_MASK, xfeatures_mask); + xsetbv(XCR_XFEATURE_ENABLED_MASK, xfeatures_mask_user()); } /* @@ -840,10 +854,9 @@ void *get_xsave_addr(struct xregs_state *xsave, int xfeature_nr) /* * We should not ever be requesting features that we - * have not enabled. Remember that xfeatures_mask is - * what we write to the XCR0 register. + * have not enabled. */ - WARN_ONCE(!(xfeatures_mask & BIT_ULL(xfeature_nr)), + WARN_ONCE(!(xfeatures_mask_all & BIT_ULL(xfeature_nr)), "get of unsupported state"); /* * This assumes the last 'xsave*' instruction to @@ -996,7 +1009,7 @@ int copy_xstate_to_kernel(void *kbuf, struct xregs_state *xsave, unsigned int of */ memset(&header, 0, sizeof(header)); header.xfeatures = xsave->header.xfeatures; - header.xfeatures &= XFEATURE_MASK_USER_SUPPORTED; + header.xfeatures &= xfeatures_mask_user(); /* * Copy xregs_state->header: @@ -1080,7 +1093,7 @@ int copy_xstate_to_user(void __user *ubuf, struct xregs_state *xsave, unsigned i */ memset(&header, 0, sizeof(header)); header.xfeatures = xsave->header.xfeatures; - header.xfeatures &= XFEATURE_MASK_USER_SUPPORTED; + header.xfeatures &= xfeatures_mask_user(); /* * Copy xregs_state->header: -- 2.21.0