[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-Id: <20180118134800.711245485@infradead.org>
Date: Thu, 18 Jan 2018 14:48:00 +0100
From: Peter Zijlstra <peterz@...radead.org>
To: David Woodhouse <dwmw2@...radead.org>,
Thomas Gleixner <tglx@...utronix.de>,
Josh Poimboeuf <jpoimboe@...hat.com>
Cc: linux-kernel@...r.kernel.org, Dave Hansen <dave.hansen@...el.com>,
Ashok Raj <ashok.raj@...el.com>,
Tim Chen <tim.c.chen@...ux.intel.com>,
Andy Lutomirski <luto@...nel.org>,
Linus Torvalds <torvalds@...ux-foundation.org>,
Greg KH <gregkh@...uxfoundation.org>,
Andrea Arcangeli <aarcange@...hat.com>,
Andi Kleen <ak@...ux.intel.com>,
Arjan Van De Ven <arjan.van.de.ven@...el.com>,
Dan Williams <dan.j.williams@...el.com>,
Paolo Bonzini <pbonzini@...hat.com>,
Jun Nakajima <jun.nakajima@...el.com>,
Asit Mallick <asit.k.mallick@...el.com>,
Jason Baron <jbaron@...mai.com>,
Peter Zijlstra <peterz@...radead.org>
Subject: [PATCH 00/35] jump_label, objtool, IBRS and IBPB
Lots of patches..
They include:
- objtool validation of jump_label/static_cpu_has
Allows asserting that the code block following a jump_label/static_cpu_has
is indeed unconditional. Ensures GCC doesn't generate particularly stupid
code which would re-insert a dynamic test.
- objtool validation of retpoline
Looks for indirect JMP/CALL sites when build with a retpoline enabled
compiler. Has already spotted a bunch of sites that need fixing, see
below.
- makes x86 hard rely on asm-goto to ensure we can indeed use static_cpu_has
to avoid dynamic branches (and thus speculation).
- The IBRS/IBPB patches from Thomas that use static_cpu_has()
These hard rely on the above; we must not speculate across the IBRS/IBPB
MSR writes otherwise that would totally defeat the point. Prior patches had
LFENCE crud in the else-clause, which then makes the primitives
unconditionally expensive.
- Rebased the IBRS/IBPB-KVM patches from Ashok on top
- Random odd fixes for various things encountered while doing the above.
Please have a look and sorry for this many patches.
---
Output of a x86_64-allmodconfig -KCOV -KASAN build:
arch/x86/entry/.tmp_entry_64.o: warning: objtool: .entry.text+0x1cb2: indirect call found in RETPOLINE build
arch/x86/entry/.tmp_entry_64.o: warning: objtool: .entry.text+0x1cc7: indirect call found in RETPOLINE build
arch/x86/hyperv/.tmp_mmu.o: warning: objtool: hyperv_flush_tlb_others()+0x30c: indirect call found in RETPOLINE build
arch/x86/hyperv/.tmp_mmu.o: warning: objtool: hyperv_flush_tlb_others()+0x3b0: indirect call found in RETPOLINE build
arch/x86/hyperv/.tmp_mmu.o: warning: objtool: hyperv_flush_tlb_others_ex()+0x3a1: indirect call found in RETPOLINE build
arch/x86/hyperv/.tmp_mmu.o: warning: objtool: hyperv_flush_tlb_others_ex()+0x45c: indirect call found in RETPOLINE build
arch/x86/xen/.tmp_multicalls.o: warning: objtool: xen_mc_flush()+0x1da: indirect call found in RETPOLINE build
arch/x86/kvm/.tmp_emulate.o: warning: objtool: fastop()+0x54: indirect call found in RETPOLINE build
arch/x86/kvm/.tmp_emulate.o: warning: objtool: em_loop()+0xcc: indirect call found in RETPOLINE build
arch/x86/kvm/.tmp_emulate.o: warning: objtool: x86_emulate_insn()+0xbd6: indirect call found in RETPOLINE build
arch/x86/kvm/.tmp_emulate.o: warning: objtool: x86_emulate_insn()+0xc1a: indirect call found in RETPOLINE build
arch/x86/kvm/.tmp_emulate.o: warning: objtool: x86_emulate_insn()+0xc66: indirect call found in RETPOLINE build
arch/x86/kvm/.tmp_vmx.o: warning: objtool: vmx_handle_external_intr()+0x50: indirect call found in RETPOLINE build
arch/x86/mm/.tmp_mem_encrypt_boot.o: warning: objtool: sme_encrypt_execute()+0x48: indirect call found in RETPOLINE build
drivers/hv/.tmp_hv.o: warning: objtool: hv_post_message()+0x72: indirect call found in RETPOLINE build
drivers/hv/.tmp_connection.o: warning: objtool: vmbus_set_event()+0x33: indirect call found in RETPOLINE build
drivers/pci/host/.tmp_pci-hyperv.o: warning: objtool: hv_irq_unmask()+0x22b: indirect call found in RETPOLINE build
drivers/xen/.tmp_privcmd.o: warning: objtool: privcmd_ioctl()+0xcf: indirect call found in RETPOLINE build
drivers/watchdog/.tmp_hpwdt.o: warning: objtool: .text+0x24: indirect call found in RETPOLINE build
Powered by blists - more mailing lists