From: Thomas Garnier <thgarnie@google.com> To: Brian Gerst <brgerst@gmail.com> Cc: "Herbert Xu" <herbert@gondor.apana.org.au>, "David S . Miller" <davem@davemloft.net>, "Thomas Gleixner" <tglx@linutronix.de>, "Ingo Molnar" <mingo@redhat.com>, "H . Peter Anvin" <hpa@zytor.com>, "Peter Zijlstra" <peterz@infradead.org>, "Josh Poimboeuf" <jpoimboe@redhat.com>, "Arnd Bergmann" <arnd@arndb.de>, "Matthias Kaehlcke" <mka@chromium.org>, "Boris Ostrovsky" <boris.ostrovsky@oracle.com>, "Juergen Gross" <jgross@suse.com>, "Paolo Bonzini" <pbonzini@redhat.com>, "Radim Krčmář" <rkrcmar@redhat.com>, "Joerg Roedel" <joro@8bytes.org>, "Andy Lutomirski" <luto@kernel.org>, "Borislav Petkov" <bp@alien8.de>, "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>, "Borislav Petkov" <bp@suse.de>, "Christian Borntraeger" <borntraeger@de.ibm.com>, "Rafael J . Wysocki" <rjw@rjwysocki.net>, "Len Brown" <len.brown@intel.com>, "Pavel Machek" <pavel@ucw.cz>, "Tejun Heo" <tj@kernel.org> Subject: Re: [RFC 16/22] x86/percpu: Adapt percpu for PIE support Date: Wed, 19 Jul 2017 11:26:32 -0700 [thread overview] Message-ID: <CAJcbSZFXrDZikh9P5M81ztkiMv7EhO4x0bzBdYE8RYC=HMZgqg@mail.gmail.com> (raw) In-Reply-To: <CAMzpN2g5YkFZTY7yfvG03QUKc-=asKMZbqke9g4e2oT_pgg7Yw@mail.gmail.com> On Tue, Jul 18, 2017 at 8:08 PM, Brian Gerst <brgerst@gmail.com> wrote: > On Tue, Jul 18, 2017 at 6:33 PM, Thomas Garnier <thgarnie@google.com> wrote: >> Perpcu uses a clever design where the .percu ELF section has a virtual >> address of zero and the relocation code avoid relocating specific >> symbols. It makes the code simple and easily adaptable with or without >> SMP support. >> >> This design is incompatible with PIE because generated code always try to >> access the zero virtual address relative to the default mapping address. >> It becomes impossible when KASLR is configured to go below -2G. This >> patch solves this problem by removing the zero mapping and adapting the GS >> base to be relative to the expected address. These changes are done only >> when PIE is enabled. The original implementation is kept as-is >> by default. > > The reason the per-cpu section is zero-based on x86-64 is to > workaround GCC hardcoding the stack protector canary at %gs:40. So > this patch is incompatible with CONFIG_STACK_PROTECTOR. Ok, that make sense. I don't want this feature to not work with CONFIG_CC_STACKPROTECTOR*. One way to fix that would be adding a GDT entry for gs so gs:40 points to the correct memory address and gs:[rip+XX] works correctly through the MSR. Given the separate discussion on mcmodel, I am going first to check if we can move from PIE to PIC with a mcmodel=small or medium that would remove the percpu change requirement. I tried before without success but I understand better percpu and other components so maybe I can make it work. Thanks a lot for the feedback. > > -- > Brian Gerst -- Thomas
WARNING: multiple messages have this Message-ID (diff)
From: Thomas Garnier <thgarnie@google.com> To: Brian Gerst <brgerst@gmail.com> Cc: "Herbert Xu" <herbert@gondor.apana.org.au>, "David S . Miller" <davem@davemloft.net>, "Thomas Gleixner" <tglx@linutronix.de>, "Ingo Molnar" <mingo@redhat.com>, "H . Peter Anvin" <hpa@zytor.com>, "Peter Zijlstra" <peterz@infradead.org>, "Josh Poimboeuf" <jpoimboe@redhat.com>, "Arnd Bergmann" <arnd@arndb.de>, "Matthias Kaehlcke" <mka@chromium.org>, "Boris Ostrovsky" <boris.ostrovsky@oracle.com>, "Juergen Gross" <jgross@suse.com>, "Paolo Bonzini" <pbonzini@redhat.com>, "Radim Krčmář" <rkrcmar@redhat.com>, "Joerg Roedel" <joro@8bytes.org>, "Andy Lutomirski" <luto@kernel.org>, "Borislav Petkov" <bp@alien8.de>, "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>, "Borislav Petkov" <bp@suse.de>, "Christian Borntraeger" <borntraeger@de.ibm.com>, "Rafael J . Wysocki" <rjw@rjwysocki.net>, "Len Brown" <len.brown@intel.com>, "Pavel Machek" <pavel@ucw.cz>, "Tejun Heo" <tj@kernel.org>, "Christoph Lameter" <cl@linux.com>, "Kees Cook" <keescook@chromium.org>, "Paul Gortmaker" <paul.gortmaker@windriver.com>, "Chris Metcalf" <cmetcalf@mellanox.com>, "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>, "Andrew Morton" <akpm@linux-foundation.org>, "Christopher Li" <sparse@chrisli.org>, "Dou Liyang" <douly.fnst@cn.fujitsu.com>, "Masahiro Yamada" <yamada.masahiro@socionext.com>, "Daniel Borkmann" <daniel@iogearbox.net>, "Markus Trippelsdorf" <markus@trippelsdorf.de>, "Peter Foley" <pefoley2@pefoley.com>, "Steven Rostedt" <rostedt@goodmis.org>, "Tim Chen" <tim.c.chen@linux.intel.com>, "Ard Biesheuvel" <ard.biesheuvel@linaro.org>, "Catalin Marinas" <catalin.marinas@arm.com>, "Matthew Wilcox" <mawilcox@microsoft.com>, "Michal Hocko" <mhocko@suse.com>, "Rob Landley" <rob@landley.net>, "Jiri Kosina" <jkosina@suse.cz>, "H . J . Lu" <hjl.tools@gmail.com>, "Paul Bolle" <pebolle@tiscali.nl>, "Baoquan He" <bhe@redhat.com>, "Daniel Micay" <danielmicay@gmail.com>, "the arch/x86 maintainers" <x86@kernel.org>, "linux-crypto@vger.kernel.org" <linux-crypto@vger.kernel.org>, "Linux Kernel Mailing List" <linux-kernel@vger.kernel.org>, xen-devel@lists.xenproject.org, "kvm list" <kvm@vger.kernel.org>, linux-pm <linux-pm@vger.kernel.org>, linux-arch <linux-arch@vger.kernel.org>, linux-sparse@vger.kernel.org, "Kernel Hardening" <kernel-hardening@lists.openwall.com> Subject: [kernel-hardening] Re: [RFC 16/22] x86/percpu: Adapt percpu for PIE support Date: Wed, 19 Jul 2017 11:26:32 -0700 [thread overview] Message-ID: <CAJcbSZFXrDZikh9P5M81ztkiMv7EhO4x0bzBdYE8RYC=HMZgqg@mail.gmail.com> (raw) In-Reply-To: <CAMzpN2g5YkFZTY7yfvG03QUKc-=asKMZbqke9g4e2oT_pgg7Yw@mail.gmail.com> On Tue, Jul 18, 2017 at 8:08 PM, Brian Gerst <brgerst@gmail.com> wrote: > On Tue, Jul 18, 2017 at 6:33 PM, Thomas Garnier <thgarnie@google.com> wrote: >> Perpcu uses a clever design where the .percu ELF section has a virtual >> address of zero and the relocation code avoid relocating specific >> symbols. It makes the code simple and easily adaptable with or without >> SMP support. >> >> This design is incompatible with PIE because generated code always try to >> access the zero virtual address relative to the default mapping address. >> It becomes impossible when KASLR is configured to go below -2G. This >> patch solves this problem by removing the zero mapping and adapting the GS >> base to be relative to the expected address. These changes are done only >> when PIE is enabled. The original implementation is kept as-is >> by default. > > The reason the per-cpu section is zero-based on x86-64 is to > workaround GCC hardcoding the stack protector canary at %gs:40. So > this patch is incompatible with CONFIG_STACK_PROTECTOR. Ok, that make sense. I don't want this feature to not work with CONFIG_CC_STACKPROTECTOR*. One way to fix that would be adding a GDT entry for gs so gs:40 points to the correct memory address and gs:[rip+XX] works correctly through the MSR. Given the separate discussion on mcmodel, I am going first to check if we can move from PIE to PIC with a mcmodel=small or medium that would remove the percpu change requirement. I tried before without success but I understand better percpu and other components so maybe I can make it work. Thanks a lot for the feedback. > > -- > Brian Gerst -- Thomas
next prev parent reply other threads:[~2017-07-19 18:26 UTC|newest] Thread overview: 151+ messages / expand[flat|nested] mbox.gz Atom feed top 2017-07-18 22:33 x86: PIE support and option to extend KASLR randomization Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 01/22] x86/crypto: Adapt assembly for PIE support Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` Thomas Garnier 2017-07-18 22:33 ` [RFC 02/22] x86: Use symbol name on bug table " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 03/22] x86: Use symbol name in jump " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 04/22] x86: Add macro to get symbol address " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 05/22] xen: Adapt assembly " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 06/22] kvm: " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-19 2:49 ` Brian Gerst 2017-07-19 2:49 ` Brian Gerst 2017-07-19 2:49 ` [kernel-hardening] " Brian Gerst 2017-07-19 15:40 ` Thomas Garnier 2017-07-19 15:40 ` [kernel-hardening] " Thomas Garnier 2017-07-19 22:27 ` H. Peter Anvin 2017-07-19 22:27 ` H. Peter Anvin 2017-07-19 22:27 ` [kernel-hardening] " H. Peter Anvin 2017-07-19 22:44 ` Thomas Garnier 2017-07-19 22:44 ` [kernel-hardening] " Thomas Garnier 2017-07-19 22:44 ` Thomas Garnier 2017-07-19 22:58 ` Ard Biesheuvel 2017-07-19 22:58 ` Ard Biesheuvel 2017-07-19 22:58 ` [kernel-hardening] " Ard Biesheuvel 2017-07-19 23:47 ` H. Peter Anvin 2017-07-19 23:47 ` H. Peter Anvin 2017-07-19 23:47 ` H. Peter Anvin 2017-07-19 23:47 ` H. Peter Anvin 2017-07-19 23:47 ` H. Peter Anvin 2017-07-19 23:47 ` H. Peter Anvin 2017-07-19 23:47 ` H. Peter Anvin 2017-07-19 15:40 ` Thomas Garnier 2017-07-18 22:33 ` [RFC 07/22] x86: relocate_kernel - " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-19 22:58 ` H. Peter Anvin 2017-07-19 22:58 ` [kernel-hardening] " H. Peter Anvin 2017-07-19 23:23 ` Thomas Garnier 2017-07-19 23:23 ` [kernel-hardening] " Thomas Garnier 2017-07-19 23:23 ` Thomas Garnier 2017-07-19 22:58 ` H. Peter Anvin 2017-07-18 22:33 ` [RFC 08/22] x86/entry/64: " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 09/22] x86: pm-trace - " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 10/22] x86/CPU: " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 11/22] x86/acpi: " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 12/22] x86/boot/64: " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 13/22] x86/power/64: " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-19 18:41 ` Pavel Machek 2017-07-19 18:41 ` [kernel-hardening] " Pavel Machek 2017-07-19 18:41 ` Pavel Machek 2017-07-18 22:33 ` [RFC 14/22] x86/paravirt: " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 15/22] x86/boot/64: Use _text in a global " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 16/22] x86/percpu: Adapt percpu " Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-19 3:08 ` Brian Gerst 2017-07-19 3:08 ` [kernel-hardening] " Brian Gerst 2017-07-19 18:26 ` Thomas Garnier [this message] 2017-07-19 18:26 ` Thomas Garnier 2017-07-19 23:33 ` H. Peter Anvin 2017-07-19 23:33 ` [kernel-hardening] " H. Peter Anvin 2017-07-20 2:21 ` H. Peter Anvin 2017-07-20 2:21 ` [kernel-hardening] " H. Peter Anvin 2017-07-20 3:03 ` H. Peter Anvin 2017-07-20 3:03 ` H. Peter Anvin 2017-07-20 3:03 ` [kernel-hardening] " H. Peter Anvin 2017-07-20 2:21 ` H. Peter Anvin 2017-07-20 14:26 ` Thomas Garnier 2017-07-20 14:26 ` [kernel-hardening] " Thomas Garnier 2017-08-02 16:42 ` Thomas Garnier 2017-08-02 16:42 ` [kernel-hardening] " Thomas Garnier 2017-08-02 16:56 ` Kees Cook 2017-08-02 16:56 ` [kernel-hardening] " Kees Cook 2017-08-02 18:05 ` Thomas Garnier 2017-08-02 18:05 ` Thomas Garnier 2017-08-02 18:05 ` [kernel-hardening] " Thomas Garnier 2017-08-02 16:56 ` Kees Cook 2017-08-02 16:42 ` Thomas Garnier 2017-07-20 14:26 ` Thomas Garnier 2017-07-19 23:33 ` H. Peter Anvin 2017-07-19 18:26 ` Thomas Garnier 2017-07-19 3:08 ` Brian Gerst 2017-07-18 22:33 ` [RFC 17/22] compiler: Option to default to hidden symbols Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 18/22] x86/relocs: Handle DYN relocations for PIE support Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 19/22] x86/pie: Add option to build the kernel as PIE for x86_64 Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 20/22] x86/relocs: Add option to generate 64-bit relocations Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-19 22:33 ` H. Peter Anvin 2017-07-19 22:33 ` H. Peter Anvin 2017-07-19 22:33 ` [kernel-hardening] " H. Peter Anvin 2017-07-19 22:47 ` Thomas Garnier 2017-07-19 22:47 ` [kernel-hardening] " Thomas Garnier 2017-07-19 23:08 ` H. Peter Anvin 2017-07-19 23:08 ` [kernel-hardening] " H. Peter Anvin 2017-07-19 23:25 ` Thomas Garnier 2017-07-19 23:25 ` Thomas Garnier 2017-07-19 23:25 ` [kernel-hardening] " Thomas Garnier 2017-07-19 23:45 ` H. Peter Anvin 2017-07-19 23:45 ` H. Peter Anvin 2017-07-19 23:45 ` H. Peter Anvin 2017-07-19 23:45 ` H. Peter Anvin 2017-07-19 23:45 ` H. Peter Anvin 2017-07-19 23:45 ` H. Peter Anvin 2017-07-19 23:45 ` H. Peter Anvin 2017-07-19 23:08 ` H. Peter Anvin 2017-07-19 22:47 ` Thomas Garnier 2017-07-18 22:33 ` [RFC 21/22] x86/module: Add support for mcmodel large and PLTs Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-19 1:35 ` H. Peter Anvin 2017-07-19 1:35 ` H. Peter Anvin 2017-07-19 1:35 ` [kernel-hardening] " H. Peter Anvin 2017-07-19 3:59 ` Brian Gerst 2017-07-19 3:59 ` Brian Gerst 2017-07-19 3:59 ` [kernel-hardening] " Brian Gerst 2017-07-19 15:58 ` Thomas Garnier 2017-07-19 15:58 ` Thomas Garnier 2017-07-19 15:58 ` [kernel-hardening] " Thomas Garnier 2017-07-19 17:34 ` Brian Gerst 2017-07-19 17:34 ` Brian Gerst 2017-07-19 17:34 ` [kernel-hardening] " Brian Gerst 2017-07-24 16:32 ` Thomas Garnier 2017-07-24 16:32 ` Thomas Garnier 2017-07-24 16:32 ` [kernel-hardening] " Thomas Garnier 2017-07-18 22:33 ` [RFC 22/22] x86/kaslr: Add option to extend KASLR range from 1GB to 3GB Thomas Garnier 2017-07-18 22:33 ` [kernel-hardening] " Thomas Garnier 2017-07-19 12:10 ` Baoquan He 2017-07-19 12:10 ` Baoquan He 2017-07-19 12:10 ` [kernel-hardening] " Baoquan He 2017-07-19 13:49 ` Baoquan He 2017-07-19 13:49 ` [kernel-hardening] " Baoquan He 2017-07-19 13:49 ` Baoquan He 2017-07-19 14:08 ` x86: PIE support and option to extend KASLR randomization Christopher Lameter 2017-07-19 14:08 ` Christopher Lameter 2017-07-19 14:08 ` [kernel-hardening] " Christopher Lameter 2017-07-19 19:21 ` Kees Cook 2017-07-19 19:21 ` Kees Cook 2017-07-19 19:21 ` [kernel-hardening] " Kees Cook
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='CAJcbSZFXrDZikh9P5M81ztkiMv7EhO4x0bzBdYE8RYC=HMZgqg@mail.gmail.com' \ --to=thgarnie@google.com \ --cc=arnd@arndb.de \ --cc=boris.ostrovsky@oracle.com \ --cc=borntraeger@de.ibm.com \ --cc=bp@alien8.de \ --cc=bp@suse.de \ --cc=brgerst@gmail.com \ --cc=davem@davemloft.net \ --cc=herbert@gondor.apana.org.au \ --cc=hpa@zytor.com \ --cc=jgross@suse.com \ --cc=joro@8bytes.org \ --cc=jpoimboe@redhat.com \ --cc=kirill.shutemov@linux.intel.com \ --cc=len.brown@intel.com \ --cc=luto@kernel.org \ --cc=mingo@redhat.com \ --cc=mka@chromium.org \ --cc=pavel@ucw.cz \ --cc=pbonzini@redhat.com \ --cc=peterz@infradead.org \ --cc=rjw@rjwysocki.net \ --cc=rkrcmar@redhat.com \ --cc=tglx@linutronix.de \ --cc=tj@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: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.