From: Charlie Jenkins <charlie@rivosinc.com>
To: Deepak Gupta <debug@rivosinc.com>
Cc: paul.walmsley@sifive.com, rick.p.edgecombe@intel.com,
broonie@kernel.org, Szabolcs.Nagy@arm.com, kito.cheng@sifive.com,
keescook@chromium.org, ajones@ventanamicro.com,
conor.dooley@microchip.com, cleger@rivosinc.com,
atishp@atishpatra.org, alex@ghiti.fr, bjorn@rivosinc.com,
alexghiti@rivosinc.com, samuel.holland@sifive.com,
conor@kernel.org, linux-doc@vger.kernel.org,
linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org,
devicetree@vger.kernel.org, linux-mm@kvack.org,
linux-arch@vger.kernel.org, linux-kselftest@vger.kernel.org,
corbet@lwn.net, palmer@dabbelt.com, aou@eecs.berkeley.edu,
robh+dt@kernel.org, krzysztof.kozlowski+dt@linaro.org,
oleg@redhat.com, akpm@linux-foundation.org, arnd@arndb.de,
ebiederm@xmission.com, Liam.Howlett@oracle.com, vbabka@suse.cz,
lstoakes@gmail.com, shuah@kernel.org, brauner@kernel.org,
andy.chiu@sifive.com, jerry.shih@sifive.com,
hankuan.chen@sifive.com, greentime.hu@sifive.com,
evan@rivosinc.com, xiao.w.wang@intel.com,
apatel@ventanamicro.com, mchitale@ventanamicro.com,
dbarboza@ventanamicro.com, sameo@rivosinc.com,
shikemeng@huaweicloud.com, willy@infradead.org,
vincent.chen@sifive.com, guoren@kernel.org,
samitolvanen@google.com, songshuaishuai@tinylab.org,
gerg@kernel.org, heiko@sntech.de, bhe@redhat.com,
jeeheng.sia@starfivetech.com, cyy@cyyself.name,
maskray@google.com, ancientmodern4@gmail.com,
mathis.salmen@matsal.de, cuiyunhui@bytedance.com,
bgray@linux.ibm.com, mpe@ellerman.id.au, baruch@tkos.co.il,
alx@kernel.org, david@redhat.com, catalin.marinas@arm.com,
revest@chromium.org, josh@joshtriplett.org, shr@devkernel.io,
deller@gmx.de, omosnace@redhat.com, ojeda@kernel.org,
jhubbard@nvidia.com
Subject: Re: [PATCH v3 10/29] riscv/mm : ensure PROT_WRITE leads to VM_READ | VM_WRITE
Date: Mon, 13 May 2024 14:26:31 -0700 [thread overview]
Message-ID: <ZkKFh2YJh3AcUTfN@ghost> (raw)
In-Reply-To: <ZkJe3ivq7m4NptHd@debug.ba.rivosinc.com>
On Mon, May 13, 2024 at 11:41:34AM -0700, Deepak Gupta wrote:
> On Mon, May 13, 2024 at 11:36:49AM -0700, Charlie Jenkins wrote:
> > On Mon, May 13, 2024 at 10:47:25AM -0700, Deepak Gupta wrote:
> > > On Fri, May 10, 2024 at 02:02:54PM -0700, Charlie Jenkins wrote:
> > > > On Wed, Apr 03, 2024 at 04:34:58PM -0700, Deepak Gupta wrote:
> > > > > `arch_calc_vm_prot_bits` is implemented on risc-v to return VM_READ |
> > > > > VM_WRITE if PROT_WRITE is specified. Similarly `riscv_sys_mmap` is
> > > > > updated to convert all incoming PROT_WRITE to (PROT_WRITE | PROT_READ).
> > > > > This is to make sure that any existing apps using PROT_WRITE still work.
> > > > >
> > > > > Earlier `protection_map[VM_WRITE]` used to pick read-write PTE encodings.
> > > > > Now `protection_map[VM_WRITE]` will always pick PAGE_SHADOWSTACK PTE
> > > > > encodings for shadow stack. Above changes ensure that existing apps
> > > > > continue to work because underneath kernel will be picking
> > > > > `protection_map[VM_WRITE|VM_READ]` PTE encodings.
> > > > >
> > > > > Signed-off-by: Deepak Gupta <debug@rivosinc.com>
> > > > > ---
> > > > > arch/riscv/include/asm/mman.h | 24 ++++++++++++++++++++++++
> > > > > arch/riscv/include/asm/pgtable.h | 1 +
> > > > > arch/riscv/kernel/sys_riscv.c | 11 +++++++++++
> > > > > arch/riscv/mm/init.c | 2 +-
> > > > > mm/mmap.c | 1 +
> > > > > 5 files changed, 38 insertions(+), 1 deletion(-)
> > > > > create mode 100644 arch/riscv/include/asm/mman.h
> > > > >
> > > > > diff --git a/arch/riscv/include/asm/mman.h b/arch/riscv/include/asm/mman.h
> > > > > new file mode 100644
> > > > > index 000000000000..ef9fedf32546
> > > > > --- /dev/null
> > > > > +++ b/arch/riscv/include/asm/mman.h
> > > > > @@ -0,0 +1,24 @@
> > > > > +/* SPDX-License-Identifier: GPL-2.0 */
> > > > > +#ifndef __ASM_MMAN_H__
> > > > > +#define __ASM_MMAN_H__
> > > > > +
> > > > > +#include <linux/compiler.h>
> > > > > +#include <linux/types.h>
> > > > > +#include <uapi/asm/mman.h>
> > > > > +
> > > > > +static inline unsigned long arch_calc_vm_prot_bits(unsigned long prot,
> > > > > + unsigned long pkey __always_unused)
> > > > > +{
> > > > > + unsigned long ret = 0;
> > > > > +
> > > > > + /*
> > > > > + * If PROT_WRITE was specified, force it to VM_READ | VM_WRITE.
> > > > > + * Only VM_WRITE means shadow stack.
> > > > > + */
> > > > > + if (prot & PROT_WRITE)
> > > > > + ret = (VM_READ | VM_WRITE);
> > > > > + return ret;
> > > > > +}
> > > > > +#define arch_calc_vm_prot_bits(prot, pkey) arch_calc_vm_prot_bits(prot, pkey)
> > > > > +
> > > > > +#endif /* ! __ASM_MMAN_H__ */
> > > > > diff --git a/arch/riscv/include/asm/pgtable.h b/arch/riscv/include/asm/pgtable.h
> > > > > index 6066822e7396..4d5983bc6766 100644
> > > > > --- a/arch/riscv/include/asm/pgtable.h
> > > > > +++ b/arch/riscv/include/asm/pgtable.h
> > > > > @@ -184,6 +184,7 @@ extern struct pt_alloc_ops pt_ops __initdata;
> > > > > #define PAGE_READ_EXEC __pgprot(_PAGE_BASE | _PAGE_READ | _PAGE_EXEC)
> > > > > #define PAGE_WRITE_EXEC __pgprot(_PAGE_BASE | _PAGE_READ | \
> > > > > _PAGE_EXEC | _PAGE_WRITE)
> > > > > +#define PAGE_SHADOWSTACK __pgprot(_PAGE_BASE | _PAGE_WRITE)
> > > > >
> > > > > #define PAGE_COPY PAGE_READ
> > > > > #define PAGE_COPY_EXEC PAGE_READ_EXEC
> > > > > diff --git a/arch/riscv/kernel/sys_riscv.c b/arch/riscv/kernel/sys_riscv.c
> > > > > index f1c1416a9f1e..846c36b1b3d5 100644
> > > > > --- a/arch/riscv/kernel/sys_riscv.c
> > > > > +++ b/arch/riscv/kernel/sys_riscv.c
> > > > > @@ -8,6 +8,8 @@
> > > > > #include <linux/syscalls.h>
> > > > > #include <asm/cacheflush.h>
> > > > > #include <asm-generic/mman-common.h>
> > > > > +#include <vdso/vsyscall.h>
> > > > > +#include <asm/mman.h>
> > > > >
> > > > > static long riscv_sys_mmap(unsigned long addr, unsigned long len,
> > > > > unsigned long prot, unsigned long flags,
> > > > > @@ -17,6 +19,15 @@ static long riscv_sys_mmap(unsigned long addr, unsigned long len,
> > > > > if (unlikely(offset & (~PAGE_MASK >> page_shift_offset)))
> > > > > return -EINVAL;
> > > > >
> > > > > + /*
> > > > > + * If only PROT_WRITE is specified then extend that to PROT_READ
> > > > > + * protection_map[VM_WRITE] is now going to select shadow stack encodings.
> > > > > + * So specifying PROT_WRITE actually should select protection_map [VM_WRITE | VM_READ]
> > > > > + * If user wants to create shadow stack then they should use `map_shadow_stack` syscall.
> > > > > + */
> > > > > + if (unlikely((prot & PROT_WRITE) && !(prot & PROT_READ)))
> > > >
> > > > The comments says that this should extend to PROT_READ if only
> > > > PROT_WRITE is specified. This condition instead is checking if
> > > > PROT_WRITE is selected but PROT_READ is not. If prot is (VM_EXEC |
> > > > VM_WRITE) then it would be extended to (VM_EXEC | VM_WRITE | VM_READ).
> > > > This will not currently cause any issues because these both map to the
> > > > same value in the protection_map PAGE_COPY_EXEC, however this seems to
> > > > be not the intention of this change.
> > > >
> > > > prot == PROT_WRITE better suits the condition explained in the comment.
> > >
> > > If someone specifies this (PROT_EXEC | PROT_WRITE) today, it works because
> > > of the way permissions are setup in `protection_map`. On risc-v there is no
> > > way to have a page which is execute and write only. So expectation is that
> > > if some apps were using `PROT_EXEC | PROT_WRITE` today, they were working
> > > because internally it was translating to read, write and execute on page
> > > permissions level. This patch make sure that, it stays same from page
> > > permissions perspective.
> > >
> > > If someone was using PROT_EXEC, it may translate to execute only and this change
> > > doesn't impact that.
> > >
> > > Patch simply looks for presence of `PROT_WRITE` and absence of `PROT_READ` in
> > > protection flags and if that condition is satisfied, it assumes that caller assumed
> > > page is going to be read allowed as well.
> >
> > The purpose of this change is for compatibility with shadow stack pages
> > but this affects flags for pages that are not shadow stack pages.
> > Adding PROT_READ to the other cases is redundant as protection_map
> > already handles that mapping. Permissions being strictly PROT_WRITE is
> > the only case that needs to be handled, and is the only case that is
> > called out in the commit message and in the comment.
>
> Yeah that's fine.
> I can change the commit message or just strictly check for PROT_WRITE.
> It doesn't change bottomline, I am fine with either option.
>
> Let me know your preference.
I would prefer the strict check. This is not critical though so I will
support whatever you decide!
- Charlie
>
> >
> > - Charlie
> >
> > >
> > >
> > > >
> > > > > + prot |= PROT_READ;
> > > > > +
> > > > > return ksys_mmap_pgoff(addr, len, prot, flags, fd,
> > > > > offset >> (PAGE_SHIFT - page_shift_offset));
> > > > > }
> > > > > diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c
> > > > > index fa34cf55037b..98e5ece4052a 100644
> > > > > --- a/arch/riscv/mm/init.c
> > > > > +++ b/arch/riscv/mm/init.c
> > > > > @@ -299,7 +299,7 @@ pgd_t early_pg_dir[PTRS_PER_PGD] __initdata __aligned(PAGE_SIZE);
> > > > > static const pgprot_t protection_map[16] = {
> > > > > [VM_NONE] = PAGE_NONE,
> > > > > [VM_READ] = PAGE_READ,
> > > > > - [VM_WRITE] = PAGE_COPY,
> > > > > + [VM_WRITE] = PAGE_SHADOWSTACK,
> > > > > [VM_WRITE | VM_READ] = PAGE_COPY,
> > > > > [VM_EXEC] = PAGE_EXEC,
> > > > > [VM_EXEC | VM_READ] = PAGE_READ_EXEC,
> > > > > diff --git a/mm/mmap.c b/mm/mmap.c
> > > > > index d89770eaab6b..57a974f49b00 100644
> > > > > --- a/mm/mmap.c
> > > > > +++ b/mm/mmap.c
> > > > > @@ -47,6 +47,7 @@
> > > > > #include <linux/oom.h>
> > > > > #include <linux/sched/mm.h>
> > > > > #include <linux/ksm.h>
> > > > > +#include <linux/processor.h>
> > > >
> > > > It doesn't seem like this is necessary for this patch.
> > >
> > > Thanks. Yeah it looks like I forgot to remove this over the churn.
> > > Will fix it.
> > >
> > > >
> > > > - Charlie
> > > >
> > > > >
> > > > > #include <linux/uaccess.h>
> > > > > #include <asm/cacheflush.h>
> > > > > --
> > > > > 2.43.2
> > > > >
next prev parent reply other threads:[~2024-05-13 21:26 UTC|newest]
Thread overview: 79+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-04-03 23:34 [PATCH v3 00/29] riscv control-flow integrity for usermode Deepak Gupta
2024-04-03 23:34 ` [PATCH v3 01/29] riscv: envcfg save and restore on task switching Deepak Gupta
2024-05-09 0:10 ` Charlie Jenkins
2024-05-09 19:00 ` Deepak Gupta
2024-04-03 23:34 ` [PATCH v3 02/29] riscv: define default value for envcfg for task Deepak Gupta
2024-05-10 22:33 ` Charlie Jenkins
2024-05-13 18:33 ` Deepak Gupta
2024-04-03 23:34 ` [PATCH v3 03/29] riscv/Kconfig: enable HAVE_EXIT_THREAD for riscv Deepak Gupta
2024-05-10 22:36 ` Charlie Jenkins
2024-04-03 23:34 ` [PATCH v3 04/29] riscv: zicfilp / zicfiss in dt-bindings (extensions.yaml) Deepak Gupta
2024-04-10 11:58 ` Rob Herring
2024-04-10 21:37 ` Deepak Gupta
2024-04-15 19:41 ` Rob Herring
[not found] ` <Zh6c0FH2OvrfDLje@debug.ba.rivosinc.com>
2024-05-09 18:14 ` Conor Dooley
2024-05-09 18:46 ` Deepak Gupta
2024-05-09 20:32 ` Conor Dooley
2024-05-09 23:26 ` Deepak Gupta
2024-04-03 23:34 ` [PATCH v3 05/29] riscv: zicfiss / zicfilp enumeration Deepak Gupta
2024-05-09 0:00 ` Andy Chiu
2024-05-09 0:07 ` Charlie Jenkins
2024-04-03 23:34 ` [PATCH v3 06/29] riscv: zicfiss / zicfilp extension csr and bit definitions Deepak Gupta
2024-05-10 22:37 ` Charlie Jenkins
2024-04-03 23:34 ` [PATCH v3 07/29] riscv: usercfi state for task and save/restore of CSR_SSP on trap entry/exit Deepak Gupta
2024-05-10 22:51 ` Charlie Jenkins
2024-04-03 23:34 ` [PATCH v3 08/29] mm: Define VM_SHADOW_STACK for RISC-V Deepak Gupta
[not found] ` <8fb37319-288c-4f77-9cd7-92f17bb567ee@redhat.com>
2024-04-04 19:04 ` Mark Brown
2024-04-04 19:15 ` David Hildenbrand
2024-04-04 19:21 ` Deepak Gupta
2024-04-03 23:34 ` [PATCH v3 09/29] mm: abstract shadow stack vma behind `vma_is_shadow_stack` Deepak Gupta
[not found] ` <c438ea3a-24bc-470b-a2eb-6e7517bd4362@redhat.com>
2024-04-04 21:39 ` Deepak Gupta
2024-04-03 23:34 ` [PATCH v3 10/29] riscv/mm : ensure PROT_WRITE leads to VM_READ | VM_WRITE Deepak Gupta
2024-05-10 21:02 ` Charlie Jenkins
2024-05-13 17:47 ` Deepak Gupta
2024-05-13 18:36 ` Charlie Jenkins
2024-05-13 18:41 ` Deepak Gupta
2024-05-13 21:26 ` Charlie Jenkins [this message]
2024-05-12 16:24 ` Alexandre Ghiti
2024-05-13 18:29 ` Deepak Gupta
2024-04-03 23:34 ` [PATCH v3 11/29] riscv mm: manufacture shadow stack pte Deepak Gupta
2024-05-12 16:26 ` Alexandre Ghiti
2024-04-03 23:35 ` [PATCH v3 12/29] riscv mmu: teach pte_mkwrite to manufacture shadow stack PTEs Deepak Gupta
2024-05-12 16:28 ` Alexandre Ghiti
2024-05-13 17:33 ` Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 13/29] riscv mmu: write protect and shadow stack Deepak Gupta
2024-05-12 16:31 ` Alexandre Ghiti
2024-05-13 17:32 ` Deepak Gupta
2024-05-23 14:59 ` Alexandre Ghiti
2024-05-24 7:16 ` Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 14/29] riscv/mm: Implement map_shadow_stack() syscall Deepak Gupta
2024-05-12 16:50 ` Alexandre Ghiti
2024-05-13 17:25 ` Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 15/29] riscv/shstk: If needed allocate a new shadow stack on clone Deepak Gupta
2024-05-12 17:05 ` Alexandre Ghiti
2024-05-13 17:10 ` Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 16/29] prctl: arch-agnostic prctl for shadow stack Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 17/29] prctl: arch-agnostic prctl for indirect branch tracking Deepak Gupta
2024-05-10 23:29 ` Charlie Jenkins
2024-05-13 18:31 ` Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 18/29] riscv: Implements arch agnostic shadow stack prctls Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 19/29] riscv: Implements arch agnostic indirect branch tracking prctls Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 20/29] riscv/kernel: update __show_regs to print shadow stack register Deepak Gupta
2024-05-12 17:10 ` Alexandre Ghiti
2024-04-03 23:35 ` [PATCH v3 21/29] riscv/traps: Introduce software check exception Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 22/29] riscv sigcontext: adding cfi state field in sigcontext Deepak Gupta
2024-05-24 9:46 ` Andy Chiu
2024-05-24 19:11 ` Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 23/29] riscv signal: Save and restore of shadow stack for signal Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 24/29] riscv/ptrace: riscv cfi status and state via ptrace and in core files Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 25/29] riscv/hwprobe: zicfilp / zicfiss enumeration in hwprobe Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 26/29] riscv: create a config for shadow stack and landing pad instr support Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 27/29] riscv: Documentation for landing pad / indirect branch tracking Deepak Gupta
2024-05-10 20:30 ` Charlie Jenkins
2024-05-13 17:07 ` Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 28/29] riscv: Documentation for shadow stack on riscv Deepak Gupta
2024-04-03 23:35 ` [PATCH v3 29/29] kselftest/riscv: kselftest for user mode cfi Deepak Gupta
2024-05-09 18:21 ` Charlie Jenkins
2024-05-09 19:16 ` Deepak Gupta
2024-05-10 1:20 ` Charlie Jenkins
2024-05-09 0:33 ` [PATCH v3 00/29] riscv control-flow integrity for usermode Charlie Jenkins
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=ZkKFh2YJh3AcUTfN@ghost \
--to=charlie@rivosinc.com \
--cc=Liam.Howlett@oracle.com \
--cc=Szabolcs.Nagy@arm.com \
--cc=ajones@ventanamicro.com \
--cc=akpm@linux-foundation.org \
--cc=alex@ghiti.fr \
--cc=alexghiti@rivosinc.com \
--cc=alx@kernel.org \
--cc=ancientmodern4@gmail.com \
--cc=andy.chiu@sifive.com \
--cc=aou@eecs.berkeley.edu \
--cc=apatel@ventanamicro.com \
--cc=arnd@arndb.de \
--cc=atishp@atishpatra.org \
--cc=baruch@tkos.co.il \
--cc=bgray@linux.ibm.com \
--cc=bhe@redhat.com \
--cc=bjorn@rivosinc.com \
--cc=brauner@kernel.org \
--cc=broonie@kernel.org \
--cc=catalin.marinas@arm.com \
--cc=cleger@rivosinc.com \
--cc=conor.dooley@microchip.com \
--cc=conor@kernel.org \
--cc=corbet@lwn.net \
--cc=cuiyunhui@bytedance.com \
--cc=cyy@cyyself.name \
--cc=david@redhat.com \
--cc=dbarboza@ventanamicro.com \
--cc=debug@rivosinc.com \
--cc=deller@gmx.de \
--cc=devicetree@vger.kernel.org \
--cc=ebiederm@xmission.com \
--cc=evan@rivosinc.com \
--cc=gerg@kernel.org \
--cc=greentime.hu@sifive.com \
--cc=guoren@kernel.org \
--cc=hankuan.chen@sifive.com \
--cc=heiko@sntech.de \
--cc=jeeheng.sia@starfivetech.com \
--cc=jerry.shih@sifive.com \
--cc=jhubbard@nvidia.com \
--cc=josh@joshtriplett.org \
--cc=keescook@chromium.org \
--cc=kito.cheng@sifive.com \
--cc=krzysztof.kozlowski+dt@linaro.org \
--cc=linux-arch@vger.kernel.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-riscv@lists.infradead.org \
--cc=lstoakes@gmail.com \
--cc=maskray@google.com \
--cc=mathis.salmen@matsal.de \
--cc=mchitale@ventanamicro.com \
--cc=mpe@ellerman.id.au \
--cc=ojeda@kernel.org \
--cc=oleg@redhat.com \
--cc=omosnace@redhat.com \
--cc=palmer@dabbelt.com \
--cc=paul.walmsley@sifive.com \
--cc=revest@chromium.org \
--cc=rick.p.edgecombe@intel.com \
--cc=robh+dt@kernel.org \
--cc=sameo@rivosinc.com \
--cc=samitolvanen@google.com \
--cc=samuel.holland@sifive.com \
--cc=shikemeng@huaweicloud.com \
--cc=shr@devkernel.io \
--cc=shuah@kernel.org \
--cc=songshuaishuai@tinylab.org \
--cc=vbabka@suse.cz \
--cc=vincent.chen@sifive.com \
--cc=willy@infradead.org \
--cc=xiao.w.wang@intel.com \
/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
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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 read-only IMAP folder(s) and NNTP newsgroup(s).