From: Jisheng Zhang <jszhang3@mail.ustc.edu.cn> To: Andreas Schwab <schwab@linux-m68k.org> Cc: "Paul Walmsley" <paul.walmsley@sifive.com>, "Palmer Dabbelt" <palmer@dabbelt.com>, "Albert Ou" <aou@eecs.berkeley.edu>, "Andrey Ryabinin" <ryabinin.a.a@gmail.com>, "Alexander Potapenko" <glider@google.com>, "Andrey Konovalov" <andreyknvl@gmail.com>, "Dmitry Vyukov" <dvyukov@google.com>, "Björn Töpel" <bjorn@kernel.org>, "Alexei Starovoitov" <ast@kernel.org>, "Daniel Borkmann" <daniel@iogearbox.net>, "Andrii Nakryiko" <andrii@kernel.org>, "Martin KaFai Lau" <kafai@fb.com>, "Song Liu" <songliubraving@fb.com>, "Yonghong Song" <yhs@fb.com>, "John Fastabend" <john.fastabend@gmail.com>, "KP Singh" <kpsingh@kernel.org>, "Luke Nelson" <luke.r.nels@gmail.com>, "Xi Wang" <xi.wang@gmail.com>, linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, kasan-dev@googlegroups.com, netdev@vger.kernel.org, bpf@vger.kernel.org Subject: Re: [PATCH 7/9] riscv: bpf: Avoid breaking W^X Date: Mon, 14 Jun 2021 01:05:46 +0800 [thread overview] Message-ID: <20210614010546.7a0d5584@xhacker> (raw) In-Reply-To: <87bl8cqrpv.fsf@igel.home> Hi, On Fri, 11 Jun 2021 18:41:16 +0200 Andreas Schwab <schwab@linux-m68k.org> wrote: > On Jun 12 2021, Jisheng Zhang wrote: > > > I reproduced an kernel panic with the defconfig on qemu, but I'm not sure whether > > this is the issue you saw, I will check. > > > > 0.161959] futex hash table entries: 512 (order: 3, 32768 bytes, linear) > > [ 0.167028] pinctrl core: initialized pinctrl subsystem > > [ 0.190727] Unable to handle kernel paging request at virtual address ffffffff81651bd8 > > [ 0.191361] Oops [#1] > > [ 0.191509] Modules linked in: > > [ 0.191814] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.13.0-rc5-default+ #3 > > [ 0.192179] Hardware name: riscv-virtio,qemu (DT) > > [ 0.192492] epc : __memset+0xc4/0xfc > > [ 0.192712] ra : skb_flow_dissector_init+0x22/0x86 > > Yes, that's the same. > > Andreas. > I think I found the root cause: commit 2bfc6cd81bd ("move kernel mapping outside of linear mapping") moves BPF JIT region after the kernel: #define BPF_JIT_REGION_START PFN_ALIGN((unsigned long)&_end) The &_end is unlikely aligned with PMD SIZE, so the front bpf jit region sits with kernel .data section in one PMD. But kenrel is mapped in PMD SIZE, so when bpf_jit_binary_lock_ro() is called to make the first bpf jit prog ROX, we will make part of kernel .data section RO too, so when we write, for example memset the .data section, MMU will trigger store page fault. To fix the issue, we need to make the bpf jit region PMD size aligned by either patch BPF_JIT_REGION_START to align on PMD size rather than PAGE SIZE, or something as below patch to move the BPF region before modules region: diff --git a/arch/riscv/include/asm/pgtable.h b/arch/riscv/include/asm/pgtable.h index 9469f464e71a..997b894edbc2 100644 --- a/arch/riscv/include/asm/pgtable.h +++ b/arch/riscv/include/asm/pgtable.h @@ -31,8 +31,8 @@ #define BPF_JIT_REGION_SIZE (SZ_128M) #ifdef CONFIG_64BIT /* KASLR should leave at least 128MB for BPF after the kernel */ -#define BPF_JIT_REGION_START PFN_ALIGN((unsigned long)&_end) -#define BPF_JIT_REGION_END (BPF_JIT_REGION_START + BPF_JIT_REGION_SIZE) +#define BPF_JIT_REGION_START (BPF_JIT_REGION_END - BPF_JIT_REGION_SIZE) +#define BPF_JIT_REGION_END (MODULES_VADDR) #else #define BPF_JIT_REGION_START (PAGE_OFFSET - BPF_JIT_REGION_SIZE) #define BPF_JIT_REGION_END (VMALLOC_END) @@ -40,8 +40,8 @@ /* Modules always live before the kernel */ #ifdef CONFIG_64BIT -#define MODULES_VADDR (PFN_ALIGN((unsigned long)&_end) - SZ_2G) #define MODULES_END (PFN_ALIGN((unsigned long)&_start)) +#define MODULES_VADDR (MODULES_END - SZ_128M) #endif can you please try it? Per my test, the issue is fixed. Thanks
WARNING: multiple messages have this Message-ID (diff)
From: Jisheng Zhang <jszhang3@mail.ustc.edu.cn> To: Andreas Schwab <schwab@linux-m68k.org> Cc: "Paul Walmsley" <paul.walmsley@sifive.com>, "Palmer Dabbelt" <palmer@dabbelt.com>, "Albert Ou" <aou@eecs.berkeley.edu>, "Andrey Ryabinin" <ryabinin.a.a@gmail.com>, "Alexander Potapenko" <glider@google.com>, "Andrey Konovalov" <andreyknvl@gmail.com>, "Dmitry Vyukov" <dvyukov@google.com>, "Björn Töpel" <bjorn@kernel.org>, "Alexei Starovoitov" <ast@kernel.org>, "Daniel Borkmann" <daniel@iogearbox.net>, "Andrii Nakryiko" <andrii@kernel.org>, "Martin KaFai Lau" <kafai@fb.com>, "Song Liu" <songliubraving@fb.com>, "Yonghong Song" <yhs@fb.com>, "John Fastabend" <john.fastabend@gmail.com>, "KP Singh" <kpsingh@kernel.org>, "Luke Nelson" <luke.r.nels@gmail.com>, "Xi Wang" <xi.wang@gmail.com>, linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, kasan-dev@googlegroups.com, netdev@vger.kernel.org, bpf@vger.kernel.org Subject: Re: [PATCH 7/9] riscv: bpf: Avoid breaking W^X Date: Mon, 14 Jun 2021 01:05:46 +0800 [thread overview] Message-ID: <20210614010546.7a0d5584@xhacker> (raw) In-Reply-To: <87bl8cqrpv.fsf@igel.home> Hi, On Fri, 11 Jun 2021 18:41:16 +0200 Andreas Schwab <schwab@linux-m68k.org> wrote: > On Jun 12 2021, Jisheng Zhang wrote: > > > I reproduced an kernel panic with the defconfig on qemu, but I'm not sure whether > > this is the issue you saw, I will check. > > > > 0.161959] futex hash table entries: 512 (order: 3, 32768 bytes, linear) > > [ 0.167028] pinctrl core: initialized pinctrl subsystem > > [ 0.190727] Unable to handle kernel paging request at virtual address ffffffff81651bd8 > > [ 0.191361] Oops [#1] > > [ 0.191509] Modules linked in: > > [ 0.191814] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.13.0-rc5-default+ #3 > > [ 0.192179] Hardware name: riscv-virtio,qemu (DT) > > [ 0.192492] epc : __memset+0xc4/0xfc > > [ 0.192712] ra : skb_flow_dissector_init+0x22/0x86 > > Yes, that's the same. > > Andreas. > I think I found the root cause: commit 2bfc6cd81bd ("move kernel mapping outside of linear mapping") moves BPF JIT region after the kernel: #define BPF_JIT_REGION_START PFN_ALIGN((unsigned long)&_end) The &_end is unlikely aligned with PMD SIZE, so the front bpf jit region sits with kernel .data section in one PMD. But kenrel is mapped in PMD SIZE, so when bpf_jit_binary_lock_ro() is called to make the first bpf jit prog ROX, we will make part of kernel .data section RO too, so when we write, for example memset the .data section, MMU will trigger store page fault. To fix the issue, we need to make the bpf jit region PMD size aligned by either patch BPF_JIT_REGION_START to align on PMD size rather than PAGE SIZE, or something as below patch to move the BPF region before modules region: diff --git a/arch/riscv/include/asm/pgtable.h b/arch/riscv/include/asm/pgtable.h index 9469f464e71a..997b894edbc2 100644 --- a/arch/riscv/include/asm/pgtable.h +++ b/arch/riscv/include/asm/pgtable.h @@ -31,8 +31,8 @@ #define BPF_JIT_REGION_SIZE (SZ_128M) #ifdef CONFIG_64BIT /* KASLR should leave at least 128MB for BPF after the kernel */ -#define BPF_JIT_REGION_START PFN_ALIGN((unsigned long)&_end) -#define BPF_JIT_REGION_END (BPF_JIT_REGION_START + BPF_JIT_REGION_SIZE) +#define BPF_JIT_REGION_START (BPF_JIT_REGION_END - BPF_JIT_REGION_SIZE) +#define BPF_JIT_REGION_END (MODULES_VADDR) #else #define BPF_JIT_REGION_START (PAGE_OFFSET - BPF_JIT_REGION_SIZE) #define BPF_JIT_REGION_END (VMALLOC_END) @@ -40,8 +40,8 @@ /* Modules always live before the kernel */ #ifdef CONFIG_64BIT -#define MODULES_VADDR (PFN_ALIGN((unsigned long)&_end) - SZ_2G) #define MODULES_END (PFN_ALIGN((unsigned long)&_start)) +#define MODULES_VADDR (MODULES_END - SZ_128M) #endif can you please try it? Per my test, the issue is fixed. Thanks _______________________________________________ linux-riscv mailing list linux-riscv@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-riscv
next prev parent reply other threads:[~2021-06-13 17:12 UTC|newest] Thread overview: 63+ messages / expand[flat|nested] mbox.gz Atom feed top 2021-03-29 18:21 [PATCH 0/9] riscv: improve self-protection Jisheng Zhang 2021-03-29 18:21 ` Jisheng Zhang 2021-03-29 18:22 ` [PATCH 1/9] riscv: add __init section marker to some functions Jisheng Zhang 2021-03-29 18:22 ` Jisheng Zhang 2021-03-29 18:22 ` [PATCH 2/9] riscv: Mark some global variables __ro_after_init Jisheng Zhang 2021-03-29 18:22 ` Jisheng Zhang 2021-03-29 18:23 ` [PATCH 3/9] riscv: Constify sys_call_table Jisheng Zhang 2021-03-29 18:23 ` Jisheng Zhang 2021-03-29 18:23 ` [PATCH 4/9] riscv: Constify sbi_ipi_ops Jisheng Zhang 2021-03-29 18:23 ` Jisheng Zhang 2021-03-29 18:24 ` [PATCH 5/9] riscv: kprobes: Implement alloc_insn_page() Jisheng Zhang 2021-03-29 18:24 ` Jisheng Zhang 2021-03-29 18:24 ` [PATCH 6/9] riscv: bpf: Move bpf_jit_alloc_exec() and bpf_jit_free_exec() to core Jisheng Zhang 2021-03-29 18:24 ` Jisheng Zhang 2021-03-29 20:41 ` Luke Nelson 2021-03-29 20:41 ` Luke Nelson 2021-03-29 18:25 ` [PATCH 7/9] riscv: bpf: Avoid breaking W^X Jisheng Zhang 2021-03-29 18:25 ` Jisheng Zhang 2021-03-30 16:39 ` kernel test robot 2021-06-11 14:10 ` Andreas Schwab 2021-06-11 14:10 ` Andreas Schwab 2021-06-11 16:23 ` Jisheng Zhang 2021-06-11 16:23 ` Jisheng Zhang 2021-06-11 16:41 ` Andreas Schwab 2021-06-11 16:41 ` Andreas Schwab 2021-06-13 17:05 ` Jisheng Zhang [this message] 2021-06-13 17:05 ` Jisheng Zhang 2021-06-13 19:50 ` Andreas Schwab 2021-06-13 19:50 ` Andreas Schwab 2021-06-14 16:49 ` [PATCH] riscv: Ensure BPF_JIT_REGION_START aligned with PMD size Jisheng Zhang 2021-06-14 16:49 ` Jisheng Zhang 2021-06-15 12:29 ` Daniel Borkmann 2021-06-15 12:29 ` Daniel Borkmann 2021-06-15 18:54 ` Alex Ghiti 2021-06-15 18:54 ` Alex Ghiti 2021-06-16 0:03 ` Jisheng Zhang 2021-06-16 0:03 ` Jisheng Zhang 2021-06-17 7:23 ` Alex Ghiti 2021-06-17 7:23 ` Alex Ghiti 2021-06-17 17:17 ` Jisheng Zhang 2021-06-17 17:17 ` Jisheng Zhang 2021-06-17 7:30 ` Palmer Dabbelt 2021-06-17 7:30 ` Palmer Dabbelt 2021-06-17 8:09 ` Alex Ghiti 2021-06-17 8:09 ` Alex Ghiti 2021-06-17 14:18 ` Alex Ghiti 2021-06-17 14:18 ` Alex Ghiti 2021-06-17 17:27 ` Jisheng Zhang 2021-06-17 17:27 ` Jisheng Zhang 2021-06-17 17:46 ` Jisheng Zhang 2021-06-17 17:46 ` Jisheng Zhang 2021-06-17 18:10 ` Jisheng Zhang 2021-06-17 18:10 ` Jisheng Zhang 2021-06-17 18:15 ` [PATCH v2] " Jisheng Zhang 2021-06-17 18:15 ` Jisheng Zhang 2021-06-18 6:48 ` Alex Ghiti 2021-06-18 6:48 ` Alex Ghiti 2021-03-29 18:25 ` [PATCH 8/9] riscv: module: Create module allocations without exec permissions Jisheng Zhang 2021-03-29 18:25 ` Jisheng Zhang 2021-03-29 18:26 ` [PATCH 9/9] riscv: Set ARCH_HAS_STRICT_MODULE_RWX if MMU Jisheng Zhang 2021-03-29 18:26 ` Jisheng Zhang 2021-04-23 1:48 ` [PATCH 0/9] riscv: improve self-protection Palmer Dabbelt 2021-04-23 1:48 ` Palmer Dabbelt
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=20210614010546.7a0d5584@xhacker \ --to=jszhang3@mail.ustc.edu.cn \ --cc=andreyknvl@gmail.com \ --cc=andrii@kernel.org \ --cc=aou@eecs.berkeley.edu \ --cc=ast@kernel.org \ --cc=bjorn@kernel.org \ --cc=bpf@vger.kernel.org \ --cc=daniel@iogearbox.net \ --cc=dvyukov@google.com \ --cc=glider@google.com \ --cc=john.fastabend@gmail.com \ --cc=kafai@fb.com \ --cc=kasan-dev@googlegroups.com \ --cc=kpsingh@kernel.org \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-riscv@lists.infradead.org \ --cc=luke.r.nels@gmail.com \ --cc=netdev@vger.kernel.org \ --cc=palmer@dabbelt.com \ --cc=paul.walmsley@sifive.com \ --cc=ryabinin.a.a@gmail.com \ --cc=schwab@linux-m68k.org \ --cc=songliubraving@fb.com \ --cc=xi.wang@gmail.com \ --cc=yhs@fb.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: 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.