linux-kernel - Re: [PATCH v2] x86/cpu: replacing the open-coded shift with BIT(x)

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <eed4512b-ac6e-bcec-2f66-7ab45aef9be9@huawei.com>
Date:   Tue, 1 Nov 2022 21:37:02 +0800
From:   cuigaosheng <cuigaosheng1@...wei.com>
To:     "Jason A. Donenfeld" <Jason@...c4.com>
CC:     <tglx@...utronix.de>, <mingo@...hat.com>, <bp@...en8.de>,
        <dave.hansen@...ux.intel.com>, <x86@...nel.org>, <hpa@...or.com>,
        <puwen@...on.cn>, <TonyWWang-oc@...oxin.com>,
        <peterz@...radead.org>, <gregkh@...uxfoundation.org>,
        <andrew.cooper3@...rix.com>, <tony.luck@...el.com>,
        <mario.limonciello@....com>, <pawan.kumar.gupta@...ux.intel.com>,
        <chenyi.qiang@...el.com>, <rdunlap@...radead.org>,
        <jithu.joseph@...el.com>, <rafael.j.wysocki@...el.com>,
        <paulmck@...nel.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2] x86/cpu: replacing the open-coded shift with BIT(x)

> Others might feel differently and that's fine, but I always found the
> BIT() thing so much less clear than doing 1<<n, which is not only a
> pattern that I recognize as builtin to my brain, but also provides a
> direct description of what's happening, "shift a 1 over n times",
> leaving no off-by-one ambiguity about it. If anything I'd like to see
> the BIT() macro expanded throughout and then removed entirely.
>
> Probably just me though. You can safely ignore my opinion.

Thanks for taking time to review the patch, I submit the patch to remove
the UBSAN warning, even it's not a bug, for example, when I am testing the
kernel, I get some logs as follows, maybe it's better to avoid this?

> [ 0.951719][ T0] 
> ================================================================================ 
> 215 [ 0.953146][ T0] UBSAN: shift-out-of-bounds in mm/shmem.c:3749:18 
> 216 [ 0.953863][ T0] left shift of 1 by 31 places cannot be 
> represented in type 'int' 217 [ 0.955067][ T0] CPU: 0 PID: 0 Comm: 
> swapper/0 Not tainted 6.1.0-rc2-00062-ga970174d7a10 #5 218 [ 
> 0.956400][ T0] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), 
> BIOS rel-1.13.0-48-gd9c812dda519-prebuilt.qemu.org 04/01/2014 219 [ 
> 0.958278][ T0] Call Trace: 220 [ 0.958777][ T0] <TASK> 221 [ 
> 0.959224][ T0] dump_stack_lvl+0x8d/0xcf 222 [ 0.959922][ T0] 
> ubsan_epilogue+0xa/0x44 223 [ 0.960599][ T0] 
> __ubsan_handle_shift_out_of_bounds+0x1e7/0x208 224 [ 0.961575][ T0] ? 
> __kmem_cache_alloc_node+0x167/0x290 225 [ 0.962434][ T0] ? 
> shmem_fill_super+0x2e/0x2e0 226 [ 0.963187][ T0] ? 
> rcu_read_lock_held_common+0x9/0x40 227 [ 0.963857][ T0] ? 
> shmem_alloc_hugefolio+0x110/0x110 228 [ 0.963857][ T0] ? 
> shmem_fill_super+0x2cc/0x2e0 229 [ 0.963857][ T0] 
> shmem_fill_super+0x2cc/0x2e0 230 [ 0.963857][ T0] 
> vfs_get_super+0x78/0x160 231 [ 0.963857][ T0] vfs_get_tree+0x28/0x100 
> 232 [ 0.963857][ T0] fc_mount+0x12/0x60 233 [ 0.963857][ T0] 
> vfs_kern_mount.part.38+0xa5/0xc0 234 [ 0.963857][ T0] 
> kern_mount+0x2e/0x60 235 [ 0.963857][ T0] shmem_init+0x63/0xef 236 [ 
> 0.963857][ T0] mnt_init+0x159/0x2e0 237 [ 0.963857][ T0] ? 
> trace_init_perf_perm_irq_work_exit+0xe/0xe 238 [ 0.963857][ T0] 
> vfs_caches_init+0xd4/0xde 239 [ 0.963857][ T0] 
> start_kernel+0x837/0x8a4 240 [ 0.963857][ T0] 
> secondary_startup_64_no_verify+0xce/0xdb 241 [ 0.963857][ T0] </TASK> 
> 242 [ 0.963860][ T0] 
> ================================================================================ 
> 243 [ 0.965288][ T0] Kernel panic - not syncing: panic_on_warn set ... 
> 244 [ 0.966299][ T0] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 
> 6.1.0-rc2-00062-ga970174d7a10 #5 245 [ 0.967645][ T0] Hardware name: 
> QEMU Standard PC (i440FX + PIIX, 1996), BIOS 
> rel-1.13.0-48-gd9c812dda519-prebuilt.qemu.org 04/01/2014 246 [ 
> 0.969548][ T0] Call Trace: 247 [ 0.970050][ T0] <TASK> 248 [ 
> 0.970499][ T0] dump_stack_lvl+0x8d/0xcf 249 [ 0.971195][ T0] 
> panic+0x182/0x387 250 [ 0.971797][ T0] ? ubsan_epilogue+0x33/0x44 251 
> [ 0.972539][ T0] ubsan_epilogue+0x3f/0x44 252 [ 0.973237][ T0] 
> __ubsan_handle_shift_out_of_bounds+0x1e7/0x208 253 [ 0.973857][ T0] ? 
> __kmem_cache_alloc_node+0x167/0x290 254 [ 0.973857][ T0] ? 
> shmem_fill_super+0x2e/0x2e0 255 [ 0.973857][ T0] ? 
> rcu_read_lock_held_common+0x9/0x40 256 [ 0.973857][ T0] ? 
> shmem_alloc_hugefolio+0x110/0x110 257 [ 0.973857][ T0] ? 
> shmem_fill_super+0x2cc/0x2e0 258 [ 0.973857][ T0] 
> shmem_fill_super+0x2cc/0x2e0 259 [ 0.973857][ T0] 
> vfs_get_super+0x78/0x160 260 [ 0.973857][ T0] vfs_get_tree+0x28/0x100 
> 261 [ 0.973857][ T0] fc_mount+0x12/0x60 262 [ 0.973857][ T0] 
> vfs_kern_mount.part.38+0xa5/0xc0 263 [ 0.973857][ T0] 
> kern_mount+0x2e/0x60 264 [ 0.973857][ T0] shmem_init+0x63/0xef 265 [ 
> 0.973857][ T0] mnt_init+0x159/0x2e0 266 [ 0.973857][ T0] ? 
> trace_init_perf_perm_irq_work_exit+0xe/0xe 267 [ 0.973857][ T0] 
> vfs_caches_init+0xd4/0xde 268 [ 0.973857][ T0] 
> start_kernel+0x837/0x8a4 269 [ 0.973857][ T0] 
> secondary_startup_64_no_verify+0xce/0xdb 270 [ 0.973857][ T0] </TASK> 
> 271 [ 0.973857][ T0] Rebooting in 86400 seconds..

On 2022/11/1 19:34, Jason A. Donenfeld wrote:
> Others might feel differently and that's fine, but I always found the
> BIT() thing so much less clear than doing 1<<n, which is not only a
> pattern that I recognize as builtin to my brain, but also provides a
> direct description of what's happening, "shift a 1 over n times",
> leaving no off-by-one ambiguity about it. If anything I'd like to see
> the BIT() macro expanded throughout and then removed entirely.
>
> Probably just me though. You can safely ignore my opinion:).