[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <87edqy1jzf.ffs@tglx>
Date: Fri, 10 Feb 2023 00:45:24 +0100
From: Thomas Gleixner <tglx@...utronix.de>
To: Gregory Price <gourry.memverge@...il.com>,
linux-kernel@...r.kernel.org
Cc: linux-doc@...r.kernel.org, oleg@...hat.com, avagin@...il.com,
peterz@...radead.org, luto@...nel.org, krisman@...labora.com,
corbet@....net, shuah@...nel.org,
Gregory Price <gregory.price@...verge.com>
Subject: Re: [PATCH v8 1/1] ptrace,syscall_user_dispatch: checkpoint/restore
support for SUD
On Tue, Jan 31 2023 at 09:44, Gregory Price wrote:
>
> +static inline int syscall_user_dispatch_get_config(struct task_struct *task, unsigned long size,
> + void __user *data)
https://www.kernel.org/doc/html/latest/process/maintainer-tip.html#line-breaks
All over the place.
> index 195ae64a8c87..6d2f3b86f932 100644
> --- a/include/uapi/linux/ptrace.h
> +++ b/include/uapi/linux/ptrace.h
> @@ -112,6 +112,15 @@ struct ptrace_rseq_configuration {
> __u32 pad;
> };
>
> +#define PTRACE_SET_SYSCALL_USER_DISPATCH_CONFIG 0x4210
> +#define PTRACE_GET_SYSCALL_USER_DISPATCH_CONFIG 0x4211
> +struct syscall_user_dispatch_config {
Can you please visibly separate the defines from the struct definition
by a newline? Glueing that stuff together is just horrible to read.
> + __u64 mode;
> + __s8 *selector;
> + __u64 offset;
> + __u64 len;
https://www.kernel.org/doc/html/latest/process/maintainer-tip.html#struct-declarations-and-initializers
Please add proper documentation to this struct. It's user space ABI and
it's not the job of the man page maintainers to figure out what this
actually means.
> +int syscall_user_dispatch_set_config(struct task_struct *task, unsigned long size,
> + void __user *data)
> +{
> + struct syscall_user_dispatch_config config;
> +
> + if (size != sizeof(struct syscall_user_dispatch_config))
> + return -EINVAL;
> +
> + if (copy_from_user(&config, data, sizeof(config)))
> + return -EFAULT;
> +
> + return set_syscall_user_dispatch(config.mode, config.offset, config.len,
> + config.selector);
How is this supposed to work? This is called from the ptracer to set the
user dispatch mode on the ptracee, i.e. on @task.
But set_syscall_user_dispatch() operates on current, which is the
ptracer itself.
Clearly well tested with the non-existant selftest, which is part of
this submission.
So please fix the above issues, add a selftest and proper documentation.
I'm neither impressed by this patch nor by the reviews.
Sigh
tglx
Powered by blists - more mailing lists