Linux 单用户模式patch解析

刺骨的言语ヽ痛彻心扉 2022-05-13 21:32 261阅读 0赞

在我之前文章提到Linux 4.1内核支持单用户模式（传送门：[https://blog.csdn.net/cui841923894/article/details/81568351][https_blog.csdn.net_cui841923894_article_details_81568351]），此模式下用户UID和GID均为0同时不再区分用户权限（类root权限），应用于在某些小系统（例如嵌入式系统）。  
接下来我们看下这个patch是如何实现内核单用户的。

## 内核patch解析 ##

patch查看地址：[https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=2813893f8b197a14f1e1ddb04d99bce46817c84a][https_git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_commit_id_2813893f8b197a14f1e1ddb04d99bce46817c84a]

1.commit说明

kernel: conditionally support non-root users, groups and capabilities
    There are a lot of embedded systems that run most or all of their
    functionality in init, running as root:root.  For these systems,
    supporting multiple users is not necessary.
    在很多嵌入式系统中，他们始终使用root:root用户进行操作。这些系统中，多用户功能显得不是很必需（鸡肋了~）。
    
    This patch adds a new symbol, CONFIG_MULTIUSER, that makes support for
    non-root users, non-root groups, and capabilities optional.  It is enabled
    under CONFIG_EXPERT menu.
    这个patch添加了新的CONFIG_MULTIUSER内核开关，支持non-root users，, non-root groups, and capabilities。
    
    When this symbol is not defined, UID and GID are zero in any possible case
    and processes always have all capabilities.
    当CONFIG_MULTIUSER关闭（关闭多用户模式），UID和GID均是0，进程拥有所有capabilities拥有的功能。
    
    The following syscalls are compiled out: setuid, setregid, setgid,
    setreuid, setresuid, getresuid, setresgid, getresgid, setgroups,
    getgroups, setfsuid, setfsgid, capget, capset.
    同时系统调用setuid, setregid, setgid,
    setreuid, setresuid, getresuid, setresgid, getresgid, setgroups,
    getgroups, setfsuid, setfsgid, capget, capset将不再编译（和支持）。
    
    Also, groups.c is compiled out completely.
    同时group.c文件不再编译。
    
    In kernel/capability.c, capable function was moved in order to avoid
    adding two ifdef blocks.
    kernel/capability.c中的capable相关函数也将移除（其实是采用#ifdef来判断进入正常处理还是直接返回）。
    
    This change saves about 25 KB on a defconfig build.  The most minimal
    kernels have total text sizes in the high hundreds of kB rather than
    low MB.  (The 25k goes down a bit with allnoconfig, but not that much.
    这项修改在使用defconfig（内核的默认config）可以节省25KB的内核二进制大小。在小内核的config场景可以节省数百KB空间（小于1MB）。在allnoconfig下节省稍微小于25KB的空间。
    
    The kernel was booted in Qemu.  All the common functionalities work.
    Adding users/groups is not possible, failing with -ENOSYS.
    在虚拟机启动的系统（验证），基本系统调用都可以正常运行，所有设计添加users/groups的操作都无效，返回-ENOSYS。
    
    Bloat-o-meter output:
    add/remove: 7/87 grow/shrink: 19/397 up/down: 1675/-26325 (-24650)
    
    [akpm@linux-foundation.org: coding-style fixes]
    Signed-off-by: Iulia Manda <iulia.manda21@gmail.com>
    Reviewed-by: Josh Triplett <josh@joshtriplett.org>
    Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>
    Tested-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
    Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
    Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
    Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

2.patch修改内容解析  
因为patch涉及修改行多，并且很多目的相同，所以挑重点介绍。

a.某些功能和架构中添加对MULTIUSER config的支持：

diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
    index a5ced5c..de2726a 100644
    --- a/arch/s390/Kconfig
    +++ b/arch/s390/Kconfig
    @@ -328,6 +328,7 @@ config COMPAT
        select COMPAT_BINFMT_ELF if BINFMT_ELF
        select ARCH_WANT_OLD_COMPAT_IPC
        select COMPAT_OLD_SIGACTION
    +   depends on MULTIUSER
    
    diff --git a/drivers/staging/lustre/lustre/Kconfig b/drivers/staging/lustre/lustre/Kconfig
    index 6725467..62c7bba 100644
    --- a/drivers/staging/lustre/lustre/Kconfig
    +++ b/drivers/staging/lustre/lustre/Kconfig
    @@ -10,6 +10,7 @@ config LUSTRE_FS
        select CRYPTO_SHA1
        select CRYPTO_SHA256
        select CRYPTO_SHA512
    +   depends on MULTIUSER
    …

b.通过\#ifdef CONFIG\_MULTIUSER设置函数分支

diff --git a/include/linux/capability.h b/include/linux/capability.h
    index aa93e5e..af9f0b9 100644
    --- a/include/linux/capability.h
    +++ b/include/linux/capability.h
    @@ -205,6 +205,7 @@ static inline kernel_cap_t cap_raise_nfsd_set(const kernel_cap_t a,
                   cap_intersect(permitted, __cap_nfsd_set));
     }
    
    +#ifdef CONFIG_MULTIUSER //如果定义多用户，则执行正常功能函数
     extern bool has_capability(struct task_struct *t, int cap);
     extern bool has_ns_capability(struct task_struct *t,
                      struct user_namespace *ns, int cap);
    @@ -213,6 +214,34 @@ extern bool has_ns_capability_noaudit(struct task_struct *t,
                          struct user_namespace *ns, int cap);
     extern bool capable(int cap);
     extern bool ns_capable(struct user_namespace *ns, int cap);
    +#else // 如果non-root模式，则capability等操作不支持
    +static inline bool has_capability(struct task_struct *t, int cap)
    +{
    +   return true;
    +}
    …
    +static inline bool ns_capable(struct user_namespace *ns, int cap)
    +{
    +   return true;
    +}
    +#endif /* CONFIG_MULTIUSER *
    
    diff --git a/include/linux/cred.h b/include/linux/cred.h
    index 2fb2ca2..8b6c083 100644
    --- a/include/linux/cred.h
    +++ b/include/linux/cred.h
    @@ -62,9 +62,27 @@ do {                         \
            groups_free(group_info);        \
     } while (0)
    
    -extern struct group_info *groups_alloc(int);
     extern struct group_info init_groups;
    +#ifdef CONFIG_MULTIUSER //non-root模式屏蔽in_group_p和in_egroup_p等函数
    +extern struct group_info *groups_alloc(int);
     extern void groups_free(struct group_info *);
    +
    +extern int in_group_p(kgid_t);
    +extern int in_egroup_p(kgid_t);
    +#else
    +static inline void groups_free(struct group_info *group_info)
    +{
    +}
    +
    +static inline int in_group_p(kgid_t grp)
    +{
    +        return 1;
    +}
    +static inline int in_egroup_p(kgid_t grp)
    +{
    +        return 1;
    +}
    +#endif
    
    diff --git a/include/linux/uidgid.h b/include/linux/uidgid.h
    index 2d1f9b6..0ee05da 100644
    --- a/include/linux/uidgid.h
    +++ b/include/linux/uidgid.h
    @@ -29,6 +29,7 @@ typedef struct {
     #define KUIDT_INIT(value) (kuid_t){ value }
     #define KGIDT_INIT(value) (kgid_t){ value }
    
    +#ifdef CONFIG_MULTIUSER  //屏蔽__kuid_val和__kuid_val
     static inline uid_t __kuid_val(kuid_t uid)
     {
        return uid.val;
    @@ -38,6 +39,17 @@ static inline gid_t __kgid_val(kgid_t gid)
     {
        return gid.val;
     }
    +#else
    +static inline uid_t __kuid_val(kuid_t uid)
    +{
    +   return 0;
    +}
    +
    +static inline gid_t __kgid_val(kgid_t gid)
    +{
    +   return 0;
    +}
    +#endif

c. init/Kconfig添加MULTIUSER支持，这样内核make menuconfig可以看到MULTIUSER

…
    +config MULTIUSER
    +   bool "Multiple users, groups and capabilities support" if EXPERT
    +   default y
    +   help
    +     This option enables support for non-root users, groups and
    +     capabilities.
    +
    +     If you say N here, all processes will run with UID 0, GID 0, and all
    +     possible capabilities.  Saying N here also compiles out support for
    +     system calls related to UIDs, GIDs, and capabilities, such as setuid,
    +     setgid, and capset.
    +
    +     If unsure, say Y here.
    +

d. kernel/Makefile添加MULTIUSER支持

diff --git a/kernel/Makefile b/kernel/Makefile
    index 1408b33..0f8f8b0 100644
    --- a/kernel/Makefile
    +++ b/kernel/Makefile
    @@ -9,7 +9,9 @@ obj-y     = fork.o exec_domain.o panic.o \
            extable.o params.o \
            kthread.o sys_ni.o nsproxy.o \
            notifier.o ksysfs.o cred.o reboot.o \
    -       async.o range.o groups.o smpboot.o
    +       async.o range.o smpboot.o
    +
    +obj-$(CONFIG_MULTIUSER) += groups.o //这里，选择CONFIG_MULTIUSER后才会编译group.c

e.这里在capability.c中，第35行添加ifdef CONFIG\_MULTIUSER，第386行添加+\#endif /\* CONFIG\_MULTIUSER \*/，说明只有选择CONFIG\_MULTIUSER，文件第35行——386行中包括的函数，才可以生效（定义，实现）。

diff --git a/kernel/capability.c b/kernel/capability.c
    index 989f5bf..45432b5 100644
    --- a/kernel/capability.c
    +++ b/kernel/capability.c
    @@ -35,6 +35,7 @@ static int __init file_caps_disable(char *str)
     }
     __setup("no_file_caps", file_caps_disable);
    
    +#ifdef CONFIG_MULTIUSER
     /*
      * More recent versions of libcap are available from:
      *
    @@ -386,6 +387,24 @@ bool ns_capable(struct user_namespace *ns, int cap)
     }
     EXPORT_SYMBOL(ns_capable);
    
    +
    +/**
    + * capable - Determine if the current task has a superior capability in effect
    + * @cap: The capability to be tested for
    + *
    + * Return true if the current task has the given superior capability currently
    + * available for use, false if not.
    + *
    + * This sets PF_SUPERPRIV on the task if the capability is available on the
    + * assumption that it's about to be used.
    + */
    +bool capable(int cap)
    +{
    +   return ns_capable(&init_user_ns, cap);
    +}
    +EXPORT_SYMBOL(capable);
    +#endif /* CONFIG_MULTIUSER */

f.sys\_ni.c中添加以上处理函数。  
这里提一下sys\_ni.c作用，如果一个系统调用被淘汰，它所对应的服务例程就要被指定为sys\_ni\_syscall。sys\_ni\_syscall中的”ni”即表示”not implemented（没有实现）”。

diff --git a/kernel/sys_ni.c b/kernel/sys_ni.c
    index 5adcb0a..7995ef5 100644
    --- a/kernel/sys_ni.c
    +++ b/kernel/sys_ni.c
    @@ -159,6 +159,20 @@ cond_syscall(sys_uselib);
     cond_syscall(sys_fadvise64);
     cond_syscall(sys_fadvise64_64);
     cond_syscall(sys_madvise);
    +cond_syscall(sys_setuid);
    +cond_syscall(sys_setregid);
    +cond_syscall(sys_setgid);
    +cond_syscall(sys_setreuid);
    +cond_syscall(sys_setresuid);
    +cond_syscall(sys_getresuid);
    +cond_syscall(sys_setresgid);
    +cond_syscall(sys_getresgid);
    +cond_syscall(sys_setgroups);
    +cond_syscall(sys_getgroups);
    +cond_syscall(sys_setfsuid);
    +cond_syscall(sys_setfsgid);
    +cond_syscall(sys_capget);
    +cond_syscall(sys_capset);

以上，patch简单来说，就是实现了：

> The following syscalls are compiled out: setuid, setregid, setgid,  
> setreuid, setresuid, getresuid, setresgid, getresgid, setgroups,  
> getgroups, setfsuid, setfsgid, capget, capset.
> 
> Also, groups.c is compiled out completely.
> 
> In kernel/capability.c, capable function was moved in order to avoid  
> adding two ifdef blocks.

## 运行效果 ##

1.使用v4.18内核编译bzImage

#git branch
    * (HEAD detached at v4.18)
    
    #cp arch/x86/configs/x86_64_defconfig ./.config
    #make menuconfig (关闭MULTIUSER)
    #make bzImage -j8

![这里写图片描述][70]  
编译好后，内核在目录arch/x86/boot/bzImage

2.使用qemu启动

/ # adduser cuibixuan  （这里为什么还能添加用户？）
    adduser: /home/cuibixuan: No such file or directory
    passwd: unknown uid 0
    / # su cuibixuan
    su: can't set groups: Function not implemented

可以看到，groups相关操作，已经” Function not implemented”。说明添加到kernel/sys\_ni.c的函数sys\_setgroups已经生效（+cond\_syscall(sys\_setgroups);）。

## 后续 ##

Linux对single-user system的支持，个人认为仅仅不支持uid/gid、group和等capability等相关函数是不够的。比如，启动前fs已经配置多个用户(/etc/passwd和/etc/group)怎么处理；以及某些（安全相关）系统调用建议运行在个人用户权限下怎么办?以及[https://lwn.net/Articles/631853/][https_lwn.net_Articles_631853]讨论提到：  
multiple processes, scheduling等问题：

> Come to think of it, I look forward to the next tinification patch  
> that removes support for multiple processes, scheduling, and makes the  
> only running process always have pid 1.

或者针对threads讨论：

> The problem is then that the single userspace task can prevent  
> necessary kernel threads from running.

![这里写代码片][70 1]

[https_blog.csdn.net_cui841923894_article_details_81568351]: https://blog.csdn.net/cui841923894/article/details/81568351
[https_git.kernel.org_pub_scm_linux_kernel_git_torvalds_linux.git_commit_id_2813893f8b197a14f1e1ddb04d99bce46817c84a]: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=2813893f8b197a14f1e1ddb04d99bce46817c84a
[70]: /images/20220514/301f4685692f4eb1a305658f2fbb0752.png
[https_lwn.net_Articles_631853]: https://lwn.net/Articles/631853/
[70 1]: /images/20220514/424bf4051b964ff2a1086c88b71df75d.png