From mboxrd@z Thu Jan 1 00:00:00 1970 From: Marek Szyprowski Subject: Re: [PATCH V37 27/29] tracefs: Restrict tracefs when the kernel is locked down Date: Tue, 13 Aug 2019 08:10:51 +0200 Message-ID: References: <20190731221617.234725-1-matthewgarrett@google.com> <20190731221617.234725-28-matthewgarrett@google.com> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 8bit Return-path: In-Reply-To: <20190731221617.234725-28-matthewgarrett@google.com> Content-Language: en-US Sender: linux-kernel-owner@vger.kernel.org To: Matthew Garrett , jmorris@namei.org Cc: linux-security-module@vger.kernel.org, linux-kernel@vger.kernel.org, linux-api@vger.kernel.org, Matthew Garrett , Steven Rostedt , Krzysztof Kozlowski List-Id: linux-api@vger.kernel.org Hi On 2019-08-01 00:16, Matthew Garrett wrote: > Tracefs may release more information about the kernel than desirable, so > restrict it when the kernel is locked down in confidentiality mode by > preventing open(). > > Signed-off-by: Matthew Garrett > Reviewed-by: Steven Rostedt (VMware) This patch causes the following regression on various Samsung Exynos SoC based boards (ARM 32bit): [   15.364422] Unable to handle kernel NULL pointer dereference at virtual address 00000000 [   15.368775] pgd = a530ddbe [   15.371447] [00000000] *pgd=bcd7c831 [   15.374993] Internal error: Oops: 80000007 [#1] PREEMPT SMP ARM [   15.380890] Modules linked in: [   15.383929] CPU: 0 PID: 1393 Comm: perf Not tainted 5.2.0-00027-g757ff7244358-dirty #6459 [   15.392086] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree) [   15.398164] PC is at 0x0 [   15.400687] LR is at do_dentry_open+0x22c/0x3b0 [   15.405193] pc : [<00000000>]    lr : []    psr: 60000053 [   15.411442] sp : e7317dd8  ip : 00000000  fp : 00000000 [   15.416650] r10: c0187e6c  r9 : c041f8cc  r8 : e72123c8 [   15.421858] r7 : e7317ec0  r6 : e7d89630  r5 : 00000000  r4 : e72123c0 [   15.428368] r3 : 00000000  r2 : 5ba370f3  r1 : e72123c0  r0 : e7d89630 [   15.434880] Flags: nZCv  IRQs on  FIQs off  Mode SVC_32  ISA ARM  Segment none [   15.442083] Control: 10c5387d  Table: 6726404a  DAC: 00000051 [   15.447812] Process perf (pid: 1393, stack limit = 0x17621431) [   15.453628] Stack: (0xe7317dd8 to 0xe7318000) ... [   15.604842] [] (do_dentry_open) from [] (path_openat+0x5a0/0x1004) [   15.612735] [] (path_openat) from [] (do_filp_open+0x6c/0xd8) [   15.620200] [] (do_filp_open) from [] (do_sys_open+0x130/0x1f4) [   15.627839] [] (do_sys_open) from [] (ret_fast_syscall+0x0/0x28) [   15.635560] Exception stack(0xe7317fa8 to 0xe7317ff0) [   15.640596] 7fa0:                   0022dc0b 001deee0 ffffff9c beb6d764 00020000 00000000 [   15.648756] 7fc0: 0022dc0b 001deee0 0022dba8 00000142 001ba044 00241d68 001a13d8 beb6e78c [   15.656913] 7fe0: b6f7e000 beb6c6f8 9a27c600 b6f69504 [   15.661952] Code: bad PC value [   15.665105] ---[ end trace 7e8b864582108f4a ]--- This is standard ARM 32bit kernel with arch/arm/configs/exynos_defconfig. It is enough to run "perf list" command. > --- > fs/tracefs/inode.c | 40 +++++++++++++++++++++++++++++++++++- > include/linux/security.h | 1 + > security/lockdown/lockdown.c | 1 + > 3 files changed, 41 insertions(+), 1 deletion(-) > > diff --git a/fs/tracefs/inode.c b/fs/tracefs/inode.c > index 1387bcd96a79..12a325fb4cbd 100644 > --- a/fs/tracefs/inode.c > +++ b/fs/tracefs/inode.c > @@ -21,6 +21,7 @@ > #include > #include > #include > +#include > > #define TRACEFS_DEFAULT_MODE 0700 > > @@ -28,6 +29,23 @@ static struct vfsmount *tracefs_mount; > static int tracefs_mount_count; > static bool tracefs_registered; > > +static int default_open_file(struct inode *inode, struct file *filp) > +{ > + struct dentry *dentry = filp->f_path.dentry; > + struct file_operations *real_fops; > + int ret; > + > + if (!dentry) > + return -EINVAL; > + > + ret = security_locked_down(LOCKDOWN_TRACEFS); > + if (ret) > + return ret; > + > + real_fops = dentry->d_fsdata; real_fops are NULL in my test case. > + return real_fops->open(inode, filp); > +} > + > static ssize_t default_read_file(struct file *file, char __user *buf, > size_t count, loff_t *ppos) > { > @@ -210,6 +228,12 @@ static int tracefs_apply_options(struct super_block *sb) > return 0; > } > > +static void tracefs_destroy_inode(struct inode *inode) > +{ > + if (S_ISREG(inode->i_mode)) > + kfree(inode->i_fop); > +} > + > static int tracefs_reconfigure(struct fs_context *fc) > { > struct super_block *sb = fc->root->d_sb; > @@ -236,6 +260,7 @@ static int tracefs_show_options(struct seq_file *m, struct dentry *root) > > static const struct super_operations tracefs_super_operations = { > .statfs = simple_statfs, > + .destroy_inode = tracefs_destroy_inode, > .show_options = tracefs_show_options, > }; > > @@ -372,6 +397,7 @@ struct dentry *tracefs_create_file(const char *name, umode_t mode, > struct dentry *parent, void *data, > const struct file_operations *fops) > { > + struct file_operations *proxy_fops; > struct dentry *dentry; > struct inode *inode; > > @@ -387,8 +413,20 @@ struct dentry *tracefs_create_file(const char *name, umode_t mode, > if (unlikely(!inode)) > return failed_creating(dentry); > > + proxy_fops = kzalloc(sizeof(struct file_operations), GFP_KERNEL); > + if (unlikely(!proxy_fops)) { > + iput(inode); > + return failed_creating(dentry); > + } > + > + if (!fops) > + fops = &tracefs_file_operations; > + > + dentry->d_fsdata = (void *)fops; > + memcpy(proxy_fops, fops, sizeof(*proxy_fops)); > + proxy_fops->open = default_open_file; > inode->i_mode = mode; > - inode->i_fop = fops ? fops : &tracefs_file_operations; > + inode->i_fop = proxy_fops; > inode->i_private = data; > d_instantiate(dentry, inode); > fsnotify_create(dentry->d_parent->d_inode, dentry); > diff --git a/include/linux/security.h b/include/linux/security.h > index d92323b44a3f..807dc0d24982 100644 > --- a/include/linux/security.h > +++ b/include/linux/security.h > @@ -121,6 +121,7 @@ enum lockdown_reason { > LOCKDOWN_KPROBES, > LOCKDOWN_BPF_READ, > LOCKDOWN_PERF, > + LOCKDOWN_TRACEFS, > LOCKDOWN_CONFIDENTIALITY_MAX, > }; > > diff --git a/security/lockdown/lockdown.c b/security/lockdown/lockdown.c > index 88064ce1c844..173191562047 100644 > --- a/security/lockdown/lockdown.c > +++ b/security/lockdown/lockdown.c > @@ -36,6 +36,7 @@ static char *lockdown_reasons[LOCKDOWN_CONFIDENTIALITY_MAX+1] = { > [LOCKDOWN_KPROBES] = "use of kprobes", > [LOCKDOWN_BPF_READ] = "use of bpf to read kernel RAM", > [LOCKDOWN_PERF] = "unsafe use of perf", > + [LOCKDOWN_TRACEFS] = "use of tracefs", > [LOCKDOWN_CONFIDENTIALITY_MAX] = "confidentiality", > }; > Best regards -- Marek Szyprowski, PhD Samsung R&D Institute Poland