afs: Work around strnlen() oops with CONFIG_FORTIFIED_SOURCE=y
Subsystems (2)
David Howells
maintainerRFC PATCH 1/1 afs: Work around strnlen() oops with CONFIG_FORTIFIED_SOURCE=y
[cc'ing mailing lists]
Hi Daniel,
CONFIG_FORTIFIED_SOURCE=y now causes an oops in strnlen() from afs (see attached patch for an explanation). Is replacing the use with memchr() the right approach? Or should I be calling __real_strnlen() or whatever it's called?
David
---
From: David Howells <[email protected]>
afs: Work around strnlen() oops with CONFIG_FORTIFIED_SOURCE=y
AFS has a structured layout in its directory contents (AFS dirs are downloaded as files and parsed locally by the client for lookup/readdir). The slots in the directory are defined by union afs_xdr_dirent. This, however, only directly allows a name of a length that will fit into that union. To support a longer name, the next 1-8 contiguous entries are annexed to the first one and the name flows across these.
afs_dir_iterate_block() uses strnlen(), limited to the space to the end of the page, to find out how long the name is. This worked fine until 6a39e62abbaf. With that commit, the compiler determines the size of the array and asserts that the string fits inside that array. This is a problem for AFS because we *expect* it to overflow one or more arrays.
A similar problem also occurs in afs_dir_scan_block() when a directory file is being locally edited to avoid the need to redownload it. There strlen() was being used safely because each page has the last byte set to 0 when the file is downloaded and validated (in afs_dir_check_page()).
Fix this by using memchr() instead and hoping no one changes that to check the object size.
The issue can be triggered by something like:
touch /afs/example.com/thisisaveryveryverylongname
and it generates a report that looks like:
detected buffer overflow in strnlen
------------[ cut here ]------------
kernel BUG at lib/string.c:1149!
...
RIP: 0010:fortify_panic+0xf/0x11
...
Call Trace:
afs_dir_iterate_block+0x12b/0x35b
afs_dir_iterate+0x14e/0x1ce
afs_do_lookup+0x131/0x417
afs_lookup+0x24f/0x344
lookup_open.isra.0+0x1bb/0x27d
open_last_lookups+0x166/0x237
path_openat+0xe0/0x159
do_filp_open+0x48/0xa4
? kmem_cache_alloc+0xf5/0x16e
? __clear_close_on_exec+0x13/0x22
? _raw_spin_unlock+0xa/0xb
do_sys_openat2+0x72/0xde
do_sys_open+0x3b/0x58
do_syscall_64+0x2d/0x3a
entry_SYSCALL_64_after_hwframe+0x44/0xa9fs/afs/dir.c +10 -3
@@ -350,6 +350,7 @@ static int afs_dir_iterate_block(struct afs_vnode *dvnode, unsigned blkoff) {union afs_xdr_dirent *dire; +const u8 *p;unsigned offset, next, curr;size_t nlen;int tmp;
@@ -378,9 +379,15 @@ static int afs_dir_iterate_block(struct afs_vnode *dvnode,/* got a valid entry */ dire = &block->dirents[offset]; - nlen =strnlen (dire->u.name, -sizeof (*block) - - offset *sizeof (union afs_xdr_dirent )); + p =memchr (dire->u.name,0 , +sizeof (*block) - offset *sizeof (union afs_xdr_dirent )); +if (!p) { +_debug ("ENT[%zu.%u]: %u unterminated dirent name" , + blkoff /sizeof (union afs_xdr_dir_block ), + offset, next); +return afs_bad (dvnode, afs_file_error_dir_over_end); + } + nlen = p - dire->u.name; _debug("ENT[%zu.%u]: %s %zu \"%s\"", blkoff / sizeof(union afs_xdr_dir_block), offset,
fs/afs/dir_edit.c +7 -1
@@ -111,6 +111,8 @@ static int afs_dir_scan_block(union afs_xdr_dir_block *block, struct qstr *name, unsigned int blocknum) {union afs_xdr_dirent *de; +const u8 *p; +unsigned long offset;u64 bitmap;int d, len, n;
@@ -135,7 +137,11 @@ static int afs_dir_scan_block(union afs_xdr_dir_block *block, struct qstr *name,continue ;/* The block was NUL-terminated by afs_dir_check_page(). */ - len =strlen (de->u.name); + offset = (unsigned long )de->u.name & (PAGE_SIZE -1 ); + p =memchr (de->u.name,0 ,PAGE_SIZE - offset); +if (!p) +return -1 ; + len = p - de->u.name;if (len == name->len &&memcmp (de->u.name, name->name, name->len) ==0 )return d;
Quoted 5 lines · David Howells
CONFIG_FORTIFIED_SOURCE=ynow causes an oops instrnlen()from afs (see attached patch for an explanation). Is replacing the use withmemchr()the right approach? Or should I be calling__real_strnlen()or whatever it's called?
Ugh. No.
Quoted 6 lines
AFS has a structured layout in its directory contents (AFS dirs are downloaded as files and parsed locally by the client for lookup/readdir). The slots in the directory are defined by union
afs_xdr_dirent. This, however, only directly allows a name of a length that will fit into that union. To support a longer name, the next 1-8 contiguous entries are annexed to the first one and the name flows across these.
I htink the right fix would be to try to create a type that actually describes that.
IOW, maybe the afs_xdr_dirent union could be written something like
union afs_xdr_dirent {
struct {
u8 valid;
u8 unused[1];
__be16 hash_next;
__be32 vnode;
__be32 unique;
u8 name[];
} u;
u8 extended_name[32];
} __packed;instead, and have a big comment about how "name[]" is that "16+overflow+next entries" thing?
I didn't check how you currently use that ->name thing (not a good identifier to grep for..), so you might want some other model - like using a separate union case for this "unconstrained name" case.
In fact, maybe that separate union struct is a better model anyway, to act as even more of documentation about the different cases..
Linus
Quoted 2 lines · Linus Torvalds
I htink the right fix would be to try to create a type that actually describes that.
How about the attached, then? It's basically what you suggested.
I realised whilst doing this that I'm not quite getting the extended-name handling correct, but I'll deal with that in a separate patch.
David
---
commit 26982a89cad77c0efc1c0c79bee0e3d75e9281d4
Author: David Howells <[email protected]> Date: Mon Dec 21 22:37:58 2020 +0000
afs: Work around strnlen() oops with CONFIG_FORTIFIED_SOURCE=y
AFS has a structured layout in its directory contents (AFS dirs are downloaded as files and parsed locally by the client for lookup/readdir). The slots in the directory are defined by union afs_xdr_dirent. This, however, only directly allows a name of a length that will fit into that union. To support a longer name, the next 1-8 contiguous entries are annexed to the first one and the name flows across these.
afs_dir_iterate_block() uses strnlen(), limited to the space to the end of the page, to find out how long the name is. This worked fine until 6a39e62abbaf. With that commit, the compiler determines the size of the array and asserts that the string fits inside that array. This is a problem for AFS because we *expect* it to overflow one or more arrays.
A similar problem also occurs in afs_dir_scan_block() when a directory file is being locally edited to avoid the need to redownload it. There strlen() was being used safely because each page has the last byte set to 0 when the file is downloaded and validated (in afs_dir_check_page()).
Fix this by changing the afs_xdr_dirent union name field to an indeterminate-length array and dropping the overflow field.
(Note that whilst looking at this, I realised that the calculation of the number of slots a dirent used is non-standard and not quite right, but I'll address that in a separate patch.)
The issue can be triggered by something like:
touch /afs/example.com/thisisaveryveryverylongname
and it generates a report that looks like:
detected buffer overflow in strnlen ------------[ cut here ]------------ kernel BUG at lib/string.c:1149!
...
RIP: 0010:fortify_panic+0xf/0x11
...
Call Trace:
afs_dir_iterate_block+0x12b/0x35b afs_dir_iterate+0x14e/0x1ce afs_do_lookup+0x131/0x417 afs_lookup+0x24f/0x344 lookup_open.isra.0+0x1bb/0x27d open_last_lookups+0x166/0x237 path_openat+0xe0/0x159 do_filp_open+0x48/0xa4
? kmem_cache_alloc+0xf5/0x16e
? __clear_close_on_exec+0x13/0x22 ? _raw_spin_unlock+0xa/0xb do_sys_openat2+0x72/0xde do_sys_open+0x3b/0x58 do_syscall_64+0x2d/0x3a entry_SYSCALL_64_after_hwframe+0x44/0xa9
fs/afs/xdr_fs.h +9 -4
@@ -54,10 +54,15 @@ union afs_xdr_dirent {__be16 hash_next;__be32 vnode;__be32 unique; -u8 name[16 ]; -u8 overflow[4 ];/* if any char of the name (inc -* NUL) reaches here, consume -* the next dirent too */ +u8 name[]; +/* When determining the number of dirent slots needed to +* represent a directory entry, name should be assumed to be 16 +* bytes, due to a now-standardised (mis)calculation, but it is +* in fact 20 bytes in size. +* +* For names longer than (16 or) 20 bytes, extra slots should +* be annexed to this one using the extended_name format. +*/ } u;u8 extended_name[32 ]; } __packed;
Quoted 2 lines · David Howells
How about the attached, then? I
That looks like the right thing, but I didn't check how the name[] array (or the overflow[] one) is actually used. But I assume you've tested this.
Do you want me to apply the patch as-is, or will I be getting a pull request with this (and the number-of-slots calculation thing you mention in the commit message)?
Linus
David Howells
maintainerQuoted 6 lines · Linus Torvalds
array (or the overflow[] one) is actually used. But I assume you've tested this.
Do you want me to apply the patch as-is, or will I be getting a pull request with this (and the number-of-slots calculation thing you mention in the commit message)?
I can give you a pull req for them as a pair. I don't know if Daniel wants to comment on the first patch.
David
Hi David,
Quoted 4 lines
CONFIG_FORTIFIED_SOURCE=ynow causes an oops instrnlen()from afs (see attached patch for an explanation). Is replacing the use withmemchr()the right approach? Or should I be calling__real_strnlen()or whatever it's called?
You certainly shouldn't be calling __real_strnlen().
memchr() is probably the right answer if you want a small fix.
However, as Linus suggested, the 'right' answer is to re-engineer your data structures so that the string operations you do don't overflow their arrays.
Kind regards,
Daniel
Quoted 119 lines
David
---
From: David Howells <[email protected]>afs: Work around
strnlen()oops withCONFIG_FORTIFIED_SOURCE=yAFS has a structured layout in its directory contents (AFS dirs are downloaded as files and parsed locally by the client for lookup/readdir). The slots in the directory are defined by union
afs_xdr_dirent. This, however, only directly allows a name of a length that will fit into that union. To support a longer name, the next 1-8 contiguous entries are annexed to the first one and the name flows across these.
afs_dir_iterate_block()usesstrnlen(), limited to the space to the end of the page, to find out how long the name is. This worked fine until6a39e62abbaf. With that commit, the compiler determines the size of the array and asserts that the string fits inside that array. This is a problem for AFS because we *expect* it to overflow one or more arrays.A similar problem also occurs in
afs_dir_scan_block()when a directory file is being locally edited to avoid the need to redownload it. Therestrlen()was being used safely because each page has the last byte set to 0 when the file is downloaded and validated (inafs_dir_check_page()).Fix this by using
memchr()instead and hoping no one changes that to check the object size.The issue can be triggered by something like:
touch /afs/example.com/thisisaveryveryverylongname
and it generates a report that looks like:
detected buffer overflow in strnlen ------------[ cut here ]------------ kernel BUG at lib/string.c:1149! ... RIP: 0010:fortify_panic+0xf/0x11 ... Call Trace: afs_dir_iterate_block+0x12b/0x35b afs_dir_iterate+0x14e/0x1ce afs_do_lookup+0x131/0x417 afs_lookup+0x24f/0x344 lookup_open.isra.0+0x1bb/0x27d open_last_lookups+0x166/0x237 path_openat+0xe0/0x159 do_filp_open+0x48/0xa4 ? kmem_cache_alloc+0xf5/0x16e ? __clear_close_on_exec+0x13/0x22 ? _raw_spin_unlock+0xa/0xb do_sys_openat2+0x72/0xde do_sys_open+0x3b/0x58 do_syscall_64+0x2d/0x3a entry_SYSCALL_64_after_hwframe+0x44/0xa9Fixes 6a39e62abbaf ("lib: string.h: detect intra-object overflow in fortified string functions")Reported-by Marc Dionne <[email protected]>Signed-off-by David Howells <[email protected]>cc Daniel Axtens <[email protected]>fs/afs/dir.c +10
-3@@ -350,6 +350,7 @@ static int afs_dir_iterate_block(struct afs_vnode *dvnode, unsigned blkoff) {union afs_xdr_dirent *dire; +const u8 *p;unsigned offset, next, curr;size_t nlen;int tmp;@@ -378,9 +379,15 @@ static int afs_dir_iterate_block(struct afs_vnode *dvnode,/* got a valid entry */ dire = &block->dirents[offset]; - nlen =strnlen (dire->u.name, -sizeof (*block) - - offset *sizeof (union afs_xdr_dirent )); + p =memchr (dire->u.name,0 , +sizeof (*block) - offset *sizeof (union afs_xdr_dirent )); +if (!p) { +_debug ("ENT[%zu.%u]: %u unterminated dirent name" , + blkoff /sizeof (union afs_xdr_dir_block ), + offset, next); +return afs_bad (dvnode, afs_file_error_dir_over_end); + } + nlen = p - dire->u.name; _debug("ENT[%zu.%u]: %s %zu \"%s\"", blkoff / sizeof(union afs_xdr_dir_block), offset,fs/afs/dir_edit.c +7
-1@@ -111,6 +111,8 @@ static int afs_dir_scan_block(union afs_xdr_dir_block *block, struct qstr *name, unsigned int blocknum) {union afs_xdr_dirent *de; +const u8 *p; +unsigned long offset;u64 bitmap;int d, len, n;@@ -135,7 +137,11 @@ static int afs_dir_scan_block(union afs_xdr_dir_block *block, struct qstr *name,continue ;/* The block was NUL-terminated by afs_dir_check_page(). */ - len =strlen (de->u.name); + offset = (unsigned long )de->u.name & (PAGE_SIZE -1 ); + p =memchr (de->u.name,0 ,PAGE_SIZE - offset); +if (!p) +return -1 ; + len = p - de->u.name;if (len == name->len &&memcmp (de->u.name, name->name, name->len) ==0 )return d;