libceph: Amend checking to fix `make W=1` build breakage
PATCH 1/1 libceph: Amend checking to fix `make W=1` build breakage
In a few cases the code compares 32-bit value to a SIZE_MAX derived constant which is much higher than that value on 64-bit platforms, Clang, in particular, is not happy about this
net/ceph/osdmap.c:1441:10: error: result of comparison of constant 4611686018427387891 with expression of type 'u32' (aka 'unsigned int') is always false [-Werror,-Wtautological-constant-out-of-range-compare]
1441 | if (len > (SIZE_MAX - sizeof(*pg)) / sizeof(u32))
| ~~~ ^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
net/ceph/osdmap.c:1624:10: error: result of comparison of constant 2305843009213693945 with expression of type 'u32' (aka 'unsigned int') is always false [-Werror,-Wtautological-constant-out-of-range-compare]
1624 | if (len > (SIZE_MAX - sizeof(*pg)) / (2 * sizeof(u32)))
| ~~~ ^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~fs/ceph/snap.c:377:10: error: result of comparison of constant 2305843009213693948 with expression of type 'u32' (aka 'unsigned int') is always false [-Werror,-Wtautological-constant-out-of-range-compare]
377 | if (num > (SIZE_MAX - sizeof(*snapc)) / sizeof(u64))
| ~~~ ^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Fix this by casting to size_t. Note, that possible replacement of SIZE_MAX by U32_MAX may lead to the behaviour changes on the corner cases.
--- net/ceph/osdmap.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
net/ceph/osdmap.c +2 -2
@@ -1438,7 +1438,7 @@ static struct ceph_pg_mapping *__decode_pg_temp(void **p, void *end,ceph_decode_32_safe (p, end, len, e_inval);if (len ==0 && incremental)return NULL ;/* new_pg_temp: [] to remove */ -if (len > (SIZE_MAX -sizeof (*pg)) /sizeof (u32)) +if ((size_t )len > (SIZE_MAX -sizeof (*pg)) /sizeof (u32))return ERR_PTR (-EINVAL );ceph_decode_need (p, end, len *sizeof (u32), e_inval);
@@ -1621,7 +1621,7 @@ static struct ceph_pg_mapping *__decode_pg_upmap_items(void **p, void *end,u32 len, i;ceph_decode_32_safe (p, end, len, e_inval); -if (len > (SIZE_MAX -sizeof (*pg)) / (2 *sizeof (u32))) +if ((size_t )len > (SIZE_MAX -sizeof (*pg)) / (2 *sizeof (u32)))return ERR_PTR (-EINVAL );ceph_decode_need (p, end,2 * len *sizeof (u32), e_inval);
Signature · 1 line
2.50.1
Quoted 37 lines · Andy Shevchenko
In a few cases the code compares 32-bit value to a
SIZE_MAXderived constant which is much higher than that value on 64-bit platforms, Clang, in particular, is not happy about this
net/ceph/osdmap.c:1441:10: error: result of comparison of constant 4611686018427387891 with expression of type 'u32' (aka 'unsigned int') is always false [-Werror,-Wtautological-constant-out-of-range-compare]
1441 | if (len > (SIZE_MAX-sizeof(*pg)) / sizeof(u32))| ~~~ ^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ net/ceph/osdmap.c:1624:10: error: result of comparison of constant 2305843009213693945 with expression of type 'u32' (aka 'unsigned int') is always false [-Werror,-Wtautological-constant-out-of-range-compare] 1624 | if (len > (SIZE_MAX - sizeof(*pg)) / (2 * sizeof(u32))) | ~~~ ^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
fs/ceph/snap.c:377:10: error: result of comparison of constant 2305843009213693948 with expression of type 'u32' (aka 'unsigned int') is always false [-Werror,-Wtautological-constant-out-of-range-compare]
377 | if (num > (SIZE_MAX-sizeof(*snapc)) / sizeof(u64))| ~~~ ^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Fix this by casting to
size_t. Note, that possible replacement ofSIZE_MAXbyU32_MAXmay lead to the behaviour changes on the corner cases.Signed-off-by Andy Shevchenko <[email protected]>--- net/ceph/osdmap.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)net/ceph/osdmap.c +1
-1@@ -1438,7 +1438,7 @@ static struct ceph_pg_mapping *__decode_pg_temp(void **p, void *end,ceph_decode_32_safe (p, end, len, e_inval);if (len ==0 && incremental)return NULL ;/* new_pg_temp: [] to remove */ -if (len > (SIZE_MAX -sizeof (*pg)) /sizeof (u32)) +if ((size_t )len > (SIZE_MAX -sizeof (*pg)) /sizeof (u32))return ERR_PTR (-EINVAL );ceph_decode_need (p, end, len *sizeof (u32), e_inval);
net/ceph/osdmap.c static struct ceph_pg_mapping *__decode_pg_upmap_items(void **p, void *end, · line 1621
Quoted 1 line
net/ceph/osdmap.c +0
-0@@ -1621,7 +1621,7 @@ static struct ceph_pg_mapping *__decode_pg_upmap_items(void **p, void *end,u32 len, i;
I am guessing... What if we change the declaration of len on size_t, then could it be more clear solution here? For example, let's consider this for both cases:
size_t len, i;
Could it eliminate the issue and to make the Clang happy? Or could it introduce another warnings/issues?
Thanks,
Slava.
Quoted 7 lines
net/ceph/osdmap.c +1
-1ceph_decode_32_safe (p, end, len, e_inval); -if (len > (SIZE_MAX -sizeof (*pg)) / (2 *sizeof (u32))) +if ((size_t )len > (SIZE_MAX -sizeof (*pg)) / (2 *sizeof (u32)))return ERR_PTR (-EINVAL );ceph_decode_need (p, end,2 * len *sizeof (u32), e_inval);
Quoted 1 line · Viacheslav Dubeyko
On Mon, 2025-11-10 at 15:46 +0100, Andy Shevchenko wrote:
...
Quoted 12 lines
net/ceph/osdmap.c static struct ceph_pg_mapping *__decode_pg_temp(void **p, void *end, · line 1444
Quoted 8 lines
net/ceph/osdmap.c +1
-1ceph_decode_32_safe (p, end, len, e_inval);if (len ==0 && incremental)return NULL ;/* new_pg_temp: [] to remove */ -if (len > (SIZE_MAX -sizeof (*pg)) /sizeof (u32)) +if ((size_t )len > (SIZE_MAX -sizeof (*pg)) /sizeof (u32))return ERR_PTR (-EINVAL );ceph_decode_need (p, end, len *sizeof (u32), e_inval);I am guessing... What if we change the declaration of len on
size_t, then could it be more clear solution here? For example, let's consider this for both cases:
Quoted 4 lines
size_tlen, i;Could it eliminate the issue and to make the Clang happy? Or could it introduce another warnings/issues?
Probably, but the code is pierced with the sizeof(u32) and alike, moreover size_t is architecture-dependent type, while the set of macros in decode.h seems to operate on the fixed-width type. That said, I prefer my way of fixing this. But if you find another, better one, I am all ears!
*Also note, I'm not familiar with the guts of the ceph, so maybe your solution is the best, but I want more people to confirm this.
Signature · 2 lines
With Best Regards, Andy Shevchenko
Viacheslav Dubeyko
Reviewed-byQuoted 22 lines · On Mon, 2025-11-10 at 21:36 +0200, [email protected] wrote:
Quoted 1 line · Viacheslav Dubeyko
On Mon, 2025-11-10 at 15:46 +0100, Andy Shevchenko wrote:
...
Quoted 12 lines
net/ceph/osdmap.c static struct ceph_pg_mapping *__decode_pg_temp(void **p, void *end, · line 1444
Quoted 8 lines
net/ceph/osdmap.c +1
-1ceph_decode_32_safe (p, end, len, e_inval);if (len ==0 && incremental)return NULL ;/* new_pg_temp: [] to remove */ -if (len > (SIZE_MAX -sizeof (*pg)) /sizeof (u32)) +if ((size_t )len > (SIZE_MAX -sizeof (*pg)) /sizeof (u32))return ERR_PTR (-EINVAL );ceph_decode_need (p, end, len *sizeof (u32), e_inval);I am guessing... What if we change the declaration of len on
size_t, then could it be more clear solution here? For example, let's consider this for both cases:Quoted 4 lines
size_tlen, i;Could it eliminate the issue and to make the Clang happy? Or could it introduce another warnings/issues?
Quoted 7 lines
Probably, but the code is pierced with the sizeof(u32) and alike, moreover
size_tis architecture-dependent type, while the set of macros indecode.hseems to operate on the fixed-width type. That said, I prefer my way of fixing this. But if you find another, better one, I am all ears!*Also note, I'm not familiar with the guts of the ceph, so maybe your solution is the best, but I want more people to confirm this.
I think the patch looks good as it is. And we can take it. If we find the better way
of fixing this, then we can do it anytime.
Thanks,
Slava.
Quoted 3 lines · Viacheslav Dubeyko
Quoted 2 lines · On Mon, 2025-11-10 at 21:36 +0200, [email protected] wrote:
Quoted 1 line · Viacheslav Dubeyko
On Mon, 2025-11-10 at 15:46 +0100, Andy Shevchenko wrote:
...
Quoted 22 lines
Quoted 22 lines
Quoted 12 lines
net/ceph/osdmap.c static struct ceph_pg_mapping *__decode_pg_temp(void **p, void *end, · line 1444
Quoted 8 lines
net/ceph/osdmap.c +1
-1ceph_decode_32_safe (p, end, len, e_inval);if (len ==0 && incremental)return NULL ;/* new_pg_temp: [] to remove */ -if (len > (SIZE_MAX -sizeof (*pg)) /sizeof (u32)) +if ((size_t )len > (SIZE_MAX -sizeof (*pg)) /sizeof (u32))return ERR_PTR (-EINVAL );ceph_decode_need (p, end, len *sizeof (u32), e_inval);I am guessing... What if we change the declaration of len on
size_t, then could it be more clear solution here? For example, let's consider this for both cases:Quoted 4 lines
size_tlen, i;Could it eliminate the issue and to make the Clang happy? Or could it introduce another warnings/issues?
Probably, but the code is pierced with the sizeof(u32) and alike, moreover
size_tis architecture-dependent type, while the set of macros indecode.hseems to operate on the fixed-width type. That said, I prefer my way of fixing this. But if you find another, better one, I am all ears!
Quoted 8 lines
Quoted 2 lines
*Also note, I'm not familiar with the guts of the ceph, so maybe your solution is the best, but I want more people to confirm this.
I think the patch looks good as it is. And we can take it. If we find the better way
of fixing this, then we can do it anytime.Reviewed-by Viacheslav Dubeyko <[email protected]>
Thanks, can this be applied? My builds are still broken.
Signature · 2 lines
With Best Regards, Andy Shevchenko
Quoted 32 lines · On Mon, 2025-11-24 at 17:09 +0200, [email protected] wrote:
Quoted 3 lines · Viacheslav Dubeyko
Quoted 2 lines · On Mon, 2025-11-10 at 21:36 +0200, [email protected] wrote:
Quoted 1 line · Viacheslav Dubeyko
On Mon, 2025-11-10 at 15:46 +0100, Andy Shevchenko wrote:
...
Quoted 25 lines
Quoted 24 lines
Quoted 12 lines
net/ceph/osdmap.c static struct ceph_pg_mapping *__decode_pg_temp(void **p, void *end, · line 1444
Quoted 8 lines
net/ceph/osdmap.c +1
-1ceph_decode_32_safe (p, end, len, e_inval);if (len ==0 && incremental)return NULL ;/* new_pg_temp: [] to remove */ -if (len > (SIZE_MAX -sizeof (*pg)) /sizeof (u32)) +if ((size_t )len > (SIZE_MAX -sizeof (*pg)) /sizeof (u32))return ERR_PTR (-EINVAL );ceph_decode_need (p, end, len *sizeof (u32), e_inval);I am guessing... What if we change the declaration of len on
size_t, then could it be more clear solution here? For example, let's consider this for both cases:Quoted 4 lines
size_tlen, i;Could it eliminate the issue and to make the Clang happy? Or could it introduce another warnings/issues?
Probably, but the code is pierced with the sizeof(u32) and alike, moreover
size_tis architecture-dependent type, while the set of macros indecode.hseems to operate on the fixed-width type. That said, I prefer my way of fixing this. But if you find another, better one, I am all ears!*Also note, I'm not familiar with the guts of the ceph, so maybe your solution is the best, but I want more people to confirm this.
Quoted 7 lines
Quoted 5 lines
I think the patch looks good as it is. And we can take it. If we find the better way
of fixing this, then we can do it anytime.Reviewed-by Viacheslav Dubeyko <[email protected]>Thanks, can this be applied? My builds are still broken.
The patchset has been applied on testing branch already [1].
Ilya, when are we planning to send this patchset upstream?
Thanks,
Slava.
Quoted 2 lines · Viacheslav Dubeyko
Quoted 1 line · On Mon, 2025-11-24 at 17:09 +0200, [email protected] wrote:
On Mon, Nov 10, 2025 at 08:39:49PM +0000, Viacheslav Dubeyko wrote:
...
Quoted 3 lines
Quoted 1 line
Thanks, can this be applied? My builds are still broken.
The patchset has been applied on testing branch already [1].
Thanks again, and same recommendation as per previous email.
Quoted 3 lines
Ilya, when are we planning to send this patchset upstream?
Signature · 2 lines
With Best Regards, Andy Shevchenko