mm: Guard a use of node_reclaim_distance with CONFIFG_NUMA
Subsystems (2)
PATCH 1/2 mm: Guard a use of node_reclaim_distance with CONFIFG_NUMA
From: Palmer Dabbelt <[email protected]>
This is only useful under CONFIG_NUMA. IIUC skipping the check is the right thing to do here, as without CONFIG_NUMA there will never be any large node distances on non-NUMA systems.
I expected this to manifest as a link failure under (!CONFIG_NUMA && CONFIG_TRANSPARENT_HUGE_PAGES), but I'm not actually seeing that. I think the reference is just getting pruned before it's checked, but I didn't get that from reading the code so I'm worried I'm missing something.
Either way, this is necessary to guard the definition of node_reclaim_distance with CONFIG_NUMA.
--- mm/khugepaged.c | 2 ++ 1 file changed, 2 insertions(+)
mm/khugepaged.c +2 -0
@@ -819,8 +819,10 @@ static bool khugepaged_scan_abort(int nid)for (i =0 ; i <MAX_NUMNODES ; i++) {if (!khugepaged_node_load[i])continue ; +#ifdef CONFIG_NUMA if (node_distance (nid, i) > node_reclaim_distance)return true; +#endif }return false; }
Signature · 1 line
2.30.1.766.gb4fecdf3b7-goog
PATCH 2/2 topology: Guard node_reclaim_distance with CONFIFG_NUMA
From: Palmer Dabbelt <[email protected]>
This is only defined (and useful) for CONFIG_NUMA, but isn't conditionally declared. This makes users a bit more of a headache to track down.
--- include/linux/topology.h | 2 ++ 1 file changed, 2 insertions(+)
include/linux/topology.h +2 -0
@@ -73,7 +73,9 @@ int arch_update_cpu_topology(void);* *improves* if allowed to reclaim memory and load balance tasks * between NUMA nodes 2-hops apart. */ +#ifdef CONFIG_NUMA extern int __read_mostly node_reclaim_distance; +#endif #ifndef PENALTY_FOR_NODE_WITH_CPUS #define PENALTY_FOR_NODE_WITH_CPUS (1)
Signature · 1 line
2.30.1.766.gb4fecdf3b7-goog
Andrew Morton
maintainerQuoted 25 lines · On Fri, 26 Feb 2021 12:17:20 -0800 Palmer Dabbelt <[email protected]> wrote:
From: Palmer Dabbelt <[email protected]>
This is only useful under
CONFIG_NUMA. IIUC skipping the check is the right thing to do here, as withoutCONFIG_NUMAthere will never be any large node distances on non-NUMA systems.I expected this to manifest as a link failure under (!CONFIG_NUMA &&
CONFIG_TRANSPARENT_HUGE_PAGES), but I'm not actually seeing that. I think the reference is just getting pruned before it's checked, but I didn't get that from reading the code so I'm worried I'm missing something.Either way, this is necessary to guard the definition of
node_reclaim_distancewithCONFIG_NUMA.Signed-off-by Palmer Dabbelt <[email protected]>--- mm/khugepaged.c | 2 ++ 1 file changed, 2 insertions(+)mm/khugepaged.c +0
-0
mm/khugepaged.c static bool khugepaged_scan_abort(int nid) · line 828
Quoted 10 lines
mm/khugepaged.c +2
-0@@ -819,8 +819,10 @@ static bool khugepaged_scan_abort(int nid)for (i =0 ; i <MAX_NUMNODES ; i++) {if (!khugepaged_node_load[i])continue ; +#ifdef CONFIG_NUMA if (node_distance (nid, i) > node_reclaim_distance)return true; +#endif }return false; }
This makes the entire loop a no-op. Perhaps Kirill can help take a look at removing unnecessary code in khugepaged.c when CONFIG_NUMA=n?
Quoted 31 lines · Andrew Morton
Quoted 30 lines · On Fri, 26 Feb 2021 12:17:20 -0800 Palmer Dabbelt <[email protected]> wrote:
From: Palmer Dabbelt <[email protected]>
This is only useful under
CONFIG_NUMA. IIUC skipping the check is the right thing to do here, as withoutCONFIG_NUMAthere will never be any large node distances on non-NUMA systems.I expected this to manifest as a link failure under (!CONFIG_NUMA &&
CONFIG_TRANSPARENT_HUGE_PAGES), but I'm not actually seeing that. I think the reference is just getting pruned before it's checked, but I didn't get that from reading the code so I'm worried I'm missing something.Either way, this is necessary to guard the definition of
node_reclaim_distancewithCONFIG_NUMA.Signed-off-by Palmer Dabbelt <[email protected]>--- mm/khugepaged.c | 2 ++ 1 file changed, 2 insertions(+)mm/khugepaged.c +1
-0@@ -819,8 +819,10 @@ static bool khugepaged_scan_abort(int nid)for (i =0 ; i <MAX_NUMNODES ; i++) {if (!khugepaged_node_load[i])continue ; +#ifdef CONFIG_NUMA if (node_distance (nid, i) > node_reclaim_distance)
Quoted 8 lines
mm/khugepaged.c static bool khugepaged_scan_abort(int nid) · line 828
Quoted 5 lines
mm/khugepaged.c +1
-0return true; +#endif }return false; }This makes the entire loop a no-op. Perhaps Kirill can help take a look at removing unnecessary code in
khugepaged.cwhenCONFIG_NUMA=n?
First lines of khugepaged_scan_abort() say
if (!node_reclaim_mode) return false;
And include/linux/swap.h says
#ifdef CONFIG_NUMA
extern int node_reclaim_mode;
extern int sysctl_min_unmapped_ratio;
extern int sysctl_min_slab_ratio;
#else
#define node_reclaim_mode 0
#endif
So, no need for an #ifdef CONFIG_NUMA inside khugepaged_scan_abort().
Hugh
Quoted 54 lines · On Fri, 26 Feb 2021 17:31:40 PST (-0800), [email protected] wrote:
Quoted 31 lines · Andrew Morton
Quoted 30 lines · On Fri, 26 Feb 2021 12:17:20 -0800 Palmer Dabbelt <[email protected]> wrote:
From: Palmer Dabbelt <[email protected]>
This is only useful under
CONFIG_NUMA. IIUC skipping the check is the right thing to do here, as withoutCONFIG_NUMAthere will never be any large node distances on non-NUMA systems.I expected this to manifest as a link failure under (!CONFIG_NUMA &&
CONFIG_TRANSPARENT_HUGE_PAGES), but I'm not actually seeing that. I think the reference is just getting pruned before it's checked, but I didn't get that from reading the code so I'm worried I'm missing something.Either way, this is necessary to guard the definition of
node_reclaim_distancewithCONFIG_NUMA.Signed-off-by Palmer Dabbelt <[email protected]>--- mm/khugepaged.c | 2 ++ 1 file changed, 2 insertions(+)mm/khugepaged.c +1
-0@@ -819,8 +819,10 @@ static bool khugepaged_scan_abort(int nid)for (i =0 ; i <MAX_NUMNODES ; i++) {if (!khugepaged_node_load[i])continue ; +#ifdef CONFIG_NUMA if (node_distance (nid, i) > node_reclaim_distance)Quoted 8 lines
mm/khugepaged.c static bool khugepaged_scan_abort(int nid) · line 828
Quoted 5 lines
mm/khugepaged.c +1
-0return true; +#endif }return false; }This makes the entire loop a no-op. Perhaps Kirill can help take a look at removing unnecessary code in
khugepaged.cwhenCONFIG_NUMA=n?First lines of
khugepaged_scan_abort()sayif (!node_reclaim_mode) return false;And
include/linux/swap.hsays
#ifdefCONFIG_NUMA
extern intnode_reclaim_mode;
extern intsysctl_min_unmapped_ratio;
extern intsysctl_min_slab_ratio;
#else
#definenode_reclaim_mode0
#endif
Quoted 1 line
So, no need for an #ifdef
CONFIG_NUMAinsidekhugepaged_scan_abort().
Ah, thanks, I hadn't seen that. That certainly explains the lack of an undefined reference.
That said: do we generally rely on DCE to prune references to undefined symbols? This particular one seems like it'd get reliably deleted, but it seems like a fragile thing to do in general. This kind of stuff would certainly make some code easier to write, though.
I don't really care all that much, though, as I was just sending this along due to some build failure report from a user that I couldn't reproduce. It looked like they had some out-of-tree stuff, so in this case I'm fine on fixing this being their problem.
Quoted 58 lines · Palmer Dabbelt
Quoted 55 lines · On Fri, 26 Feb 2021 17:31:40 PST (-0800), [email protected] wrote:
Quoted 32 lines · Andrew Morton
Quoted 30 lines · On Fri, 26 Feb 2021 12:17:20 -0800 Palmer Dabbelt <[email protected]> wrote:
From: Palmer Dabbelt <[email protected]>
This is only useful under
CONFIG_NUMA. IIUC skipping the check is the right thing to do here, as withoutCONFIG_NUMAthere will never be any large node distances on non-NUMA systems.I expected this to manifest as a link failure under (!CONFIG_NUMA &&
CONFIG_TRANSPARENT_HUGE_PAGES), but I'm not actually seeing that. I think the reference is just getting pruned before it's checked, but I didn't get that from reading the code so I'm worried I'm missing something.Either way, this is necessary to guard the definition of
node_reclaim_distancewithCONFIG_NUMA.Signed-off-by Palmer Dabbelt <[email protected]>--- mm/khugepaged.c | 2 ++ 1 file changed, 2 insertions(+)mm/khugepaged.c +1
-0@@ -819,8 +819,10 @@ static bool khugepaged_scan_abort(int nid)for (i =0 ; i <MAX_NUMNODES ; i++) {if (!khugepaged_node_load[i])continue ; +#ifdef CONFIG_NUMA if (node_distance (nid, i) > node_reclaim_distance)Quoted 8 lines
mm/khugepaged.c static bool khugepaged_scan_abort(int nid) · line 828
Quoted 5 lines
mm/khugepaged.c +1
-0return true; +#endif }return false; }This makes the entire loop a no-op. Perhaps Kirill can help take a look at removing unnecessary code in
khugepaged.cwhenCONFIG_NUMA=n?First lines of
khugepaged_scan_abort()sayif (!node_reclaim_mode) return false;And
include/linux/swap.hsays
#ifdefCONFIG_NUMA
extern intnode_reclaim_mode;
extern intsysctl_min_unmapped_ratio;
extern intsysctl_min_slab_ratio;
#else
#definenode_reclaim_mode0
#endifQuoted 1 line
So, no need for an #ifdef
CONFIG_NUMAinsidekhugepaged_scan_abort().
Quoted 7 lines
Ah, thanks, I hadn't seen that. That certainly explains the lack of an undefined reference.
That said: do we generally rely on DCE to prune references to undefined symbols? This particular one seems like it'd get reliably deleted, but it seems like a fragile thing to do in general. This kind of stuff would certainly make some code easier to write, though.
Yes, the kernel build very much depends on the optimizer eliminating dead code, in many many places. We do prefer to keep the #ifdefs to the header files as much as possible.
Quoted 5 lines
I don't really care all that much, though, as I was just sending this along due to some build failure report from a user that I couldn't reproduce. It looked like they had some out-of-tree stuff, so in this case I'm fine on fixing this being their problem.
I didn't see your 2/2 at the time; but wouldn't be surprised if that needs 1/2, to avoid an error on undeclared node_reclaim_distance before the optimizer comes into play. If so, best just to drop 2/2 too.
Hugh
Quoted 62 lines · On Fri, 26 Feb 2021 19:41:40 PST (-0800), [email protected] wrote:
Quoted 61 lines · Palmer Dabbelt
Quoted 55 lines · On Fri, 26 Feb 2021 17:31:40 PST (-0800), [email protected] wrote:
Quoted 32 lines · Andrew Morton
Quoted 30 lines · On Fri, 26 Feb 2021 12:17:20 -0800 Palmer Dabbelt <[email protected]> wrote:
From: Palmer Dabbelt <[email protected]>
This is only useful under
CONFIG_NUMA. IIUC skipping the check is the right thing to do here, as withoutCONFIG_NUMAthere will never be any large node distances on non-NUMA systems.I expected this to manifest as a link failure under (!CONFIG_NUMA &&
CONFIG_TRANSPARENT_HUGE_PAGES), but I'm not actually seeing that. I think the reference is just getting pruned before it's checked, but I didn't get that from reading the code so I'm worried I'm missing something.Either way, this is necessary to guard the definition of
node_reclaim_distancewithCONFIG_NUMA.Signed-off-by Palmer Dabbelt <[email protected]>--- mm/khugepaged.c | 2 ++ 1 file changed, 2 insertions(+)mm/khugepaged.c +1
-0@@ -819,8 +819,10 @@ static bool khugepaged_scan_abort(int nid)for (i =0 ; i <MAX_NUMNODES ; i++) {if (!khugepaged_node_load[i])continue ; +#ifdef CONFIG_NUMA if (node_distance (nid, i) > node_reclaim_distance)Quoted 8 lines
mm/khugepaged.c static bool khugepaged_scan_abort(int nid) · line 828
Quoted 5 lines
mm/khugepaged.c +1
-0return true; +#endif }return false; }This makes the entire loop a no-op. Perhaps Kirill can help take a look at removing unnecessary code in
khugepaged.cwhenCONFIG_NUMA=n?First lines of
khugepaged_scan_abort()sayif (!node_reclaim_mode) return false;And
include/linux/swap.hsays
#ifdefCONFIG_NUMA
extern intnode_reclaim_mode;
extern intsysctl_min_unmapped_ratio;
extern intsysctl_min_slab_ratio;
#else
#definenode_reclaim_mode0
#endifQuoted 1 line
So, no need for an #ifdef
CONFIG_NUMAinsidekhugepaged_scan_abort().Ah, thanks, I hadn't seen that. That certainly explains the lack of an undefined reference.
Quoted 8 lines
Quoted 4 lines
That said: do we generally rely on DCE to prune references to undefined symbols? This particular one seems like it'd get reliably deleted, but it seems like a fragile thing to do in general. This kind of stuff would certainly make some code easier to write, though.
Yes, the kernel build very much depends on the optimizer eliminating dead code, in many many places. We do prefer to keep the #ifdefs to the header files as much as possible.
OK, makes sense. Thanks!
Quoted 8 lines
Quoted 4 lines
I don't really care all that much, though, as I was just sending this along due to some build failure report from a user that I couldn't reproduce. It looked like they had some out-of-tree stuff, so in this case I'm fine on fixing this being their problem.
I didn't see your 2/2 at the time; but wouldn't be surprised if that needs 1/2, to avoid an error on undeclared
node_reclaim_distancebefore the optimizer comes into play. If so, best just to drop 2/2 too.
Ya, definitely. Sorry for the noise!