* Suren Baghdasaryan <surenb@xxxxxxxxxx> [210528 13:31]:
> On Wed, Apr 28, 2021 at 8:36 AM Liam Howlett <liam.howlett@xxxxxxxxxx> wrote:
> >
>
> I know you have v2 for the first part of this patchset, I'm just going
> over the whole thing... There should be some description here of what
> the new struct member and new function are for. Ideally you would also
> split it in two because it introduces two seemingly independent
> additions: non_kernel and kmem_cache_get_alloc.
Your comments are still valid and appreciated.
I did add a description to the patch:
--------------------------------
radix tree test suite: Add kmem_cache enhancements and pr_err
Add kmem_cache_set_non_kernel(), a mechanism to allow a certain number
of kmem_cache_alloc requests to succeed even when GFP_KERNEL is not set
in the flags.
Add kmem_cache_get_alloc() to see the size of the allocated kmem_cache.
Add a define of pr_err to printk.
--------------------------------
I did group these two changes together as they were both affecting
kmem_cache. I will reorganize them into separate commits.
>
> > Signed-off-by: Liam R. Howlett <Liam.Howlett@xxxxxxxxxx>
> > Signed-off-by: Matthew Wilcox (Oracle) <willy@xxxxxxxxxxxxx>
> > ---
> > tools/testing/radix-tree/linux.c | 16 +++++++++++++++-
> > tools/testing/radix-tree/linux/kernel.h | 1 +
> > 2 files changed, 16 insertions(+), 1 deletion(-)
> >
> > diff --git a/tools/testing/radix-tree/linux.c b/tools/testing/radix-tree/linux.c
> > index 2d9c59df60de..93f7de81fbe8 100644
> > --- a/tools/testing/radix-tree/linux.c
> > +++ b/tools/testing/radix-tree/linux.c
> > @@ -24,15 +24,28 @@ struct kmem_cache {
> > int nr_objs;
> > void *objs;
> > void (*ctor)(void *);
> > + unsigned int non_kernel;
> > };
> >
> > +void kmem_cache_set_non_kernel(struct kmem_cache *cachep, unsigned int val)
> > +{
> > + cachep->non_kernel = val;
> > +}
> > +
> > +unsigned long kmem_cache_get_alloc(struct kmem_cache *cachep)
> > +{
> > + return cachep->size * nr_allocated;
>
> IIUC nr_allocated is incremented/decremented every time memory is
> allocated/freed from *any* kmem_cache. Each kmem_cache has its own
> size. So, nr_allocated counts allocated objects of potentially
> different sizes. If that is so then I'm unclear what the result of
> this multiplication would represent.
This is intended to only be used for testing with one kmem_cache, so it
hasn't been an issue. Having this variable exist external to the
kmem_cache struct allows for checking if any allocations remain outside
the scope of the kmem_cache (ie: threads). I think putting it in the
struct would cause issues with the IDR testing. I could make this a new
variable and increment them together but this variable existed for the
IDR testing and I didn't need to support that additional functionality
for my testing. I should at least add a comment about this limitation
though.
>
> > +}
> > void *kmem_cache_alloc(struct kmem_cache *cachep, int gfp)
> > {
> > void *p;
> >
> > - if (!(gfp & __GFP_DIRECT_RECLAIM))
> > + if (!(gfp & __GFP_DIRECT_RECLAIM) && !cachep->non_kernel)
> > return NULL;
> >
> > + if (!(gfp & __GFP_DIRECT_RECLAIM))
> > + cachep->non_kernel--;
> > +
> > pthread_mutex_lock(&cachep->lock);
> > if (cachep->nr_objs) {
> > struct radix_tree_node *node = cachep->objs;
> > @@ -116,5 +129,6 @@ kmem_cache_create(const char *name, unsigned int size, unsigned int align,
> > ret->nr_objs = 0;
> > ret->objs = NULL;
> > ret->ctor = ctor;
> > + ret->non_kernel = 0;
> > return ret;
> > }
> > diff --git a/tools/testing/radix-tree/linux/kernel.h b/tools/testing/radix-tree/linux/kernel.h
> > index 39867fd80c8f..c5c9d05f29da 100644
> > --- a/tools/testing/radix-tree/linux/kernel.h
> > +++ b/tools/testing/radix-tree/linux/kernel.h
> > @@ -14,6 +14,7 @@
> > #include "../../../include/linux/kconfig.h"
> >
> > #define printk printf
> > +#define pr_err printk
> > #define pr_info printk
> > #define pr_debug printk
> > #define pr_cont printk
> > --
> > 2.30.2
