Hi all,
I decided that Pasha's patch[1] was a good excuse to tackle a bit more
cleanup of the existing IOVA code, so it can fit in more neatly. Thus
to save time and effort I've taken the liberty of putting together this
mini-series as the version I'd like to merge.
Cheers,
Robin.
[1] https://lore.kernel.org/linux-iommu/[email protected]/
Pasha Tatashin (1):
iommu/iova: use named kmem_cache for iova magazines
Robin Murphy (2):
iommu/iova: Tidy up iova_cache_get() failure
iommu/iova: Reorganise some code
drivers/iommu/iova.c | 143 +++++++++++++++++++++++--------------------
1 file changed, 76 insertions(+), 67 deletions(-)
--
2.39.2.101.g768bb238c484.dirty
Failure handling in iova_cache_get() is a little messy, and we'd like
to add some more to it, so let's tidy up a bit first. By leaving the
hotplug handler until last we can take advantage of kmem_cache_destroy()
being NULL-safe to have a single cleanup label. We can also improve the
error reporting, noting that kmem_cache_create() already screams if it
fails, so that one is redundant.
Signed-off-by: Robin Murphy <[email protected]>
---
drivers/iommu/iova.c | 33 ++++++++++++++++-----------------
1 file changed, 16 insertions(+), 17 deletions(-)
diff --git a/drivers/iommu/iova.c b/drivers/iommu/iova.c
index d30e453d0fb4..cf95001d85c0 100644
--- a/drivers/iommu/iova.c
+++ b/drivers/iommu/iova.c
@@ -254,26 +254,20 @@ static void free_iova_mem(struct iova *iova)
int iova_cache_get(void)
{
+ int err = -ENOMEM;
+
mutex_lock(&iova_cache_mutex);
if (!iova_cache_users) {
- int ret;
+ iova_cache = kmem_cache_create("iommu_iova", sizeof(struct iova), 0,
+ SLAB_HWCACHE_ALIGN, NULL);
+ if (!iova_cache)
+ goto out_err;
- ret = cpuhp_setup_state_multi(CPUHP_IOMMU_IOVA_DEAD, "iommu/iova:dead", NULL,
- iova_cpuhp_dead);
- if (ret) {
- mutex_unlock(&iova_cache_mutex);
- pr_err("Couldn't register cpuhp handler\n");
- return ret;
- }
-
- iova_cache = kmem_cache_create(
- "iommu_iova", sizeof(struct iova), 0,
- SLAB_HWCACHE_ALIGN, NULL);
- if (!iova_cache) {
- cpuhp_remove_multi_state(CPUHP_IOMMU_IOVA_DEAD);
- mutex_unlock(&iova_cache_mutex);
- pr_err("Couldn't create iova cache\n");
- return -ENOMEM;
+ err = cpuhp_setup_state_multi(CPUHP_IOMMU_IOVA_DEAD, "iommu/iova:dead",
+ NULL, iova_cpuhp_dead);
+ if (err) {
+ pr_err("IOVA: Couldn't register cpuhp handler: %pe\n", ERR_PTR(err));
+ goto out_err;
}
}
@@ -281,6 +275,11 @@ int iova_cache_get(void)
mutex_unlock(&iova_cache_mutex);
return 0;
+
+out_err:
+ kmem_cache_destroy(iova_cache);
+ mutex_unlock(&iova_cache_mutex);
+ return err;
}
EXPORT_SYMBOL_GPL(iova_cache_get);
--
2.39.2.101.g768bb238c484.dirty
On Mon, 5 Feb 2024, Robin Murphy wrote:
> Failure handling in iova_cache_get() is a little messy, and we'd like
> to add some more to it, so let's tidy up a bit first. By leaving the
> hotplug handler until last we can take advantage of kmem_cache_destroy()
> being NULL-safe to have a single cleanup label. We can also improve the
> error reporting, noting that kmem_cache_create() already screams if it
> fails, so that one is redundant.
>
> Signed-off-by: Robin Murphy <[email protected]>
Much easier to follow :)
Acked-by: David Rientjes <[email protected]>
On 05/02/2024 15:32, Robin Murphy wrote:
> Failure handling in iova_cache_get() is a little messy, and we'd like
> to add some more to it, so let's tidy up a bit first. By leaving the
> hotplug handler until last we can take advantage of kmem_cache_destroy()
> being NULL-safe to have a single cleanup label. We can also improve the
> error reporting, noting that kmem_cache_create() already screams if it
> fails, so that one is redundant.
>
> Signed-off-by: Robin Murphy <[email protected]>
Regardless of a couple of minor comments, below, FWIW:
Reviewed-by: John Garry <[email protected]>
> ---
> drivers/iommu/iova.c | 33 ++++++++++++++++-----------------
> 1 file changed, 16 insertions(+), 17 deletions(-)
>
> diff --git a/drivers/iommu/iova.c b/drivers/iommu/iova.c
> index d30e453d0fb4..cf95001d85c0 100644
> --- a/drivers/iommu/iova.c
> +++ b/drivers/iommu/iova.c
> @@ -254,26 +254,20 @@ static void free_iova_mem(struct iova *iova)
>
> int iova_cache_get(void)
> {
> + int err = -ENOMEM;
> +
> mutex_lock(&iova_cache_mutex);
> if (!iova_cache_users) {
> - int ret;
> + iova_cache = kmem_cache_create("iommu_iova", sizeof(struct iova), 0,
> + SLAB_HWCACHE_ALIGN, NULL);
Maybe can use KMEM_CACHE(), but it would mean that the name would change.
> + if (!iova_cache)
> + goto out_err;
>
> - ret = cpuhp_setup_state_multi(CPUHP_IOMMU_IOVA_DEAD, "iommu/iova:dead", NULL,
> - iova_cpuhp_dead);
> - if (ret) {
> - mutex_unlock(&iova_cache_mutex);
> - pr_err("Couldn't register cpuhp handler\n");
> - return ret;
> - }
> -
> - iova_cache = kmem_cache_create(
> - "iommu_iova", sizeof(struct iova), 0,
> - SLAB_HWCACHE_ALIGN, NULL);
> - if (!iova_cache) {
> - cpuhp_remove_multi_state(CPUHP_IOMMU_IOVA_DEAD);
> - mutex_unlock(&iova_cache_mutex);
> - pr_err("Couldn't create iova cache\n");
> - return -ENOMEM;
> + err = cpuhp_setup_state_multi(CPUHP_IOMMU_IOVA_DEAD, "iommu/iova:dead",
> + NULL, iova_cpuhp_dead);
> + if (err) {
> + pr_err("IOVA: Couldn't register cpuhp handler: %pe\n", ERR_PTR(err));
Maybe can use pr_fmt with "iova".
> + goto out_err;
> }
> }
>
> @@ -281,6 +275,11 @@ int iova_cache_get(void)
> mutex_unlock(&iova_cache_mutex);
>
> return 0;
> +
> +out_err:
> + kmem_cache_destroy(iova_cache);
> + mutex_unlock(&iova_cache_mutex);
> + return err;
> }
> EXPORT_SYMBOL_GPL(iova_cache_get);
>
On 06/02/2024 11:01 am, John Garry wrote:
> On 05/02/2024 15:32, Robin Murphy wrote:
>> Failure handling in iova_cache_get() is a little messy, and we'd like
>> to add some more to it, so let's tidy up a bit first. By leaving the
>> hotplug handler until last we can take advantage of kmem_cache_destroy()
>> being NULL-safe to have a single cleanup label. We can also improve the
>> error reporting, noting that kmem_cache_create() already screams if it
>> fails, so that one is redundant.
>>
>> Signed-off-by: Robin Murphy <[email protected]>
>
> Regardless of a couple of minor comments, below, FWIW:
> Reviewed-by: John Garry <[email protected]>
Thanks!
>> ---
>> drivers/iommu/iova.c | 33 ++++++++++++++++-----------------
>> 1 file changed, 16 insertions(+), 17 deletions(-)
>>
>> diff --git a/drivers/iommu/iova.c b/drivers/iommu/iova.c
>> index d30e453d0fb4..cf95001d85c0 100644
>> --- a/drivers/iommu/iova.c
>> +++ b/drivers/iommu/iova.c
>> @@ -254,26 +254,20 @@ static void free_iova_mem(struct iova *iova)
>> int iova_cache_get(void)
>> {
>> + int err = -ENOMEM;
>> +
>> mutex_lock(&iova_cache_mutex);
>> if (!iova_cache_users) {
>> - int ret;
>> + iova_cache = kmem_cache_create("iommu_iova", sizeof(struct
>> iova), 0,
>> + SLAB_HWCACHE_ALIGN, NULL);
>
> Maybe can use KMEM_CACHE(), but it would mean that the name would change.
Oh, I never came across that before. On reflection, I am inclined to
think that the "iommu_" names are usefully informative to userspace, and
it's not worth churning the structure names themselves just to save a
couple of lines here.
>> + if (!iova_cache)
>> + goto out_err;
>> - ret = cpuhp_setup_state_multi(CPUHP_IOMMU_IOVA_DEAD,
>> "iommu/iova:dead", NULL,
>> - iova_cpuhp_dead);
>> - if (ret) {
>> - mutex_unlock(&iova_cache_mutex);
>> - pr_err("Couldn't register cpuhp handler\n");
>> - return ret;
>> - }
>> -
>> - iova_cache = kmem_cache_create(
>> - "iommu_iova", sizeof(struct iova), 0,
>> - SLAB_HWCACHE_ALIGN, NULL);
>> - if (!iova_cache) {
>> - cpuhp_remove_multi_state(CPUHP_IOMMU_IOVA_DEAD);
>> - mutex_unlock(&iova_cache_mutex);
>> - pr_err("Couldn't create iova cache\n");
>> - return -ENOMEM;
>> + err = cpuhp_setup_state_multi(CPUHP_IOMMU_IOVA_DEAD,
>> "iommu/iova:dead",
>> + NULL, iova_cpuhp_dead);
>> + if (err) {
>> + pr_err("IOVA: Couldn't register cpuhp handler: %pe\n",
>> ERR_PTR(err));
>
> Maybe can use pr_fmt with "iova".
That one I did consider, but since we now have just this one print and
no imminent reason to add more, I figured I'd just stick it inline, and
we can factor out a pr_fmt in future if we ever have cause to.
Cheers,
Robin.
>
>> + goto out_err;
>> }
>> }
>> @@ -281,6 +275,11 @@ int iova_cache_get(void)
>> mutex_unlock(&iova_cache_mutex);
>> return 0;
>> +
>> +out_err:
>> + kmem_cache_destroy(iova_cache);
>> + mutex_unlock(&iova_cache_mutex);
>> + return err;
>> }
>> EXPORT_SYMBOL_GPL(iova_cache_get);
>
On Mon, Feb 05, 2024 at 03:32:39PM +0000, Robin Murphy wrote:
> Failure handling in iova_cache_get() is a little messy, and we'd like
> to add some more to it, so let's tidy up a bit first. By leaving the
> hotplug handler until last we can take advantage of kmem_cache_destroy()
> being NULL-safe to have a single cleanup label. We can also improve the
> error reporting, noting that kmem_cache_create() already screams if it
> fails, so that one is redundant.
>
> Signed-off-by: Robin Murphy <[email protected]>
> ---
> drivers/iommu/iova.c | 33 ++++++++++++++++-----------------
> 1 file changed, 16 insertions(+), 17 deletions(-)
>
Reviewed-by: Jerry Snitselaar <[email protected]>
On Mon, Feb 05, 2024 at 03:32:38PM +0000, Robin Murphy wrote:
> Pasha Tatashin (1):
> iommu/iova: use named kmem_cache for iova magazines
>
> Robin Murphy (2):
> iommu/iova: Tidy up iova_cache_get() failure
> iommu/iova: Reorganise some code
>
> drivers/iommu/iova.c | 143 +++++++++++++++++++++++--------------------
> 1 file changed, 76 insertions(+), 67 deletions(-)
Applied, thanks Robin.