Hello Michael,

Here are the changes in v2 of the patch

1. added some wording to change log to specify why we are seeing the issue now. Also added "CC: stable@vger.kernel.org"

2. changed "limit" to unsigned long. I have kept "rpages" as "long"

Thanks,

Gaurav

On 5/25/23 9:34 AM, Gaurav Batra wrote:

As of now, in tce_freemulti_pSeriesLP(), there is no limit on how many TCEs are passed to H_STUFF_TCE hcall. This was not an issue until now. Newer firmware releases have started enforcing this requirement.

The interface has been in it's current form since the beginning.

Cc: stable@vger.kernel.org

Signed-off-by: Gaurav Batra gbatra@linux.vnet.ibm.com Reviewed-by: Brian King brking@linux.vnet.ibm.com

---

arch/powerpc/platforms/pseries/iommu.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/platforms/pseries/iommu.c b/arch/powerpc/platforms/pseries/iommu.c index c74b71d4733d..f159a195101d 100644 --- a/arch/powerpc/platforms/pseries/iommu.c +++ b/arch/powerpc/platforms/pseries/iommu.c @@ -306,13 +306,22 @@ static void tce_free_pSeriesLP(unsigned long liobn, long tcenum, long tceshift, static void tce_freemulti_pSeriesLP(struct iommu_table *tbl, long tcenum, long npages) { u64 rc;

• long rpages = npages;

• unsigned long limit;

  if (!firmware_has_feature(FW_FEATURE_STUFF_TCE)) return tce_free_pSeriesLP(tbl->it_index, tcenum, tbl->it_page_shift, npages);

• rc = plpar_tce_stuff((u64)tbl->it_index,

• 	     (u64)tcenum << tbl->it_page_shift, 0, npages);
  

• do {

• limit = min_t(unsigned long, rpages, 512);
  

• rc = plpar_tce_stuff((u64)tbl->it_index,
  

• 	     	(u64)tcenum << tbl->it_page_shift, 0, limit);
  

• rpages -= limit;
  

• tcenum += limit;
  

• } while (rpages > 0 && !rc);

  if (rc && printk_ratelimit()) { printk("tce_freemulti_pSeriesLP: plpar_tce_stuff failed\n");