[PATCH v3 1/1] powerpc/iommu: Enable remaining IOMMU Pagesizes present in LoPAR

[Leonardo Bras <leobras.c@gmail.com>]

According to LoPAR, ibm,query-pe-dma-window output named "IO Page Sizes"
will let the OS know all possible pagesizes that can be used for creating a
new DDW.

Currently Linux will only try using 3 of the 8 available options:
4K, 64K and 16M. According to LoPAR, Hypervisor may also offer 32M, 64M,
128M, 256M and 16G.

Enabling bigger pages would be interesting for direct mapping systems
with a lot of RAM, while using less TCE entries.

Signed-off-by: Leonardo Bras <leobras.c@gmail.com>
---
Changes since v2:
 - Restore 'int array & shift' strategy
 - Remove defines for RTAS "IO Page Size" output of ibm,query-pe-dma-window
 - Added/Improved comments
Link: http://patchwork.ozlabs.org/project/linuxppc-dev/patch/20210407195613.131140-1-leobras.c@gmail.com/
Changes since v1:
- Remove page shift defines, replace by __builtin_ctzll(SZ_XXX)
- Add bit field defines for RTAS "IO Page Shift" output of ibm,query-pe-dma-window
- Use struct array instead of int array to be more explicit on pagesizes
Link: http://patchwork.ozlabs.org/project/linuxppc-dev/patch/20210322190943.715368-1-leobras.c@gmail.com/ 

 arch/powerpc/platforms/pseries/iommu.c | 37 +++++++++++++++++++++-----
 1 file changed, 30 insertions(+), 7 deletions(-)

diff --git a/arch/powerpc/platforms/pseries/iommu.c b/arch/powerpc/platforms/pseries/iommu.c
index 9fc5217f0c8e..67c9953a6503 100644
--- a/arch/powerpc/platforms/pseries/iommu.c
+++ b/arch/powerpc/platforms/pseries/iommu.c
@@ -1099,6 +1099,33 @@ static void reset_dma_window(struct pci_dev *dev, struct device_node *par_dn)
 			 ret);
 }
 
+/* Return largest page shift based on "IO Page Sizes" output of ibm,query-pe-dma-window. */
+static int iommu_get_page_shift(u32 query_page_size)
+{
+	/* Supported IO page-sizes according to LoPAR */
+	const int shift[] = {
+		__builtin_ctzll(SZ_4K),   __builtin_ctzll(SZ_64K), __builtin_ctzll(SZ_16M),
+		__builtin_ctzll(SZ_32M),  __builtin_ctzll(SZ_64M), __builtin_ctzll(SZ_128M),
+		__builtin_ctzll(SZ_256M), __builtin_ctzll(SZ_16G)
+	};
+
+	int i = ARRAY_SIZE(shift) - 1;
+
+	/*
+	 * On LoPAR, ibm,query-pe-dma-window outputs "IO Page Sizes" using a bit field:
+	 * - bit 31 means 4k pages are supported,
+	 * - bit 30 means 64k pages are supported, and so on.
+	 * Larger pagesizes map more memory with the same amount of TCEs, so start probing them.
+	 */
+	for (; i >= 0 ; i--) {
+		if (query_page_size & (1 << i))
+			return shift[i];
+	}
+
+	/* No valid page size found. */
+	return 0;
+}
+
 /*
  * If the PE supports dynamic dma windows, and there is space for a table
  * that can map all pages in a linear offset, then setup such a table,
@@ -1206,13 +1233,9 @@ static u64 enable_ddw(struct pci_dev *dev, struct device_node *pdn)
 			goto out_failed;
 		}
 	}
-	if (query.page_size & 4) {
-		page_shift = 24; /* 16MB */
-	} else if (query.page_size & 2) {
-		page_shift = 16; /* 64kB */
-	} else if (query.page_size & 1) {
-		page_shift = 12; /* 4kB */
-	} else {
+
+	page_shift = iommu_get_page_shift(query.page_size);
+	if (!page_shift) {
 		dev_dbg(&dev->dev, "no supported direct page size in mask %x",
 			  query.page_size);
 		goto out_failed;
-- 
2.30.2