crypto: sun8i-ce - wrap accesses to descriptor address fields

The Allwinner H616 (and later) SoCs support more than 32 bits worth of
physical addresses. To accommodate the larger address space, the CE task
descriptor fields holding addresses are now encoded as "word addresses",
so take the actual address divided by four.
This is true for the fields within the descriptor, but also for the
descriptor base address, in the CE_TDA register.

Wrap all accesses to those fields in a function, which will do the
required division if needed. For now this in unused, so there should be
no change in behaviour.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

drivers/crypto/allwinner/sun8i-ce/sun8i-ce-cipher.c

@@ -190,7 +190,7 @@ static int sun8i_ce_cipher_prepare(struct crypto_engine *engine, void *async_req
 		err = -EFAULT;
 		goto theend;
 	}
-	cet->t_key = cpu_to_le32(rctx->addr_key);
+	cet->t_key = desc_addr_val_le32(ce, rctx->addr_key);
 
 	ivsize = crypto_skcipher_ivsize(tfm);
 	if (areq->iv && crypto_skcipher_ivsize(tfm) > 0) {
@@ -208,7 +208,7 @@ static int sun8i_ce_cipher_prepare(struct crypto_engine *engine, void *async_req
 			err = -ENOMEM;
 			goto theend_iv;
 		}
-		cet->t_iv = cpu_to_le32(rctx->addr_iv);
+		cet->t_iv = desc_addr_val_le32(ce, rctx->addr_iv);
 	}
 
 	if (areq->src == areq->dst) {
@@ -236,7 +236,7 @@ static int sun8i_ce_cipher_prepare(struct crypto_engine *engine, void *async_req
 
 	len = areq->cryptlen;
 	for_each_sg(areq->src, sg, nr_sgs, i) {
-		cet->t_src[i].addr = cpu_to_le32(sg_dma_address(sg));
+		cet->t_src[i].addr = desc_addr_val_le32(ce, sg_dma_address(sg));
 		todo = min(len, sg_dma_len(sg));
 		cet->t_src[i].len = cpu_to_le32(todo / 4);
 		dev_dbg(ce->dev, "%s total=%u SG(%d %u off=%d) todo=%u\n", __func__,
@@ -251,7 +251,7 @@ static int sun8i_ce_cipher_prepare(struct crypto_engine *engine, void *async_req
 
 	len = areq->cryptlen;
 	for_each_sg(areq->dst, sg, nr_sgd, i) {
-		cet->t_dst[i].addr = cpu_to_le32(sg_dma_address(sg));
+		cet->t_dst[i].addr = desc_addr_val_le32(ce, sg_dma_address(sg));
 		todo = min(len, sg_dma_len(sg));
 		cet->t_dst[i].len = cpu_to_le32(todo / 4);
 		dev_dbg(ce->dev, "%s total=%u SG(%d %u off=%d) todo=%u\n", __func__,

drivers/crypto/allwinner/sun8i-ce/sun8i-ce-core.c

@@ -172,7 +172,7 @@ int sun8i_ce_run_task(struct sun8i_ce_dev *ce, int flow, const char *name)
 	writel(v, ce->base + CE_ICR);
 
 	reinit_completion(&ce->chanlist[flow].complete);
-	writel(ce->chanlist[flow].t_phy, ce->base + CE_TDQ);
+	writel(desc_addr_val(ce, ce->chanlist[flow].t_phy), ce->base + CE_TDQ);
 
 	ce->chanlist[flow].status = 0;
 	/* Be sure all data is written before enabling the task */

drivers/crypto/allwinner/sun8i-ce/sun8i-ce-hash.c

@@ -403,7 +403,7 @@ int sun8i_ce_hash_run(struct crypto_engine *engine, void *breq)
 
 	len = areq->nbytes;
 	for_each_sg(areq->src, sg, nr_sgs, i) {
-		cet->t_src[i].addr = cpu_to_le32(sg_dma_address(sg));
+		cet->t_src[i].addr = desc_addr_val_le32(ce, sg_dma_address(sg));
 		todo = min(len, sg_dma_len(sg));
 		cet->t_src[i].len = cpu_to_le32(todo / 4);
 		len -= todo;
@@ -414,7 +414,7 @@ int sun8i_ce_hash_run(struct crypto_engine *engine, void *breq)
 		goto theend;
 	}
 	addr_res = dma_map_single(ce->dev, result, digestsize, DMA_FROM_DEVICE);
-	cet->t_dst[0].addr = cpu_to_le32(addr_res);
+	cet->t_dst[0].addr = desc_addr_val_le32(ce, addr_res);
 	cet->t_dst[0].len = cpu_to_le32(digestsize / 4);
 	if (dma_mapping_error(ce->dev, addr_res)) {
 		dev_err(ce->dev, "DMA map dest\n");
@@ -445,7 +445,7 @@ int sun8i_ce_hash_run(struct crypto_engine *engine, void *breq)
 	}
 
 	addr_pad = dma_map_single(ce->dev, buf, j * 4, DMA_TO_DEVICE);
-	cet->t_src[i].addr = cpu_to_le32(addr_pad);
+	cet->t_src[i].addr = desc_addr_val_le32(ce, addr_pad);
 	cet->t_src[i].len = cpu_to_le32(j);
 	if (dma_mapping_error(ce->dev, addr_pad)) {
 		dev_err(ce->dev, "DMA error on padding SG\n");

drivers/crypto/allwinner/sun8i-ce/sun8i-ce-prng.c

@@ -132,10 +132,10 @@ int sun8i_ce_prng_generate(struct crypto_rng *tfm, const u8 *src,
 	cet->t_sym_ctl = cpu_to_le32(sym);
 	cet->t_asym_ctl = 0;
 
-	cet->t_key = cpu_to_le32(dma_iv);
-	cet->t_iv = cpu_to_le32(dma_iv);
+	cet->t_key = desc_addr_val_le32(ce, dma_iv);
+	cet->t_iv = desc_addr_val_le32(ce, dma_iv);
 
-	cet->t_dst[0].addr = cpu_to_le32(dma_dst);
+	cet->t_dst[0].addr = desc_addr_val_le32(ce, dma_dst);
 	cet->t_dst[0].len = cpu_to_le32(todo / 4);
 	ce->chanlist[flow].timeout = 2000;
 

drivers/crypto/allwinner/sun8i-ce/sun8i-ce-trng.c

@@ -77,7 +77,7 @@ static int sun8i_ce_trng_read(struct hwrng *rng, void *data, size_t max, bool wa
 	cet->t_sym_ctl = 0;
 	cet->t_asym_ctl = 0;
 
-	cet->t_dst[0].addr = cpu_to_le32(dma_dst);
+	cet->t_dst[0].addr = desc_addr_val_le32(ce, dma_dst);
 	cet->t_dst[0].len = cpu_to_le32(todo / 4);
 	ce->chanlist[flow].timeout = todo;
 

drivers/crypto/allwinner/sun8i-ce/sun8i-ce.h

@@ -149,6 +149,7 @@ struct ce_variant {
 	bool hash_t_dlen_in_bits;
 	bool prng_t_dlen_in_bytes;
 	bool trng_t_dlen_in_bytes;
+	bool needs_word_addresses;
 	struct ce_clock ce_clks[CE_MAX_CLOCKS];
 	int esr;
 	unsigned char prng;
@@ -241,6 +242,20 @@ struct sun8i_ce_dev {
 #endif
 };
 
+static inline u32 desc_addr_val(struct sun8i_ce_dev *dev, dma_addr_t addr)
+{
+	if (dev->variant->needs_word_addresses)
+		return addr / 4;
+
+	return addr;
+}
+
+static inline __le32 desc_addr_val_le32(struct sun8i_ce_dev *dev,
+					dma_addr_t addr)
+{
+	return cpu_to_le32(desc_addr_val(dev, addr));
+}
+
 /*
  * struct sun8i_cipher_req_ctx - context for a skcipher request
  * @op_dir:		direction (encrypt vs decrypt) for this request