@@ -12,6 +12,7 @@
#include <linux/module.h>
#include <linux/of_address.h>
#include <linux/of.h>
+#include <linux/sort.h>
#include <linux/of_platform.h>
#include <linux/rpmsg.h>
#include <linux/scatterlist.h>
@@ -104,6 +105,15 @@ struct fastrpc_invoke_rsp {
int retval; /* invoke return value */
};
+struct fastrpc_buf_overlap {
+ u64 start;
+ u64 end;
+ int raix;
+ u64 mstart;
+ u64 mend;
+ u64 offset;
+};
+
struct fastrpc_buf {
struct fastrpc_user *fl;
struct dma_buf *dmabuf;
@@ -156,6 +166,7 @@ struct fastrpc_invoke_ctx {
struct fastrpc_map **maps;
struct fastrpc_buf *buf;
struct fastrpc_invoke_args *args;
+ struct fastrpc_buf_overlap *olaps;
struct fastrpc_channel_ctx *cctx;
};
@@ -300,6 +311,7 @@ static void fastrpc_context_free(struct kref *ref)
spin_unlock_irqrestore(&cctx->lock, flags);
kfree(ctx->maps);
+ kfree(ctx->olaps);
kfree(ctx);
}
@@ -321,6 +333,55 @@ static void fastrpc_context_put_wq(struct work_struct *work)
fastrpc_context_put(ctx);
}
+#define CMP(aa, bb) ((aa) == (bb) ? 0 : (aa) < (bb) ? -1 : 1)
+static int olaps_cmp(const void *a, const void *b)
+{
+ struct fastrpc_buf_overlap *pa = (struct fastrpc_buf_overlap *)a;
+ struct fastrpc_buf_overlap *pb = (struct fastrpc_buf_overlap *)b;
+ /* sort with lowest starting buffer first */
+ int st = CMP(pa->start, pb->start);
+ /* sort with highest ending buffer first */
+ int ed = CMP(pb->end, pa->end);
+
+ return st == 0 ? ed : st;
+}
+
+static void fastrpc_get_buff_overlaps(struct fastrpc_invoke_ctx *ctx)
+{
+ u64 max_end = 0;
+ int i;
+
+ for (i = 0; i < ctx->nbufs; ++i) {
+ ctx->olaps[i].start = ctx->args[i].ptr;
+ ctx->olaps[i].end = ctx->olaps[i].start + ctx->args[i].length;
+ ctx->olaps[i].raix = i;
+ }
+
+ sort(ctx->olaps, ctx->nbufs, sizeof(*ctx->olaps), olaps_cmp, NULL);
+
+ for (i = 0; i < ctx->nbufs; ++i) {
+ /* Falling inside previous range */
+ if (ctx->olaps[i].start < max_end) {
+ ctx->olaps[i].mstart = max_end;
+ ctx->olaps[i].mend = ctx->olaps[i].end;
+ ctx->olaps[i].offset = max_end - ctx->olaps[i].start;
+
+ if (ctx->olaps[i].end > max_end) {
+ max_end = ctx->olaps[i].end;
+ } else {
+ ctx->olaps[i].mend = 0;
+ ctx->olaps[i].mstart = 0;
+ }
+
+ } else {
+ ctx->olaps[i].mend = ctx->olaps[i].end;
+ ctx->olaps[i].mstart = ctx->olaps[i].start;
+ ctx->olaps[i].offset = 0;
+ max_end = ctx->olaps[i].end;
+ }
+ }
+}
+
static struct fastrpc_invoke_ctx *fastrpc_context_alloc(
struct fastrpc_user *user, u32 kernel, u32 sc,
struct fastrpc_invoke_args *args)
@@ -347,7 +408,15 @@ static struct fastrpc_invoke_ctx *fastrpc_context_alloc(
kfree(ctx);
return ERR_PTR(-ENOMEM);
}
+ ctx->olaps = kcalloc(ctx->nscalars,
+ sizeof(*ctx->olaps), GFP_KERNEL);
+ if (!ctx->olaps) {
+ kfree(ctx->maps);
+ kfree(ctx);
+ return ERR_PTR(-ENOMEM);
+ }
ctx->args = args;
+ fastrpc_get_buff_overlaps(ctx);
}
ctx->sc = sc;
@@ -380,6 +449,7 @@ static struct fastrpc_invoke_ctx *fastrpc_context_alloc(
list_del(&ctx->node);
spin_unlock(&user->lock);
kfree(ctx->maps);
+ kfree(ctx->olaps);
kfree(ctx);
return ERR_PTR(ret);
@@ -598,8 +668,11 @@ static u64 fastrpc_get_payload_size(struct fastrpc_invoke_ctx *ctx, int metalen)
size = ALIGN(metalen, FASTRPC_ALIGN);
for (i = 0; i < ctx->nscalars; i++) {
if (ctx->args[i].fd == 0 || ctx->args[i].fd == -1) {
- size = ALIGN(size, FASTRPC_ALIGN);
- size += ctx->args[i].length;
+
+ if (ctx->olaps[i].offset == 0)
+ size = ALIGN(size, FASTRPC_ALIGN);
+
+ size += (ctx->olaps[i].mend - ctx->olaps[i].mstart);
}
}
@@ -637,12 +710,11 @@ static int fastrpc_get_args(u32 kernel, struct fastrpc_invoke_ctx *ctx)
struct fastrpc_remote_arg *rpra;
struct fastrpc_invoke_buf *list;
struct fastrpc_phy_page *pages;
- int inbufs, i, err = 0;
- u64 rlen, pkt_size;
+ int inbufs, i, oix, err = 0;
+ u64 len, rlen, pkt_size;
uintptr_t args;
int metalen;
-
inbufs = REMOTE_SCALARS_INBUFS(ctx->sc);
metalen = fastrpc_get_meta_size(ctx);
pkt_size = fastrpc_get_payload_size(ctx, metalen);
@@ -665,8 +737,11 @@ static int fastrpc_get_args(u32 kernel, struct fastrpc_invoke_ctx *ctx)
rlen = pkt_size - metalen;
ctx->rpra = rpra;
- for (i = 0; i < ctx->nbufs; ++i) {
- u64 len = ctx->args[i].length;
+ for (oix = 0; oix < ctx->nbufs; ++oix) {
+ int mlen;
+
+ i = ctx->olaps[oix].raix;
+ len = ctx->args[i].length;
rpra[i].pv = 0;
rpra[i].len = len;
@@ -690,16 +765,25 @@ static int fastrpc_get_args(u32 kernel, struct fastrpc_invoke_ctx *ctx)
vma->vm_start;
} else {
- rlen -= ALIGN(args, FASTRPC_ALIGN) - args;
- args = ALIGN(args, FASTRPC_ALIGN);
- if (rlen < len)
+
+ if (ctx->olaps[oix].offset == 0) {
+ rlen -= ALIGN(args, FASTRPC_ALIGN) - args;
+ args = ALIGN(args, FASTRPC_ALIGN);
+ }
+
+ mlen = ctx->olaps[oix].mend - ctx->olaps[oix].mstart;
+
+ if (rlen < mlen)
goto bail;
- rpra[i].pv = args;
- pages[i].addr = ctx->buf->phys + (pkt_size - rlen);
+ rpra[i].pv = args - ctx->olaps[oix].offset;
+ pages[i].addr = ctx->buf->phys -
+ ctx->olaps[oix].offset +
+ (pkt_size - rlen);
pages[i].addr = pages[i].addr & PAGE_MASK;
- args = args + len;
- rlen -= len;
+
+ args = args + mlen;
+ rlen -= mlen;
}
if (i < inbufs && !ctx->maps[i]) {
Argument buffers that are passed could be derived from a big buffer, and some of the arguments buffers could overlap each other. Take care of such instanaces. This is optimization that DSP expects while sending buffers which overlap. So make the DSP happy doing it. Without which DSP seems to crash. Fixes: c68cfb718c8f ("misc: fastrpc: Add support for context Invoke method") Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> --- drivers/misc/fastrpc.c | 112 +++++++++++++++++++++++++++++++++++------ 1 file changed, 98 insertions(+), 14 deletions(-) -- 2.21.0