output/vcd: assume packed input data image, unbreak 2nd+ channel

[libsigrok.git] / src / output / vcd.c

diff --git a/src/output/vcd.c b/src/output/vcd.c
index db77a7adeddf09cb37c6c6be9b7745a4de69d6c4..d0b14a04d26f1ba661d57949a52fcbb0a9aba3b3 100644 (file)
--- a/src/output/vcd.c
+++ b/src/output/vcd.c
@@ -200,7 +200,16 @@ static int receive(const struct sr_output *o, const struct sr_datafeed_packet *p
                        timestamp_written = FALSE;
 
                        for (p = 0; p < ctx->num_enabled_channels; p++) {
-                               index = ctx->channel_index[p];
+                               /*
+                                * TODO Check whether the mapping from
+                                * data image positions to channel numbers
+                                * is required. Experiments suggest that
+                                * the data image "is dense", and packs
+                                * bits of enabled channels, and leaves no
+                                * room for positions of disabled channels.
+                                */
+                               /* index = ctx->channel_index[p]; */
+                               index = p;
 
                                curbit = ((unsigned)sample[index / 8]
                                                >> (index % 8)) & 1;