Send commitlog mailing list submissions to
commitlog@lists.openmoko.org
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.openmoko.org/mailman/listinfo/commitlog
or, via email, send a message with subject or body 'help' to
commitlog-requ...@lists.openmoko.org
You can reach the person managing the list at
commitlog-ow...@lists.openmoko.org
When replying, please edit your Subject line so it is more specific
than "Re: Contents of commitlog digest..."
Today's Topics:
1. r5445 - trunk/eda/fped (wer...@docs.openmoko.org)
2. r5446 - trunk/eda/fped (wer...@docs.openmoko.org)
3. r5447 - trunk/gta02-core/docs/ecn (wer...@docs.openmoko.org)
4. r5448 - trunk/gta02-core/docs/ecn (wer...@docs.openmoko.org)
5. r5449 - trunk/gta02-core/docs/ecn (wer...@docs.openmoko.org)
--- Begin Message ---
Author: werner
Date: 2009-08-14 17:53:21 +0200 (Fri, 14 Aug 2009)
New Revision: 5445
Modified:
trunk/eda/fped/inst.c
Log:
- when selecting an item in the canvas, we now check that the underlying object
is still connected. This prevents surprises when making extensive changes
during instantiation failure.
Modified: trunk/eda/fped/inst.c
===================================================================
--- trunk/eda/fped/inst.c 2009-08-14 14:39:17 UTC (rev 5444)
+++ trunk/eda/fped/inst.c 2009-08-14 15:53:21 UTC (rev 5445)
@@ -40,7 +40,10 @@
static unsigned long active_set = 0;
+static struct inst_ops vec_ops;
+static struct inst_ops frame_ops;
+
#define IS_ACTIVE ((active_set & 1))
@@ -87,13 +90,44 @@
}
-/* ----- selection --------------------------------------------------------- */
+/* ----- check connectedness ----------------------------------------------- */
-static struct inst_ops vec_ops;
-static struct inst_ops frame_ops;
+/*
+ * After an instantiation failure, the instances can get out of sync with the
+ * object tree, and attempts to select an item on the canvas can cause accesses
+ * to objects that aren't there anymore. So we need to check if we can still
+ * reach the corresponding object.
+ *
+ * Note: even this isn't bullet-proof. Theoretically, we may get a new object
+ * in the old place. However, this probably doesn't do any serious damage.
+ */
+static int inst_connected(const struct inst *inst)
+{
+ const struct frame *frame;
+ const struct vec *vec;
+ const struct obj *obj;
+
+ for (frame = frames; frame; frame = frame->next) {
+ if (inst->ops == &vec_ops) {
+ for (vec = frame->vecs; vec; vec = vec->next)
+ if (vec == inst->vec)
+ return 1;
+ } else {
+ for (obj = frame->objs; obj; obj = obj->next)
+ if (obj == inst->obj)
+ return 1;
+ }
+ }
+ return 0;
+}
+
+
+/* ----- selection --------------------------------------------------------- */
+
+
static void set_path(int on)
{
struct inst *inst;
@@ -140,6 +174,8 @@
for (inst = insts[prio]; inst; inst = inst->next) {
if (!inst->active || !inst->ops->distance)
continue;
+ if (!inst_connected(inst))
+ continue;
dist = inst->ops->distance(inst, pos, draw_ctx.scale);
if (dist >= 0 && (!selected_inst || best_dist > dist)) {
selected_inst = inst;
@@ -158,6 +194,8 @@
for (inst = insts[ip_vec]; inst; inst = inst->next) {
if (!inst->active)
continue;
+ if (!inst_connected(inst))
+ continue;
dist = gui_dist_vec_fallback(inst, pos, draw_ctx.scale);
if (dist >= 0 && (!selected_inst || best_dist > dist)) {
selected_inst = inst;
--- End Message ---
--- Begin Message ---
Author: werner
Date: 2009-08-15 00:12:16 +0200 (Sat, 15 Aug 2009)
New Revision: 5446
Modified:
trunk/eda/fped/fpd.y
trunk/eda/fped/gui_frame.c
Log:
- KiCad is very liberal when it comes to valid part names, and so are we now
(i.e., we now accept anything but control and non-ASCII characters)
Modified: trunk/eda/fped/fpd.y
===================================================================
--- trunk/eda/fped/fpd.y 2009-08-14 15:53:21 UTC (rev 5445)
+++ trunk/eda/fped/fpd.y 2009-08-14 22:12:16 UTC (rev 5446)
@@ -207,7 +207,7 @@
YYABORT;
}
for (p = $2; *p; *p++)
- if (!is_id_char(*p, 0)) {
+ if (*p < 32 || *p > 126) {
yyerrorf("invalid part name");
YYABORT;
}
Modified: trunk/eda/fped/gui_frame.c
===================================================================
--- trunk/eda/fped/gui_frame.c 2009-08-14 15:53:21 UTC (rev 5445)
+++ trunk/eda/fped/gui_frame.c 2009-08-14 22:12:16 UTC (rev 5446)
@@ -1120,9 +1120,11 @@
{
if (!*s)
return 0;
- while (*s)
- if (!is_id_char(*s++, 0))
+ while (*s) {
+ if (*s < 32 || *s > 126)
return 0;
+ s++;
+ }
return 1;
}
--- End Message ---
--- Begin Message ---
Author: werner
Date: 2009-08-15 01:16:40 +0200 (Sat, 15 Aug 2009)
New Revision: 5447
Modified:
trunk/gta02-core/docs/ecn/README
trunk/gta02-core/docs/ecn/STATUS
trunk/gta02-core/docs/ecn/ecn0009.txt
trunk/gta02-core/docs/ecn/ecn0012.txt
Log:
- ecn0009 moves from Discuss to Execute. Also got reviewed by Joerg.
- README: added new state "Done". Until now, the only way to stop working
on an ECNs was to cancel or defer it :-)
Modified: trunk/gta02-core/docs/ecn/README
===================================================================
--- trunk/gta02-core/docs/ecn/README 2009-08-14 22:12:16 UTC (rev 5446)
+++ trunk/gta02-core/docs/ecn/README 2009-08-14 23:16:40 UTC (rev 5447)
@@ -102,6 +102,9 @@
- Execute: ECN is being implemented and reviewed.
+- Done: ECN has been implemented and the implementation has been
+ reviewed.
+
- Defer: ECN will not be implemented in the current hardware revision
but may be considered for future revisions.
Modified: trunk/gta02-core/docs/ecn/STATUS
===================================================================
--- trunk/gta02-core/docs/ecn/STATUS 2009-08-14 22:12:16 UTC (rev 5446)
+++ trunk/gta02-core/docs/ecn/STATUS 2009-08-14 23:16:40 UTC (rev 5447)
@@ -8,7 +8,7 @@
0006 Discuss Either remove or populate R1701 (PMU.ADAPTSNS pull-down)
0007 Discuss Remove KEEPACT net and connect PMU.KEEPACT directly to IO_3V3
0008 Edit Remove external GPS antenna connector and circuit (CON7602)
-0009 Discuss Remove LED transistors (Q1501 and Q1502)
+0009 Execute Remove LED transistors (Q1501 and Q1502)
0010 Discuss Remove audio amplifier (U4101)
0011 Edit Reduce number of varistors on USB_VBUS (C4902, C4909)
0012 Defer Add fuse to USB_VBUS
Modified: trunk/gta02-core/docs/ecn/ecn0009.txt
===================================================================
--- trunk/gta02-core/docs/ecn/ecn0009.txt 2009-08-14 22:12:16 UTC (rev
5446)
+++ trunk/gta02-core/docs/ecn/ecn0009.txt 2009-08-14 23:16:40 UTC (rev
5447)
@@ -49,3 +49,4 @@
Author: Werner Almesberger <wer...@openmoko.org>
Review: Dave Ball <openm...@daveball.org.uk>, rev5338, reverse the logic
for gta02-core, less hassle than sourcing new LED's.
+Review: Joerg Reisenweber <jo...@openmoko.org>, SVN 5446
Modified: trunk/gta02-core/docs/ecn/ecn0012.txt
===================================================================
--- trunk/gta02-core/docs/ecn/ecn0012.txt 2009-08-14 22:12:16 UTC (rev
5446)
+++ trunk/gta02-core/docs/ecn/ecn0012.txt 2009-08-14 23:16:40 UTC (rev
5447)
@@ -4,3 +4,6 @@
The GTA03-6410 design added a small fuse of USB_VBUS.
To keep sourcing for gta02-core simple, we should defer this change.
+
+
+Author: Werner Almesberger <wer...@openmoko.org>
--- End Message ---
--- Begin Message ---
Author: werner
Date: 2009-08-15 01:32:47 +0200 (Sat, 15 Aug 2009)
New Revision: 5448
Added:
trunk/gta02-core/docs/ecn/ecn0028.txt
Modified:
trunk/gta02-core/docs/ecn/STATUS
trunk/gta02-core/docs/ecn/ecn0025.txt
Log:
- ECN0025 moves to Done. First one to cross the finishing line.
- added ECN0028, proposing removal of U4905
Modified: trunk/gta02-core/docs/ecn/STATUS
===================================================================
--- trunk/gta02-core/docs/ecn/STATUS 2009-08-14 23:16:40 UTC (rev 5447)
+++ trunk/gta02-core/docs/ecn/STATUS 2009-08-14 23:32:47 UTC (rev 5448)
@@ -24,6 +24,7 @@
0022 Discuss Apply bass fix
0023 Edit Fix #1024 (C1009)
0024 Edit Improve GPS ground and PCB layout
-0025 Discuss Connect CLKOUT1 to UEXTCLK (U1501)
+0025 Done Connect CLKOUT1 to UEXTCLK (U1501)
0026 Discuss Improve bus termination at sd-card
0027 Edit Renamed some nets for clarity
+0028 Edit Remove USB host power switch
Modified: trunk/gta02-core/docs/ecn/ecn0025.txt
===================================================================
--- trunk/gta02-core/docs/ecn/ecn0025.txt 2009-08-14 23:16:40 UTC (rev
5447)
+++ trunk/gta02-core/docs/ecn/ecn0025.txt 2009-08-14 23:32:47 UTC (rev
5448)
@@ -12,3 +12,4 @@
Author: Rask Ingemann Lambertsen <ccc94...@vip.cybercity.dk>
Review: Dave Ball <openm...@daveball.org.uk>, rev5338.
Commit: SVN 5339, Connect CLKOUT1 to UEXTCLK
+Review: Werner Almesberger <wer...@openmoko.org>, SVN 5447
Added: trunk/gta02-core/docs/ecn/ecn0028.txt
===================================================================
--- trunk/gta02-core/docs/ecn/ecn0028.txt (rev 0)
+++ trunk/gta02-core/docs/ecn/ecn0028.txt 2009-08-14 23:32:47 UTC (rev
5448)
@@ -0,0 +1,23 @@
+Remove USB "charge path" power switch U4905
+
+
+We can just disable USB host power at the PMU, so U4905 has no real
+purpose. Note that an overvoltage-limiting component should be used in
+the future, as explained in ECN0014.
+
+Hardware changes:
+
+- remove U4905
+- eliminate net USB_VBUS and just call it VBUS (i.e., merge the two)
+- remove C4911, C4913, R4920
+- remove C4909 (see also ECN0011)
+- keep R1918 but move it to the up-converter delivering host mode power
+- eliminate net nUSB_OC, freeing GPG9
+
+Software changes:
+
+- when switching to USB host mode, make sure the PMU MBC is in USB
+ suspend mode (MBCC7.usbdevstat = 3)
+
+
+Author: Werner Almesberger <wer...@openmoko.org>
--- End Message ---
--- Begin Message ---
Author: werner
Date: 2009-08-15 01:34:25 +0200 (Sat, 15 Aug 2009)
New Revision: 5449
Modified:
trunk/gta02-core/docs/ecn/STATUS
Log:
Oops, title of ECN0028 was wrong in STATUS.
Modified: trunk/gta02-core/docs/ecn/STATUS
===================================================================
--- trunk/gta02-core/docs/ecn/STATUS 2009-08-14 23:32:47 UTC (rev 5448)
+++ trunk/gta02-core/docs/ecn/STATUS 2009-08-14 23:34:25 UTC (rev 5449)
@@ -27,4 +27,4 @@
0025 Done Connect CLKOUT1 to UEXTCLK (U1501)
0026 Discuss Improve bus termination at sd-card
0027 Edit Renamed some nets for clarity
-0028 Edit Remove USB host power switch
+0028 Edit Remove USB "charge path" power switch U4905
--- End Message ---
_______________________________________________
commitlog mailing list
commitlog@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/commitlog
