diff for duplicates of <87hb86a5mm.fsf@ti.com> diff --git a/a/1.txt b/N1/1.txt index 636b8a9..d7673e0 100644 --- a/a/1.txt +++ b/N1/1.txt @@ -27,7 +27,7 @@ Tomi Valkeinen <tomi.valkeinen@ti.com> writes: > > Yes, the counter. I don't think > -> if (dispc.runtime_count++ = 0) +> if (dispc.runtime_count++ == 0) > > is thread safe. @@ -35,7 +35,7 @@ OK, if it's just the counter, you can drop the mutex and use an atomic variable and use atomic_inc(), atomic_dec() etc. Then it will be clear from reading what exactly is protected. ->> > + if (dispc.runtime_count++ = 0) { +>> > + if (dispc.runtime_count++ == 0) { >> >> You shouldn't need your own use-counting here. The runtime PM core is >> already doing usage counting. @@ -64,7 +64,7 @@ Sorry about that. Just to clarify, what exactly does the opt clock have to be enabled for? -From what I can gather, this particular opt clock has to be enabled for +>From what I can gather, this particular opt clock has to be enabled for the hwmod itself to be enabled (or disabled), correct? This has been a known issue for reset[1], but I wasn't aware that it was @@ -122,7 +122,7 @@ clocks in the hwmod enable/idle path. We're doing it for reset[1], but not for anything else. And based on the changelog there[1], it doesn't even sound like we fully understand the exact requirements around reset. -From the contortions you've had to go through though, it sounds like the +>From the contortions you've had to go through though, it sounds like the same optional clocks that are needed for reset are needed for a "normal" module enable/disable. @@ -189,7 +189,7 @@ Now I'll wait for Benoit/Paul to enlighten us. >> > +{ >> > + mutex_lock(&dispc.runtime_lock); >> > + ->> > + if (--dispc.runtime_count = 0) { +>> > + if (--dispc.runtime_count == 0) { >> > + int r; >> >> Same problem here. @@ -243,7 +243,7 @@ Date: Tue Sep 21 18:57:58 2010 +0200 [2] -From d2806a4148fbed869eff8703b1137cd35d16ef53 Mon Sep 17 00:00:00 2001 +>From d2806a4148fbed869eff8703b1137cd35d16ef53 Mon Sep 17 00:00:00 2001 From: Kevin Hilman <khilman@ti.com> Date: Fri, 3 Jun 2011 15:33:25 -0700 Subject: [PATCH] OMAP2+: omap_hwmod: selectively manage optional clocks in enable/disable diff --git a/a/content_digest b/N1/content_digest index 99747ab..6613243 100644 --- a/a/content_digest +++ b/N1/content_digest @@ -4,7 +4,7 @@ "ref\01307122985.2016.72.camel@deskari\0" "From\0Kevin Hilman <khilman@ti.com>\0" "Subject\0Re: [PATCH 19/27] OMAP: DSS2: Use PM runtime & HWMOD support\0" - "Date\0Fri, 03 Jun 2011 22:53:53 +0000\0" + "Date\0Fri, 03 Jun 2011 15:53:53 -0700\0" "To\0Tomi Valkeinen <tomi.valkeinen@ti.com>\0" "Cc\0linux-omap@vger.kernel.org" linux-fbdev@vger.kernel.org @@ -41,7 +41,7 @@ ">\n" "> Yes, the counter. I don't think\n" ">\n" - "> if (dispc.runtime_count++ = 0)\n" + "> if (dispc.runtime_count++ == 0)\n" ">\n" "> is thread safe.\n" "\n" @@ -49,7 +49,7 @@ "variable and use atomic_inc(), atomic_dec() etc. Then it will be clear\n" "from reading what exactly is protected.\n" "\n" - ">> > +\tif (dispc.runtime_count++ = 0) {\n" + ">> > +\tif (dispc.runtime_count++ == 0) {\n" ">> \n" ">> You shouldn't need your own use-counting here. The runtime PM core is\n" ">> already doing usage counting.\n" @@ -78,7 +78,7 @@ "\n" "Just to clarify, what exactly does the opt clock have to be enabled for?\n" "\n" - "From what I can gather, this particular opt clock has to be enabled for\n" + ">From what I can gather, this particular opt clock has to be enabled for\n" "the hwmod itself to be enabled (or disabled), correct?\n" "\n" "This has been a known issue for reset[1], but I wasn't aware that it was\n" @@ -136,7 +136,7 @@ "not for anything else. And based on the changelog there[1], it doesn't\n" "even sound like we fully understand the exact requirements around reset.\n" "\n" - "From the contortions you've had to go through though, it sounds like the\n" + ">From the contortions you've had to go through though, it sounds like the\n" "same optional clocks that are needed for reset are needed for a \"normal\"\n" "module enable/disable.\n" "\n" @@ -203,7 +203,7 @@ ">> > +{\n" ">> > +\tmutex_lock(&dispc.runtime_lock);\n" ">> > +\n" - ">> > +\tif (--dispc.runtime_count = 0) {\n" + ">> > +\tif (--dispc.runtime_count == 0) {\n" ">> > +\t\tint r;\n" ">> \n" ">> Same problem here. \n" @@ -257,7 +257,7 @@ "\n" "\n" "[2]\n" - "From d2806a4148fbed869eff8703b1137cd35d16ef53 Mon Sep 17 00:00:00 2001\n" + ">From d2806a4148fbed869eff8703b1137cd35d16ef53 Mon Sep 17 00:00:00 2001\n" "From: Kevin Hilman <khilman@ti.com>\n" "Date: Fri, 3 Jun 2011 15:33:25 -0700\n" "Subject: [PATCH] OMAP2+: omap_hwmod: selectively manage optional clocks in enable/disable\n" @@ -304,4 +304,4 @@ "-- \n" 1.7.4 -d1c928a420c4cfb2f413c5cbb910edef1d4eaaecdb53e73c78f39126251bddb4 +a8ffb69931e68f6124154aaaba9ef6d16490f32d2f93822d714f4ede39e8fb82
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.