From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Antonino A. Daplas" Subject: Re: Re: Linux 2.6.10-rc2 Date: Thu, 18 Nov 2004 08:01:59 +0800 Message-ID: <200411180802.01209.adaplas@hotpop.com> References: <200411172233.28606.andrew@walrond.org> Reply-To: linux-fbdev-devel@lists.sourceforge.net Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: Received: from sc8-sf-mx2-b.sourceforge.net ([10.3.1.12] helo=sc8-sf-mx2.sourceforge.net) by sc8-sf-list1.sourceforge.net with esmtp (Exim 4.30) id 1CUZkd-0008Jo-TP for linux-fbdev-devel@lists.sourceforge.net; Wed, 17 Nov 2004 16:02:15 -0800 Received: from smtp-out.hotpop.com ([38.113.3.71]) by sc8-sf-mx2.sourceforge.net with esmtp (Exim 4.41) id 1CUZkc-0007vl-Ld for linux-fbdev-devel@lists.sourceforge.net; Wed, 17 Nov 2004 16:02:15 -0800 Received: from hotpop.com (kubrick.hotpop.com [38.113.3.103]) by smtp-out.hotpop.com (Postfix) with SMTP id C7E0511ED68C for ; Thu, 18 Nov 2004 00:02:08 +0000 (UTC) In-Reply-To: Content-Disposition: inline Sender: linux-fbdev-devel-admin@lists.sourceforge.net Errors-To: linux-fbdev-devel-admin@lists.sourceforge.net List-Unsubscribe: , List-Id: List-Post: List-Help: List-Subscribe: , List-Archive: Content-Type: text/plain; charset="us-ascii" To: Linus Torvalds , Andrew Walrond Cc: linux-fbdev-devel@lists.sourceforge.net On Thursday 18 November 2004 07:12, Linus Torvalds wrote: > On Wed, 17 Nov 2004, Andrew Walrond wrote: > > And I can confirm that the problem exists in both 32bit and 64bit > > kernels. > > Ok. None of the code should actually have changed at all, but I notice one > bug: > > please undo those parts (by hand, I've messed up white-space), since they > make "burst" be one too big. Looks like the reason for doing that in the > first place was an old sparse warning that no longer exists. > > Antonino? Any I missed? Except for the above, I don't think so. The problem seems to be restricted to the monochrome expansion and rectangle fill functions. The only changes in those functions are the use of the NV_{WR|RD} instead of direct loads. Andrew's problem looks like that it might be related to the "burst" bug, although I'm surprised it did not show up in my hardware (tried on 3 different nvidia cards, NV3, NV11, NV32). Here's a patch that reverts to old code. Tony diff -Nru a/drivers/video/riva/riva_hw.c b/drivers/video/riva/riva_hw.c --- a/drivers/video/riva/riva_hw.c 2004-11-10 07:14:37 +08:00 +++ b/drivers/video/riva/riva_hw.c 2004-11-18 07:45:23 +08:00 @@ -641,10 +641,8 @@ { int b = fifo_data.graphics_burst_size >> 4; *burst = 0; - while (b) { + while (b >>= 1) (*burst)++; - b >>= 1; - } *lwm = fifo_data.graphics_lwm >> 3; } else @@ -834,10 +832,8 @@ { int b = fifo_data.graphics_burst_size >> 4; *burst = 0; - while (b) { + while (b >>= 1) (*burst)++; - b >>= 1; - } *lwm = fifo_data.graphics_lwm >> 3; } } @@ -1101,10 +1097,8 @@ { int b = fifo_data.graphics_burst_size >> 4; *burst = 0; - while (b) { + while (b >>= 1) (*burst)++; - b >>= 1; - } *lwm = fifo_data.graphics_lwm >> 3; } } @@ -1155,10 +1149,8 @@ { int b = fifo_data.graphics_burst_size >> 4; *burst = 0; - while (b) { + while (b >>= 1) (*burst)++; - b >>= 1; - } *lwm = fifo_data.graphics_lwm >> 3; } } ------------------------------------------------------- This SF.Net email is sponsored by: InterSystems CACHE FREE OODBMS DOWNLOAD - A multidimensional database that combines robust object and relational technologies, making it a perfect match for Java, C++,COM, XML, ODBC and JDBC. www.intersystems.com/match8