From mboxrd@z Thu Jan 1 00:00:00 1970 From: Rene Herman Subject: Re: snd-es1968 (+ snd-cs4236) trouble Date: Tue, 13 Dec 2005 08:09:05 +0100 Message-ID: <439E7391.60408@keyaccess.nl> References: <439B36E0.3090709@keyaccess.nl> <439DFF60.1020105@keyaccess.nl> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="------------090306040801080509020308" Return-path: Received: from smtpq2.groni1.gr.home.nl (smtpq2.groni1.gr.home.nl [213.51.130.201]) by alsa.jcu.cz (ALSA's E-mail Delivery System) with ESMTP id 7D133132 for ; Tue, 13 Dec 2005 08:05:15 +0100 (MET) In-Reply-To: <439DFF60.1020105@keyaccess.nl> Sender: alsa-devel-admin@lists.sourceforge.net Errors-To: alsa-devel-admin@lists.sourceforge.net List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , List-Archive: To: Takashi Iwai Cc: alsa-devel@alsa-project.org List-Id: alsa-devel@alsa-project.org This is a multi-part message in MIME format. --------------090306040801080509020308 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Rene Herman wrote: > I'll try to see if I can find anything in the 1969 or 1978 docs. > Please beat me to it though... :-) The minimal patch I need to have things functional, is not enabling TDMA in sound/pci/es1968.c: @@ -2197,10 +2197,7 @@ static void snd_es1968_chip_init(es1968_ /* Config Reg A */ pci_read_config_word(pci, ESM_CONFIG_A, &w); - /* Use TDMA for now. TDMA works on all boards, so while its - * not the most efficient its the simplest. */ w &= ~DMA_CLEAR; /* Clear DMA bits */ - w |= DMA_TDMA; /* TDMA on */ With that applied, things seem to be working fine again. With the bits cleared, it's set to "Distributed DMA" which is disabled via the DDMA control register (0x60), so this should be okay? The comment's a bit puzzling but doesn't this stuff apply only to the legacy DMA, which isn't used in this driver, anyway? While looking there, I also noticed two bits that are reversed between documentation and source. One of them in fact the "Enable Legacy Audio" bit, so I expected that to be the trouble at first, although it wasn't. In the second attached patch, this bit is set in accordance with the documentation. Works for me... with the bit set either way, I don't seem to be experiencing conflicts with the CS4236 (which also has an enabled SBPro emulation) so I just trusted the documentation. The other one is the "Emulate ISA timing on PCI bit". The code said and commented that it enabled the emulation, but according to the datasheet it was disabling it instead. I've left it the same, but changed the name of the define and the comment. This card has Canyon-3D (ES1970MS-3D) written on the main chip, but it advertises itself as a 125d:1978, rev 10, which should be a Maestro-2E, so I've been looking at that datasheet... As said, on this TerraTec SoundSystem DMX, version 1.2, I need the first patch at minimum. The second one also does the bit reversals, and works for me. Please apply? Rene. --------------090306040801080509020308 Content-Type: text/plain; name="linux-2.6.14.3_es1968_tdma.diff" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="linux-2.6.14.3_es1968_tdma.diff" --- linux-2.6.14.3-local-wc/sound/pci/es1968.c.orig 2005-12-13 07:32:43.000000000 +0100 +++ linux-2.6.14.3-local-wc/sound/pci/es1968.c 2005-12-13 07:32:49.000000000 +0100 @@ -2197,10 +2197,7 @@ static void snd_es1968_chip_init(es1968_ /* Config Reg A */ pci_read_config_word(pci, ESM_CONFIG_A, &w); - /* Use TDMA for now. TDMA works on all boards, so while its - * not the most efficient its the simplest. */ w &= ~DMA_CLEAR; /* Clear DMA bits */ - w |= DMA_TDMA; /* TDMA on */ w &= ~(PIC_SNOOP1 | PIC_SNOOP2); /* Clear Pic Snoop Mode Bits */ w &= ~SAFEGUARD; /* Safeguard off */ w |= POST_WRITE; /* Posted write */ --------------090306040801080509020308 Content-Type: text/plain; name="linux-2.6.14.3_es1968_tdma_la_iem.diff" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="linux-2.6.14.3_es1968_tdma_la_iem.diff" --- linux-2.6.14.3-local-wc/sound/pci/es1968.c.orig 2005-12-13 07:25:47.000000000 +0100 +++ linux-2.6.14.3-local-wc/sound/pci/es1968.c 2005-12-13 07:26:04.000000000 +0100 @@ -176,7 +176,7 @@ MODULE_PARM_DESC(joystick, "Enable joyst /* Values for the ESM_LEGACY_AUDIO_CONTROL */ -#define ESS_ENABLE_AUDIO 0x8000 +#define ESS_DISABLE_AUDIO 0x8000 #define ESS_ENABLE_SERIAL_IRQ 0x4000 #define IO_ADRESS_ALIAS 0x0020 #define MPU401_IRQ_ENABLE 0x0010 @@ -195,7 +195,7 @@ MODULE_PARM_DESC(joystick, "Enable joyst #define DMA_TDMA 0x0100 #define DMA_PCPCI 0x0200 #define POST_WRITE 0x0080 -#define ISA_TIMING 0x0040 +#define PCI_TIMING 0x0040 #define SWAP_LR 0x0020 #define SUBTR_DECODE 0x0002 @@ -2197,14 +2197,11 @@ static void snd_es1968_chip_init(es1968_ /* Config Reg A */ pci_read_config_word(pci, ESM_CONFIG_A, &w); - /* Use TDMA for now. TDMA works on all boards, so while its - * not the most efficient its the simplest. */ w &= ~DMA_CLEAR; /* Clear DMA bits */ - w |= DMA_TDMA; /* TDMA on */ w &= ~(PIC_SNOOP1 | PIC_SNOOP2); /* Clear Pic Snoop Mode Bits */ w &= ~SAFEGUARD; /* Safeguard off */ w |= POST_WRITE; /* Posted write */ - w |= ISA_TIMING; /* ISA timing on */ + w |= PCI_TIMING; /* PCI timing on */ /* XXX huh? claims to be reserved.. */ w &= ~SWAP_LR; /* swap left/right seems to only have effect on SB @@ -2245,7 +2242,7 @@ static void snd_es1968_chip_init(es1968_ pci_read_config_word(pci, ESM_LEGACY_AUDIO_CONTROL, &w); - w &= ~ESS_ENABLE_AUDIO; /* Disable Legacy Audio */ + w |= ESS_DISABLE_AUDIO; /* Disable Legacy Audio */ w &= ~ESS_ENABLE_SERIAL_IRQ; /* Disable SIRQ */ w &= ~(0x1f); /* disable mpu irq/io, game port, fm, SB */ --------------090306040801080509020308-- ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click