Re: SOUND: OPTI 931 troubles in 2.1.103 - patch

Itai Nahshon (nahshon@actcom.co.il)
Sun, 31 May 1998 05:52:46 +0300



(new patch)
Content-Type: multipart/mixed; boundary="------------4D6477DA30BDE85A9B23965B"

This is a multi-part message in MIME format.
--------------4D6477DA30BDE85A9B23965B
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Maciej,
Thanks for the fix, it's working for me (in pre-2.1.101-1)
but it is probably not good for everyone. Your patch reverses
the change done in linux-2.1.102.

The CVS comment for this change is:
revision 1.19
date: 1998/04/29 22:29:42; author: alan; state: Exp; lines: +3 -3
MAD16 fixes I hope. Looks like there might be mutually incompatible
variants of the Opti chips

Looking at the docs, bit 2 of MC4 reversed its meaning between
the C930 and the C931.

Alan,
Can you verify that the change was done for the C930?
In that case, my patch should cover both cases (I tried it only
with the C931).

-- 
Itai Nahshon   nahshon@actcom.co.il
        Also   nahshon@vnet.ibm.com
--------------4D6477DA30BDE85A9B23965B
Content-Type: text/plain; charset=us-ascii; name="mad16.diff"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="mad16.diff"

--- mad16.c.ORIG Sun May 31 05:49:24 1998 +++ mad16.c Sun May 31 05:49:40 1998 @@ -405,13 +405,17 @@ /* MC2 is CD configuration. Don't touch it. */ mad_write(MC3_PORT, 0); /* Disable SB mode IRQ and DMA */ + + /* bit 2 of MC4 reverses it's meaning between the C930 + and the C931. */ + cfg = c931_detected ? 0x04 : 0x00; #ifdef MAD16_CDSEL if(MAD16_CDSEL & 0x20) - mad_write(MC4_PORT, 0x62); /* opl4 */ + mad_write(MC4_PORT, 0x62|cfg); /* opl4 */ else - mad_write(MC4_PORT, 0x52); /* opl3 */ + mad_write(MC4_PORT, 0x52|cfg); /* opl3 */ #else - mad_write(MC4_PORT, 0x52); + mad_write(MC4_PORT, 0x52|cfg); #endif mad_write(MC5_PORT, 0x3C); /* Init it into mode2 */ mad_write(MC6_PORT, 0x02); /* Enable WSS, Disable MPU and SB */

--------------4D6477DA30BDE85A9B23965B--

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu