Found the reason: the chip is AD1980, and Dell uses the "Miscellanous bit" (0x76) from that codec in a different way from most AD1980 implementations. I found in linux kernel 2.4 there is a solution: /* Switch the inputs/outputs over (from Dell code) */ control = codec->codec_read(codec, AC97_AD_MISC); codec->codec_write(codec, AC97_AD_MISC, control | 0x4420); I tried. It works! > > >Hi Hakon, >Thanks for the information. > >My purpose is to have demo of our previous product, the product works on >Intel845GE chipset (ICH4) >But that motherboard is out of date, even can not be found it in Fry's :) > >So I have to use another motherboard which uses Intel 865G chipset (ICH5) >It's a quite normal DELL desktop PC. >Actually the motherboard has a driver provided by ADI (soundMax). >So I don't know where they shareing jacks or not. > >But every time I query codecs ready or not, the GLOB_STA register always >return 0x10700000, >that meas it has tertiary codec, no primary codec.. >quite strange.. > >> >>Dear Rai, >> >>I am not sure this is the cause of your issue but the AC-97 spec is >>essentially broken from a unified driver POV due to jack sharing wiring >>techniques introduced around the time ICH5 appeared (but it also happens >>on board designs with AC-97 codecs attached to other chipsets). >> >>Background: >>At some point the industry required a jack sharing system as some system >>vendors wanted to ship a 5.1 codec system but didn't want the expense of >>populating two 3-jack stacks on their Motherboards. I believe a >>prominent chipset vendor's board design division came up with a jack >>sharing arrangement on one 3 jack stack as follows: >> >>Pink Jack: Primary function: MIC In. Secondary function C/LFE Out Green >>Jack: Line Out with HP drive, no input function Blue Jack: Primary >>Function: Line In. Secondary function Surround Out. >> >>From a hardware perspective the input/output pins were wired in parallel >>or there was a multiplexer in the codec to switch one set of pins >>between input and the output in the codec. >> >>The jack sharing combinations had system design specific variances (e.g. >>MIC shared with Surround instead of C/LFE and Line In shared with MIC >>instead of Surround) this "3 jack sharing scheme" became pretty much the >>defacto standard for channel motherboard products. >> >>In any case all shipped with IHV provided drivers/software to make this >>all work correctly platform by platform, delineated in the INF via >>SSID's. And so the spec was no longer suitable for unified chipset based >>drivers. >> >>You could try plugging your headset or speakers into one of the other >>jacks to see if this is the issue you are seeing. >> >>Microsoft created the UAA initiative which is requiring audio HW >>solutions (like HD Audio) that can report the system design dependent >>wiring of the codec to the OS in part to avoid this kind of problem in >>the future. >> >>Sincerely, >> >>Hakon Strande >>PM Integrated, Internal, External, and Wireless Audio Devices >>MediaTech/DMD/Windows Client/Microsoft >> >>-----Original Message----- >>From: wdmaudiodev-bounce@xxxxxxxxxxxxx >>[mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx] On Behalf Of rai >>Sent: Thursday, June 01, 2006 4:09 PM >>To: wdmaudiodev >>Subject: [wdmaudiodev] ac97+ICH5 can not playback sound. >> >> >>Hi All, >>I am working on a AC97 driver for WinXP. >>I started from the sample driver code provided by XDDK. >> >>First time, I installed the driver on an Intel 845GE machine (ICH4), >>the driver reported that : PrimaryCodecReady timed out. >>then I modified >>NTSTATUS CAdapterCommon::PrimaryCodecReady (void), >>make sure that the function will always return STATUS_SUCCESS. >>then, the driver can basically work (play/record) on ICH4. >> >>After that I tried the driver on ICH5, it also report primary codec is >>not ready, >>after I did the same thing(modify the function), the driver can be >>installed. >>Device manager also report that the driver works properly. >>But I can not hear any sound from the speaker jak. >> >>I think there must be some difference between ICH4 and ICH5 on >>controlling AC97 codec, >>but I am not familiar with AC97 standard. >>Does anybody here can help or has the same experience? >>Thanks. >> >>Rai. >>2006-06-01 >> >> >> >> >> >> >> >>****************** >> >>WDMAUDIODEV addresses: >>Post message: mailto:wdmaudiodev@xxxxxxxxxxxxx >>Subscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=subscribe >>Unsubscribe: >>mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=unsubscribe >>Moderator: mailto:wdmaudiodev-moderators@xxxxxxxxxxxxx >> >>URL to WDMAUDIODEV page: >>http://www.wdmaudiodev.com/ >> >> >>****************** >> >>WDMAUDIODEV addresses: >>Post message: mailto:wdmaudiodev@xxxxxxxxxxxxx >>Subscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=subscribe >>Unsubscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=unsubscribe >>Moderator: mailto:wdmaudiodev-moderators@xxxxxxxxxxxxx >> >>URL to WDMAUDIODEV page: >>http://www.wdmaudiodev.com/ >> >> >> > > > >****************** > >WDMAUDIODEV addresses: >Post message: mailto:wdmaudiodev@xxxxxxxxxxxxx >Subscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=subscribe >Unsubscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=unsubscribe >Moderator: mailto:wdmaudiodev-moderators@xxxxxxxxxxxxx > >URL to WDMAUDIODEV page: >http://www.wdmaudiodev.com/ > > > ****************** WDMAUDIODEV addresses: Post message: mailto:wdmaudiodev@xxxxxxxxxxxxx Subscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=subscribe Unsubscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=unsubscribe Moderator: mailto:wdmaudiodev-moderators@xxxxxxxxxxxxx URL to WDMAUDIODEV page: http://www.wdmaudiodev.com/