Maybe the base processing code evaluates the tag name as a variable or in a key/value pair, In which case its overwritten by the next duplicate entry.

Many Thanks

Yeah there are several multi-dimensional arrays that are used and copied / collated / sorted and some are local functions and some are global. it is quite involved to go back and forth from module to module to subroutine... etc. but the bottom line is the same... no duplicate tags or the GIGO error occurs. and reading a BASE log in debug mode is a long read too .

DP_MassStorage_wnt5_x86-32_1108r1
Download link

@mr_smartepants Could you post your fixes for the H7 and H9 so that i can modify my qsc ini please.

Many Thanks

why is this anyway... why not

[IT1]
ms_count=1
ms_1_deviceName="ITE IT8211 ATA"
ms_1_tag="iteatapi"
ms_1_sysFile="iteatapi.sys"
ms_1_hwids="PCI\VEN_1283&DEV_8211"
ms_1_isBusExtender=false
ms_1_exc_skipIfOS=""

Because the .inf, .cat, .sys files are all different for each OS.  They have to be kept separate.

I split the IT* files into separate folders (shortens the file-path length.)

[IT1]
ms_count=1
ms_1_deviceName="ITE IT8211 ATA XP"
ms_1_tag="iteatapi"
ms_1_sysFile="iteatapi.sys"
ms_1_hwids="PCI\VEN_1283&DEV_8211"
ms_1_isBusExtender=false
ms_1_exc_skipIfOS="w2k3,w2k"

[IT2]
ms_count=1
ms_1_deviceName="ITE IT8211 ATA Win2000"
ms_1_tag="iteatap2"
ms_1_sysFile="iteatap2.sys"
ms_1_hwids="PCI\VEN_1283&DEV_8211"
ms_1_isBusExtender=false
ms_1_exc_skipIfOS="w2k3,wxp"

[IT3]
ms_count=1
ms_1_deviceName="ITE IT8211 ATA Win2003"
ms_1_tag="iteatap3"
ms_1_sysFile="iteatap3.sys"
ms_1_hwids="PCI\VEN_1283&DEV_8211"
ms_1_isBusExtender=false
ms_1_exc_skipIfOS="wxp,w2k"

If QSC is on all files are output to the QSC folder (XP, 2K3 and 2K) regardless of which OS is the current target. Programatically this needs to be done because we don't know which OS will be the target the next time BASE is run . So if the tag name is reused only the last file copied to the cache will exist. In this particular case of ITE it would in theory be no issue because the source file is the same for all three. however if the source were not the same only the last file copied would be the correct one. (the xp file overwrites the 2k3 file or whatever)

Something to do with QSC? - As all drivers are pre compressed.

PG

nope because i never use QSC when i am testing ... it might skew my results
I tend to use the same file name over and over again...
if the file name don't change the cache is used instead of the new file. So any changes would not show up in my results.

Plus QSC is only useful if one intends to slipstream the same mass storage pack multiple times. I never do .
Additionaly using or not using QSC will not alter the final result (unless the driverpack is altered after using QSC)... the files are simply output to a temp folder and then copied to the destination. the same exact code is used in both instances the only difference is the output folder. (and a copy if QSC is on)

@Paul That's a valid guess for someone who doesn't have the code in front of them .

Also the source file is the same for all three... so the cache is valid in any case. The issue is with how the mass INI is processed by BASE. The mass storage code is over a thousand lines in several modules in at least three different files... (it's pretty complicated) somewhere in that mess is the answer... i could not find it quickly... but the fix is simple and the same as it has been for the last 7 years... don't duplicate the tag name .

Hmmm... this makes me think...
Note to self and smartepants:
If the files won't be changed and only the mass INI needs to be altered one may speed up testing by simply altering the contents of the mass INI in the QSC folder !

Thanks Paul your question has taught me something

[IT1]
ms_count=1
ms_1_deviceName="ITE IT8211 ATA XP"
ms_1_tag="iteatap1"
ms_1_sysFile="iteatapi.sys"
ms_1_hwids="PCI\VEN_1283&DEV_8211"
ms_1_isBusExtender=false
ms_1_exc_skipIfOS="w2k3,w2k"

[IT1-2k]
ms_count=1
ms_1_deviceName="ITE IT8211 ATA Win2000"
ms_1_tag="iteatap0"
ms_1_sysFile="iteatapi.sys"
ms_1_hwids="PCI\VEN_1283&DEV_8211"
ms_1_isBusExtender=false
ms_1_exc_skipIfOS="w2k3,wxp"

[IT1-2k3]
ms_count=1
ms_1_deviceName="ITE IT8211 ATA Win2003"
ms_1_tag="iteatap2"
ms_1_sysFile="iteatapi.sys"
ms_1_hwids="PCI\VEN_1283&DEV_8211"
ms_1_isBusExtender=false
ms_1_exc_skipIfOS="wxp,w2k"

yep... can't have a duplicate name even if it is technically skipped ...

the above code works (GIGO Erik)

[IT1]
ms_count=3
ms_1_deviceName="ITE IT8211 ATA XP"
ms_1_tag="iteatap0"
ms_1_sysFile="iteatapi.sys"
ms_1_hwids="PCI\VEN_1283&DEV_8211"
ms_1_isBusExtender=false
ms_1_exc_skipIfOS="w2k3,w2k"
ms_2_deviceName="ITE IT8211 ATA Win2000"
ms_2_tag="iteatap1"
ms_2_sysFile="iteatapi.sys"
ms_2_hwids="PCI\VEN_1283&DEV_8211"
ms_2_isBusExtender=false
ms_2_exc_skipIfOS="w2k3,wxp"
ms_3_deviceName="ITE IT8211 ATA Win2003"
ms_3_tag="iteatap2"
ms_3_sysFile="iteatapi.sys"
ms_3_hwids="PCI\VEN_1283&DEV_8211"
ms_3_isBusExtender=false
ms_3_exc_skipIfOS="wxp,w2k"

also fails

why is this anyway... why not

[IT1]
ms_count=1
ms_1_deviceName="ITE IT8211 ATA"
ms_1_tag="iteatapi"
ms_1_sysFile="iteatapi.sys"
ms_1_hwids="PCI\VEN_1283&DEV_8211"
ms_1_isBusExtender=false
ms_1_exc_skipIfOS=""

somebody broke the oldest rule in the book "no duplicate tag names"
Although as i looked through the code i fail to see why it wouldn't work

PG

although the rule has always been no tag name may be duplicated.

that may also apply to sections that are technically excluded...

BRB

[IT1]
ms_count=1
ms_1_deviceName="ITE IT8211 ATA XP"
ms_1_tag="iteatapi"
ms_1_sysFile="iteatapi.sys"
ms_1_hwids="PCI\VEN_1283&DEV_8211"
ms_1_isBusExtender=false
ms_1_exc_skipIfOS="w2k3,w2k"

[IT1-2k]
ms_count=1
ms_1_deviceName="ITE IT8211 ATA Win2000"
ms_1_tag="iteatapi"
ms_1_sysFile="iteatapi.sys"
ms_1_hwids="PCI\VEN_1283&DEV_8211"
ms_1_isBusExtender=false
ms_1_exc_skipIfOS="w2k3,wxp"

[IT1-2k3]
ms_count=1
ms_1_deviceName="ITE IT8211 ATA Win2003"
ms_1_tag="iteatapi"
ms_1_sysFile="iteatapi.sys"
ms_1_hwids="PCI\VEN_1283&DEV_8211"
ms_1_isBusExtender=false
ms_1_exc_skipIfOS="wxp,w2k"

Notice in the above extract from the dpms .ini that there is a OS discriminator built in.
ms_1_exc_skipIfOS=
This line means the inclusive section is "skipped" if the OS is (insert OS here).
IT1 is for XP.
The other two are similar.  H7 is for XP, H7S is for (S)erver.

*****
Edit:
Oops, Now I see what you mean.  There are no corresponding entries in the txtsetup.sif for XP.  I see it now.
I renamed your thread title to be more descriptive.
Thanks for reporting.

IT1 shares "iteatapi" with IT1-2k and IT1-2K3
H7 Shares "rr2340" with H7S
H9 Shares "rr172x" with H9S

Are these typos or have i overlooked a selection method.

Many Thanks
Paul