Timeline



Jul 16, 2011:

9:22 PM Changeset [235] by Lars Erdmann
the computation of the number of necessary microFrames to use in split …
9:17 PM Changeset [234] by Lars Erdmann
The back pointer of an siTD needs to have ALL 5 lower bits set to zero …
1:15 AM Changeset [233] by Lars Erdmann
this is 10.171

Jul 15, 2011:

7:47 PM Changeset [232] by Lars Erdmann
need to increase ITD structure size from 96 bytes to 128 bytes: EHCI …

Jul 14, 2011:

8:35 PM Changeset [231] by Lars Erdmann
split transactions: add routine "setSplitMask" to properly compute the …

Jul 13, 2011:

4:44 PM Changeset [230] by Lars Erdmann
again, reworked "CheckIsoBandwidth?" and "RecalculateIsoBand?": it …
1:49 PM Changeset [229] by Lars Erdmann
removed "CheckIso_LS_FS_PeriodBand" from implementation and also trace …
1:29 AM Changeset [228] by Lars Erdmann
"CheckIsoBandwidth?": updated the bandwidth computation for HS. It now …

Jul 11, 2011:

5:53 PM Changeset [227] by Lars Erdmann
fix bandwidth computation (not yet done), prepare to move over to a …

Jul 8, 2011:

5:39 PM Changeset [226] by Lars Erdmann
use pushf,cli,...,popf instead of setting sti explicitely, we don't …

Jul 6, 2011:

11:14 PM Changeset [225] by Lars Erdmann
changed the HC initialization sequence, also, moved setting of 4GB …
10:29 PM Changeset [224] by Lars Erdmann
removing 32-bit shift routines, replacing "multiple of page" …
9:52 PM Changeset [223] by Lars Erdmann
changing "LockYield?" to use DevHelp_ProcBlock instead of DevHelp_Yield …

Jul 5, 2011:

9:47 PM Changeset [222] by Lars Erdmann
Commented out "CheckIsoBandwidth?" until it is fixed (and we have …

Jul 3, 2011:

8:27 PM Changeset [221] by Lars Erdmann
Found yet some other places that needed fixes because of 32-bit offsets
4:32 PM Changeset [220] by Lars Erdmann
"IsoFillTDs": fix computation of offset into the (s)iTD array, again …

Jul 1, 2011:

12:15 AM Changeset [219] by Lars Erdmann
remove unnecessary register pushes/pops (these routines interface with C)

Jun 30, 2011:

11:56 PM Changeset [218] by Lars Erdmann
serializing concurrent "iso open": protecting the update of isoinfo …

Jun 29, 2011:

1:35 AM Changeset [217] by Lars Erdmann
updated access to config.sys commandline as we are no longer reporting …
1:33 AM Changeset [216] by Lars Erdmann
updated makefiles to specify PRELOAD attribute for data and code segments

Jun 28, 2011:

11:37 PM Changeset [215] by Lars Erdmann
fixed the device attributes/capabilities bit strip

Jun 22, 2011:

10:53 PM Changeset [214] by Lars Erdmann
separated the AttachDD tables for the OEMHLP and USBD drivers

Jun 21, 2011:

11:03 PM Changeset [213] by Lars Erdmann
"DetachIsoTDs": after the "frame index cycling" loop I was calling …
8:48 PM Changeset [212] by Lars Erdmann
update makefiles
12:45 AM Changeset [211] by Lars Erdmann
"AccNonIsoReq?": we better quit the function with a memory error if …
12:44 AM Changeset [210] by Lars Erdmann
"AddIsoTDsToSchedule": fixed the computation of 32-bit offset (again …

Jun 18, 2011:

11:38 PM Changeset [209] by Lars Erdmann
using DevHelp_ProcBlock instead of DevHelp_Yield for wherever we need …
1:24 AM Changeset [208] by Lars Erdmann
fixed "GetCurrentFrame?": the mask for the frame index is VARIABLE and …

Jun 16, 2011:

9:11 PM Changeset [207] by Lars Erdmann
removed the check if periodic schedule is executing, make sure in …

Jun 15, 2011:

2:15 PM Ticket #4 (usbehcd init complete delay) created by Joseph Shrago
Gigabyte P55-UD3L Four USBUHCD & two USBEHCD After the all drivers …

Jun 11, 2011:

10:18 PM Changeset [206] by Lars Erdmann
coded 32-bit shift operations as _fastcall assembler routines to …
12:30 AM Changeset [205] by Lars Erdmann
fixed a slight shortcoming in "memlset": make sure that each BYTE is …

Jun 10, 2011:

11:39 PM Changeset [204] by Lars Erdmann
to increase clarity I now moved the the computation of the "Mult" …
6:53 PM Changeset [203] by Lars Erdmann
"ReleaseIsoStruct?": I shouldn't have touched the code to save physical …
6:55 AM Changeset [202] by Lars Erdmann
"AllocateHugeMemory?": fixed a bug in computing the actual number of …
5:41 AM Changeset [201] by Lars Erdmann
for all the isochronous handling code dealing with iTDs and siTDs, …
5:33 AM Changeset [200] by Lars Erdmann
Removed DevHelp_Beep from PCIPMPowerDown functions: it was supposed to …
5:31 AM Changeset [199] by Lars Erdmann
updated makefiles to support symbolic debugging (for ICAT)

Jun 6, 2011:

1:08 AM Changeset [198] by Lars Erdmann
Modified code so it's no longer necessary to link objects from slibcep.lib.

Jun 5, 2011:

9:51 PM Changeset [197] by Lars Erdmann
don't automatically copy trace files to the TRACE subdirectory
9:39 PM Changeset [196] by Lars Erdmann
updated makefiles and build scripts to support symbol debugging, for …
2:55 PM Changeset [195] by Lars Erdmann
"CreateIsoStruct?": added a HACK so that the number of iTDs will be 512 …

May 31, 2011:

10:32 PM Changeset [194] by Lars Erdmann
removed some traps in tracing code: changed a couple of places where …
3:38 PM Ticket #3 (Problem with USBOHCD.SYS from usbxhcd8.zip) created by losepete
This was originally posted in comp.os.os2.misc/Update to USB host …

May 30, 2011:

11:38 PM Changeset [193] by Lars Erdmann
Changed "DetachIsoTDs" and "AddIsoTDsToSchedule": fixed a bug in …

May 29, 2011:

9:44 PM Changeset [192] by Lars Erdmann
backed out a change in "CreateIsoStruct?": the update to the number of …

May 25, 2011:

12:48 AM Changeset [191] by Lars Erdmann
"CreateIsoStruct?": kicked out superfluous CLI/STI, "ReleaseIso?": added …

May 24, 2011:

9:20 PM Changeset [190] by Lars Erdmann
updated tracing, changed to new major trace codes 0xE0,0xE1,0xE2 (for …

May 22, 2011:

11:19 PM Changeset [189] by Lars Erdmann
updated tracing

May 21, 2011:

12:18 AM Changeset [188] by Lars Erdmann
adding tracing

May 20, 2011:

11:03 PM Changeset [187] by Lars Erdmann
adding tracing
10:52 PM Changeset [186] by Lars Erdmann
add tracing related file
8:57 PM Changeset [185] by Lars Erdmann
Bug in "CalculatePollRate?" for (bInterval - 1) < 3 (or bInterval < 4 …

May 19, 2011:

9:40 PM Changeset [184] by Lars Erdmann
updates to "AddIsoTDsToSchedule": lets update the COMPLETE frame list …

May 18, 2011:

9:49 PM Changeset [183] by Lars Erdmann
in the interrupt routine, clear ALL interrupt status flags right after …
9:48 PM Changeset [182] by Lars Erdmann
remove generation for every TD of a Bulk IN, instead have an IOC …
9:37 PM Changeset [181] by Lars Erdmann
Routine "AccIsoReq?" would not enable periodic list processing, routine …
9:35 PM Changeset [180] by Lars Erdmann
Routine "SetEnableFlags?" would not enable periodic processing (typo)

May 17, 2011:

10:09 PM Changeset [179] by Lars Erdmann
"AddIsoTDsToSchedule": only duplicate the (s)ITDs starting at offset …
10:06 PM Changeset [178] by Lars Erdmann
yet another fix to properly set the toggle bit. Since I kicked out the …

May 16, 2011:

6:57 PM Changeset [177] by Lars Erdmann
better be safe than sorry: mask out 11 bits of the …

May 13, 2011:

11:21 PM Changeset [176] by Lars Erdmann
backing out the changes to preset the packet size to the maximum value …
10:39 PM Changeset [175] by Lars Erdmann
better be safe than sorry: restore the direction flag with pushf/popf …
10:29 PM Changeset [174] by Lars Erdmann
getting rid of the dependency on drive letter
10:27 PM Changeset [173] by Lars Erdmann
clarify that the "Mult" value in the USBRB2 req. packet is the value …

May 12, 2011:

9:23 PM Changeset [172] by Lars Erdmann
backed out the changes I did in checking the correct size for …

May 11, 2011:

7:35 PM Changeset [171] by Lars Erdmann
modified macro GET_ITD_MULT so that it requires a pointer to a USBRB2 …
12:21 AM Changeset [170] by Lars Erdmann
Wim Brul reports that the USBEHCD.SYS driver does not respect the …

May 10, 2011:

10:43 PM Changeset [169] by Lars Erdmann
adding the full speed max packet sizes for the various transfer types …
10:02 PM Changeset [168] by Lars Erdmann
to be in line with USBEHCD.SYS, readded the generation of interrupts …
9:35 PM Changeset [167] by Lars Erdmann
Wim Brul reports that the "Mult" value that specifies how many …

May 9, 2011:

4:23 PM Changeset [166] by Lars Erdmann
"ProcessIsoIRQ": an attempt to make this routine cycle through ALL …
4:20 PM Changeset [165] by Lars Erdmann
reworked setting the default packet size when creating the list of …

May 5, 2011:

6:11 PM Changeset [164] by Lars Erdmann
Rudi Ihle reports problems with isochronous devices compared to …

May 4, 2011:

9:04 PM Changeset [163] by Lars Erdmann
"FinishIO","ResendPacket?": "setmem" irqRB structure (defensive …
8:55 PM Changeset [162] by Lars Erdmann
fix a problem in "CreateTDList": if ioRB would pass in a maxPacketSize …
12:03 AM Changeset [161] by Lars Erdmann
backed out the obscure manipulation of the data toggle bit. It is now …

May 3, 2011:

11:45 PM Changeset [160] by Lars Erdmann
fixed a minor problem in "ResendPacket?": for the GenIOCTL packet, …
11:43 PM Changeset [159] by Lars Erdmann
Wim Brul reports problems with the size of buffers used for send AND …

May 2, 2011:

10:37 PM Changeset [158] by Lars Erdmann
Wim Brul has problems with bulk transfer where it works in IBM 10.162: …
10:36 PM Changeset [157] by Lars Erdmann
did some minor changes to FinishIO to improve clarity, in particular …

Apr 30, 2011:

11:40 PM Changeset [156] by Lars Erdmann
crash reported by Wim Brul: in "CreateIsoStruct?", properly compute the …
11:31 PM Changeset [155] by Lars Erdmann
commented out the "LockYield?"/"UnLockYield?" for serializing multiple …
11:25 PM Changeset [154] by Lars Erdmann
removed drive dependency in build script (I have changed the build …
11:23 PM Changeset [153] by Lars Erdmann
added workaround to prevent the C-optimizer to generate code that …

Apr 22, 2011:

6:24 PM Ticket #2 (WD Book not working) created by rbri
Hi Lars, hi Greg, first of all many thanks for your hard work. I have …

Apr 19, 2011:

10:46 PM Changeset [152] by Lars Erdmann
removed "gRootReqStatus" variable and handling, it was an optimization …
10:22 AM Changeset [151] by Lars Erdmann
Context arming: changed unlock routine to reset lock count and keep …
Note: See TracTimeline for information about the timeline view.