First public contribution.
1 THIS RELEASE NOTE NO LONGER IN USE
3 See \E32\RELEASE.TXT for changes to E32TEST
6 ============================
10 1. DEFECT FIX: DEF138318: SMP + Data Paging T_Chunk fails: t_chunk.exe::Local-c811fe18 Panic KERN-EXEC
11 Modified to use thread pool in chunk-close race condition test to prevent KERN-EXEC 3 when stack is accessed
12 after 50 CloseChunk threads have been created over and over again for 800 times until the backing store is
13 exhausted and there is no storage left to commit to the stack page.
16 ============================
17 (Made by DjordjeK, 18/08/2004)
20 1) Fix DEF045247 - T_IDRV assumes that the RAM drive is on drive 0
23 1. Fix DEF046246 - Check in t_serial TimedCheckedWrite() sometimes loses the race
24 2. Fix DEF047403 - T_SERIAL does not exercise Hardware Handshaking
27 ============================
28 (Made by KamranN, 04/08/2004)
31 1) Implement CR - LKUH-5WJMEA: Correct and improve EKA2 Stop-Mode Debug API
34 ============================
35 (Made by AndrewJ, 21/07/2004
39 1) REQ2731.3 (PREQ728): Descriptors API enhancement (RBuf)
40 Added new t_rbuf test.
41 2) DEF045758 T_CHUNK shouldn't depend on shared io buffers
42 - Removed D_CHUNK project.
43 - T_Chunk does not check shared io buffers any more.
44 - D_Sharedio now checks the buffer is initialised to all 0x03s every time it creates it.
47 1) Fixed bug in T_HEAP which caused failures on the emulator.
50 ===========================
51 (Made by AndrewH, 06/07/2004)
54 1) Fixed defect DEF046550 - T_USB sends wrong buffer size for non-max size transfers:
55 We now subtract 1 from the reported buffer size only for 64k transfers (=0xFFFF).
59 ===========================
60 (Made by Michaelp, 17/06/2004)
63 1) Changes for CR HWIE-5YBETE Networking, GT0194 IP Networking, Configuration tool for Etherdrv
66 1) REQ2731.1 RArray API enhancement & REQ2731.2 RMessagePtr2 API enhancement:
67 The following tests are changed:
68 T_ARRAY - Added test cases to check leaving interface of TPointerArray<T>,
69 TArray<T>, TArray<TInt> & TArray<TUint>.
70 T_CAPSUP - Added test cases to check leaving interface of RMessagePtr2.
71 T_MESSGE - Added test cases to check leaving interface of RMessagePtr2.
75 ===========================
76 (Made by JonathanM, 26/05/2004)
79 1) Fixed defect DEF045263 - "T_USBAPI to be made more generic for autotesting":
80 a) We don't check any longer at the beginning of the test suite whether we're
81 running on a 'UsbPlatform': If we're not, then the subsequent attempt to load the
82 USB LDD will fail anyway.
83 b) We now execute the endpoint stall tests only if we're running on a known
84 supported platform, otherwise we just skip those tests.
86 1) Fix for 'DEF043213 : Build dependency introduced between e32test t_lfsdrv2 and
91 ===========================
92 (Made by AndrewJ, 28/04/2004)
95 1) Added new test case to T_TEXTNOTIFIER for defect DEF044341 - Notifier framework
96 (ewsrv.dll) generated KErrNoMemory
99 1) Fixed defect DEF044040 - "iMemMemPslInfo in DDmaTestChannel not set correctly":
100 We now assign the DMA PSL cookie to the kernel side channel object in
101 DDmaTestChannel::DoCreate() for a user side RTestDma::Open() call (instead of
102 for an RTestDma::GetInfo() call).
105 ===========================
106 (Made by JonathanM, 23/04/2004
108 1. JonathanM and KeithR
109 1) Updated T_CAPSUP to test new Platform Security APIs.
112 ===========================
113 (Made by CarlosF, 14/04/2004)
116 1) Added new test cases to T_SOUND which test playing tiny buffers.
119 ===========================
120 (Made by JonathanM, 31/03/2004)
123 1) Fixed defect INC041473 - T_TLOCL fails in WINS - deleting the test as it's considered
124 worthless (see defect notes in TT)
127 1) Fixed defect DEF042556 - "Wrong IPR categorisation on usbrflct binaries":
128 Changed IPR category of USB test prgram USBRFLCT (in directory
129 /e32test/win32/usbrflct_distribution) from B to E. Permission to do so was
130 granted by way of accepted Source Categorisation Change Request
131 CANY-5W8J8E - "Inconsistent USBRFLCT binary IPR classifications".
132 2) Fixed defect DEF043596 - "T_USB/USBRFLCT don't test zero-bytes transfers":
133 We now start the loop transfers with 0 bytes instead of with 4. Upped
134 USBRFLCT version number to 1.3.1.
137 ===========================
138 (Made by JonathanM, 19/03/2004)
141 1. Implemented the requirements:
142 REQ2632 - Configurability of the enforcement of capabilities
143 REQ2633 - Security violation diagnostic
144 REQ3142 - Associate Platform Security information with an executable.
145 See /cedar/generic/base/documentation/Base_How_To_Configure_Platform_Security_Settings.doc
148 ===========================
149 (Made by JonathanM, 26/02/2004)
152 1. Modified T_CAPSUP to test new Platform Security classes: TCapabilitySet and TSecurityInfo
153 2. Added T_PLATSECCONFIG which tests use of the PlatSecDisabledCaps keywork in OBY files
154 and Emulator INI files.
158 (Made by AndrewHi, 23/01/2004)
161 1. Fixed DEF040971 Benchmark suite test code optimised out by RVCT in ARM5 builds
164 1. Modified t_lfsdrv2 to conditionally apply Read-While_Write tests for FLASH devices.
165 The condition is dependent on the value of the device identifier read from the chip.
166 Currently, these tests are only applied for TYAX flash devices.
170 (Made by StefanW, 09/01/2004)
173 1. IBUG-5QNJZC "Implement CR IBUG-5N2DZE on Cedar": Modified T_SERIAL to test the new
178 (Made by StefanW, 19/12/2003)
181 1. Updated USBRFLCT to use USBIO v2.0: Upped USBRFLCT version to 1.2.0 and recompiled
182 it against USBIOLIB 2.0. Updated USBRFLCT.SYS to version 2.0. Added separate driver
183 version for Win98 (USBRFL98.SYS, renamed from USBIO98.SYS) and added entries to the
184 usbrflct.inf file so it gets installed on Win98SE.
185 2. Deleted USB host side GUI test program USBTEST. This program has been discontinued.
186 3. Added complimentary version of USBRFLCT:
187 USBRFLCTD.EXE is a version of USBRFLCT.EXE with lots of verbose output
188 activated; otherwise it's exactly the same program.
189 The output, most importantly the error messages, can be used to track down
190 problems when trying to get USBRFLCT and T_USB to work together, either for the
191 first time (after a Base/PSL port), or after installation changes on the PC.
195 (Made by JonathanM, 05/12/2003)
198 1. Fixed Defect DEF040185 - Some e32tests make assumptions about the time and date formatting
199 Set the locale before the test to the locale info used in the base test roms. Reset it back
200 to the current locale at the end.
203 1. Added new USB test program T_USBAPI: This program tests device side USB API
204 functionality that is not dependent upon a physical USB connection; it requires
205 no manual intervention and can thus be run as part of an automated test suite.
206 These tests used to be part of T_USB but they have been removed from there.
209 1. Modified T_TIMER on Lubbock so it doesn't attempt to power off.
210 (Because the Base Port doesn't support this.)
214 (Made by StephanGu, 12/11/2003)
216 1. - Renamed Win32 USB test program USBTEST(cmdline) to USBRFLCT, also restructured it.
217 - Updated T_USB to use new PID 0x1111.
218 - Added USBRFLCT binary distribution, complete with .sys driver and .inf file.
219 2. - Added new Win32 USB test program USBTEST (was internally: USBPCTester).
220 - Updated T_USB2 to use new PID 0x2222.
221 - Added USBTEST binary distribution, complete with .sys driver and .inf file.
225 (Made by JonathanM, 31/10/2003)
228 1. Added test for RNotifier::Notify to T_TEXTNOTIFIER
233 (Made by AndrewJ, 06/10/2003)
236 1. Added Shared Io Buffers tests.
240 (Made by JonathanM, 24/09/2003)
243 1. Added testing of USB endpoint status, device status ep0 max packet sizes and device control
247 1. Added test case for DEF037709: NULL deref when setting a text RProperty to KNullDesC.
251 (Made by JonathanM, 05/08/2003)
254 1. Extended DMA tests to exercise per-request max transfer sizes.
255 2. Extended DMA tests to exercise "missed interrupts" case.
258 1. CR CANY-5JQFM3 "Enable e32/f32 test suite to be used from a CustKit or DevKit installation".
259 For Typhoon most of the changes were winc specific files.
262 1. Extended T_DES to reproduce defect DEF023302 - Invalid test in
263 HBufC8/16::__DbgTestInvariant(). Invariant functions are now
264 called by the test harness.
267 1. Extended T_ARRAY to test new binary search functions.
270 1) Added testing of USB endpoint status, device status ep0 max packet sizes and device control
275 (Made by JonathanM, 19/05/2003)
277 0. JonathanM and ChrisM
278 1. Changes resulting from making E32 compatible with with
279 Symbian OS version 7.0
282 1. Added test case for fixed defect INC022668 - "Modifier example
283 crashes the emulator"
286 1. Modified D_LDD to test global constructors and destructors for LDDs.
287 Modified T_TLDD to test LDD unloading.
288 Modified T_KHEAP to test OOM handling on LDD loading.
290 2. Modified T_SVR5 to do some more testing of asynchronous messages
294 1. Introduced T_EVENTTRACKER to exercise resource tracking events.
295 See usage instruction at top of file.
296 2. Introduced T_CONTEXT to exercise get/set user context APIs,
297 hardware/software exception trapping and thread death trapping.
298 3. Modified T_SCHEDHOOK to share some code with T_CONTEXT and
299 use the new event handler scheme.
302 1. Introduced T_TRACEREDIRECT to exercise user trace redirection
308 (Made by ChrisM, 17/03/2003)
311 1. Increased maximum loop size for T_MWAIT from 32K to 64K.
314 1. Added missing critical section region in d_dma(sim).ldd
315 2. Added missing critical section region in d_schedhook.ldd
316 3. Migrated fix for Typhoon defect INC017119 - T_TLOCL fails
317 cause test assumes locale UK. The UK-specific part of
318 this test now runs only if the default locale is UK.
319 4. Fixed bugs in test DMA LDD:
320 a. DMA buffers were deallocated before stopping transfers
321 b. DMA requests and channels were closed without cancelling
322 transfers in progress.
323 5. Fixed bugs in T_TRAP, T_THREAD and T_EUNIW: '=' instead of
324 '==' in test assertions.
327 1. Modified T_SOUND to provide a test case for DEF016645
328 - "Sound driver returns incorect length of data when recording"
329 2. Integrate T_MATCH test code changes for lost defect fix
330 BAD-58CDM4 - "Locale change has no effect on contact filtering."
331 and some "Extra folding tests added to T_Match"
335 (Made by JonathanM, 14/01/2003)
338 1. Reorganised how tests get built: All Variant specific tests are
339 now build by the variant. See relevent \variant_name\TEST\BLD.INF
341 2. Made E32TEST code work when binaries are stored in \System\Bin
342 3. Extended T_FAIL to test use of ReAlloc. (Test case for defect DEF015827
343 - "OOM testing works incorrectly for ReAlloc")
344 4. Fixed ARM/Thumb interworking problems with T_MWAIT, also made this
346 5. Fixed ARM/Thumb interworking problems with T_IMB.
349 1. Removed useless T_HUF test case
350 2. Added test T_CACTW for the CActiveSchedulerWait class
353 1. Added T_ENV and T_ENV_CHILD to test parameter passing into
357 1. Added EVENTPRINTER kernel extension in ..\debug. Intentionally
358 not in bld.inf as it is useful only for manually testing event
360 2. Migrated fix for Typhoon defect INC014337 - T_TIMER locked
366 (Made by AndrewT, 29/10/2002)
369 1. Made capability checking test code only test using capabilities which
371 2. Removed all usage of RProcess::SecureApi and User::RemoveCapabilities.
372 3. Fixed T_TIMER so it is more robust and accurate. (Applies same fix as
373 for Typhoon defect INC002103)
376 1. Fixed T_TIMER defect introduced by new power model: use new APIs instead
377 of UserHal::SwitchOff. The test now passes on WINS as well.
380 1. T_DMA and T_DMASIM do not fail anymore when passed an unexpected
385 (Made by JonathanM, 21/10/2002)
388 1. T_MMCPW, T_MSGTXT, T_KILL: set capabilites to none.
390 1. Made T_SHAL an Auto test
392 1. Fixed DEF002050 t_dceutl panic on startup. Also removed build warnings
393 & made secure kernel api changes.
399 (Made by JonathanM, 20/09/2002)
402 1. Split all ARM inline assembler into separate .cia files.
404 1. Changed E32TESTs for secure APIs to use SETCAP.EXE.
406 1. Made all test code work with secure kernel APIs and added
407 CAPABILITY statements to MMP files.
411 (Made by Nicolas, 16/09/2002)
414 1. Implemented Pub/Sub Properties Tests.
416 1) Fixed DEF000090 - E32TEST MMP files should not specify
417 chkstk.obj for WINSCW builds.
419 1) Changes to d_latncy.cpp, d_tick and d_mstim to support Integrator
420 2) Released bm_mi920_pdd PDD for the BM_SUITE test on Integrator
422 1) Added Huffman test buffer\t_huff.cpp & mmp group\t_huff.mmp
427 (Made by Dennis, 16/08/2002)
430 1) Modified T_LDRTST to use Kern::CodeAccess() and
431 Kern::EndCodeAccess().
432 2) Added EXC, an utility which generates various kinds
433 exceptions (data abort, ...).
436 (Made by AndrewT, 12/06/2002)
439 1) Update t_float.exe in line with the fix for migrated Huricane defect
440 (TLex::Val() should not give error for missing exponent)
441 2) Added t_emul.exe to test emulator specific functionality, in particular
442 the 'escape' mechanism.
446 (Made by AndrewT, 23/05/2002)
448 First release of EKA2 into Symbian OS mainline.
453 (Made by Dennis, 04/04/2002)
456 1) Submitted Jon's changes to T_KHEAP. This test now
457 checks that there is no memory leak if an OOM occurs
458 while expanding the handle array.
462 (Made by CarlosF, 13/02/2002)
465 1) In T_KHEAP, enabled again the process-related code
466 which was commented out. This helped find an
467 untrapped leave in the process creation code which has
469 2) In T_KHEAP, enabled again the chunk-related code which
470 was commented out. A comment said that the code
471 (setting the next kernel heap allocation to fail and
472 check that the chunk size can be adjusted) was
473 disabled because it could not possibly work. Actually
474 it does because the memory used when adjusting a chunk
475 is taken straight from the physical memory pool, not
476 from the kernel heap.
477 3) Silenced warnings in T_KHEAP.
481 (Made by ChrisM, 06/03/2002)
484 1) Extended T_PWSTR to test E32 fix CAN-579MLN.
487 1) Changes to T_SERIAL to introduce a new test to test the configurable power
488 down timeout feature introduced on the serial driver LDD.
492 (Made by MarkCa, 15/01/2002)
495 1) Fixed e32test\win32\usbrflct\posusb\brutus_pos.inf
496 [SYMBIAN] section had a missing ']'.
497 Removed references to 'ER6.2'
498 2) Renamed brutus_pos.inf to symbian_pos.inf
501 1) Fixed FRS-54YG25 T_SERIAL fails on Assabet.
505 (Made by MichaelP, 20/11/2001)
508 1) Modified e32test\win32\usbrflct\posusb\brutus_pos.inf
509 to use official vendor id
512 1) Updated T_ATADRV to unlock the secure area when it is run on
516 1) Fixed MOU-543FVR HAL attribute ECaseSwitch returns wrong value for Assabet.
517 The following files were modified:
518 \e32test\pccd\t_lfsdrv.cpp.
519 Fixed FRS-534EWW T_SERIAL fails on Assabet. The following files were
521 \e32test\device\t_serial.cpp.
525 (Made by Dennis, 26/10/2001)
528 1) Moved some definitions from P32MMC.H into T_PWSTR.CPP because it can
529 only be included into ASSP-specific builds. See \e32\release.txt.
530 2) Updated T_MMCDRV to print read position only every 32k to speed up test.
533 1) Added tests to T_TIME to verify defects FER-4XGDVT and FER-4XDK9G
536 1) Fixed MOS-53LEVN t_lfsdrv uninitialised var on WINS.
539 Migrated the fixes for the following 6.1 defects:
540 1) Fixed SCE-534FSY "Serial power handling tests disabled in T_SERIAL"
541 - Renabled the power handling tests and modified them to conform to the
542 serial driver 'idle off' power scheme.
543 2) Fixed PAS-4WRMQS "Requeing a read() after PowerOn Panics the Driver/Kernel"
544 - Added tests to T_PCCD1 for the new KPccdConfigOverrideNormalPwrEvents option
545 now supported by the PC Card Controller function:
546 DPcCardController::RequestConfig().
548 3) Fixed MAY-4ZGH8S: "t_serial shoots itself in the foot, and can never pass".
552 (Made by ChrisM, 05/10/2001)
554 1) Fixed ROS-4Y3CUU (T_MATH and T_FLOAT rely on the compiler generating denormal constants)
555 Defined the denormals used in the test in t_vals.h. Removed redundant repeated code in t_vals.h
556 2) Fixed MOS-52BH6B (t_math loop problem)
559 1) Removed example device driver PDD - SEROMAWD.PDD.
560 2) Removed T_MEDCH and D_MEDCH.
561 3) Removed T_BEXFIQ and D_BEXFIQ.
562 4) Removed T_EXTFIQ and D_EXTFIQ.
563 5) Removed T_CRMDRV and T_CRRDRV
567 (Made by Pete, 06/08/2001)
570 1) Moved T_PWSTR from f32test to e32test. F32 API tested by T_MMC.
573 1) Removed \e32test\group\mnt.bat which was redundant.
574 2) Fixed PSCE-4YTR5G "Update SEROUT example driver":-
575 Improved the power handling in the example driver SEROUT.LDD - making
576 it more representative of other device drivers.
577 Modified the driver so that it supports 19200bps+9600bps rather than
578 9600bps+300bps (together with corresponding changes to the test program
580 Removed simulation of emergency power down from the driver and test.
581 Modified test program so that during the XON/XOFF test it runs until a
582 key is pressed rather than running for a fixed period of time.
583 Modified test program so it unloads the Brutus PDD rather than the Cogent
585 3) Added new tests for functions TBusLocalDrive::SetMountInfo() and
586 TBusLocalDrive::ForceRemount() to T_MEDDRV.EXE. Corresponding changes to
587 MEDT1.PDD and MEDT2.PDD which now require the correct
588 mount info descriptor to be supplied before opening.
591 1) Added tests for folding of Latin extension characters to T_MATCH
595 (Made by Michael, 21/06/2001)
598 1) Added and modified files for the MHELEN ASSP base port.
601 1) Fixed test programs t_atadrv and t_mmcdrv to work with e32 fixes
602 for EDNHLJT-4WZGTD and EDNHAAO-4WZFYR.
603 2) Added debug password notifier tests to t_mmcpw.
607 (Made by Carlos, 31/05/2001)
610 1) T_Char can now check the graphical/printable attributes. Tests have
611 been added to check the assigned private use area characters.
612 2) T_Match has now more stringent checks on matching folded, and tests
613 new internal functions for matching folded.
614 3) Indic vowel swapping tests have been removed, as collation no longer
615 has this functionality.
619 (Made by Pete, 10/05/2001)
622 1) Minor change to T_PCCD3 to allow it to be re-run. Prior to
623 this, it passes the first time run but then subsequently fails
626 1) Added T_PROC3.EXE to test RProcess::GetMemoryInfo().
627 2) Modified T_DEBUGGER and T_DEBUGGEE_1 to test the validity of the
628 text section base address provided by RDebug.
632 (Made by Nicolas, 04/05/2001)
635 1) Merged 6.1 release branch back into the mainline.
639 (Made by Dennis, 20/04/2001)
642 1) Corrected cases where delete was used on pointers obtained with new[] - these
643 should use delete[] instead. This only affects T_ARRAY, and was picked up by the
644 CodeWarrior compiler which handles new and new[] differently.
646 1) Added T_SILENT to test 'silent running' functionality
647 2) Updated all tests that call UserHal::SwitchOff() to turn the screen back on
652 (Made by RobertJ, 26/03/2001)
655 1) Removed media change tests from t_atadrv for l7200.
656 2) Added more thorough checking of multi-block reads to t_mmcdrv.
657 3) Increased WINS PDD name buffer size for t_dtenot.
660 1) In test T_CPER defer all printing until after the timing-critical phase.
661 2) T_TIMER doesn't test power off on IQ80310.
665 (Made by Morgan, 08/03/2001)
668 1) Added test program T_CP0 to test the XScale DSP coprocessor.
669 2) Added test program CRASH to crash the system and get it into the
671 3) Modified test T_PS to use default screen size instead of hardcoded
672 78*18. This fixes defect HAS-4TVQ6D (T_PS crashes Brutus).
673 4) Modified test T_WWINS to take account of the screen size when setting
674 up windows. This is part of the fix for HAS-4U7N92.
678 (Made by Carlos, 20/02/2001)
681 1) Fixed defect EDNJHAS-4MKF4J (T_KILLER support test fails) by
682 implementing suggested fix. Also fixed two bugs preventing the test
684 2) Adding test case in T_SHADOW to verify code patching (following
685 defect EDNNTHY-4PMKJW).
686 3) D_SHADOW.LPP now flushes caches after modifying shadow pages.
689 1) Fixed T_DTENOT for Assabet (see comments in t_dtenot.cpp)
690 2) Modified T_SERIAL for Assabet (see comments in t_serial.cpp)
694 (Made by Morgan, 09/04/2001)
697 1) Fixed BRM-4V3E23 "T_NAME Colourful language".
700 1) Fixed EDNPSCE-4K9MZJ "T_PCCD3 panics on Brutus"
704 (Made by Nicolas, 27/02/2001)
707 1) Removed media change tests from t_atadrv for l7200.
708 2) Added more thorough checking of multi-block reads to t_mmcdrv.
709 3) Increased WINS PDD name buffer size for t_dtenot.
712 1) Modified test T_PS to use default screen size instead of hardcoded
713 78*18. This fixes defect HAS-4TVQ6D (T_PS crashes Brutus).
714 2) Modified test T_WWINS to take account of the screen size when setting
715 up windows. This is part of the fix for HAS-4U7N92.
718 1) Modified romutil.mke so that D_DRVIF and D_PCCDIF ldds are build
724 (Made by Markdo, 07/02/2001)
727 1) Added test to check the fix for EDNTBAD-4Q8M2F
728 "CBufSeg can leak memory" works.
731 1) Fixed defect EDNJHAS-4MKF4J (T_KILLER support test fails) by
732 implementing suggested fix. Also fixed two bugs preventing the test
738 (Made by Pete, 19/01/2001)
741 1) Added test code for RMessagePtr::IsNull to T_MESSGE
746 (Made by RobertJ, 13/12/2000)
749 1) Added new test, t_thread2, because t_thread was becoming quite big.
750 Currently this test just checks that deadlock doesn't occur when a
751 thread suspends itself.
754 1) Disabled serial logging over time critical parts of T_CPER for
755 platforms with slow serial devices
758 1) Fixed defect EDNRFID-4PXFJ8 (Not enough output in T_Y2K).
762 Made by Nicolas, 19/10/2000
765 1) Disabled T_PCCD1, T_PCCD2 and T_PCCD3 in WINS builds
766 since WINS is now built with an MMC Peripheral bus controller rather
767 than an PC Card MMC Controller. (These tests will be
768 re-enabled in WINS when multiple Peripheral Bus Controllers are
773 Made by MarkCa, 29/09/2000
776 1) Adjusted MMP files to match E32 301 include file re-org.
779 1) Added MLNK to list of platforms supported in bld.inf.
783 Made by Dennis, 13/09/2000
786 1) Modified T_FUNC, T_LEX, T_TIME and T_BUF to get rid of GCC warnings
787 "decimal integer constant is so large that it is unsigned". The fix is
788 to put the "u" qualifier on the constants involved.
792 Made by Alastair, 2/8/2000
795 1) Completed fix to EDNCMOS-4M8F52 "t_thread undertaker test failure".
800 Made by RobertJ, 20/7/00
803 1) Added locale specific test code to to T_LEX
806 1) Fixed EDNCMOS-4M8F52 "t_thread undertaker test failure"
807 2) Fixed EDNCMOS-4M8FDR "t_euniw fails if logged on as a user"
810 1) Modified T_CPER to print the times at which the locked timers expire.
811 2) Modified T_INACT and T_TIMER to test that inactivity timers are
812 unaffected by system time changes.
816 Made by Morgan, 5/7/00
819 1) Modified t_fir2 to be more rigorous. Now writes the sent length into
820 the first 2 bytes of the buffer so that the receiver can see if the
821 correct length was received. Also modified text menus, added a nominal
822 1ms delay and changed 2051 bytes option to 2042 so that 2051 is covered
823 at the upper limit of the test. Text report now shows the number of frames,
824 the number of errors and the error %ge
829 Made by Bob 29-06-2000
832 1) Modified serial tests so that it is possible to open
833 a channel on Com4 for Brutus. Changes to:-
841 Made by Dennis 22-06-2000
844 1) Removed nasty hacky memory snoopoing from T_CALIB
847 1) Changes to T_BUF and T_CHAR to test TDesC16::MatchC.
852 (Made by MarkCa, 16/05/00)
855 1) Modifications to T_SERIAL and T_DTENOT to allow it to
856 work with PCMCIA serial ports.
857 2) Updated T_MODEM1 and removed T_MODEM2.
860 1) Fixed EDNABRY-4K2LQA
861 "Test t_tdll fails under WINS"
862 2) Fixed EDNABRY-4K2LBC
863 "Test T_Messge failing under WINS UDEB"
867 (Made by Morgan, 7/4/2000)
870 1) Fixed EDNABRY-4JHLV3
871 "Test T_Thread failing intermittently under WINS UDEB"
876 (Made by Morgan, 14.04.2000)
879 1) Removed calls of UserPcCardCntrl::PwrDown();
880 2) Modified serial tests so that it is possible to
881 overide the default serial port designated to be used
882 by that test by supplying a serial port number on the
890 1) Removed T_DBG and T_BREAK. application debugging tests are
891 now in F32TEST because F32 is needed to copy support
892 files to the RAM-drive.
895 1) \E32TEST\WINDOW\T_MMCPW.CPP: no longer tests that
896 KMaxMediaPassword % 4 == 0.
899 1) Copied k32snd.* and d32snd.* from \e32\include into e32test
900 so the sound test programs build. These files are no longer
904 1) Updated t_euniw test to cope with CreateMutex() and OpenMutex().
909 (Made by Robert, 17/03/00)
912 1) Added User::ResetInactivityTime tests to T_TIMER.CPP
915 1) \E32TEST\WINDOW\T_MMCPW.CPP: minor change to password notifier
920 (Made by MarkCa, 03/03/00)
923 1) Minor changes to T_SEROUT.
924 2) Added T_MMCPW.CPP to test new password notifier.
927 1) Alterations to T_MODEM2 to work with CF card serial
933 (Made by Pete, 24/02/00)
936 1) Change to T_COMMSK to work with MAWD CF card modem driver.
937 2) Completed T_DCE test program to test DCE Serial driver on
939 3) Exteneded T_DTENOT to cover signal change notification for
940 each of the CTS/DSR/DCD signals becoming MARK or SPACE.
941 Test now passes on WINS,MAWD,MISA,MEIG,MCGA.
944 1) Added Slow IR test T_SLOWIR
949 (Made by Dennis, 14-02-2000)
952 1) Replaced most UserHal:: calls with corresponding HAL:: calls (if they
956 1) Added T_EUNIW to test unicode wrapper functions in EUNIW.DLL.
959 1) Added T_SEROUT test program to test SEROMAWD.PDD, SEROWINS.PDD and
960 SEROUT.LDD in EXAMPLE dir.
963 1) Further changes to T_MODEM2 to work with MAWD CF card modem driver.
968 (Made by Simon, 03-02-2000)
971 1) Modified T_MODEM2 to work with MAWD CF card modem driver.
972 2) Modified T_DCE to take into account Cogent Serial PDD re-org.
975 1) Replaced '/e32/inc' paths with '/epoc32/include /epoc32/include/kernel' in
976 a number of mmp files.
980 (Made by Dennis, 19/01/00)
983 1) Removed ARM3 and MARM targets from BLD.INF.
984 2) Minor fix to T_EXC so that it works under WINC.
985 2) Minor change to T_WBC to format test output correctly.
988 1) Changes due to the removal in E32 of the classes
989 TLocalDrive, RLogicalChannel, RDevComm.
992 1) Removed tests for functions which no longer exist.
993 2) Changed T_BUF, T_R32 and T_FLOAT to use TRealX instead of TReal96.
994 3) Added chkstk.obj to T_FLOAT so it works on WINS.
995 4) Removed T_R96 since TReal96 no longer exists.
996 5) Temporarily removed T_R64 since it made extensive use of TReal96 and
997 simply changing over to TRealX doesn't work since the extra precision
998 of TRealX makes the tests fail. What is really needed is a set of test
1000 6) Increased stack size for T_BUF to 12K.
1005 (Made by Peter, 21st December 1999)
1008 1) Added T_DTENOT to test DTE serial driver change notifiers.
1009 2) Added T_DCE and T_DCEUTL to test DCE serial driver.
1013 (Made by Morgan, 3/12/99)
1016 1) Fixed T_MESSGE problem caused by different panic code following IPC
1018 2) Modified tests so they can run from any drive, not just Z:. This
1019 basically involved removing explicit references to Z: - we use
1020 RProcess().FileName() to determine the path instead.
1021 3) Added test program T_WBC to test that write back cache is flushed
1022 when memory is unmapped.
1023 4) Changed T_SVR3 and T_SVR6 to support since they don't terminate.
1026 1) Added T_CLI to test RProcess::CommandLine()
1030 (Made by Simon 25-11-1999)
1033 1) Changed t_fir2 to test config and added t_fir.
1039 1) Fixed minor typo in T_BMA.
1044 (Made by Morgan, 26/8/99)
1047 1) Added E32SDRV.PDD and E32SOUND.LDD, previously part of
1049 2) Added new keywords to various .MMP files so that .IBY files can be
1050 generated by ABLD TEST ROMFILE, also added extension makefile
1052 3) Fixed WINS compilation problem in PCCD\T_PCCD3.CPP.
1053 4) Added extension makefile GROUP\ROMUTIL.MKE, to incorporate required
1054 executables from E32UTILS into ROMS. This is a temporary measure and
1055 should be undone once the executables in question are moved from
1056 E32UTILS into E32TEST.
1059 1) Changes to T_PCCD3 to allow it to work on Brutus.
1062 1) Added TestContiguousAllocation() to t_bma to test
1063 ImpMmu::RemoveRamPages().
1065 3) Tidied up T_Serial. Tested for Wins and Brutus.
1068 1) Fixed structure packing problems in T_FARRAY with new GCC.
1069 Use 4-character arrays instead of 3 in the offending code.
1070 2) Modified T_TLDD and D_LDD to test initialisation of .bss.
1071 3) Various fixes to T_TPARSE.
1072 4) Fixed operator precedence errors in T_TPARSE, T_ATADRV,
1073 T_ATDRV2, T_CRRDRV, T_IDRV, T_CHNOT.
1074 5) Added __SWITCH_TO_ARM macros to some tests so that they can build
1078 1) Renamed t_fircon2.cpp to (8.3) t_firco2.cpp
1083 (Made by Alastair, 2.8.1999)
1087 1) Added ARMI, ARM4 and THUMB platforms to BLD.INF file.
1092 (Made by Dennis, 13.7.1999)
1095 1) Added dopcvs.pl and lock.pl to GROUP directory for use with MNT.BAT.
1098 1) Modified T_MESSGE, T_ARRAY to work with new GCC/THUMB.
1099 2) Added tests to T_BUF to test variable precision TDes::Format() and
1100 precision applied to strings.
1101 3) Modified MMP files for T_CALIB, T_COLOUR, T_PS, T_WJPWIN, T_WSIMP
1102 and T_WWINS to link against EWSRV.LIB instead of ECONS.LIB.
1103 4) Removed tests of the form (boolean expr)==TRUE from T_R96.CPP.
1104 5) Changed checked panic category in T_TRAP from USER-Plat to
1106 6) Took updated version of T_CHAR from Graham Asher to match new unicode
1108 7) Changed status of some tests to 'support'.
1113 (Made by Alastair, 9.6.99)
1116 1) Moved the *.rel files to e*.rel to match e32 and f32
1121 1) Modification to T_PCCDSK to allow the soak test buffer size to be input
1122 in units of a sector. Prior to this, the buffer size had to be input in
1123 units of 1Kbyte, with 8K being the largest size.
1124 2) Modified T_ATADRV - added additional tests to verify the operation of the
1125 functon TBusLocalDrive::Format(TInt aPos,TInt aLength). This allows just part
1126 of a disk to be formatted and was added to E32-181.
1127 3) Added the new test program T_LFSDRV.CPP to test the LFS media driver on
1129 4) Added general media driver test T_MEDDRV and two associated media drivers
1130 - MEDT1.PDD and MEDT2.PDD.
1134 1) Changed zips of releasables put onto the network. There is now one zip
1135 file for each platform containing full \Epoc32 paths to each item.
1136 There is also a zip file for exported header files. Hence, the contents
1137 of S:\e32test\zip includes wins.<ver>, meig.<ver>, misa.<ver>, marm.<ver>
1138 mawd.<ver> and winc.<ver>.
1139 The files within these zip files now also have full pathnames so ensure
1140 you unzip them into the root of your EPOC drive with any flags required
1141 to recreate the directory structure.
1142 Added new command, MNT CHECKREL, to check that all releasables are
1143 present before doing an MNT PUTREL.
1144 Before doing MNT CHECKREL or MNT PUTREL call BLDMAKE -v RELFILES to
1145 create the lists of releasables.
1146 2) Changed D_LDDNS.MMP and added BMARM\D_LDDNS.DEF to avoid new
1147 makmake warnings arising from LDD targettypes having the wrong second
1149 3) Replaced all B[plat].PRJ files for use with BLDMAKE with BLDMAKE.PRJ in
1150 accordance with ER5u bldmake changes. No MAWD or SARM tests are now
1151 built in addition to the MARM tests - there doesn't seem to be any need
1152 since the __MAWD__ and __SINGLE__ macros haven't been used anywhere to
1153 build the test differently from the old standard MARM build. Where
1154 tests do behave differently according to the h/w platform, if they do,
1155 then this should be tested at runtime and the behaviour of the test
1156 changed accordingly.
1157 4) Removed superseded \E32TEST\GROUP\RUN.BAT.
1158 5) Removed the batch files used for running tests. These batch files are
1159 now created by BLDMAKE in \Epoc32\Bldmake\E32test\[platform]\, and are
1160 now called AUTO.BAT and MANUAL.BAT.
1161 6) ETIMR and T_DEVC have been moved into this project from E32.
1162 7) Removed all *.REL files from GROUP directory and changed MNT.BAT so that
1163 MNT PUTREL looks for .REL files generated by BLDMAKE instead. Updated
1164 BLDMAKE.PRJ accordingly.
1165 8) Changed #defines in the following files now that __MEIG__ is being used
1166 to mean __MARM__ and __MARM__ is being used to mean ARM CPU:
1167 SYSTEM\T_EXC.CPP, SYSTEM\T_DBG.CPP.
1168 9) Updated MNT GETBLD to fetch the SDMODELS component which replaces the
1169 SWINS component that E32 used to release.
1170 10) Added the following .DEF files:-
1172 D_TLSDLL.DEF, T_DLL.DEF, T_DLL1.DEF, T_DLL2.DEF,
1173 T_START1.DEF, T_START2.DEF, T_START3.DEF, T_DUID.DEF,
1175 T_DLL.DEF, T_DLL1.DEF, T_DLL2.DEF, T_DLLA1.DEF, T_DLLA3.DEF,
1176 T_START1.DEF, T_START2.DEF, T_START3.DEF, T_DUID.DEF
1177 11) Added \EPOC32\INCLUDE\KERNEL\ as a systeminclude path to those .MMP
1179 12) Specified EKERN.LIB as an ASSPLIBRARY in those .MMP files requiring it.
1184 (Made by Jonathan, 23rd February 1999)
1187 1) Boilerplated source.
1190 1) Commented out tests in T_REALX involving assigning a NaN from a float
1191 or double to a TRealX in MSVC builds. This is because MSVC insists on
1192 passing the NaN through the X86 FPU which quietens the NaNs and causes
1196 2) Added test to check that switch-off is disabled for 5 seconds before
1200 1) Withdrew T_RTIM test.
1201 2) Made T_DBG build for MAWD as well as MARM. MAWD version doesn't
1202 contain assumptions about Eiger registers.
1207 (Made by Jonathan, 10th February 1999)
1210 1) Conditioned compilation of functions in T_PROT related to
1211 instruction exceptions on __EPOC32__ to prevent WINS/C warnings and
1212 fixed cast in ExceptionThread() to keep MSVC compiler happy.
1216 2) Added D_LDDNS.LDD which is an LDD with dodgy UID's.
1219 1) Modified BMAWD.PRJ to remove EIGER specific test programs.
1224 (Made by Pete, 3rd February 1999)
1227 1) More tests in T_TDLLA to test calling of Dll entrypoints with
1228 EProcessAttach and EThreadAttach
1229 2) Added case switch on/off test
1232 1) Updated T_PROT.CPP to test that data abort, prefetch abort and
1233 undefined instruction exceptions in user mode all result in
1234 KERN-EXEC 3 panics rather than killing the kernel.
1237 1) Added new tests to T_DBG to check that RDebug::ReadMemory and
1238 RDebug::WriteMemory return errors (and don't panic the kernel -
1239 problem EDN671244) when reading/writing to/from unmapped address
1243 1) Changed over to use the TBusLocalDrive class rather than TLocalDrive
1244 for local drive access. This in turn required converting calls to
1245 TBusLocalDrive::Caps() to take a TLocalDriveCapsBuf parameter rather
1246 than a TLocalDriveCaps.
1247 2) Added T_ATDRV2.CPP to test TLocalDrive class specifically.
1253 (Made by Morgan, 27th January 1999)
1256 1) Modified T_SOUND.CPP so it can be run on a Snowdrop.
1261 (Made by Pete, 21st December 1998)
1264 1) Updated D_PATCH*.MMP to specify KPatchLddUid as third uid, in line with
1265 changes to F32 version 123.
1266 2) Moved T_LDLL to F32TEST.
1267 3) Minor modifications to MNT.BAT.
1268 4) Minor change to T_REALX to avoid VC5 compilation problem.
1274 1) Modified T_PCCD1, T_PCCD2, T_PCCD3 to work with E32UTILS 025.
1275 2) Removed a call to a UserPcCardCntrl function which was removed from
1276 EUSER in E32-158. (The call was already redundant since the function
1277 has been withered for a while).
1278 3) Released MAWD releasables.
1283 (Made by Jonathan, 27th October 1998)
1286 1) Added code from Andrew Thoelke to T_DES.CPP to test compiler-generated
1287 literal descriptors.
1290 1) Updated .MMP files so that UNICODE builds are available in line with
1291 UNICODE changes to E32TOOLP 095.
1296 (Made by Morgan, 21st October 1998)
1299 1) Changed WINS T_SERIAL.CPP to link to a new dll,
1300 D_TLSDLL.DLL, the entrypoint of which sets thread local storage.
1301 This will test whether the deadlock problem described by
1302 Epoc Proposal SW1 - 19 is avoided. T_SOUND.CPP should link to
1304 2) Changed all the .rel files so that they specify T_DUID.DLL
1305 rather that non-existent T_DUID.EXE.
1306 3) Removed T_ARRAY from AWINC.BAT since it isn't built or
1307 released under WINC.
1308 4) Turned off just-in-time debugging in T_THREAD.CPP and T_PROC1.CPP where these
1309 tests panic threads and then check their status in WINS debug builds.
1310 Now that code jumps into the debugger whenever a thread is panicked (E32 152),
1311 rather than just whenever a thread is panicked with RThread::Panic(),
1312 these tests would have jumped into the debugger rather than completing successfully.
1315 1) Added new test program:
1316 T_Y2K.EXE - tests the Year 2000 compliance of E32
1321 (Made by Pete, 20/08/1998)
1324 1) Added new test programs:
1325 T_ARRAY.EXE - tests new RArray<> classes
1326 T_MWAIT.EXE - tests core clock and memory access speeds
1327 T_PROF.EXE - retrieves and displays thread profiling information
1328 D_PROF.LDD - LDD for use with T_PROF.EXE
1329 T_CTXSW.EXE - measures raw context-switch speed (not message passing)
1330 T_PAGES.EXE - dumps all MMU page tables to serial port
1331 T_PANIC.EXE - prints out a threads context when it dies
1332 2) Modified T_CHUNK to test RChunk::IsReadable() and RChunk::IsWritable().
1333 3) Added another test to T_MATH2.CPP to test the accuracy of Math::Mod() using
1334 the Java IEEERemainder test vector data.
1337 1) Turned off just in time debugging in T_THREAD, T_MESSGE, and T_EXC
1338 2) Massaged T_TDLL.CPP for Dll MODULE_HANDLE changes
1341 1) Turned off just-in-time debugging in T_SOUND, and removed WINS test which
1342 is only valid when the driver does dummy recording.
1343 2) Added test to T_SOUND to check that memory allocated in prepare functions
1345 3) Changed T_SOUND so that it passes (with warnings) on hardware which
1346 doesn't support all possible capabilities.
1347 4) Switched off just-in-time debugging in T_TIMER.
1350 1) Changed .mmp files D_LDLL, D_RTIM, T_DLLA1 and T_DLLA3
1353 1) Changed t_reset test so that it checks that 3 dummy patch ldds
1354 are reloaded after a warm reset. Added dummy patch ldds D_PATCH1.SYS,
1355 D_PATCH2.SYS and D_PATCH3.SYS.
1356 2) Minor changes to T_SVR.CPP, T_MSTIM.CPP to remedy compilation
1358 3) Change to T_CHUNK2.CPP for !WINC builds to allow for E32 change in total
1359 RAM size to 64Mb under WINS in E32 141.
1360 4) Updated T_CHAR.CPP to take account of Graham Asher's changes to E32 141.
1361 5) Fixed MNT GETSRC so that it handles version numbers correctly.
1362 6) Minor corrections to T_THREAD.CPP.
1365 1) Added T_PCCDSR to PCCD. This is a program for
1366 repeatly writing to a fixed number of sectors on a CF card - to test
1367 if the wear leveling scheme on the card causes any problems.
1368 2) Modified T_PCCD1.CPP,T_PCCD2.CPP and T_PCCD3.CPP to
1369 use D_PCCDIF.LDD for accessing the PC Card Contoller rather than
1371 3) Modified T_CRMDRV.CPP to use D_DRVIF.LDD for direct
1372 access of the media device rather than UserPcCardCntrl.
1373 4) Modified T_PCCDSK.CPP to use
1374 UserSvr::ForceRemountMedia() rather than UserPcCardCntrl::PwrDown();
1375 5) Added further tests to T_ATADRV - testing reads and writes
1376 which are longer than the media drivers local buffer
1377 - particularly required after the changes to the ATA
1378 media driver in E32-143.
1383 (Made by Graham, 26th March 1998)
1386 1) Added T_TLDD and D_LDD to test LDD static data. EPOC only.
1387 2) Modified Dll static data tests (T_TDLLA). Added MiniRomLoader
1391 1) Changed MNT.BAT to retrieve E32TOOLS 091 and E32TOOLP 091.
1394 1) Changes Test6() and Test8() in T_COBJ to use less stack and
1395 hence build on WINS/WINC without __chkstk errors.
1397 2) Changed D_RTIM and D_BEXFIQ interrupt binding to use narrow
1398 descriptors, in order to work with E32 127 and higher.
1400 3) Removed User::After(999999) from T_CHNOT, to speed it up. This
1401 wait is not needed, since midnight-crossover changes now happen
1402 earlier than they used to.
1404 4) Changed T_CHUNK2 so that it did something more sensible on WINC,
1405 where the concept of "free ram" is different.
1407 5) Changed T_REGRAM so that is disables auto switch-off while running.
1409 6) Fixed a few incorrect tests in T_THREAD.
1414 (Made by Jonathan, 10th February 1998)
1417 1) Changed t_sec.cpp to test new security function for E32. The
1418 function will only be tested if the version of E32 being used
1419 supports the new function, so that the new version of t_sec.cpp
1420 will still work with old versions of E32.
1423 1) Added the test program T_COLOUR to test the new
1424 colour text window server.
1427 1) Reoved kernel heap checking fro a number of tests. We need
1428 to rethink our strategy for this since the current system
1431 2) Removed superfluous operator attendance from T_TIMER,
1434 3) Added AMARM.BAT, AWINS.BAT, etc which run all the
1435 automatic tests fro a given platform.
1440 (Made by Graham, 23rd January 1998)
1442 E32TEST contains those test programs which used to be a part of E32
1445 drvbench -> t_pccdbm, drvsoak -> t_pccdsk and more
1448 t_svr2 fails due to heap debugging code errors on MARM
1449 t_cobj won't compile for WINS - needs stack use reducing
1450 t_tdll,t_rtim and t_irqs seem to fail