In order for the Eject command to properly release the image and reset the status,
we need to override the Eject function with the appropriate logic
The other removable disk drive devices inherit the Disk class,
and hence do not need this
Mode Sense (6) supports a 32-bit number for the maximum block count, where Mode Sense (10) supports a 64-bit number for the count. The Seagate SCSI Commands Reference Manual from 2016, has this to say about a "Short LBA mode parameter block descriptor" as returned by Mode Sense (6):
> On a MODE SENSE command, if the number of logical blocks on the
> medium exceeds the maximum value that is able to be specified in
> the NUMBER OF LOGICAL BLOCKS field, the device server shall return
> a value of FFFFFFFh.
Similarly, the Read Capacity (10) command should return FFFFFFFFh if the capacity of the drive exceeds 2^32 sectors, and that a driver should then know to use Read Capacity (16) if it supports that command.
There may be an unexpected side-effect if presenting a drive of more than 2^32 sectors to a legacy host that does not support devices that large. The Read Capacity commands specify the value returned as the last addressable sector on the device. This means that typically, an application which uses the value returned by that command is going to add 1 to it to get the actual number of blocks on a given device. If the program is not aware of Read Capacity (16), and is not using greater than 32-bit math, then it might report to the user that the total number of sectors on the drive as 0. I don't view this as a huge problem, however. In that case, the legacy driver wouldn't correctly support the capacity of the drive anyway, so providing that driver with a device that is 2 ^ 32 sectors or larger wouldn't make sense from the user perspective.
The scsictl command parser looks for single letter shorthands for each device type;
this adds 't' for SCTP and changes SCLP to 'p'
Now you can do '-t tape' and '-t printer' for an easy to remember command
Also moves the SCTP default product/vendor into the device factory,
updates the unit tests, and an overhauls the man pages
The X68000 host bridge and associated host file system module
has not been functional for many years, as reported by multiple users
Removing the code will reduce complexity for users -
the SCBR device has caused some misunderstandings in the past -
while eliminating code that has been flagged as insecure by SonarQube
* Revert "Don't ResizeCache on sector change if no filename is defined (#1438)"
This reverts commit dd9a3296d4.
* Revert "Add ModeSense page 0x25 (DEC special function control page) (#1412)"
This reverts commit 1121b8d9d6.
* Revert "DiskCache needs a size"
This reverts commit 7cc8df271c.
* Revert "Honor sector size change via ModeSelect6 in scsicd (#1406)"
This reverts commit b7f65d33e2.
* Revert "Multiple fixes for ModeSelect (#1405)"
This reverts commit ad5eae93e7.
* Add tests for issue#1426
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* add test without medium
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* Don't resize DiskCache without a file
DiskCache computes its size from a file.
Skip ResizeCache Without file (i.e. 'empty' CD)
Fixes#1426
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* Add ModeSense page 0x25 (DECSpecialFunctionControlPage)
VAXServer 3100 (CPU KA41-E) console firmware issues ModeSense(6)
page 0x25 when probing for disks. A disk won't be recognized if
this returns an error.
The DEC SCSI specification[1], section 8.5 documents this page.
[1] https://manx-docs.org/collections/antonio/dec/dec-scsi.pdfFixes#1410
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* Make ModeSelect() non-const
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* Implement ModeSelect for scsicd
and honor sector size setting.
DEC's VMS can't handle 2k sector sizes and uses ModeSelect6 to set the
sector size to 512 bytes.
Fixes#1397
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* Test sector size setting via ModeSelect in SCSICD
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* Re-calculate total blocks when sector size changes
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* Reset CD data tracks after sector size change
The track calculation is based on sector size and must be reset after
change of sector size.
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* resize cache after change of sector size
The disk cache is based on sector size and must be resized when the
sector size changes.
Disk::ResizeCache needs a `raw` parameter, make this value accessible
from the current cache.
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* Make GetRawMode const
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
---------
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* Allow 'empty' ModeSelect6
tl;dr
Treat a computed length of 0 as `has_valid_page_code`.
Details:
The SRM console (aka 'BIOS') of DEC Alpha sends an empty
ModeSelect6 with the following data:
~~~
ModeSelect6, CDB $151000000c00
~~~
That makes 12 byte(s) as follows
~~~
0 1 2 3 4 5 6 7 8 9 10 11
00 00 00 08 00 00 00 00 00 00 02 00
~~~
decoding it (accoring to [1], Section 8.3.3, Table 94) gives us
Mode Data Length 0
Medium Type 0
Device-specific 0
Block desc len 8
Density Code 0
Number of blks 0
Reserved 0
Block length 512
`scsi_command_util::ModeSelect` computes
~~~
offset = 4 + buf[3];
~~~
giving 12 and
~~~
length -= offset;
~~~
giving 0.
Thus it never enters the `while` loop and `has_valid_page_code` stays
`false`, raising an error.
[1] [Small Computer System Interface - 2 rev 10L.pdf](https://dn790004.ca.archive.org/0/items/SCSISpecificationDocumentsSCSIDocuments/Small%20Computer%20System%20Interface%20-%202%20rev%2010L.pdf)
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* Allow ModeSelect with page code 1
OpenVMS Alpha (the operating system, not the SRM BIOS) uses
ModeSelect6 with a page code of 1.
The semantics are unknown, just accepting it works for me.
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* Fix page length computation in ModeSelect
tl;dr
The 'skip to next ModeSelect page' computation was off-by-one, either not
taking the page code itself into account or missing the fact that the
page length is given as `n - 1`.
Fix:
Add 1 to the computed length.
Details:
OpenVMS Alpha sends a ModeSelect6 as follows
~~~
command:
ModeSelect6, CDB $151000001900
payload:
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
00 00 00 08 00 00 00 00 00 00 02 00 01 0a 24 00 00 00 00 00 00 00 00 00 00
~~~
This translates to (accoring to [1], Section 8.3.3)
~~~
Mode Data Length 0
Medium Type 0
Device-specific 0
Block desc len 8
~~~
with the following offset / length computation _before_ the `while` loop
~~~
offset = 12
length = 13
~~~
The first payload section is
~~~
4 5 6 7 8 9 10 11
00 00 00 00 00 00 02 00
~~~
translating to
~~~
Density Code 0
Number of blks 0
Reserved 0
Block length 0x200 512
~~~
Then follows a pagecode 1 as
~~~
12 13 14 15 16 17 18 19 20 21 22 23 24
01 0a 24 00 00 00 00 00 00 00 00 00 00
~~~
translating to
~~~~
Page code 1
Page length -1 10
Mode parameters 24 00 00 00 00 00 00 00 00 00 00
~~~
computing (inside the `while` loop, as `// Advance to the next page`)
~~~
size = 10 + 2 = 12
~~~
followed by new `offset` and `length` values
~~~
offset = 25
length = 1
~~~
So it stays in the `while` loop (and has a larger-than-buffer `offset`
value)
Signed-off-by: Klaus Kämpf <kkaempf@gmail.com>
* Fix missing logging for a LUN when the LUN is explicitly specified
* Do not suppress controller messages when LUN is specified
* Fix misleading logging for DaynaPort
* Rebrand project to PiSCSI
- rascsi ->piscsi
- rasctl -> scsictl
- rasdump -> scsidump
- ras* -> piscsi* (rasutil -> piscsi_util, etc.)
* Refined the formatting and wording of the app startup banner
* Kept some references to rascsi and rasctl where backwards compatibility is concerned
* Point to the new github repo URL
Co-authored-by: nucleogenic <nr@nucleogenic.com>
Co-authored-by: Uwe Seimet <Uwe.Seimet@seimet.de>
* Support for ID and LUN parameter for the -L option in rascsi
* Added DeviceLogger class
* Removed dupiicate code
* Fixed SonarQube issues
* Added unit tests, improved code sharing
* Fixed regression (#979)
* Moved rasdump and monitor to sub-folders, cleaned up code
* Fixes rasdump issues and added additional features like device type check, LUN support and configurable buffer size
* Cleaned up dependencies on controller manager
* Removed global fields
* Simplified setting up RascsiResponse and RascsiExecutor
* Got rid of remaining raw pointers
* Use references instead of pointers
* Improved encapsulation
* Moved rascsi/rasctl/scsimon/rasdump.cpp to classes (for better testability)
* Moved bus.* to hal folder
* Removed some global variables
* Fixed code redundancies
* Using lambdas instead of member function pointers simplifies the command dispatching and reduces the code volume
* Removed duplicate error handling
* Fix for issue #956
* Unit test updates
* Resolved SonarQube issues
* Unit test updates
* Lambda syntax cleanup
* Use new-style casts
* Use std::none_of when saving the cache
* Use to_integer instead of casts
* Use accessors for getting CDB data
* Made ctrl_t private
* Improved encapsulation
* Replaced pointers by references
* Removed all remaining occurrences of DWORD and BYTE, making os.h obsolete
- Moved C++ code to cpp/ from src/raspberrypi
- Related updates to Makefile, easyinstall.sh, and the github build rules
- Removed the native X68k C code in src/x68k from the repo
