New kernels

[xeno74]

Machine with a PowerPC build environment for crossbuilding:



GCC: powerpc-linux-gnu-gcc (Ubuntu 9.5.0-1ubuntu1) 9.5.0

X1000: 6.4-rc7 compile time:

time make -j16 CROSS_COMPILE=powerpc-linux-gnu- ARCH=powerpc vmlinux

Code: Select all

real	2m23,610s
user	32m57,752s
sys	2m22,167s

time make -j16 CROSS_COMPILE=powerpc-linux-gnu- ARCH=powerpc modules

Code: Select all

real	0m55,701s
user	13m11,215s
sys	1m5,703s

X5000: 6.4-rc7 compile time:

time make -j16 CROSS_COMPILE=powerpc-linux-gnu- ARCH=powerpc uImage

Code: Select all

real	2m24,286s
user	31m18,960s
sys	2m27,824s

time make -j16 CROSS_COMPILE=powerpc-linux-gnu- ARCH=powerpc modules

Code: Select all

real	1m14,648s
user	17m52,960s
sys	1m31,409s

[sailorMH]

In past I had problems with cfcard, both with fat and ext2. It is very often unreadable from CFE. Maybe I have to exchange my old cfcard.

Is CF more reliable than USB? I cannot get USB working from my case. It only works from my Apple keyboard if it wants to.

For me was most reliable USB stick with ext2 partition. I have in first USB port hub, where is connected keyboard and this stick.
If it not works for you, maybe you can try different hub.

Now I tested CF card again, and have problems - I formatted it under linux both fat or ext2 and under CFE it is very often unreadable. After boot to linux it works normally. So I will try to swap CF card for new one and test again.

[xeno74]

Now I tested CF card again, and have problems - I formatted it under linux both fat or ext2 and under CFE it is very often unreadable. After boot to linux it works normally. So I will try to swap CF card for new one and test again.

That's a good idea to test a new CF card. My CF card works with FAT16 without any problems.

[musa]

Hi
Re-installed Debian ppc64 and upgraded to trixie/sid.
Vmlinux-6.4_rc7 boots debian ppc64 without error.

Have a nice day

[Hypex]

For me was most reliable USB stick with ext2 partition. I have in first USB port hub, where is connected keyboard and this stick.
If it not works for you, maybe you can try different hub.

That makes sense. Ext2 avoids the problems of FAT in CFE that not even FAT16 can solve at times. I think this is mostly because AmigaOS4 lacks a way to unmount or cleanly update after a write. So the volume is in effect "invalidated" in a FAT sense. However, if writing from a "normal" OS still breaks, then that isn't the problem.

So I recalled, my keyboard is plugged into the rear, into the keyboard port and this tends to work well. On the front case I cannot get USB working fully. Detects a stick but it hangs there and never mounts.

Now I tested CF card again, and have problems - I formatted it under linux both fat or ext2 and under CFE it is very often unreadable. After boot to linux it works normally. So I will try to swap CF card for new one and test again.

The CF is permanent in the slot. Well unless you have an extension externally. So would expect it to be reliable but that seems touchy as well.

In any case, I only need USB in CFE for testing. Such as seeing if OS4 can boot off USB but I don't think it can at all despite reading it was done over ten years ago. For Linux I don't need CF or USB since I can boot Linux totally off HDD from a normal boot menu with my X1 manager.

[xeno74]

Hi
Re-installed Debian ppc64 and upgraded to trixie/sid.
Vmlinux-6.4_rc7 boots debian ppc64 without error.

Have a nice day

Thank you for testing!

[xeno74]

Hi All,

I released the RC7 of kernel 6.4 for testing today.

New:

• Linux 6.4-rc7 Released With The Quiet Cycle Continuing -- phoronix.com
• 6.4-rc7 announcement
• USB / Thunderbolt fixes
• DRM fixes (including Radeon fix)
• Sound fixes (including USB audio fixes)
• Ext4 fixes
• Linux Git log
• Phoronix articles, reviews and news stories covering Linux 6.4

Download: linux-image-6.4-rc7-X1000_X5000.tar.gz



Please test the kernels.

Thanks,
Christian

[xeno74]

Hi All,

Here is the final kernel 6.4. Thanks a lot to Skateman, Musa, sailorMH, Hypex and Trevor for testing the alphas and release candidates.

New and important for our AmigaOnes:

• X1000, X5000, and QEMU e5500: Multi-Gen LRU support (CONFIG_LRU_GEN=y and CONFIG_LRU_GEN_ENABLED=y)
• PASEMI updates in the PowerPC updates
• PASEMI ethernet and DPAA updates in the networking updates
• There are some issues with the BTTV driver for TV cards so we added the new BTTV driver originally planned for the kernel 6.5 to our kernel 6.4.
• X1000: Fix ISA mapping when "ranges" property is not present, for PASemi Nemo boards

More news:

• Linux 6.4 Released With Early Apple M2 Code, More WiFi 7, AMD Guided Autonomous Mode -- phoronix.com
• Linux Kernel 6.4 Released: Embracing Apple M2, New Hardware, and More Rust Code -- itsfoss.com
• 6.4.0 final announcement
• PowerPC fix 6.4-5
• Linux Git log
• Phoronix articles, reviews and news stories covering Linux 6.4

Download: linux-image-6.4-X1000_X5000.tar.gz



Cheers,
Christian

[xeno74]

The alpha1 of kernel 6.5 can't read the Amiga partition table via the rigid disk block (RDB) correctly anymore so it isn't possible to mount a Linux partition from a disk with an Amiga partition table. The kernel 6.4 doesn't have this issue.

Kernel 6.4.0 final:

parted -l

Code: Select all

Model: ATA ST2000DM001-9YN1 (scsi)
Disk /dev/sda: 2000GB
Sector size (logical/physical): 512B/4096B
Partition Table: amiga
Disk Flags: 

Number  Start   End     Size    File system  Name  Flags
 1      1057kB  123MB   122MB   affs7        BDH0  hidden
 2      123MB   2274MB  2150MB               DH0   boot
 3      2274MB  691GB   689GB                DH2
 4      691GB   1992GB  1301GB  ext4         dhx   boot

dmesg | grep -i sda

Code: Select all

[    0.000000] Kernel command line: root=/dev/sda4
[    3.987717] sd 0:0:0:0: [sda] 3907029168 512-byte logical blocks: (2.00 TB/1.82 TiB)
[    4.031349] sd 0:0:0:0: [sda] 4096-byte physical blocks
[    4.123773] sd 0:0:0:0: [sda] Write Protect is off
[    4.168682] sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00
[    4.279304] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
[    4.463508] sd 0:0:0:0: [sda] Preferred minimum I/O size 4096 bytes
[    4.519477]  sda: RDSK (512) sda1 (DOS^G)(res 2 spb 2) sda2 (SFS^B)(res 2 spb 1) sda3 (SFS^B)(res 2 spb 2) sda4 ((res 2 spb 1)
[    4.720896] sda: p4 size 18446744071956107760 extends beyond EOD, 
[    4.922550]  sda: RDSK (512) sda1 (DOS^G)(res 2 spb 2) sda2 (SFS^B)(res 2 spb 1) sda3 (SFS^B)(res 2 spb 2) sda4 ((res 2 spb 1)
[    4.948655] sda: p4 size 18446744071956107760 extends beyond EOD, truncated
[    4.998956] sd 0:0:0:0: [sda] Attached SCSI disk
[    8.394695] EXT4-fs (sda4): mounted filesystem 93cb7dd2-ce1b-4bf5-ba47-818cf8e8c9f4 ro with ordered data mode. Quota mode: none.
[   18.578020] EXT4-fs (sda4): re-mounted 93cb7dd2-ce1b-4bf5-ba47-818cf8e8c9f4 ro. Quota mode: none.
[   23.159524] EXT4-fs (sda4): re-mounted 93cb7dd2-ce1b-4bf5-ba47-818cf8e8c9f4 r/w. Quota mode: none.

Kernel 6.5-alpha1:

dmesg | grep -i sda

Code: Select all

[    4.208905] sd 0:0:0:0: [sda] 3907029168 512-byte logical blocks: (2.00 TB/1.82 TiB)
[    4.253995] sd 0:0:0:0: [sda] 4096-byte physical blocks
[    4.254826] sd 0:0:0:0: [sda] Write Protect is off
[    4.300069] sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00
[    4.486476] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
[    4.580507] sd 0:0:0:0: [sda] Preferred minimum I/O size 4096 bytes
[    4.712624] Dev sda: unable to read partition block 4294967295
[    4.761532]  sda: RDSK (512) sda1 (DOS^G)(res 2 spb 2) sda2 (SFS^B)(res 2 spb 1) sda3 (SFS^B)(res 2 spb 2) sda4 ((res 2 spb 1) unable to read partition table
[    4.761892] sda: partition table beyond EOD, 
[    4.861681] Dev sda: unable to read partition block 4294967295
[    4.912094]  sda: RDSK (512) sda1 (DOS^G)(res 2 spb 2) sda2 (SFS^B)(res 2 spb 1) sda3 (SFS^B)(res 2 spb 2) sda4 ((res 2 spb 1) unable to read partition table
[    4.963387] sda: partition table beyond EOD, 
[    5.014769] sd 0:0:0:0: [sda] Attached SCSI disk

[xeno74]

I think, I have the problematic commit.

- fix overflows in the Amiga partition handling code (Michael Schmitz <[email protected]>)

...

block: add overflow checks for Amiga partition support
block: fix signed int overflow in Amiga partition support

Commit: Block updates

Code: a/block/partitions/amiga.c