Re: kernel 4.* slow IO due to -mlong-calls (hppa/c3600)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 2018-03-13 10:17 AM, Carlo Pisani wrote:
hi
I am experiencing a very annoying behavior with my HPPA C3600: if I
compile the (linux) kernel with  -mlong-calls then the IO (e.g. file
copy) becomes very slow, and the PCI becomes unstable (i.e. it crashes
the machine)

kernel  gcc     binutils    with mlong    without mlong
4.15.7  4.9.3   2.25.1     13.4 MB/s    27.0 MB/s
4.15.7  6.4.0   2.25.1     13.4 MB/s    27.0 MB/s
4.15.7  6.4.0   2.29.1     14.4 MB/s    25.0 MB/s

these tests were performed with

dd if=/dev/zero of=here bs=1k count=100000
I tried the above and see more or less the same effect on systems with and without long calls.

You didn't say whether your config was 32 or 64 bits.  I would have thought long calls might be more efficient on 64-bit kernels.  I'm surprised that they have this much effect on I/O performance.
Most of my testing is done on 64-bit kernels without long calls.

My c3750 seems stable with long calls:
dave@hiauly6:~$ uname -a
Linux hiauly6 4.15.0-1-parisc64-smp #1 SMP Debian 4.15.4-1 (2018-02-18) parisc64 GNU/Linux

I'm not sure your stability issue is related to long calls.  On c8000, I have the driver for the video card disabled because enabling DRM seems to cause stability issues. This seems to be some sort
of PCI issue.

Dave

--
John David Anglin  dave.anglin@xxxxxxxx

--
To unsubscribe from this list: send the line "unsubscribe linux-parisc" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html



[Index of Archives]     [Linux SoC]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux