Xen project Mailing List

Hi,

I just need to make sure the delay is coming from xen or linux kernel.The below print is coming from decompress_kernel function in arch/arm/boot/compressed/misc.c.

DOM0: Uncompressing Linux... done, booting the kernel.

On Fri, Mar 6, 2020 at 5:10 PM Julien Grall <julien@xxxxxxx> wrote:

Hello,

You seem to have dropped xen-users@ from the CC. Is there any reason for
that?

On 06/03/2020 11:25, Subhasree P V wrote:
> Hi Julien,
> Thanks for your reply.
> Below is my current log after disabling the prints and enabling
> CONFIG_DEBUG_LL in arch/arm/configs/odroidxu3_xen_defconfig and
> arch/arm/configs/odroidxu3_defconfig.
> [2020-03-06 16:05:51.039]
> [2020-03-06 16:05:51.039] CPU: Exynos5422 @ 800 MHz
> [2020-03-06 16:05:51.039] Model: Odroid XU4 based on EXYNOS5422
> [2020-03-06 16:05:51.044] Board: Odroid XU4 based on EXYNOS5422
> [2020-03-06 16:05:51.049] Type: xu4
> [2020-03-06 16:05:51.049] DRAM: 2 GiB
> [2020-03-06 16:05:51.183] MMC: EXYNOS DWMMC: 0, EXYNOS DWMMC: 1
> [2020-03-06 16:05:51.275] MMC Device 0 ( SD ): 7.4 GiB
> [2020-03-06 16:05:51.275] mmc_init: -5, time 3
> [2020-03-06 16:05:51.288] *** Warning - bad CRC, using default environment
> [2020-03-06 16:05:51.288]
> [2020-03-06 16:05:51.447] In: serial
> [2020-03-06 16:05:51.447] Out: serial
> [2020-03-06 16:05:51.447] Err: serial
> [2020-03-06 16:05:51.916] Kernel image @ 0x50000000 [ 0x000000 - 0x101700 ]
> [2020-03-06 16:05:51.916] ## Flattened Device Tree blob at 42000000
> [2020-03-06 16:05:51.922] Booting using the fdt blob at 0x42000000
> [2020-03-06 16:05:51.932] reserving fdt memory region: addr=42000000
> size=d000
> [2020-03-06 16:05:51.932] Loading Device Tree to 4fff0000, end
> 4fffffff ... OK
> [2020-03-06 16:05:52.024]
> [2020-03-06 16:05:52.025] Starting kernel ...
> [2020-03-06 16:05:52.025]
> [2020-03-06 16:05:52.050] (XEN) Checking for initrd in /chosen
> [2020-03-06 16:05:52.573] Xen 4.6.0
> [2020-03-06 16:05:52.573] (XEN) Generic Timer IRQ: phys=30 hyp=26
> virt=27 Freq: 24000 KHz
> [2020-03-06 16:05:52.579] (XEN) GICv2 initialization:
> [2020-03-06 16:05:52.579] (XEN) gic_dist_addr=0000000010481000
> [2020-03-06 16:05:52.584] (XEN) gic_cpu_addr=0000000010482000
> [2020-03-06 16:05:52.590] (XEN) gic_hyp_addr=0000000010484000
> [2020-03-06 16:05:52.590] (XEN) gic_vcpu_addr=0000000010486000
> [2020-03-06 16:05:52.595] (XEN) gic_maintenance_irq=25
> [2020-03-06 16:05:52.603] (XEN) GICv2: 256 lines, 8 cpus, secure (IID
> 0200043b).
> [2020-03-06 16:05:52.614] (XEN) Brought up 8 CPUs
> [2020-03-06 16:05:52.614] (XEN) I/O virtualisation disabled
> [2020-03-06 16:05:52.614] (XEN) *** LOADING DOMAIN 0 ***
> [2020-03-06 16:05:52.619] (XEN) Loading kernel from boot module @
> 0000000060000000
> [2020-03-06 16:05:53.443] (XEN) Std. Loglevel: All
> [2020-03-06 16:05:53.443] (XEN) Guest Loglevel: All
> [2020-03-06 16:05:53.443] (XEN) Freed 264kB init memory.
> [2020-03-06 16:05:55.111] (XEN) DOM0: Uncompressing Linux... done,
> booting the kernel.
>
> I'm getting an additional print here now like
> *(XEN) DOM0: Uncompressing Linux... done, booting the kernel.*
> But it is coming from arch/arm/boot/compressed/misc.c of dom0 Linux and
> hangs after that print.

Don't you see any progress at all afterwards?

> Upto "Freed 264kB init memory" my boot time is*1.418*s.But delay is
> coming after that.
>
> In my previous logs without enabling CONFIG_DEBUG_LL i got prints like,
>
> [2020-03-06 10:23:31.406] Starting kernel ...
> [2020-03-06 10:23:31.406]
> [2020-03-06 10:23:31.432] (XEN) Checking for initrd in /chosen
> [2020-03-06 10:23:31.954] Xen 4.6.0
> [2020-03-06 10:23:31.954] (XEN) Generic Timer IRQ: phys=30 hyp=26
> virt=27 Freq: 24000 KHz
> [2020-03-06 10:23:31.960] (XEN) GICv2 initialization:
> [2020-03-06 10:23:31.960] (XEN) gic_dist_addr=0000000010481000
> [2020-03-06 10:23:31.965] (XEN) gic_cpu_addr=0000000010482000
> [2020-03-06 10:23:31.971] (XEN) gic_hyp_addr=0000000010484000
> [2020-03-06 10:23:31.971] (XEN) gic_vcpu_addr=0000000010486000
> [2020-03-06 10:23:31.977] (XEN) gic_maintenance_irq=25
> [2020-03-06 10:23:31.984] (XEN) GICv2: 256 lines, 8 cpus, secure (IID
> 0200043b).
> [2020-03-06 10:23:31.995] (XEN) Brought up 8 CPUs
> [2020-03-06 10:23:31.995] (XEN) I/O virtualisation disabled
> [2020-03-06 10:23:31.995] (XEN) *** LOADING DOMAIN 0 ***
> [2020-03-06 10:23:32.001] (XEN) Loading kernel from boot module @
> 0000000060000000
> [2020-03-06 10:23:32.817] (XEN) Std. Loglevel: All
> [2020-03-06 10:23:32.817] (XEN) Guest Loglevel: All
> [2020-03-06 10:23:32.817] (XEN) Freed 264kB init memory.
> [2020-03-06 10:23:36.719] [ 0.000000] [c0] Booting Linux on physical
> CPU 0x0
> [2020-03-06 10:23:36.719] [ 0.000000] [c0] Initializing cgroup subsys cpu
> [2020-03-06 10:23:36.724] [ 0.000000] [c0] Initializing cgroup subsys
> cpuacct
>
> Is that delay coming from xen or dom0 kernel.If it's from xen how could
> i reduce it..

I don't have enough data to now exactly where the bottlneck is. I would
recommend to get early prink in Linux fully working so you have more log
and timestamp to narrow it down.

--
Julien Grall

This message contains confidential information and is intended only for the individual(s) named. If you are not the intended recipient, you are notified that disclosing, copying, distributing or taking any action in reliance on the contents of this mail and attached file/s is strictly prohibited. Please notify the sender immediately and delete this e-mail from your system. E-mail transmission cannot be guaranteed to be secured or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or contain viruses. The sender therefore does not accept liability for any errors or omissions in the contents of this message, which arise as a result of e-mail transmission.

Re: [Xen-users] Xen boot time optimization