Re: [Xen-devel] [PATCH] xen/arm: io: add function swap_mmio_handler()

2019-06-27 Thread chenbaodong
: On Mon, 24 Jun 2019, Stefano Stabellini wrote: On Thu, 13 Jun 2019, chenbaodong wrote: Let me add that if you prefer to document one of the other interfaces listed above in my email, you are welcome to pick another one. For example, we are also missing a doc about the DomU memory map, listing

Re: [Xen-devel] [PATCH] xen/arm: io: add function swap_mmio_handler()

2019-06-12 Thread chenbaodong
On 6/12/19 20:21, Julien Grall wrote: Hi, On 12/06/2019 11:08, chenbaodong wrote: On 6/12/19 17:08, Julien Grall wrote: Hi, On 6/12/19 6:42 AM, Baodong Chen wrote: Swap function can be used when calling sort(). or else, the default swap function generic_swap() is used, which is a little

Re: [Xen-devel] [PATCH] xen/arm: io: add function swap_mmio_handler()

2019-06-12 Thread chenbaodong
On 6/12/19 17:08, Julien Grall wrote: Hi, On 6/12/19 6:42 AM, Baodong Chen wrote: Swap function can be used when calling sort(). or else, the default swap function generic_swap() is used, which is a little inefficient. I am not entirely convince this will be more efficient. mmio_handler doe

Re: [Xen-devel] [PATCH] xen/coverage: wrap coverage related things under 'CONFIG_COVERAGE'

2019-06-12 Thread chenbaodong
On 6/12/19 15:58, Jan Beulich wrote: On 12.06.19 at 09:36, wrote: On 6/12/19 14:34, Jan Beulich wrote: On 12.06.19 at 02:23, wrote: On 6/11/19 22:03, Jan Beulich wrote: On 11.06.19 at 08:02, wrote: --- a/xen/arch/x86/xen.lds.S +++ b/xen/arch/x86/xen.lds.S @@ -240,12 +240,14 @@ SECTIONS

Re: [Xen-devel] [PATCH] xen/coverage: wrap coverage related things under 'CONFIG_COVERAGE'

2019-06-12 Thread chenbaodong
On 6/12/19 14:34, Jan Beulich wrote: On 12.06.19 at 02:23, wrote: On 6/11/19 22:03, Jan Beulich wrote: On 11.06.19 at 08:02, wrote: --- a/xen/arch/x86/xen.lds.S +++ b/xen/arch/x86/xen.lds.S @@ -240,12 +240,14 @@ SECTIONS *(.altinstructions) __alt_instructions_end = .;

Re: [Xen-devel] [PATCH] xen: superficial clean-ups

2019-06-11 Thread chenbaodong
On 6/11/19 18:29, Juergen Gross wrote: On 11.06.19 12:27, Andrew Cooper wrote: On 11/06/2019 11:25, Juergen Gross wrote: On 11.06.19 12:18, George Dunlap wrote: On 6/11/19 10:20 AM, Baodong Chen wrote: * Remove redundant set 'DOMDYING_dead' domain_create() will set this when fail, thus no ne

Re: [Xen-devel] [PATCH] xen: superficial clean-ups

2019-06-11 Thread chenbaodong
On 6/11/19 18:53, Andrew Cooper wrote: On 11/06/2019 11:33, chenbaodong wrote: On 6/11/19 17:45, Andrew Cooper wrote: On 11/06/2019 10:20, Baodong Chen wrote: * Remove redundant set 'DOMDYING_dead' domain_create() will set this when fail, thus no need set in arch_domain_create()

Re: [Xen-devel] [PATCH] xen/coverage: wrap coverage related things under 'CONFIG_COVERAGE'

2019-06-11 Thread chenbaodong
On 6/11/19 22:03, Jan Beulich wrote: On 11.06.19 at 08:02, wrote: --- a/xen/arch/x86/xen.lds.S +++ b/xen/arch/x86/xen.lds.S @@ -240,12 +240,14 @@ SECTIONS *(.altinstructions) __alt_instructions_end = .; +#if defined(CONFIG_COVERAGE) . = ALIGN(8); __cto

Re: [Xen-devel] [PATCH] xen/scheduler: remove 'name' from 'struct scheduler'

2019-06-11 Thread chenbaodong
On 6/11/19 18:04, George Dunlap wrote: On 6/11/19 2:35 AM, Baodong Chen wrote: 'struct scheduler' already has member 'opt_name' and 'sched_id', thus 'name' is a little redundant, so remove it. Signed-off-by: Baodong Chen It's not redundant; one is a longer-form human-readable description, an

Re: [Xen-devel] [PATCH] xen: superficial clean-ups

2019-06-11 Thread chenbaodong
On 6/11/19 17:45, Andrew Cooper wrote: On 11/06/2019 10:20, Baodong Chen wrote: * Remove redundant set 'DOMDYING_dead' domain_create() will set this when fail, thus no need set in arch_domain_create(). Its not redundant.  It is necessary for correct cleanup. Hello Andrew, Thanks for your co

Re: [Xen-devel] [PATCH] xen/arm: domain: remove redundant memset for arch's saved_context when creating vcpu

2019-06-10 Thread chenbaodong
On 6/11/19 04:11, Julien Grall wrote: Hi, Thank you for the patch. The title should be at max 80 characters. So how about the following title? "xen/arm: domain: Remove redundant memset for v->arch.saved_context" Max 80 characters, roger that. On 6/10/19 6:15 AM, Baodong Chen wrote: Alrea

Re: [Xen-devel] [PATCH] xen/arm: vtimer: fix return value to void for virt_timer_[save|restore]

2019-06-10 Thread chenbaodong
On 6/11/19 04:16, Julien Grall wrote: Hi, NIT: I would use "change" instead of "fix". I feel "fix" is more when there are an actual bug. Sound good to me. On 6/10/19 6:07 AM, Baodong Chen wrote: The original type is int and not used at all so fix to void. The commit message is a bit uncl

Re: [Xen-devel] [PATCH v3] xen: make tracebuffer configurable

2019-06-09 Thread chenbaodong
On 6/5/19 18:38, George Dunlap wrote: On Jun 5, 2019, at 2:27 AM, Baodong Chen wrote: Xen internal running status(trace event) will be saved to trace memory when enabled. trace event data and config params can be read/changed by system control hypercall at run time. Can be disabled for smal

Re: [Xen-devel] [PATCH v1] xen: make tracebuffer configurable

2019-06-04 Thread chenbaodong
On 6/4/19 19:43, Jan Beulich wrote: On 04.06.19 at 12:49, wrote: On Jun 4, 2019, at 1:44 AM, Baodong Chen wrote: --- a/xen/common/Kconfig +++ b/xen/common/Kconfig @@ -368,4 +368,16 @@ config DOM0_MEM Leave empty if you are not sure what to specify. +config TRACEBUFFER + bool

Re: [Xen-devel] [PATCH v1] xen: make tracebuffer configurable

2019-06-04 Thread chenbaodong
On 6/4/19 15:10, Jan Beulich wrote: On 04.06.19 at 02:44, wrote: --- a/xen/include/xen/trace.h +++ b/xen/include/xen/trace.h @@ -21,12 +21,15 @@ #ifndef __XEN_TRACE_H__ #define __XEN_TRACE_H__ +#ifdef CONFIG_TRACEBUFFER extern int tb_init_done; +#endif If this is to stay up here (wh

Re: [Xen-devel] [PATCH] xen: make tracebuffer configurable

2019-06-03 Thread chenbaodong
On 6/3/19 22:31, Jan Beulich wrote: On 03.06.19 at 16:08, wrote: On Jun 3, 2019, at 11:54 AM, Jan Beulich wrote: On 03.06.19 at 12:41, wrote: On 6/3/19 16:31, Jan Beulich wrote: On 03.06.19 at 05:07, wrote: On 5/31/19 19:10, Jan Beulich wrote: On 30.05.19 at 12:17, wrote: Default: e

Re: [Xen-devel] [PATCH] xen: cpu: change 'cpu_hotplug_[begin|done]' to inline function

2019-06-03 Thread chenbaodong
On 6/3/19 18:40, Julien Grall wrote: Hi, On 03/06/2019 11:22, chenbaodong wrote: On 6/3/19 17:37, Julien Grall wrote: Hi, On 03/06/2019 02:52, chenbaodong wrote: On 5/31/19 18:55, Julien Grall wrote: Hi, On 5/31/19 3:46 AM, Baodong Chen wrote: Signed-off-by: Baodong Chen ---   xen

Re: [Xen-devel] [PATCH] xen: make tracebuffer configurable

2019-06-03 Thread chenbaodong
On 6/3/19 16:31, Jan Beulich wrote: On 03.06.19 at 05:07, wrote: On 5/31/19 19:10, Jan Beulich wrote: On 30.05.19 at 12:17, wrote: Default: enabled. Can be disabled for smaller code footprint. But you're aware that we're, for now at least, trying to limit the number of independently select

Re: [Xen-devel] [PATCH] xen: cpu: change 'cpu_hotplug_[begin|done]' to inline function

2019-06-03 Thread chenbaodong
On 6/3/19 17:37, Julien Grall wrote: Hi, On 03/06/2019 02:52, chenbaodong wrote: On 5/31/19 18:55, Julien Grall wrote: Hi, On 5/31/19 3:46 AM, Baodong Chen wrote: Signed-off-by: Baodong Chen ---   xen/common/cpu.c  | 10 --   xen/include/xen/cpu.h |  4 ++--   2 files changed

Re: [Xen-devel] [PATCH] xen: put cpupool's member 'n_dom' after 'cpupool_id'

2019-06-03 Thread chenbaodong
On 6/3/19 17:30, Julien Grall wrote: Hi, On 03/06/2019 02:48, chenbaodong wrote: On 5/31/19 18:52, Julien Grall wrote: Hi, On 5/31/19 4:18 AM, Baodong Chen wrote: Thus, sizeof(struct cpupool) will save 8 bytes for 64-bit system. I am happy with the change, although AFAIK cpupool is not

Re: [Xen-devel] [PATCH RESEND] xen: notifier: refine 'notifier_head', use 'list_head' directly

2019-06-03 Thread chenbaodong
On 6/3/19 17:27, Jan Beulich wrote: On 03.06.19 at 03:33, wrote: 'notifier_block' can be replaced with 'list_head' when used for 'notifier_head', this make the a little more clear. I guess you mean "... makes the code a little ..."? Yes, fixed, see v1. @@ -71,16 +71,16 @@ int notifier_call_

Re: [Xen-devel] [PATCH RESEND] xen: notifier: refine 'notifier_head', use 'list_head' directly

2019-06-03 Thread chenbaodong
On 6/3/19 17:28, Jan Beulich wrote: On 03.06.19 at 03:33, wrote: 'notifier_block' can be replaced with 'list_head' when used for 'notifier_head', this make the a little more clear. Signed-off-by: Baodong Chen Oh, and also a remark regarding the title: Why "RESEND"? This should be used only

Re: [Xen-devel] [PATCH] xen: make keyhanler configurable

2019-06-02 Thread chenbaodong
On 6/1/19 06:30, Andrew Cooper wrote: On 30/05/2019 18:58, Baodong Chen wrote: keyhandler mainly used for debug usage by developers which can dump xen module(eg. timer, scheduler) status at runtime by input character from console. Signed-off-by: Baodong Chen What is the motivation here?  I d

Re: [Xen-devel] [PATCH] xen: make keyhanler configurable

2019-06-02 Thread chenbaodong
On 5/31/19 18:53, Juergen Gross wrote: On 31/05/2019 03:58, Baodong Chen wrote: keyhandler mainly used for debug usage by developers which can dump xen module(eg. timer, scheduler) status at runtime by input character from console. Signed-off-by: Baodong Chen --- xen/arch/arm/gic.c

Re: [Xen-devel] [PATCH] xen: make keyhanler configurable

2019-06-02 Thread chenbaodong
On 5/31/19 18:39, Dario Faggioli wrote: On Fri, 2019-05-31 at 09:58 +0800, Baodong Chen wrote: keyhandler mainly used for debug usage by developers which can dump xen module(eg. timer, scheduler) status at runtime by input character from console. Signed-off-by: Baodong Chen --- --- a/xen/comm

Re: [Xen-devel] [PATCH] xen: make tracebuffer configurable

2019-06-02 Thread chenbaodong
On 5/31/19 20:58, George Dunlap wrote: On May 31, 2019, at 12:10 PM, Jan Beulich wrote: On 30.05.19 at 12:17, wrote: Default: enabled. Can be disabled for smaller code footprint. But you're aware that we're, for now at least, trying to limit the number of independently selectable config

Re: [Xen-devel] [PATCH] xen: make tracebuffer configurable

2019-06-02 Thread chenbaodong
On 5/31/19 19:10, Jan Beulich wrote: On 30.05.19 at 12:17, wrote: Default: enabled. Can be disabled for smaller code footprint. But you're aware that we're, for now at least, trying to limit the number of independently selectable config options? Ones depending on EXPERT are sort of an excepti

Re: [Xen-devel] [PATCH] xen: cpu: change 'cpu_hotplug_[begin|done]' to inline function

2019-06-02 Thread chenbaodong
On 5/31/19 19:30, Jan Beulich wrote: On 31.05.19 at 12:55, wrote: On 5/31/19 3:46 AM, Baodong Chen wrote: --- a/xen/include/xen/cpu.h +++ b/xen/include/xen/cpu.h @@ -10,8 +10,8 @@ bool_t get_cpu_maps(void); void put_cpu_maps(void); /* Safely perform CPU hotplug and update cpu_online

Re: [Xen-devel] [PATCH] xen: cpu: change 'cpu_hotplug_[begin|done]' to inline function

2019-06-02 Thread chenbaodong
On 5/31/19 18:55, Julien Grall wrote: Hi, On 5/31/19 3:46 AM, Baodong Chen wrote: Signed-off-by: Baodong Chen ---   xen/common/cpu.c  | 10 --   xen/include/xen/cpu.h |  4 ++--   2 files changed, 2 insertions(+), 12 deletions(-) diff --git a/xen/common/cpu.c b/xen/common/cpu.c ind

Re: [Xen-devel] [PATCH] xen: put cpupool's member 'n_dom' after 'cpupool_id'

2019-06-02 Thread chenbaodong
On 5/31/19 18:52, Julien Grall wrote: Hi, On 5/31/19 4:18 AM, Baodong Chen wrote: Thus, sizeof(struct cpupool) will save 8 bytes for 64-bit system. I am happy with the change, although AFAIK cpupool is not instantiated that often. Are you planning to have more instantiation of it? Cheers,

Re: [Xen-devel] [PATCH] xen: notifier: refine 'notifier_head', use 'list_head' directly

2019-06-02 Thread chenbaodong
On 5/31/19 18:22, Julien Grall wrote: Hi, Missing commit message here. Yes, will be added. On 5/31/19 3:35 AM, Baodong Chen wrote: Signed-off-by: Baodong Chen ---   xen/common/notifier.c  | 25 ++---   xen/include/xen/notifier.h | 21 +++--   2 files c

Re: [Xen-devel] [PATCH] xen: 'keyhandler' is not used in null scheduler

2019-05-30 Thread chenbaodong
Hello George, On 5/30/19 17:05, George Dunlap wrote: On May 30, 2019, at 6:47 AM, Baodong Chen wrote: So remove 'keyhandler.h' include. Also add 'static' prefix for 'schud_bull_def' Signed-off-by: Baodong Chen Thanks for the patch — these changes look good. I think the title would be bett