Re: [PATCH v3] arm64: Introduce IRQ stack

2015-10-07 Thread Jungseok Lee
On Sep 28, 2015, at 11:24 PM, Sungjinn Chung wrote: > Hi Jeongseok, Hi Sungjinn, > I expect to save memory and easily fork processes with your patchset. > > Acked-by: Sungjinn Chung > I've done re-spin the series which is version four. Could you review it since it includes stack trace suppor

Re: [PATCH v3] arm64: Introduce IRQ stack

2015-10-05 Thread Jungseok Lee
On Oct 6, 2015, at 2:24 AM, James Morse wrote: Hi James, > On 05/10/15 07:37, AKASHI Takahiro wrote: >> On 10/04/2015 11:32 PM, Jungseok Lee wrote: >>> On Oct 3, 2015, at 1:23 AM, James Morse wrote: One observed change in behaviour: Any stack-unwinding now stops at el1_irq(), which is t

Re: [PATCH v3] arm64: Introduce IRQ stack

2015-10-05 Thread James Morse
On 05/10/15 07:37, AKASHI Takahiro wrote: > On 10/04/2015 11:32 PM, Jungseok Lee wrote: >> On Oct 3, 2015, at 1:23 AM, James Morse wrote: >>> One observed change in behaviour: >>> Any stack-unwinding now stops at el1_irq(), which is the bottom of the irq >>> stack. This shows up with perf (using in

Re: [PATCH v3] arm64: Introduce IRQ stack

2015-10-04 Thread AKASHI Takahiro
On 10/04/2015 11:32 PM, Jungseok Lee wrote: On Oct 3, 2015, at 1:23 AM, James Morse wrote: Hi, Hi James, On 22/09/15 13:11, Jungseok Lee wrote: Currently, kernel context and interrupts are handled using a single kernel stack navigated by sp_el1. This forces a system to use 16KB stack, not

Re: [PATCH v3] arm64: Introduce IRQ stack

2015-10-04 Thread Jungseok Lee
On Oct 3, 2015, at 1:23 AM, James Morse wrote: > Hi, Hi James, > > On 22/09/15 13:11, Jungseok Lee wrote: >> Currently, kernel context and interrupts are handled using a single >> kernel stack navigated by sp_el1. This forces a system to use 16KB >> stack, not 8KB one. This restriction makes lo

Re: [PATCH v3] arm64: Introduce IRQ stack

2015-10-02 Thread James Morse
Hi, On 22/09/15 13:11, Jungseok Lee wrote: > Currently, kernel context and interrupts are handled using a single > kernel stack navigated by sp_el1. This forces a system to use 16KB > stack, not 8KB one. This restriction makes low memory platforms suffer > from memory pressure accompanied by perfo

Re: [PATCH v3] arm64: Introduce IRQ stack

2015-09-23 Thread Jungseok Lee
On Sep 22, 2015, at 9:11 PM, Jungseok Lee wrote: Dear all, I cannot reply to your feedbacks for about a week. I will answer your emails as soon as possible after that time. Best Regards Jungseok Lee > Currently, kernel context and interrupts are handled using a single > kernel stack navigated b

[PATCH v3] arm64: Introduce IRQ stack

2015-09-22 Thread Jungseok Lee
Currently, kernel context and interrupts are handled using a single kernel stack navigated by sp_el1. This forces a system to use 16KB stack, not 8KB one. This restriction makes low memory platforms suffer from memory pressure accompanied by performance degradation. This patch addresses the issue

Re: [PATCH v3] arm64: Introduce IRQ stack

2015-09-21 Thread Jungseok Lee
On Sep 21, 2015, at 9:19 PM, Jungseok Lee wrote: Dear all, Please ignore this. It won't work. I've typed a wrong command line in a terminal. Sorry for confusion. Best Regards Jungseok Lee > Currently, kernel context and interrupts are handled using a single > kernel stack navigated by sp_el1.

[PATCH v3] arm64: Introduce IRQ stack

2015-09-21 Thread Jungseok Lee
Currently, kernel context and interrupts are handled using a single kernel stack navigated by sp_el1. This forces a system to use 16KB stack, not 8KB one. This restriction makes low memory platforms suffer from memory pressure accompanied by performance degradation. This patch addresses the issue