[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-Id: <58DA08B4.9000503@linux.vnet.ibm.com>
Date: Tue, 28 Mar 2017 12:24:44 +0530
From: Ravi Bangoria <ravi.bangoria@...ux.vnet.ibm.com>
To: Masami Hiramatsu <mhiramat@...nel.org>
Cc: acme@...hat.com, alexis.berlemont@...il.com,
linux-kernel@...r.kernel.org, peterz@...radead.org,
mingo@...hat.com, alexander.shishkin@...ux.intel.com,
naveen.n.rao@...ux.vnet.ibm.com, mpe@...erman.id.au,
hemant@...ux.vnet.ibm.com,
Ravi Bangoria <ravi.bangoria@...ux.vnet.ibm.com>
Subject: Re: [PATCH v2 1/3] perf/sdt/x86: Add renaming logic for (missing) 8
bit registers
Hi Masami, Thanks for the review.
On Monday 27 March 2017 07:47 PM, Masami Hiramatsu wrote:
> On Mon, 27 Mar 2017 13:28:27 +0530
> Ravi Bangoria <ravi.bangoria@...ux.vnet.ibm.com> wrote:
>
>> I found couple of events using al, bl, cl and dl registers for
>> argument. These are not directly accepted by uprobe_events and
>> thus needs to be mapped to ax, bx, cx and dx respectively.
>>
>> Few ex,
>>
>> /usr/bin/qemu-system-s390x
>> css_adapter_interrupt: 1@%bl
>> css_chpid_add: 1@%cl 1@...l 1@%dl
>> dma_bdrv_io: 8@...x 8@...p -8@...4 1@%al
>>
>> /usr/bin/postgres
>> buffer__read__done: ... -1@...sh -1@%al
>> buffer__read__start: ... -1@%al
> Of course, it should be suppoted. BTW, wouldn't we take care about ah, bh ... too?
I thought about them while preparing this patch, but I couldn't find any
events using those registers. So I ignored them.
But I think no one stops compiler to use those registers for sdt argument.
And if so, they should get included.
Will send v3 for this.
Thanks,
Ravi
Powered by blists - more mailing lists