[so2] [Tema1] Problema spinlock
Marius-Cristian EŞEANU (6423)
marius.eseanu at stud.acs.upb.ro
Mon Mar 28 00:03:52 EEST 2016
Da, protejez cu acelasi spinlock.
Multumesc,
Cristi
________________________________________
From: Constantin Neagu <constantin.neagu.office at gmail.com>
Sent: Sunday, March 27, 2016 10:56 PM
To: Marius-Cristian EŞEANU (6423)
Cc: Sisteme de Operare 2
Subject: Re: [so2] [Tema1] Problema spinlock
Cand faci add/remove, protejezi lista cu acelasi spinlock ca in kmalloc si kfree ?
Daca faci acest lucru, kmalloc/kfree o sa declanseze proba si in interiorul comenzi tale de add/remove, si o sa incerce sa ia acelasi spin_lock.
> On 27 Mar 2016, at 23:52, Marius-Cristian EŞEANU (6423) <marius.eseanu at stud.acs.upb.ro> wrote:
>
> Buna seara,
>
> Nu folosesc kmalloc sau kfree in probe. Folosesc kfree/kmalloc doar pentru comenziile primite prin interfata ioctl.(add/remove process).
>
> Cristi
>
> ________________________________________
> From: Constantin Neagu <constantin.neagu.office at gmail.com>
> Sent: Sunday, March 27, 2016 10:45 PM
> To: Marius-Cristian EŞEANU (6423); Sisteme de Operare 2
> Subject: Re: [so2] [Tema1] Problema spinlock
>
> Salut,
>
> Structurile din spatele kretprobe folosesc spin_lock-uri inainte si in timpul apelari handle-urilor tale.
> Drept urmare, daca tu incerci sa faci kmalloc sau kfree in interiorul unui kretprobe o sa intri intro eroare de tipul celei descrise de tine.
>
> Solutia: Don’t use kmalloc and kfree in the probes ;)
>
> O seara buna,
>
> Neagu Constantin
>
>> On 27 Mar 2016, at 23:08, Marius-Cristian EŞEANU (6423) via so2 <so2 at cursuri.cs.pub.ro> wrote:
>>
>> Buna seara,
>>
>> Primesc urmatoarea eroare BUG: spinlock recursion on CPU#0, tracer_test/963 .
>> Am verificat de nenumarate ori codul si nu am observat niciun spinlock pus gresit .
>> Puteti sa imi dati un hint cum sa rezolv aceasta problema ?
>>
>> Multumesc,
>> Cristi
>> _______________________________________________
>> http://ocw.cs.pub.ro/courses/so2/resurse/lista-discutii
>
More information about the so2
mailing list