lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAKv+Gu8BxDdF1V9g_c44p8Dj=WH5VVpyD5GyHs=N_jeP4Epc+w@mail.gmail.com>
Date:   Fri, 9 Mar 2018 09:41:52 +0000
From:   Ard Biesheuvel <ard.biesheuvel@...aro.org>
To:     Alexander Sverdlin <alexander.sverdlin@...ia.com>
Cc:     linux-arm-kernel <linux-arm-kernel@...ts.infradead.org>,
        Ingo Molnar <mingo@...hat.com>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Steven Rostedt <rostedt@...dmis.org>,
        Russell King <linux@...linux.org.uk>
Subject: Re: [PATCH v3 2/2] ARM: ftrace: Add MODULE_PLTS support

On 9 March 2018 at 08:54, Alexander Sverdlin
<alexander.sverdlin@...ia.com> wrote:
> Hello Ard!
>
> On 09/03/18 09:50, Ard Biesheuvel wrote:
>>> +static u32 fixed_plts[] = {
>>> +       FTRACE_ADDR,
>>> +       MCOUNT_ADDR,
>>> +#ifdef CONFIG_OLD_MCOUNT
>>> +       (unsigned long)ftrace_caller_old,
>>> +       (unsigned long)mcount,
>>> +#endif
>>> +};
>>> +
>> Do we need this stuff in the PLT loader if ftrace is disabled?
>
> You are right, it probably would not even build w/o FTRACE.
> I'll fix it, thanks for looking into this!
>

Also, perhaps you can make the array const (after you put #ifdef's
around all entries)

>>>  static bool in_init(const struct module *mod, unsigned long loc)
>>>  {
>>>         return loc - (u32)mod->init_layout.base < mod->init_layout.size;
>>> @@ -31,26 +41,43 @@ u32 get_module_plt(struct module *mod, unsigned long loc, Elf32_Addr val)
>>>  {
>>>         struct mod_plt_sec *pltsec = !in_init(mod, loc) ? &mod->arch.core :
>>>                                                           &mod->arch.init;
>>> +       int idx;
>>> +       struct plt_entries *plt;
>>> +
>>> +       /* Pre-allocate entries in the first plt */
>>> +       if (!pltsec->plt_count) {
>>> +               plt = (struct plt_entries *)pltsec->plt->sh_addr;
>>> +               for (idx = 0; idx < ARRAY_SIZE(plt->ldr); ++idx)
>>> +                       plt->ldr[idx] = PLT_ENT_LDR;
>>> +               memcpy(plt->lit, fixed_plts, sizeof(fixed_plts));
>>> +               pltsec->plt_count = ARRAY_SIZE(fixed_plts);
>>> +               /*
>>> +                * cache the address,
>>> +                * ELF header is available only during module load
>>> +                */
>>> +               pltsec->plt_ent = plt;
>>> +       }
>>> +       plt = pltsec->plt_ent;
>>>
>>> -       struct plt_entries *plt = (struct plt_entries *)pltsec->plt->sh_addr;
>>> -       int idx = 0;
>>> +       idx = ARRAY_SIZE(fixed_plts);
>>> +       while (idx)
>>> +               if (plt->lit[--idx] == val)
>>> +                       return (u32)&plt->ldr[idx];
>>>

... and put all this handling in a helper function that only gets
called if ARRAY_SIZE() > 0

AFAICT, that should allow you to leave the rest unchanged.

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ