linux-kernel - Re: [PATCH v5 10/15] coresight: trbe: Workaround TRBE errata overwrite in FILL mode

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] [thread-next>] [day] [month] [year] [list]

Message-ID: <1dd61f9a-cd55-c9a5-8573-1b6a0327247b@arm.com>
Date:   Tue, 19 Oct 2021 10:06:28 +0530
From:   Anshuman Khandual <anshuman.khandual@....com>
To:     Suzuki K Poulose <suzuki.poulose@....com>,
        Mathieu Poirier <mathieu.poirier@...aro.org>
Cc:     will@...nel.org, catalin.marinas@....com, mike.leach@...aro.org,
        leo.yan@...aro.org, maz@...nel.org, coresight@...ts.linaro.org,
        linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v5 10/15] coresight: trbe: Workaround TRBE errata
 overwrite in FILL mode



On 10/19/21 2:45 AM, Suzuki K Poulose wrote:
> On 18/10/2021 16:51, Mathieu Poirier wrote:
>> On Thu, Oct 14, 2021 at 11:31:20PM +0100, Suzuki K Poulose wrote:
>>> ARM Neoverse-N2 (#2139208) and Cortex-A710(##2119858) suffers from
>>> an erratum, which when triggered, might cause the TRBE to overwrite
>>> the trace data already collected in FILL mode, in the event of a WRAP.
>>> i.e, the TRBE doesn't stop writing the data, instead wraps to the base
>>> and could write upto 3 cache line size worth trace. Thus, this could
>>> corrupt the trace at the "BASE" pointer.
>>>
>>> The workaround is to program the write pointer 256bytes from the
>>> base, such that if the erratum is triggered, it doesn't overwrite
>>> the trace data that was captured. This skipped region could be
>>> padded with ignore packets at the end of the session, so that
>>> the decoder sees a continuous buffer with some padding at the
>>> beginning. The trace data written at the base is considered
>>> lost as the limit could have been in the middle of the perf
>>> ring buffer, and jumping to the "base" is not acceptable.
>>> We set the flags already to indicate that some amount of trace
>>> was lost during the FILL event IRQ. So this is fine.
>>>
>>> One important change with the work around is, we program the
>>> TRBBASER_EL1 to current page where we are allowed to write.
>>> Otherwise, it could overwrite a region that may be consumed
>>> by the perf. Towards this, we always make sure that the
>>> "handle->head" and thus the trbe_write is PAGE_SIZE aligned,
>>> so that we can set the BASE to the PAGE base and move the
>>> TRBPTR to the 256bytes offset.
>>>
>>> Cc: Mike Leach <mike.leach@...aro.org>
>>> Cc: Mathieu Poirier <mathieu.poirier@...aro.org>
>>> Cc: Anshuman Khandual <anshuman.khandual@....com>
>>> Cc: Leo Yan <leo.yan@...aro.org>
>>> Reviewed-by: Mathieu Poirier <mathieu.poirier@...aro.org>
>>> Signed-off-by: Suzuki K Poulose <suzuki.poulose@....com>
>>> ---
>>> Changes since v2:
>>>   - Updated the ASCII art to include better description of
>>>     all the steps in the work around
>>> Change since v1:
>>>   - Updated comment with ASCII art
>>>   - Add _BYTES suffix for the space to skip for the work around.
>>> ---
>>>   drivers/hwtracing/coresight/coresight-trbe.c | 169 +++++++++++++++++--
>>>   1 file changed, 158 insertions(+), 11 deletions(-)
>>>
>>> diff --git a/drivers/hwtracing/coresight/coresight-trbe.c b/drivers/hwtracing/coresight/coresight-trbe.c
>>> index 314e5e7374c7..b56b166b2dec 100644
>>> --- a/drivers/hwtracing/coresight/coresight-trbe.c
>>> +++ b/drivers/hwtracing/coresight/coresight-trbe.c
>>> @@ -16,6 +16,7 @@
>>>   #define pr_fmt(fmt) DRVNAME ": " fmt
>>>     #include <asm/barrier.h>
>>> +#include <asm/cpufeature.h>
>>
>> Here too I get a checkpatch warning...
>>
> 
> That is a false alarm. I guess that warns for including
> linux/cpufeature.h? It is a bit odd, we include this
> for the arm64 cpucaps, not the generic linux feature

It is a bit odd, I saw that too.

> checks. (They are used for "loading modules" based
> on "features" which are more like ELF HWCAPs).

Should <asm/cpufeature.h> be renamed as <asm/arm64_cpufeature.h>
or something similar instead to differentiate it from the generic
<linux/cpufeature.h> as they are not related. Also, probably this
warning could be avoided.

> 
> As such I chose to ignore it.
> 
> Suzuki