linux-kernel - Re: [PATCH v2 3/6] firmware: qcom: scm: Handle various probe ordering for qcom_scm_assign

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: <b0db7570-5606-401f-bf1a-dd1076d2dc52@linaro.org>
Date: Wed, 8 Jan 2025 12:12:51 +0100
From: Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>
To: Bjorn Andersson <andersson@...nel.org>
Cc: Konrad Dybcio <konradybcio@...nel.org>,
 Mukesh Ojha <quic_mojha@...cinc.com>,
 Dmitry Baryshkov <dmitry.baryshkov@...aro.org>,
 Stephan Gerhold <stephan.gerhold@...aro.org>,
 Bartosz Golaszewski <bartosz.golaszewski@...aro.org>,
 Kuldeep Singh <quic_kuldsing@...cinc.com>,
 Elliot Berman <quic_eberman@...cinc.com>,
 Andrew Halaney <ahalaney@...hat.com>,
 Avaneesh Kumar Dwivedi <quic_akdwived@...cinc.com>,
 Andy Gross <andy.gross@...aro.org>, linux-arm-msm@...r.kernel.org,
 linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 3/6] firmware: qcom: scm: Handle various probe ordering
 for qcom_scm_assign_mem()

On 07/01/2025 01:42, Bjorn Andersson wrote:
>>
>> diff --git a/drivers/firmware/qcom/qcom_scm.c b/drivers/firmware/qcom/qcom_scm.c
>> index 5d91b8e22844608f35432f1ba9c08d477d4ff762..93212c8f20ad65ecc44804b00f4b93e3eaaf8d95 100644
>> --- a/drivers/firmware/qcom/qcom_scm.c
>> +++ b/drivers/firmware/qcom/qcom_scm.c
>> @@ -1075,6 +1075,9 @@ int qcom_scm_assign_mem(phys_addr_t mem_addr, size_t mem_sz,
>>  	int ret, i, b;
>>  	u64 srcvm_bits = *srcvm;
>>  
>> +	if (!qcom_scm_is_available())
>> +		return -EPROBE_DEFER;
> 
> This API is generally called from places other than probe, making the
> return of EPROBE_DEFER undesirable. While not pretty, a client depending
> on the scm driver to be probed is expected to call
> qcom_scm_is_available().
> 
> qcom_rmtfs_mem_probe() does this right before calling
> qcom_scm_assign_mem(), am I misunderstanding the case you're describing?

I tried to reproduce my NULL ptr but now I cannot and indeed if every
path is protected with qcom_scm_is_available() everything should be fine.

Let's skip this patch then.

Best regards,
Krzysztof