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: <ZMjkPLqBiWW8CpAU@gerhold.net>
Date:   Tue, 1 Aug 2023 12:53:48 +0200
From:   Stephan Gerhold <stephan@...hold.net>
To:     Konrad Dybcio <konrad.dybcio@...aro.org>
Cc:     Andy Gross <agross@...nel.org>,
        Bjorn Andersson <andersson@...nel.org>,
        Georgi Djakov <djakov@...nel.org>,
        Michael Turquette <mturquette@...libre.com>,
        Stephen Boyd <sboyd@...nel.org>,
        Marijn Suijten <marijn.suijten@...ainline.org>,
        linux-arm-msm@...r.kernel.org, linux-pm@...r.kernel.org,
        linux-kernel@...r.kernel.org, linux-clk@...r.kernel.org
Subject: Re: [PATCH v2 01/10] interconnect: qcom: icc-rpm: Add AB/IB
 calculations coefficients

On Mon, Jul 31, 2023 at 12:52:17PM +0200, Konrad Dybcio wrote:
> Presumably due to the hardware being so complex, some nodes (or busses)
> have different (usually higher) requirements for bandwidth than what
> the usual calculations would suggest.
> 
> Looking at the available downstream files, it seems like AB values are
> adjusted per-bus and IB values are adjusted per-node.
> With that in mind, introduce percentage-based coefficient struct members
> and use them in the calculations.
> 
> One thing to note is that the IB coefficient is inverse (100/ib_percent)
> which feels a bit backwards, but it's necessary for precision..
> 
> Signed-off-by: Konrad Dybcio <konrad.dybcio@...aro.org>
> ---
>  drivers/interconnect/qcom/icc-rpm.c | 14 +++++++++++---
>  drivers/interconnect/qcom/icc-rpm.h |  6 ++++++
>  2 files changed, 17 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/interconnect/qcom/icc-rpm.c b/drivers/interconnect/qcom/icc-rpm.c
> index 2c16917ba1fd..a837d20af79e 100644
> --- a/drivers/interconnect/qcom/icc-rpm.c
> +++ b/drivers/interconnect/qcom/icc-rpm.c
> @@ -298,7 +298,8 @@ static int qcom_icc_bw_aggregate(struct icc_node *node, u32 tag, u32 avg_bw,
>   */
>  static void qcom_icc_bus_aggregate(struct icc_provider *provider, u64 *agg_clk_rate)
>  {
> -	u64 agg_avg_rate, agg_rate;
> +	struct qcom_icc_provider *qp = to_qcom_provider(provider);
> +	u64 agg_avg_rate, agg_peak_rate, agg_rate;
>  	struct qcom_icc_node *qn;
>  	struct icc_node *node;
>  	int i;
> @@ -315,8 +316,15 @@ static void qcom_icc_bus_aggregate(struct icc_provider *provider, u64 *agg_clk_r
>  			else
>  				agg_avg_rate = qn->sum_avg[i];
>  
> -			agg_rate = max_t(u64, agg_avg_rate, qn->max_peak[i]);
> -			do_div(agg_rate, qn->buswidth);
> +			if (qp->ab_coeff)
> +				agg_avg_rate = mult_frac(qp->ab_coeff, agg_avg_rate, 100);

agg_avg_rate * (qp->ab_coeff / 100) would feel more logical to me (even
if it should be the same), i.e.

	agg_avg_rate = mult_frac(agg_avg_rate, qp->ab_coeff, 100);

Not sure why you swapped them.

> +
> +			if (qp->ib_coeff)
> +				agg_peak_rate = mult_frac(100, qn->max_peak[i], qp->ib_coeff);

	agg_peak_rate = mult_frac(qn->max_peak[i], 100, qp->ib_coeff);

Anyway, looks like you need to avoid mult_frac anyway for ARM32 compat :/

arm-none-eabi-ld: drivers/interconnect/qcom/icc-rpm.o: in function `qcom_icc_calc_rate':
drivers/interconnect/qcom/icc-rpm.c:310: undefined reference to `__aeabi_uldivmod'
arm-none-eabi-ld: drivers/interconnect/qcom/icc-rpm.c:312: undefined reference to `__aeabi_uldivmod'

Thanks,
Stephan

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ