Re: [PATCH V3 1/3] arm64: topology: Avoid the have_policy check
On 18-12-20, 11:01, Catalin Marinas wrote: > On Fri, Dec 18, 2020 at 09:56:02AM +0530, Viresh Kumar wrote: > > On 17-12-20, 10:55, Catalin Marinas wrote: > > > Hi Viresh, > > > > > > On Thu, Dec 17, 2020 at 01:27:32PM +0530, Viresh Kumar wrote: > > > > On 15-12-20, 11:04, Viresh Kumar wrote: > > > > > Every time I have stumbled upon this routine, I get confused with the > > > > > way 'have_policy' is used and I have to dig in to understand why is it > > > > > so. Here is an attempt to make it easier to understand, and hopefully > > > > > it > > > > > is an improvement. > > > > > > > > > > The 'have_policy' check was just an optimization to avoid writing > > > > > to amu_fie_cpus in case we don't have to, but that optimization itself > > > > > is creating more confusion than the real work. Lets just do that if > > > > > all > > > > > the CPUs support AMUs. It is much cleaner that way. > > > > > > > > > > Reviewed-by: Ionela Voinescu > > > > > Signed-off-by: Viresh Kumar > > > > > --- > > > > > V3: > > > > > - Added Reviewed by tag. > > > > > > > > Catalin, please pick the first two patches for 5.11. I will send the > > > > last one separately later on. > > > > > > I haven't figured out whether these are fixes (a cover letter would > > > help ;)). They look like generic improvements to me > > > > Right they are and since the merge window just opened I thought these > > don't really need to wait for another full cycle to get in. > > Normally we freeze the arm64 tree around the -rc6 prior to the merging > window to give the patches a bit of time in linux-next. This time > around, given the holidays, Linus even stated that if not already in > -next at 5.10, it won't be pulled: https://lkml.org/lkml/2020/12/13/290. Okay, sounds good. > So please re-post at -rc1 with the acks in place. Sure. -- viresh
Re: [PATCH V3 1/3] arm64: topology: Avoid the have_policy check
On Fri, Dec 18, 2020 at 09:56:02AM +0530, Viresh Kumar wrote: > On 17-12-20, 10:55, Catalin Marinas wrote: > > Hi Viresh, > > > > On Thu, Dec 17, 2020 at 01:27:32PM +0530, Viresh Kumar wrote: > > > On 15-12-20, 11:04, Viresh Kumar wrote: > > > > Every time I have stumbled upon this routine, I get confused with the > > > > way 'have_policy' is used and I have to dig in to understand why is it > > > > so. Here is an attempt to make it easier to understand, and hopefully it > > > > is an improvement. > > > > > > > > The 'have_policy' check was just an optimization to avoid writing > > > > to amu_fie_cpus in case we don't have to, but that optimization itself > > > > is creating more confusion than the real work. Lets just do that if all > > > > the CPUs support AMUs. It is much cleaner that way. > > > > > > > > Reviewed-by: Ionela Voinescu > > > > Signed-off-by: Viresh Kumar > > > > --- > > > > V3: > > > > - Added Reviewed by tag. > > > > > > Catalin, please pick the first two patches for 5.11. I will send the > > > last one separately later on. > > > > I haven't figured out whether these are fixes (a cover letter would > > help ;)). They look like generic improvements to me > > Right they are and since the merge window just opened I thought these > don't really need to wait for another full cycle to get in. Normally we freeze the arm64 tree around the -rc6 prior to the merging window to give the patches a bit of time in linux-next. This time around, given the holidays, Linus even stated that if not already in -next at 5.10, it won't be pulled: https://lkml.org/lkml/2020/12/13/290. So please re-post at -rc1 with the acks in place. -- Catalin
Re: [PATCH V3 1/3] arm64: topology: Avoid the have_policy check
On 17-12-20, 10:55, Catalin Marinas wrote: > Hi Viresh, > > On Thu, Dec 17, 2020 at 01:27:32PM +0530, Viresh Kumar wrote: > > On 15-12-20, 11:04, Viresh Kumar wrote: > > > Every time I have stumbled upon this routine, I get confused with the > > > way 'have_policy' is used and I have to dig in to understand why is it > > > so. Here is an attempt to make it easier to understand, and hopefully it > > > is an improvement. > > > > > > The 'have_policy' check was just an optimization to avoid writing > > > to amu_fie_cpus in case we don't have to, but that optimization itself > > > is creating more confusion than the real work. Lets just do that if all > > > the CPUs support AMUs. It is much cleaner that way. > > > > > > Reviewed-by: Ionela Voinescu > > > Signed-off-by: Viresh Kumar > > > --- > > > V3: > > > - Added Reviewed by tag. > > > > Catalin, please pick the first two patches for 5.11. I will send the > > last one separately later on. > > I haven't figured out whether these are fixes (a cover letter would > help ;)). They look like generic improvements to me Right they are and since the merge window just opened I thought these don't really need to wait for another full cycle to get in. > and given that we > are already in the 5.11 merging window, they would probably need to wait > until 5.12. Whatever you decide :) -- viresh
Re: [PATCH V3 1/3] arm64: topology: Avoid the have_policy check
Hi Viresh, On Thu, Dec 17, 2020 at 01:27:32PM +0530, Viresh Kumar wrote: > On 15-12-20, 11:04, Viresh Kumar wrote: > > Every time I have stumbled upon this routine, I get confused with the > > way 'have_policy' is used and I have to dig in to understand why is it > > so. Here is an attempt to make it easier to understand, and hopefully it > > is an improvement. > > > > The 'have_policy' check was just an optimization to avoid writing > > to amu_fie_cpus in case we don't have to, but that optimization itself > > is creating more confusion than the real work. Lets just do that if all > > the CPUs support AMUs. It is much cleaner that way. > > > > Reviewed-by: Ionela Voinescu > > Signed-off-by: Viresh Kumar > > --- > > V3: > > - Added Reviewed by tag. > > Catalin, please pick the first two patches for 5.11. I will send the > last one separately later on. I haven't figured out whether these are fixes (a cover letter would help ;)). They look like generic improvements to me and given that we are already in the 5.11 merging window, they would probably need to wait until 5.12. -- Catalin
Re: [PATCH V3 1/3] arm64: topology: Avoid the have_policy check
On 15-12-20, 11:04, Viresh Kumar wrote: > Every time I have stumbled upon this routine, I get confused with the > way 'have_policy' is used and I have to dig in to understand why is it > so. Here is an attempt to make it easier to understand, and hopefully it > is an improvement. > > The 'have_policy' check was just an optimization to avoid writing > to amu_fie_cpus in case we don't have to, but that optimization itself > is creating more confusion than the real work. Lets just do that if all > the CPUs support AMUs. It is much cleaner that way. > > Reviewed-by: Ionela Voinescu > Signed-off-by: Viresh Kumar > --- > V3: > - Added Reviewed by tag. Catalin, please pick the first two patches for 5.11. I will send the last one separately later on. -- viresh
