Re: [PATCH v2 1/1] mm: Better handling of adding new regions

2022-10-09 Thread jim945 via Grub-devel
06.10.2022 16:07, Daniel Kiper пишет: On Sun, Sep 25, 2022 at 03:59:50PM +0200, Patrick Steinhardt wrote: On Tue, Sep 13, 2022 at 01:49:52AM +0800, Zhang Boyang wrote: The code of dynamically adding new regions has two problems. First, it always invalidate disk caches, which decreases performa

Re: [PATCH v2 1/1] mm: Better handling of adding new regions

2022-10-06 Thread Daniel Kiper
On Sun, Sep 25, 2022 at 03:59:50PM +0200, Patrick Steinhardt wrote: > On Tue, Sep 13, 2022 at 01:49:52AM +0800, Zhang Boyang wrote: > > The code of dynamically adding new regions has two problems. First, it > > always invalidate disk caches, which decreases performance severely. > > Second, it requ

Re: [PATCH v2 1/1] mm: Better handling of adding new regions

2022-09-25 Thread Patrick Steinhardt
On Tue, Sep 13, 2022 at 01:49:52AM +0800, Zhang Boyang wrote: > The code of dynamically adding new regions has two problems. First, it > always invalidate disk caches, which decreases performance severely. > Second, it request exactly "size" bytes for new region, ignoring region > management overhe

Re: [PATCH v2 1/1] mm: Better handling of adding new regions

2022-09-14 Thread jim945 via Grub-devel
12.09.2022 20:49, Zhang Boyang пишет: The code of dynamically adding new regions has two problems. First, it always invalidate disk caches, which decreases performance severely. Second, it request exactly "size" bytes for new region, ignoring region management overheads. This patch makes adding

[PATCH v2 1/1] mm: Better handling of adding new regions

2022-09-12 Thread Zhang Boyang
The code of dynamically adding new regions has two problems. First, it always invalidate disk caches, which decreases performance severely. Second, it request exactly "size" bytes for new region, ignoring region management overheads. This patch makes adding new regions more priority than disk cach