Linux 2.4.21-rc6
i tried 2.4.21-rc6 as i was told it might fix the mouse stalling on heavy disk IO problem and i would like to report that it DOES fix them for the most part, even certain compiles/benchmarks/stress tests that could stall my pc for seconds now affect the mouse for mere fractions of one second, situations that used to cause short stalls are now a thing of the past
2.4.21-rc6 is the best kernel i have tried to date and i have tried many on my quest to get a smooth mouse
i dont subscribe to lkml, so if you have questions please CC me personally
hope this input is helpful
Daniel
morfic
Quoted 9 lines · Daniel Goller
i tried 2.4.21-rc6 as i was told it might fix the mouse stalling on heavy disk IO problem and i would like to report that it DOES fix them for the most part, even certain compiles/benchmarks/stress tests that could stall my pc for seconds now affect the mouse for mere fractions of one second, situations that used to cause short stalls are now a thing of the past
2.4.21-rc6 is the best kernel i have tried to date and i have tried many on my quest to get a smooth mouse
There are reports that 2.4.18 also "fixed" the problems with the mouse. Can you verify?
Quoted 8 lines · Mike Fedyk
Quoted 7 lines · Daniel Goller
i tried 2.4.21-rc6 as i was told it might fix the mouse stalling on heavy disk IO problem and i would like to report that it DOES fix them for the most part, even certain compiles/benchmarks/stress tests that could stall my pc for seconds now affect the mouse for mere fractions of one second, situations that used to cause short stalls are now a thing of the past
Quoted 5 lines
Quoted 2 lines
2.4.21-rc6 is the best kernel i have tried to date and i have tried many on my quest to get a smooth mouse
There are reports that 2.4.18 also "fixed" the problems with the mouse. Can you verify?
sorry i never ran a 2.4.18 kernel, can't comment on that
Quoted 2 lines · Daniel Goller
Quoted 1 line · Mike Fedyk
On Fri, May 30, 2003 at 02:08:51PM -0500, Daniel Goller
Quoted 9 lines · wrote:
Quoted 9 lines
Quoted 9 lines
i tried 2.4.21-rc6 as i was told it might fix the mouse stalling on heavy disk IO problem and i would like to report that it DOES fix them for the most part, even certain compiles/benchmarks/stress tests that could stall my pc for seconds now affect the mouse for mere fractions of one second, situations that used to cause short stalls are now a thing of the past
Quoted 7 lines
Quoted 6 lines
Quoted 3 lines
2.4.21-rc6 is the best kernel i have tried to date and i have tried many on my quest to get a smooth mouse
There are reports that 2.4.18 also "fixed" the problems with the mouse. Can you verify?
Yes, it performs similar to -rc6 but not nearly as good as 2.5.70.
On 2.5.70 the mouse is really smooth all the time, scrollong of large pages in opera is fairly smooth most the time also with large disk io loads such as the script i posted earlier.
Regards
Michael
Quoted 24 lines · Michael Frank
Quoted 2 lines · Daniel Goller
Quoted 1 line · Mike Fedyk
On Fri, May 30, 2003 at 02:08:51PM -0500, Daniel Goller
Quoted 9 lines · wrote:
Quoted 9 lines
Quoted 9 lines
i tried 2.4.21-rc6 as i was told it might fix the mouse stalling on heavy disk IO problem and i would like to report that it DOES fix them for the most part, even certain compiles/benchmarks/stress tests that could stall my pc for seconds now affect the mouse for mere fractions of one second, situations that used to cause short stalls are now a thing of the past
Quoted 7 lines
Quoted 6 lines
Quoted 3 lines
2.4.21-rc6 is the best kernel i have tried to date and i have tried many on my quest to get a smooth mouse
There are reports that 2.4.18 also "fixed" the problems with the mouse. Can you verify?
Yes, it performs similar to -rc6 but not nearly as good as 2.5.70.
Quoted 8 lines
On 2.5.70 the mouse is really smooth all the time, scrollong of large pages in opera is fairly smooth most the time also with large disk io loads such as the script i posted earlier.
Regards
Michael
unfortunately the radeon dri is broken in 2.5.70 so i havent tried that much, need to see if someone already suggests a fix for this unused int (it seems unused to me, after a *quick* look through the file) i guess i will have to subscribe now to lkml