Intel Haswell GT1 Graphics Have Been Busted The Past Half-Year On Linux

[Aryma]

they should give Haswell and ivy bridge some L.O.V.E too always have a problem with mpv in ubuntu because intel driver

[mrmazda]

I have two haswells:

1. i3-4150T 35W TDP
2. G3220 Pentium 53W TDP

I've tried:

• hwinfo
• lspci
• lscpu
• device-ID lookups
• inxi
• Intel's specification pages for each
• others I can't remember ATM

I've been unable to determine whether either IGP is a GT1, or ascertain that any hang is attributable to anything other than Mozillas having been up too long with too many tabs open, which is not very common. I've certainly not seen any boot time hangs. The i3 has been in this PC about 3 years, which began life 5+ years ago using the Pentium, and has run 24/7 the whole time other than for disk changes and the CPU swap. I returned the Pentium to service some months back in a test PC.

Could this have anything to do with compositing? Both these two machines are multi- multi-boot. All DEs that can be are configured with composting disabled. My 24/7 DE is KDE3. On my 20+ 64bit test boxes are a mix of KDE3, TDE and Plasma, including that with the G3220.

[Guest]

Intel Graphics has always been a regression nightmare for me, that's why I would love to see an high end laptop with an AMD APU.

For me, AMD graphics has always been a bug and regression nightmare, while Intel graphics has been excellent and pretty close to perfect.

[chocolate]

The same commit caused the regression, but the fix isn't the same. Ville's patch fixed the GT3e regression I alluded to: https://git.kernel.org/pub/scm/linux...b198fae396d660

I haven't checked if it's been backported, but it has the right tags so I assume it has been.

Good to know, thank you!

[kneekoo]

I'm on Haswell GT2 (i5-4690K). So I guess I'm lucky, huh? Anyway, I hope that at least Gen 7 will get some farewell stability updates, instead of breaking them. There's zero incentive in buying Intel GPUs when you know Intel's typical behavior is to leave the old ones in a crappy state. Let their low performance (according to the user needs) be their downfall, not some driver issues.

[Slartifartblast]

This isn't just a Linux problem, it was badly broken on a friend's Windows 10 G3220 PC, had to wind the driver back and prevent it from being updated.

[Slartifartblast]

I have two haswells:

1. i3-4150T 35W TDP
2. G3220 Pentium 53W TDP

I've tried:

• hwinfo
• lspci
• lscpu
• device-ID lookups
• inxi
• Intel's specification pages for each
• others I can't remember ATM

I've been unable to determine whether either IGP is a GT1, or ascertain that any hang is attributable to anything other than Mozillas having been up too long with too many tabs open, which is not very common. I've certainly not seen any boot time hangs. The i3 has been in this PC about 3 years, which began life 5+ years ago using the Pentium, and has run 24/7 the whole time other than for disk changes and the CPU swap. I returned the Pentium to service some months back in a test PC.

Could this have anything to do with compositing? Both these two machines are multi- multi-boot. All DEs that can be are configured with composting disabled. My 24/7 DE is KDE3. On my 20+ 64bit test boxes are a mix of KDE3, TDE and Plasma, including that with the G3220.

https://www.cpu-world.com/CPUs/Pentium_Dual-Core/Intel-Pentium%20G3220.html

GT1

https://www.cpu-world.com/CPUs/Core_i3/Intel-Core%20i3-4150T.html

GT2