ChanServ changed the topic of #linux-msm to:
aat596_2 has joined #linux-msm
Daanct12 has joined #linux-msm
enick_991 has joined #linux-msm
DylanVanAssche has joined #linux-msm
<mal> looking at the code it seems I probably need to implement that myself
danylo has quit [Ping timeout: 480 seconds]
<Daanct12> seems like some of my patches from https://patchwork.kernel.org/project/linux-arm-msm/list/?series=799073 went non-applicable, is there any other things that needs to be improved aside from reviewers changes?
<bamse> Daanct12: they are "not applicable" in the qcom tree...because they should be handled by other maintainers
marvin24 has joined #linux-msm
marvin24_ has quit [Ping timeout: 480 seconds]
AffeNull[m]1 has joined #linux-msm
Adrian[m]1 has joined #linux-msm
flamingradian[m] has joined #linux-msm
nergzd723 has joined #linux-msm
RayyanAnsari[m] has joined #linux-msm
Newbyte has joined #linux-msm
mripard has quit [Quit: mripard]
mripard has joined #linux-msm
danylo has joined #linux-msm
xtex[m] has joined #linux-msm
minecrell[m] has joined #linux-msm
vknecht[m] has joined #linux-msm
srinik has quit [Killed (NickServ (Too many failed password attempts.))]
srinik has joined #linux-msm
aat596_2 has quit [Quit: ZNC 1.8.2 - https://znc.in]
<Daanct12> oh boy the interconenct driver :D
<Daanct12> this is gonna be fun
aat596_2 has joined #linux-msm
<aka_[m]> Daanct12: I already have one for 8937
<aka_[m]> Minecrell will send 8909 for 6.8 and I want to push 8976 too
Danct12 has quit [Read error: Connection reset by peer]
<aka_[m]> You can test 8937 one and we can push all of them in single series
<Daanct12> aka_[m]: neat
<Daanct12> can i have it?
<aka_[m]> Haven't yet compile tested
<bryanodonoghue> mal
<Daanct12> where's the compat aka_[m]
<bryanodonoghue> git am ec6493bcb672be508fef4b2344fc4cb6bdef17be..a02b6a377c1b2e4d15d5c22d8f0046bec6462f0b
<bryanodonoghue> [sic] git cherry-pick ec6493bcb672be508fef4b2344fc4cb6bdef17be..a02b6a377c1b2e4d15d5c22d8f0046bec6462f0b
<Daanct12> aka_[m]: doesn't compile because of smd-rpm.h being missing, but that sounds like a copy/paste error, removed that. there's also missing of_table
<aka_[m]> Daanct12: indeed 8937 lost them, will clean it up once get to home
<aka_[m]> So in 4h
<Daanct12> after that the code compiles, still need testing
<aka_[m]> I had some really primitive generator for those so it skipped of table
<Daanct12> that aside, im a little pissed off because my computer decided to reboot :)
<Daanct12> [Hardware Error]: CPU 27: Machine Check: 0 Bank 5: baa0000000090150
<Daanct12> good ol' pbo errors
<Daanct12> now i have to clone the msm-4.9 source again :D
travmurav[m] has joined #linux-msm
<Daanct12> aka_[m]: uh how did i give it a clock
<Daanct12> downstream wants bus and bus_a, but they're all in this form..
<Daanct12> #define clk_pnoc_msmbus_clk 0x38b95c77
<Daanct12> no idea what that is
<Daanct12> ah, looks like that's for rpmcc.
<aka_[m]> Daanct12: with rework it's bit different I believe
<aka_[m]> There is dt binding in patches
<Daanct12> yay it booted
<Daanct12> into a blue screen!! :D
<aka_[m]> Clks are passed via .clk I believe
<Daanct12> qcom_icc_rpm_smd_send mas 86 error -6
<aka_[m]> Daanct12: normal
<aka_[m]> Means node not accessible
<aka_[m]> We set -1 on master/slave id
<aka_[m]> Prob one of some weird masters
<aka_[m]> Like lpass/wcss
<Daanct12> looks like mas 86 is pcnoc_int_1
<aka_[m]> However it can broke booting
<aka_[m]> Of it has mas id then set -1 on it
<aka_[m]> Leave slave if it's set
<Daanct12> but anyway it boots into a blue screen
<Daanct12> looks like the system becomes unstable
<Daanct12> when the bluescreen shows up, looks like i may have done something wrong here..
<aka_[m]> Something clks
<aka_[m]> Daanct12: developing on next or latest stable/r?
<aka_[m]> Rc
<Daanct12> stable 6.6
<minecrell> If you add interconnect scaling you also need to add the necessary paths to pretty much all consumers (cpu, sdhci, mdp, gpu, ...) otherwise it will just reduce everything to minimum and everything will be just crawling
<minecrell> blue screen = mdp doesn't manage to fetch screen contents in time
<aka_[m]> Can also be disabled dsi_phy
<Daanct12> ah yes this was all on framebuffer console
<Daanct12> simplefb*
felixwither[m] has joined #linux-msm
JIaxyga[m] has joined #linux-msm
FieryFlames[m] has joined #linux-msm
mripard has quit [Remote host closed the connection]
mripard has joined #linux-msm
Daanct12 has quit [Quit: WeeChat 4.1.1]
z3ntu has joined #linux-msm
<mal> bryanodonoghue: that CSID_CSI2_RX_CRC_ERRORS remains 0
<mal> bryanodonoghue: CSID_CSI2_RX_TOTAL_PKTS_RCVD increases all the time
<bryanodonoghue> mal what is your pipeline ?
<bryanodonoghue> and your pixel format
<mal> bryanodonoghue: SRGGB10_1X10 I have setup msm_csiphy3->msm_csid3->msm_vfe3_rdi0
<bryanodonoghue> cat /proc/interrupts | grep camss
<bryanodonoghue> i assume you've added sc7280 to the sdm845 pixel format list
<mal> yes, in camss-csiphy.c, no interrups from csiphy
<bryanodonoghue> that's expected
<bryanodonoghue> interrupts from the phy are for errors
Leandro[m]1 has joined #linux-msm
<bryanodonoghue> id trace msm_video_init_format()
<bryanodonoghue> does it complete for you ?
<mal> that function goes to end and not the error path
<mal> maybe my camera sensor driver could be broken
HappY[m] has joined #linux-msm
<mal> I see interrupts from camss_msm_vfe3 which I assume is how it should be
Danct12 has joined #linux-msm
<mal> bryanodonoghue: should csid give interrupts usually? I see one from it every time I start capture
<bryanodonoghue> vfe sounds promising
<bryanodonoghue> are they write-master completion interrupts ?
<bryanodonoghue> off the top of my head CSID gets SOF/EOF at least onc
<mal> bryanodonoghue: how can I see which interrupts those are?
<bryanodonoghue> err I had a patch once upon a time
<bryanodonoghue> maybe just add a printk here
<bryanodonoghue> drivers/media/platform/qcom/camss/camss-vfe.h:void (*wm_done)(struct vfe_device *vfe, u8 wm);
<mal> ok
<mal> bryanodonoghue: those are sof interrupts
<bryanodonoghue> if you are getting CSID::SOF but not VFE::WM then that's the error
<bryanodonoghue> getting the data out of the CSID into the VFE/RDI
<mal> vfe_isr_sof is where I added the printk
<mal> I added printk to all of the interrupt handlers in vfe
<bryanodonoghue> ok but you _need_ wm_done
<bryanodonoghue> => data written to DRAM for delivery to user-space
<mal> ok, I will have a look how that should get called
adomerle has joined #linux-msm
Nia[m] has joined #linux-msm
alikateshethey[m] has joined #linux-msm
svarbanov_ has quit [Remote host closed the connection]
svarbanov_ has joined #linux-msm
atipls_ has quit [Ping timeout: 480 seconds]
rmsilva has quit [Ping timeout: 480 seconds]
atipls has joined #linux-msm
rmsilva has joined #linux-msm