Can JTAG be remapped?

[perryc]

I see info in the tech_ref that suggests JTAG can be remapped with Direct I/O via IO_MUX. Section 4.4.2 is a bit thin but I'm not seeing alternate pads that can accept MT* functions.

Perry

[sintech]

Yes, I'm also curious about it, especially when MT pins are used for SD functions.

[ESP_Angus]

Unfortunately JTAG cannot be remapped.

[sintech]

Then maybe SD signals can be remapped to other pins?
Is there any expected performance degradation?
Thanks in advance.

[ESP_Angus]

Unfortunately SD can't be remapped either. These (& EMAC) are almost the only peripheral pins which are fixed, sorry.

[perryc]

Well this is not good news (but glad it was not a dumb question). I've got a uSD card in my design as well so what's the best practices plan here when you need a file system?
I ironically I got motivated to figure out JTAG so I could debug my attempt to get SPIFFS working. I'm thinking the best I can do is:
1) JTAG header and uSD share IO12-IO15 in next HW version
2) Get SPIFFS Working
3) When needing to do serious bug hunting, pull uSD Card, build a SPIFFS version of my firmware and then JTAG should work.

Sure hope those JTAG pins do not mind 10K pullups or I'll need to add a jumper block.

Perry

[sintech]

In my opinion, adding SPI support to sdmmc driver will partially solve this issue of SD and JTAC coexistence.
Because SPI signals can be remapped and even the default VSPI pins are not mixed with other useful signals.