[PATCH] nvme-pci: disable write zeroes for SK Hynix BC901

[hardaway.tseng at sk.com]

Hi Keith,

Hope everything you are doing is well!

These days our teams tried several way to do further verification of this disable write zero patch for BC901, let me summarize as below :

1)	The chrome image with this new write zero disable code built with 5.15, has been verified and passed (< 20min mode switching time)
2)	Due to some limitations, Google team couldn't build a new chrome image with the suggested 6.5 or later version + this disable write zero patch for verification test.
3)	Tried to build image with 5.15 + disable write zero patch + DEAC merged, system can run but the testing result is failed (> 20min mode switching, actually 27~29mins).

Above is all we could do here for the verification of this patch, could you share us your comment or suggestion?

Thank you so much!


Thanks & Regards,
Hardaway.


Hardaway Tseng/ Sr. Manager | SN-FAQE | SK hynix Semiconductor Taiwan
Address: 10F. No. 308, Zhifu Rd., Zhongshan Dist., Taipei City 10466, Taiwan
Office : +886-2-3518-2313/ Fax: +886-2-3518-2366/ Mobile: +886-919-803-458
Email: hardaway.tseng at sk.com

SK hynix Confidential
“This e-mail may contain confidential and/or privileged information. If you are not the intended recipient(or have received this e-mail in error), please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.”


-----Original Message-----
From: Hardaway (Shih Hung) Tseng SKHYT 
Sent: Wednesday, November 29, 2023 11:51 AM
To: 'Keith Busch' <kbusch at kernel.org>; Jim Lin 林志融 <jim.lin at siliconmotion.com>; David Yeh 葉寰融 <david.yeh at siliconmotion.com>
Cc: Jim.Lin <jim.chihjung.lin at gmail.com>; axboe at kernel.dk; hch at lst.de; sagi at grimberg.me; linux-nvme at lists.infradead.org; linux-kernel at vger.kernel.org; CJ Chen 陳群傑 <cj.chen at siliconmotion.com>; 한상안(HAN SANGAN) 대만법인 <sangan.han at sk.com>; Ives(Hsueh Hsien) Lu SKHYT <ives.lu at sk.com>
Subject: RE: [PATCH] nvme-pci: disable write zeroes for SK Hynix BC901

Hi Keith,

Thanks for head up Christoph's questions!
Our team will help us to answer them.

Hi Jim and David,
Thanks for your support!


Thanks & Regards,
Hardaway.


Hardaway Tseng/ Sr. Manager | SN-FAQE | SK hynix Semiconductor Taiwan
Address: 10F. No. 308, Zhifu Rd., Zhongshan Dist., Taipei City 10466, Taiwan Office : +886-2-3518-2313/ Fax: +886-2-3518-2366/ Mobile: +886-919-803-458
Email: hardaway.tseng at sk.com

SK hynix Confidential
“This e-mail may contain confidential and/or privileged information. If you are not the intended recipient(or have received this e-mail in error), please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.”


-----Original Message-----
From: Keith Busch <kbusch at kernel.org>
Sent: Wednesday, November 29, 2023 8:10 AM
To: Hardaway (Shih Hung) Tseng SKHYT <hardaway.tseng at sk.com>
Cc: Jim Lin 林志融 <jim.lin at siliconmotion.com>; Jim.Lin <jim.chihjung.lin at gmail.com>; axboe at kernel.dk; hch at lst.de; sagi at grimberg.me; linux-nvme at lists.infradead.org; linux-kernel at vger.kernel.org; CJ Chen 陳群傑 <cj.chen at siliconmotion.com>; David Yeh 葉寰融 <david.yeh at siliconmotion.com>; 한상안(HAN SANGAN) 대만법인 <sangan.han at sk.com>; Ives(Hsueh Hsien) Lu SKHYT <ives.lu at sk.com>
Subject: Re: [PATCH] nvme-pci: disable write zeroes for SK Hynix BC901

On Tue, Nov 28, 2023 at 11:51:42PM +0000, hardaway.tseng at sk.com wrote:
> Hi Keith, Jim and team,
> 
> Thanks for efforts!
> 
> This is Hardaway from SK hynix, this software change has been tested and verified by ODM Compal, the testing result is pass, therefore, we would like to apply this update.

Great, thanks for confirming. I've got this queued up internally, but want to check if Christoph's question could lead to a better outcome for everyone. For reference, asked here:

  https://lists.infradead.org/pipermail/linux-nvme/2023-November/043298.html

If you were already testing close to upstream, 6.2 or newer, then this patch is already the best option.