Re: [PATCH 4/4]AArch64 sve2: rewrite pack + NARROWB + NARROWB to NARROWB + NARROWT

2022-11-12 Thread Richard Sandiford via Gcc-patches
Richard Sandiford writes: > Tamar Christina writes: >> Hi All, >> >> This adds an RTL pattern for when two NARROWB instructions are being combined >> with a PACK. The second NARROWB is then transformed into a NARROWT. >> >> For the example: >> >> void draw_bitmap1(uint8_t* restrict pixel, uint8_

Re: [PATCH 4/4]AArch64 sve2: rewrite pack + NARROWB + NARROWB to NARROWB + NARROWT

2022-11-12 Thread Richard Sandiford via Gcc-patches
Tamar Christina writes: > Hi All, > > This adds an RTL pattern for when two NARROWB instructions are being combined > with a PACK. The second NARROWB is then transformed into a NARROWT. > > For the example: > > void draw_bitmap1(uint8_t* restrict pixel, uint8_t level, int n) > { > for (int i =

RE: [PATCH 4/4]AArch64 sve2: rewrite pack + NARROWB + NARROWB to NARROWB + NARROWT

2022-11-09 Thread Tamar Christina via Gcc-patches
4 AM > > To: gcc-patches@gcc.gnu.org > > Cc: Richard Earnshaw ; nd ; > > Richard Sandiford ; Marcus Shawcroft > > > > Subject: [PATCH 4/4]AArch64 sve2: rewrite pack + NARROWB + NARROWB > to > > NARROWB + NARROWT > > > > Hi All, > > > > This adds

RE: [PATCH 4/4]AArch64 sve2: rewrite pack + NARROWB + NARROWB to NARROWB + NARROWT

2022-10-31 Thread Tamar Christina via Gcc-patches
iford ; Marcus Shawcroft > > Subject: [PATCH 4/4]AArch64 sve2: rewrite pack + NARROWB + NARROWB to > NARROWB + NARROWT > > Hi All, > > This adds an RTL pattern for when two NARROWB instructions are being > combined with a PACK. The second NARROWB is then transformed i

[PATCH 4/4]AArch64 sve2: rewrite pack + NARROWB + NARROWB to NARROWB + NARROWT

2022-09-23 Thread Tamar Christina via Gcc-patches
Hi All, This adds an RTL pattern for when two NARROWB instructions are being combined with a PACK. The second NARROWB is then transformed into a NARROWT. For the example: void draw_bitmap1(uint8_t* restrict pixel, uint8_t level, int n) { for (int i = 0; i < (n & -16); i+=1) pixel[i] += (p